am Tl^ It) I Book • S ^ /^^<9 ELECTRICIAN'S HANDY BOOK A MODERN BOOK OF REFERENCE A condensed cyclopedia of electricity, more exhaustive than an electrical dictionary, and serving the purpose of an electrical engineer's reference book, in which the general princi- ples are fully treated in an elementary manner. A reference book for the advanced electrician and a text book for the student. BY T. O'CONOR SLOANE, A.M., E.M., Ph.D. Author of "Arithmetic of Electricity," "Standard Electrical Dictionary," Etc. FIFTH REVISED AND ENLARGED EDITION Illustrated by onjer 600 illustrations and diagrams NEW YORK THE NORMAN W. HENLEY PUBLISHING CO. 2 WEST 45th STREET 1920 Copyrighted 1919, 1913 and 1905 BY The Norman W. Henley Publishing Co. Also entered at Stationers' Hall Court, London, England All Ris'hts Reserved PRINTED IN U. S. A. MaK J i id'^0 ^ OCT Z\ «f ^-^ Composition, Electrotypin<^ and Press Work by the Publishers Printing Company, New York PREFACE TO THE FIFTH EDITION. I o" The Electricians' Handy Book has met with such favor ^ from its readers in the past that it is hoped that a new and enlarged edition will meet with the same kind reception. Since it first appeared many changes have taken place in the ^ electrical field, and it is believed that the new matter in the ^ work covers the field of new discovery and development. The ^ present work is rather practical than theoretical; the abstruse theories of the subjects treated in it are not within its scope. The work of treating the whole immense field of electrical engineering from early days to the present time would cer- tainly be an endless one; the work of writing the present book has been lightened by the fact that the progress of electrical science in its practical aspect has been in the di- rection of the survival of the fittest. This tendency has had the effect of removing from the field of engineering many most ingenious devices, whose consignment to oblivion might be a subject for regret. But this disappearance of the old makes the amount to be described and learned less, and thereby lightens the labor of author and student. It is fair to say that the development of electrical engineer- ing is largely in the direction of simplification. In early days results inferior to those attained in the present era were se- cured by the use of apparatus more elaborate than that which is now employed. The evils of complication have long been recognized, and the trend of invention has been to avoid it. One of the earliest objects of the inventor was the production of an arc lamp without mechanism; the results of these efforts have completely disappeared from view, and the simplified mechanical arc lamps of the present day are their successors. The same history can be traced for other branches of the science. Quantities of the most ingenious inventions are no 3 4 ELECTRICIAN^' HANDY BOOK. longer in use, as better machinery has taken their places. For electrical engineering is nothing if not practical, and senti- ment has no part in dictating what shall survive and what shall be forgotten. Something remains to be done in the elucidation of the theory. The very name of the science has never been ade- quately defined, although the working theory has been devel- oped to a high degree of perfection. The greater general fa- miliarity with the mere names of things electrical makes the subject seem less mysterious than formerly, when the words "ampere," *'volt,'' and the like were rarely heard outside of a college. This should not induce the student to feel that his path is any shorter than was that of his predecessors. It is a much longer one, made a little easier by the fact that it is now a better-marked one. But he has more to learn than had his predecessors, and it must be more exactly learned. The modern science cannot be trifled with. This book is sent on its way with the fullest sense of the difficulties involved in its preparation. In its writing the literature of the science and the classics of engineering litera- ture have been freely used. The author^s thanks are also due to many friends for assistance most kindly rendered. OCTOBER; 1919 CONTENTS. CHAPTER I. MATHEMATICS. Electrical Calculations — Algebra — Direct and Inverse Proportion — Percentage — Fractions — Compound Fractions — Inverted Addition and Subtraction — Multiplication and Division — Squares of Numbers — Cancela- tion — Powers of Ten or Exponential Notation — Logarithms — Angular Measurements — Radian System of Angular Measurement — Trigonometric Functions — Numerical Values of Circular Functions — Greek Letters — Useful Constants — Torque — The Dynamometer — The Prony Brake — Lumi- niferous Ether 17-40 CHAPTER II. ELECTRIC QUANTITY AND CURRENT. Electric Quantity — Storage of Electric Quantity — Condensers — Charg- ing — Meaning of Quantity of Electricity — Earthing a Condenser — Capa- city of Condensers — Single Surface Condensers — Lnit of Quantity — The Storage of Quantity of Electricity — Capacity — Dielectrics^Speciflc In- ductive Capacity or Inductivity — Examples of Capacity — Microfarad — Current and Rate Units — Conductors and Non-Conductors — Ether Waves Produced by Electricity — Action of a Conductor — Time Required to Pro- duce a Current — Production of Current — Current, Amperes, and Coulombs — Current Strength or Intensity — Analogy for the Ampere — Speed of a Current — Arrival Curve— Direction of a Current — Memoria Technica — Field of F'orce and Lines Of Force Due to Current — Electromotive Force — Production of Electromotive Force — Dynamic and Static Electricity — Electromotive Force and Energy — Conservation of Electricity — Electro- motive F'orce and the Static Charge — Electromotive Force in Thunder Clouds — Electromotive Force the Cause of Current — Drop of Potential — Analogies of Drop of Potential — Electromotive Force and Difference of Potential — Voltage 41-62 CHAPTER III. THE ELECTRIC CIRCUIT. The Electric Circuit — Constitution of a Circuit — Condensers in a Cir- cuit — Open and Closed Circuits — Circuits Without Appliances — ^^Appli- ances and Generators in Circuits — Electrolytic Conductors — Actions of a Circuit — Parallel and Shunt — Series — Series Multiple — Multiple Series — • Series and Parallel — Outer Circuit — Short Circuit — Conductibility, Con- ductance, and Conductivity — Resistance — Resistance and Energy — The Ohm — Internal and External Resistance — Circuit Without Resistance — Electrolytic Conduction 63-73 5 6 ELECTRICIANS' HANDY BOOK, CHAPTER IV. ohm's law. Three Elements in a Circuit — Ohm's Law — Examples of Ohm's Law — Five Forms of Ohm's Law — Importance of Ohm's Law — Power — Examples — Constant Current Circuit — Constant Potential Circuit — Drop and Fall of Potential — R. 1. Drop and Counter E. M. F. — Examples of Power Calculations — Calculation of Resistance of Parallel Circuits — Examples of R. I. Drop Calculations — Example of Counter Electromotive Force Drop Calculations — Kirchhoff's Laws — Conductance and Cross-Sectional Area of Conductors — Circular Mil System — Application — Area of a Cir- cular Mil — Examples — Wire Gauges — American Wire Gauge 74-87 CHAPTER V. ELECTEO-CHEMISTKY. Water- Decomposed by the Coulomb — Hydrogen Liberated by the Coulomb — Proportion of Hydrogen to Oxygen — Atomic Weights and Chem- ical Equivalents — Electro-Chemical Equivalents — Current Strength and Chemical Decomposition— Sammary — Example — Electromotive Force in Chemical Decomposition — Voltage Calculation 88-93 CHAPTER VI. PRIMARY BATTERIES. The Primary Battery Cell — Three Constituent Parts — Simple Batteries — ■^Nomenclature — Negative and Positive Plates — Cell, Couple, and Pair — Exhaustion and Polarization — Local Action and Amalgamation — Volta's Battery — Volta's Pile or the Galvanic Pile — Wallaston's Battery — Hare's Calorimeter — Zamboni's Pile — Modern Batteries — Smee's Battery — Iron Negative Plates — Aluminium Negative Plates — Grove's Battery — Carbon Negative Plates — Moving Electrodes — Bunsen's Battery — Modifications of Bunsen's Battery — Gibbs' Battery — Poggendorff's Battery — Modifications of Poggendorff's Battery — Fuller's Mercury Bichromate Battery — Cam- acho Cascade Batterv — ^Baadet Siphon Battery — Radiguet Battery — Grenet's Battery — Dip Batteries — Partz's Batterv — Depolarizing Mixtures and Exciting Solutions in Batteries of the Poggendorff Type — Mixtures of Sulphuric and Nitric Acids — Potassium Bichromate Solutions — The Daniell Battery — Modifications of Daniell's Battery — Sand Type of Dan- iell's Battery — Gravity Battery — Meidinger's Battery — ^Modification of the Gravity Cell — Caustic Alkali Batteries — Modifications of the Lalande and Chaperon Battery — Ammonium Chloride Batteries — Dry Batteries — Ar- rangement of Batteries 94-124 CHAPTER VTL STORAGE BATTERIES. The Primary Battery — Action of a Storage Battery — Regeneration — • Grove's Gas Battery — Requirements of a Storage Battery — Function — I*lante's Battery — F'orming — Storage Capacity — Faure's Battery — The Fauro-Sellon-Volckmar Battery — ChemJcal Action — Resistance — Gould Storage Battery — Helios-Upton Batterv. Philadelphia — American Stor- age Battery — Crompton-IIowell Batterv— Pasted Plates — E. P. S. Battery — Chloride Battery — Tudor Battery — Suspended Plates — Other Types of Pasted IMates — Copper Storage Batteries — Zinc Acid Storage Batteries — WaddeH-P]ntz Battery — Edison's Storage Battery — The Discharge — Dis- charge on Open Circuit — ]\ranufac^urer's Data — Determination of Dis- charge — The Charge — 'Specific Gravity Variation of Electrolyte — Hydro- meters — Gassing — (ias Evolution — The First Charge — Automatic Cut-off or Circuit Breaker — English Rule for Charging — Overcharge — Prevention of Sulphating — Short-Circuiting of Single Cells — Sediment — Buckling — CONTENTS. 7 Disintegration — Setting up a Battery — Preparing the Electrolyte — Impuri- ties in the Electrolyte and Tests — Indications from Gassing — Cadmium Plate — Connections for Charging from Lighting Circuits — The Polarity of the Circuit — Taking Out of Service — Cells — Insulation of Cells— Mak- ing Battery Connections — Practical Notes — End Cells^Counter Electro- motive Force Cells — Floating Battery — Charging Plant Operation.. 125-166 CHAPTER VIII. THE FIELD OF FORCE. The Field of Force — Ether and Current— Detection of the Field — Lines of Force Produced by a Curved Conductor — Motion of a Conductor in a Field of Force — Direction or Polarity of Lines of Force — Memoria Tech- nica for Lines of Force — Utility of the Conception of Lines of Force — Density of a Field — The Magnetic Circuit — Energy and the Magnetic Cir- cuit — Counter and Forward Electromotive Force — Building up the Field of Force — Potential Energy of the Field of P^orce — Energy and the Field of Force — Nature of the Magnetic Circuit — Permeability and Permeance — Iron and the Field of Force — Saturation — Three Factors of the Mag- netic Circuit — Magnetic Forc?s — Ampere Turns — Field Density — Permea- bility — Saturation of Iron — No Insulator of Magnetism — The Gauss- Reluctance and Reluctivity — Synonyms for B, H, and mu — B and H Curves — Interpretation — Practical Considerations — Permeability Curves — Soft Steel in Dynamos — Annealing — Determination of Curves — Relation Between Ampere Turns and Lines of Force — Leakage of Lines of Force — Stray Field — Permeance of a Magnetic Circuit — Hysteresis — Residual Magnetism — Hysteresis Curves — Loss of Energy Due to Hvsteresis — Hysteretic Constant \ . . 167-187 CHAPTER IX. MAGNETS. The Electro-Magnet — Tractive Force of the Electro-Magnet — Spreading of Lines of Force — Illustrating Lines of Force About a Magnet — Spiral Electro-Magnet — U-Shaped Electro-Magnets — Annular Chambered Magnet ■ — Electro-Magnetic Tractive Power — Multipolar Magnets — Various Arma- tures — The Natural Magnet — The Permanent Magnet— Action of Magnet Poles on Each Other — Making Magnets by Single Touch — Making Mag- nets by Double Touch — Making U-Shaped Magnets — ^Magnetizing by Coil and Electro-Magnet — Steel for Magnets — Preservation of ^Magnets — Ex- amples of Permanent Magnets — Polarized and Magnetized — Constancy of Magnetism — Mutual Action of Currents — Ampere's Theory of Magnetism — Memoria Technica — Ampere's Theory of Terrestrial Magnetism — Attrac- tion and Repulsion of :\ra2:netic Poles — ^Action of a Current on the Magnet — Ampere's Rule — ^Right-Handed Screw^ Law 188-204 CHAPTER X. INDUCTION. Electro-m.agnetic Induction — Threading, Interlinking, and Cuttting Lines of Force — Induction — Conditions for Inducing Electric Energy — Examples of Interlinking — Motionless Conductor in a F'ield of Force of Varying Density — Energy Relations — Fields of Force in Practice — Direction of Current Induced by Cutting Lines of Force — Two Systems of Induction — Generator Without Motion — Examples of Induction — Telephone Re- ceiver a Dynamo — Laws of Induction — Faraday's Law — Fleming's Rule — Ampere's Rule Adapted to Induction — Clerk-Maxwell's Rule — Lenz*s Law — Examnle of the Application of r^enz's Law — Foucault or Eddy Currents — v^iriations in Impressed Electromotive Force — Direction of Currents Induced in Coils 205-217 8 ELECTRICIANS' HANDY BOOK, CHAPTER XI. DIRECT-CUERENT GENERATORS AND MOTORS. Dynamo-Electric Generators — Interchangeability of Dynamo and Motor — Varieties of Dynamos — Elementary Idea of an Alternating Currenl Dynamo — Collecting or Slip Rings — Brushes — Elementary Idea of a Di- rect-Current Dynamo — Increasing tiie Electromotive Force by Increasing the Turns — Increasing the Electromotive Force by Adding an Armature Core — Armature and Core — Field Poles — Open-Coil Armatures — Spindle or H Armature — Closed Coil Direct-Current Armature — Cutting Lines of Force Without Change in Number of Interlinking Lines 218-224 CHAPTER XII. DIRECT-CURRENT ARMATURE WINDING. Armatures — The Pacinotti Armature — The Gramme Ring — Modern Types of Closed-Coil Armatures — Commutator Connections of Ring Arma- ture — Cores of Ring Armatures — Permeance of the Ring Core — Idle Wire — Circuit in a Ring Armature — Open-Wound Four-Part Ring Armature — Mounting of a Ring Armature — Multipolar Ring Armature — The Drum Armature — Action of the Drum Armature — Drum Armature Winding — Simple System of Armature Winding — Eight Conductor Drum Armature — Twelve Conductor Bipolar Armature — Sixteen Conductor Bipolar Arma- ture — Winding Tables — Windings for Multipolar Fields — Eighteen Con- ductor Four-Pole Armature — Circular Development — Commutator Connec- tions — Wave and Lap ATinding — Wave Winding — Lap Winding — Develop- ment of Commutator Connections — Development of Field Poles — Develop- ment of Current Induced — Straight Developments — Winding a Drum Arm- ature — General Considerations in Laying Out Drum Armature Windings — Single Layer Winding for Bipolar Field — Double Layer Winding for Bipolar Field — Commutator Connections — Multipolar Windings — Multi- polar Lap Windings — Nomenclature for Drum Armature Windings — • General Formulas — Bipolar Winding by Formula — Multipolar Winding by Formula — Lap Winding '. 225-241) CHAPTER XIII. THE Dlf^ECT-CURRENT GENERATOR. The Magneto Generator — The Modern Multipolar Dynamo — Field Wind- ing of Dynamos — Series Winding — Action of Series Winding — Shunt Wind- ing — Action of Shunt Winding — Compound Winding — Short Shunt Com- pound Winding — Long Shunt Compound Winding — Action of Short-Shunt and Long-Shunt Winding — Self-Regulation of Compound Wound Dynamos — Characteristic Curves — Over-Compounding — Example of Compound Winding Calculation — Excitation of Field Coils in Compound Dynamos — Fftect of Independent Excitation of Shunt Coil — Disconnecting or Open- ing the Shunt Coil — Separate Excitation of Shunt Coil — Exciting Series Coils from ^Main Circuit — Separately-Excited Generators — Action of the Separately-Excited Dynamo — Regulation of Separately-Excited Dvnamos and Magnetos — The Separate Circuit Dynamo — Separately and Self-Ex- cited I)ynamo — ^Multipolar Dynamo Connections — ^Conventional Repre- sentation of Machines 250-264 CHAPTER XIY. AR:MATURE REACTIONS. Armature Polarity Due to its Windings — Action of Field Poles on Armature Core —Field Distortion — Armature Reaction Diagrams — Varying Densities of Field — Neutral Points — Brush Adjustment — Demagnetizing rp,j|.jij, — Reduction of Field Density — Action of the Demagnetizing Turns — Dead Turns — Spurious Resistance — Eddy or Poucault Currents — Eddy CONTENTS. 9 Currents in Armature Cores — Eddy Currents in Core Disks — Eddy Cur- rents in Pole Pieces — End Leakage of Lines of Force in Armature — Eddy Currents in Conductors 2G5-272 CHAPTER XV. CHARACTERISTIC CURVES. Characteristic Curves — Horse-Power Lines — Types of Cliaracteristic Curves — Drooping Cliaracteristic — Interpretation of Characteristic Curves — Data for External Characteristic Curves — Data for Total Characteristic Curves — Drawing Characteristic Curves — Internal Characteristic — Termin- ology of Analytical Geometry — Line of Ohms — General Notes on Char- acteristic Curves — Critical Current — Shunt Wound Dynamo Character- istic — Critical Point of Shunt-Wound Dynamo — Total Current Character- istic in Shunt Dynamo — Total Characteristic of Shunt Dynamo — Ohm- Volt Curves 273-284 CHAPTER XVI. THE DIRECT-CURRENT MOTOR. Direct-Current Electric Motor and Torque — Reversibility of Dynamo and Motor — Generator and Motor Connected — Counter Electromotive Force — Action of Counter Electromotive Force — Relations of Speed of Generator and Motor Connected — Counter Electromotive Force and the Armature 285-288 CHAPTER XVII. OPEN-COIL GENERATORS. Open Coil Armature Winding — The Brush Dynamo — Brush Dynamo Construction — The Thomson-Houston Armature — Homopolar, Acyclic or Unipolar Dynamo — Relation of Size and Output of Dynamos — Manu- facturer's and Thompson's Rules — Deduction of Thompson's Factor — The Sixth Power Rule 289-296 CHAPTER XVIII. GENERATOR AND MOTOR CONSTRUCTION. Disks for Smooth Surface Armature Cores — Disks for Grooved Armature Cores — Formed Coils — Wire Winding — Insulation of Conductors — Core Grooves and Wooden Wedgos — Winding Armatures with Formed Coils — Pole Armatures — Disk Armature — Commutator Construction — Position of Commutator — Brushes and Brush Holders — Tangential Brushes — Trim- ming Metal Brushes — Radial Brushes — Position of Opposite Brushes — Brush Rigging — Relation of Depth of Air Gap to Sparking — Field Magnet for Multipolar Dynamos — Laminated F'ield Magnets — Sectional Laminated Field Magnets — Details of Multipolar Field Windings 297-315 CHAPIER XIX. THE ALTERNATING CURRENT. Alternating Electromotive Force — Cycle, Wave and Frequency — Elec- tromotive Force and Current Curve — Production of Alternating 'Electro- m.otive Force and Current — Length of a Wave — Form of Alternating Electromotive Force and Current — Length of Wave and Frequency — Cause of the Form of Alternating Electromotive Force and Current — Alternating Electromotive Force and Current Curves — Drawing the Elec- tromotive Force and Current Curves— Degree Svstem — The Sine Curve — Generating Circle — Interpretation of the Generating Circle — Rate of 10 ELECTRICIANS' HAXDY BOOK. Change — Graphic Representation of Rate of Change — Radius Vector and Resultant — Vector Diagram of a Sine Curve — Phase, Lag and Lead — Angle of Lag and Lead — Quadrature and Opposition — Basis of Lag and Lead — Average Values — Effective Values — Calculation of Effective Values — Form Factor — ^Formulas for Effective Values — Power Factor — Qualities of a Circuit— Resistance — Reactance — Inductance — Inductance and the Henry — Electromotive Force in an Alternating Current Circuit — Counter Electromotive Force — Forward Electromotive Force — Counter and For- ward Electromotive Force in an Alternating Current Circuit — Turns of a Circuit and Inductance — Reactance of Inductance — Ohmic Equivalent of Reactance of Inductance — Inductance Reactance in Subdivided Conductor — Capacity — Reactance of Capacity — Ohmic Equivalent of Reactance of Capacity — Impedance — Electric Resonance — Causes of Lag and Lead — Summation of Alternating Quantities — Composition of Resistance, Induct- ance and Capacity — ^Multiplication of Alternating Quantities — Power Curves — Two-Phase Current — Three-Phase Current 316-347 CHAPTER XX. ALTEEXATIXG CrEBEXT GEXEEATORS. Generation of Alternating Current — Single-Phase Armature — Multipolar Construction — Grouping of Windings — Principle of Alternating-Current Armature Winding — Drum-Armatur? Connections — Elementary Four-Pole Single-Phase Armature — Single-Phase Wave and Lap Winding — Ring Winding for Alternating Current — Conventional Representation of Col- lecting Rings — Pole Single-Phase Armature — Rotor and Stator — Inductor Alternator — Disk Windings — Two-Phase Winding — Three-Phase Winding — Six-Wire Connection of Three-Phase Alternator Winding — Y or Star Connections — Delta or Mesh Connections-Line Connections — Neutral Wire in the Y System 348-362 CHAPTER XXI. ALTEEXATIXG CUEEEXT MOTOES. The Induction Motor — The Rotary Field — Magnetic Needle in a Rotary Field — Armature in a Roiary Field — Three-Phase Induction Motor — In- duction Motors — Rotary and Revolving Field — Starting Torque — Squirrel Cage Armature— Starting Resistances — ^^Starting Compensator — Lenz's Law and the Induction Motor — Construction of Induction Motors — The Synchronous Motor — Condition of Operation — Single-Phase Synchronous Motor — Synchronous I'olyphase Motor — Self-Starting Synchronous Motor. 363-374 CHAPTER XXII. TEAXSFOEHEES. Basis of Transformer Construction — Object of a Transformer — Choking — Limitations of a Transformer — The Principle of a Transformer — Shell or Jacket Type of Transformer — Step-Up and Step-Down Transformers ■ — Ratio of Transformation — Shell-Type Transformer — Core Transformers — Disk-Wound Transformers — Pancake Coils — The Autotransformer — Action of the Transformer — Heat in Transformers — Oil Cooling — Water Cooling — Air Blast Cooling — Details of Transformer Construction — Shell- Type Transformers — Disk Winding — Constant Current Transformers — Oil for Transformers — ^Insulation in Transformers— Direct Current from Alternating Current— Rotary Converter — Use of the Rotary Converter — I*rinciples of Construction — Relation of Voltage and Current — Rotary Converter in the Three-Wire System — Starting a Rotary Converter — F'unctions of a Rotary Converter — The Rectifier — Operation of Trans- formers 375-396 CONTENTS. 11 CHAPTER XXIII. MANAGEMENT OF MOTOES AND DYNAMOS. Starting Motor;? — The Starting Boxes — Magnetic Release Starting Box — Starting-Box Connection — Changing Voltage — Motor Transformer — • Action of the Motor Transformer — Step-Down and Step-Up Transformer — Motor Transformer Practice — The Economy of Motor Transformers — Parallel Coupling of Dynamos — Parallel Coupling of Shunt Dynamos — l»arallel Coupling of Compound Dynamos — Shunt-Wound Machines in Series — Reversal of Direction of Armature Rotation — Polarity Tests — Alternators in Step — Synchronizing — Regulators or Boosters — Booster Connections— Hand Regulation of Boosters — Automatic Regulation of Boosters — Booster Construction — Motor Dynamos as Boosters — Compensa- tors — Floating Batterv — Booster and Storage Battery Connections — Crushers — The Crocker-Wheeler System of Speed Control — Accidents to Motors 397-415 CHAPTER XXIV. CARE OF DYNAMOS AND MOTORS. Reversing the Direction of Current — Stopping a Machine — Too High Speed— Loss of Magnetic Polarity — Wrong Polarity of Field — Refusal of Motor to Start — Slow Speed Without Load — Idle Motors — Speed Regulation of Motor Without Load— Starting and Stopping Motors — Bad Contacts Be- tween Winding and Commutator Bars — Temperature of Commutator — Collector Rings — Materials of Comm.utator — Loose Commutator Bars — • Oval Commutator — A Gummy or Sticky Commutator Surface — Lubricat- ing the Commutator Surface — Brushes and Brush Holders — Brush Pres- sure — Replacing Brushes — Position of Brushes — Copper Brushes — Carbon Brushes— Setting Brushes — Hard Carbon Brushes — Lifting Brushes — End Motion in an Armature Shaft — Short Circuits in Armature — Sparking at the Commutator — Starting a Machine — Starting a Dynamo — Armature Running— Balancing of Armature — Centering of the Armature — Armature Out of Center — Foucault or Eddv Currents — Heatino: of Field Coils- Break in the Field Winding — Short Circuits in Field Winding — Earthing Dynarao Frames — Short Circuits in Outer Circuits — Wrong Connections m Compound Dynamos — Turning Down a Commutator — Sandnanering and Smoothing a Commutator — To Sandpaper a Commutator — Filing a Com- mutator — Short Circuits — Short Circuits Between Armature Windings and Frame — Alternator Brushes — Trouble in Rotors of Alternators-— Self- Starting One-Phase Motor — Loral Heating of the Windinprs of the Stator — Induction Motor Rotors — Svnchronous Motors — Polvnhase Induction Motors — Field Masrnets of Alternators — Two-Phase Operntinn — Break- downs in Transformers — Care of Transformers — Oil for Filling Trans- formers — Moisture in Transformers — Inspection of Transformers— Short Circuits in Transformers 41b-4c}» CHAPTER XXV. STATION NOTES. Temperature of Dynamo or Motor — Cleaning New ISTachine— Tnter- changeabilitv of Parts— Cotton Waste — Access of Air— Oiling— Ring Oilmg —Bearings— Safetv Fuses— Insulation of Windings— Broken T^ ires— Sol- dering— Nails, Tacks, and Iron Filings — Screws in Binding Posts— Cover- ing Machines— Emersrencies and Dansrer Sienals — Forgetfuiness and Neg- ligence— Keep One Hand in Your Pocket— Treatment of ^^^^^^^^ .^J'^^^^^ CHAPTER XXVT. SWITCHBOARDS. Switchboards— Panels— Air Switches— Oil Sy^itches— Overload and Urderload Cut-Outs^Safety Fuses— Overload Circuit Breakers— Under- 12 ELECTRICIANS' HANDY BOOK. load Circuit Breakers — Mag:aetic Release Underload Circuit Breaker — ■ Mechanical Release Underload Circuit Breaker — Reverse Current Circuit Breaker — Combined Circuit Breaker — Circuit Breakers as Switches — ■ Alternating-Current Potential Regulator — Direct Current Ground Indica- tor — Ground Alarm 445-457 CHAPTER XXYII. VOLTMETERS AND AMMETERS. The Voltmeter — Weston's Voltmeter — Damping Coil — Air Vane Damp- ing — Empire Voltmeters — Graduation of Voltmeter Scales — General Isotes on Voltmeters — Cardew Voltmeter — Hot Wire Instruments — The Stanley tlot-Wire Instrument — .Ammeters — Total-Current Solenoid Ammeter — Shunted Ammeter — Transformer Ammeter — Wattmeter — Pressure Line^ or Pilot Wires — Compensated Voltmeters — Compensators — The Ohmic Compensator — The Inductance Compensator 45S-471 CHAPTER XXVIII. DISTRIBUTION. Two Distribution Systems — Arc and Incandescent Lamp Circuits — Con- stant-Current Systems — Constant-Potential Systems — Series Distribution — • Limitations — Features of Series or Constant-Current System for Arc Lamps — Calculations — Advantage of High Potential — Standard Series Lighting Current — Series Incandescent Lighting — Film Cut Out — Relief Lamps — Multiple Series System — "Municipal" Series Incandescent Light- ing — Series-Multiple System — Objections to Series Distribution — Parallel Distribution — Disadvantages of Parallel Distribution — Elementary Case of Parallel System — Potential Drop in Parallel System — Feeders, Mains and Leads — ^Classification — Loop System — Tree System — Closet System — Cylindrical and Conical Conductors — Calculation for Conical Co;nductor — Anti-Conical System — Anti-Parallel System — Individual Voltages of Lamps — Relations of Current to Drop — Uniform Potential Methods- Automatic Regulation of Voltage — Independent Circuits — Feeders — Auxil- iary Feeder Connections — Transfer Bus-Bar — Example — Feeder Economy — Three Wire System — Saving in. Copper — Two-Dynamo Three-Wire Sys- tem — Single-Dynp.mo Three-Wire System — Three-Brush Dynamo — Storage Batteries in the Three-Wire System — Storage Battery Equalizer in Three- Wire System — Balancing Dynamo — Motor and Booster — Five and Seven- Wire Svstem — High-Voltage Parallel Svstems — Alternating-Current Dis- tribution — Individual Transformers- -Choke Coils — Y Connection for Alternating Current — Delta Connection — Joints in Line Wire — Insulators. 472-512 CHAPTER XXIX. ELECTRIC METERS. Electric Meters — Wattmeter— Edison's Meter — Forbes' Meter — Thom- son's Meter — Shallenberger's Meter 513-517 CHAPTER XXX. LIGHTNING ARRESTERS. Lightning Arresters — Lightning Prot'^ctors — Comb or Saw-Tooth. Ar- rester — Magnetic Blow-Out Arrester — Non-Arcing Metal Arre^.ter — Dis- criminating Arrester — Westinghouse Lightning Arrester— Low Equivalent Alternating Current Lightning Arrester— Double-Pole Lightning ArrepJ-r • — Tank Lightning Arrester 518-;i-2 CONTENTS. 13 CHAPTER XXXI. THE IXCAXDESCENT LAMP. Incandescent Lighting — The Incandescent Lamp — Tamidine Filaments — Squirted Filaments — Carbonization — Calibration — Flashing — Occlusion of Gases by Filament — Lowering of Resistance by i^'lasuing — Making Joints by Flashing — Pasted Joints — Electroplated and Other Joints — ■ Leading-in Wires — Making the Lamps — Vacuum — Production of Vacuum — The Mercury Air Pump — Luminescence — Metallic Filaments — Oxide Filament — The Nernst Lamp — The Glower — Glower Terminals — Heaters Ballast — The Cut Out — Direct Current Lamps — Vacuum Lamps — The Efficiency of the Nernst Lamp — Distribution of Light 523-534 CHAPTER XXXII. THE AEC LAMP. The Voltaic Arc — Positive and Negative Carbons — Striking the Arc — Heat of the Arc — Voltage Drop — Counter Electromotive Force — The Re- sistance of the Arc Proper — ^Efficiency of the Arc Light — Quality of Carbons — Power Consumed in Arc — Effect of Air Blast — Effect of Magnet — Voltage Drop and Arc Length — Wearing of Carbons — Arc Light Car- bons — The Direct Current Open Arc — Distribution of Light in Direct- Current Open Arc — Commercial Rating of Arc Light — Hissing Arc — Light Given by Arc Proper — Resistance of Snort Arcs — The Resistance of Longer Arcs — Stationary State — Alternating-Current Arc — Power Factor in Alternating Current Arc — Influence of Wave Form — Distribution of Light of Alternating Current Arc Lamps — Reactance Coil or Economy Coil — Efficiency of Alternating Current Arc Lamps — Noise — Duration of Carbons — Length of Arc — Inclosed-Arc Lamps — The Action of the In- closed Arc — Globe and Carbon Holder — Inclosed Arc Lamp Carbons — • The Clutch — Tripping Platform — Carbon Feed Lamps — Concentric Mag- nets—Dash Pots — Carbon Holders — Constant Current or Series Arc Lamp — Adjusting Weight — ^Action of an Arc Lamp on a Constant Po- tential Circuit — Action of the Resistance Coil in a Constant Potential Arc Lamp — The Parallel-Circuit System of Electric Supply — Constant- Potential Arc Lamps — Management of Inclosed-Arc Carbons — Adjusting Lamps — The Inclosing Globes — Negative and Positive Connections in Inclosed-Arc Lamps — Putting a Lamp Into Service — Oil — Clutcn Stop Adjustment — Cut-Out — Carbons for Inclosed Arc Lamn — To Carbon a Lamp — Lamps Without Mechanism — The Jablochkofif Candle — The Wal- lace Lamp — The Sun Lamp — Open-Air Incandescence 535-562 CHAPTER XXXIII. PHOTOMETRY. Standards of Illuminating Power — Principle of the Photometer — Bar Photometer — Pnotometric Screens — The Bunsen Disk — The Leeson Disk — Mounting the Disks — The Lummer-Brodhun Screen — The Standard English Candle — The Apparatus — Calculating the Scale of the Bar — Tbe Observa- tion — Other Standards— Table of Photometric Standards — Shadow Photo- meter — Bouguer's Photometer — Foucault's Photometer — Direct Photo- metry of an Arc Lamp — The Luminometer — Pupillarv Photometer — Dif- fractive Photometer — Spherical Candle Power— Candle Power of Incan- descent Lamps— The Photometry of the Arc Lamp— Mechanical Equiva- lent of Light — Watts per Candle Power in Arc Lijrht — Watts per Candl'^ lower in Incandescent Lamn— Qualitv of Arc L!4-UU8 CHAPTER XXXV. ELECTRICAL MEASURING INSTRUMENTS. The Galvanometer — Simple Galvanometer- Astatic Galvanometer Fiber Suspension— Reflecting Galvanometer — Arrangement of Reflecting Galvanometer— TranslJcent Scale— Plane Mirror Reflecting Galvanometer The Thomson or Kelvin Galvanometers — Regulation of Sensibility-- The Ballistic Galvanometer — The Deprez-D'Arsonval Galvanometer— Bal- listic Measurement— Ballistic Calculation— The Tangent Galvanometer— The Sine Galvanometer — ^The Thomson or Kelvin Absolute Electrometer — Galvanometer Shunts— Compensating Resistance— Constant of a Galvano- meter—Determination of the Constant — Figure of Merit— Galvanometer Resistance— Siemens's Dynamometer— Rheostats — Resistance Coils— Re- sistance Boxes — Resistance Wire— The British Association Standard Ohm — Arranperaent of Coils— Siemens's Plan— Modern Arrangements— The Decade Plan— Details in Construction of Resistance Boxes— Metal Spools —Practical Notes— Wheatstone Bridge or Bridge Box— Operation of the Wheatstone Bridge— Null Methods— The Meter Bridge— Bridge Key- Shunt to the Galvanoscope— Proportional Coils— Galvanoscope— Condi- tions of Sensitiveness— Diroctior of Deflection— The Potentiometer— Princinle of the Potentiometer— High-Voltage Determinations with the Potentiometer— Current Measurement with the Potentiometer. . .604-641 CHAPTER XXXVI. ELECTRICAL ENGINEERING MEASUREMENTS. Voltmeter Measurement of Resistance— Voltmeter and Ammeter De- termination of Resistance— Low-Resistance ^^^^siu'ements— High Resis^^ ance :Measurements— Line Insulation Tests— Rail-Joint ^^^^^—^{^^.^^^^^^^ ment of Insulation Leakage— Insulation Resistance of ^ J^^^^Vnhle and Cable— Determination of Capacity of a Cable— Galvanoscope C'able and Tine Te^t^ Tpsts of Cable on Reels — Finding ^ ire Ends m a taoie Making BranTh^'conLoH^^^^^ in a Cable-The Telephone as a Galvano- scope-The Vibrating Magneto Bell as « Galvanometer--Vailey T^^^^^^ Test— Hand Magneto Tests— Hand Magneto Tests for Ground— Hand Magneto Test for Cross Connections— Engineering Test^^ b4- doo CHAPTER XXXVII. ELECTROPLATING. CONTENTS. 15 for Silvering- — Gold Plating- — Platinum Plating- — Tin — Steeling — Size of Conductors — Current Intensity — The Relative Position of Anode and Surface to Be Plated — Temperature of Baths — Material of Vessels — Metal Molds — Wax and Stearin Molds — Plaster Molds — Elastic Molds — Gutta Percha Molds — Preparing Molds — Varnish — Oiling — Placing Molds in the Bath — Plating on Molds — Backing Up Deposits — Plating on Glass ■ — Practical Processes 659-676 CHAPTER XXXVIII. TELEPHONY. Sound — Pitch — Fundamental Note — Overtones — Sounding Plate — The Human Voice — Principle of Telephone Receiver — The Telephone Trans- mitter — Invention of the Microphone — Hughes Microphone — The Blake Transmitter — Loose Carbon Transmitters — Running Transmitter — Edi- son's Telephone — The Solid-Back Transmitter — The Receiver — The Telephone Induction Coil — Dimensions of Telephone Induction Coils — Induction Coils in Bracket Telephones — The Telephone Magneto — Po- larized Bell — Telephone Systems — House Connections — Series Telephone Circuit — Bridged Telephone Circuit — The Hook-Switch — Common Bat- tery Systems — Stone's Common Battery System — Dean's Common Bat- tery System — Party Lines.^Polarized Bells for Party Lines — Harmonic Signal for Party Lines — Distributing Boards — Repeating Coils — The Multiple Switchboard — Operation of Switchboards — The Mechanical An- nunciator — Lamp Annunciator — Spring Jacks — Switchboard Connections — Lamp Signal System — Conduction Interference — Induction Interfer- ence — Subscribers' Pole Connections — Improvements 677-717 CHAPTER XXXIX. BELL WIRING. Bell Wiring- — Size of Wire — Fishing — Work under Floors — Racing- — Leading the Wires — Grounding Wires — Soldering — Wires — Distinguish- ing Colors 718-722 CHAPTER XL. ELECTRIC HEATING. Electric Cooking and Domestic Heating — Power Required for Cook- ing — Efficiency — Electric Furnaces — Electric Arc Blowpipe — Electric Soldering Iron — Electric Welding — Electric Incubator — Electric Radia- tor — Economy of Electric Heating 723-730 CHAPTER XLI. WIRELESS TELEGRAPHY. Wave Transmission of Signals — Hertz Receiver — Branly's Coherer — Wireless Telegraphy — Transmitting Apparatus — Receiving Apparatus — Connection of Stations — Antennae and Connections — Marconi's Coherer — Hysteresis and Other Receivers — Detectors — Italian Navy Coherer — The Lodge-Muirhead Coherer — The Stone Coherer — The Fleming- Valve Detector — Electrolytic Detectors — Heterodyne Detector — Gold- schmidts' Tone Wheel — Crystal or Contact Detectors — Spark Gaps — The Disc Discharger — Aerials or Antennse — Couplings — Conductive Coupling — -Magnetic or Inductive Coupling — 'Electric or Capacity- Coupling — Combined Couplings — Loose and Close Couplings — Marconi Sending Plant — Marconi Receiving Plant — Wireless Telephony — Sending- Arrangement for Wireless Telephony — Receiving Arrangement for V/ireless Telephony 731-748 16 CONTENTS. CHAPTER XLII. METALLIC FILAMENT INCANDESCENT LAMPS. Metallic Filament Lamps— Tungsten Filaments— Gas-filled Incan- descent Lamps— Leading in Wires in Incandescent Lamps— The Auer- ?,T.''r^fr,f?^ Attax^hmg Metallic Filaments-Molybdenum Supports of mit^fni VP^^'fiu^Tx''^ the Tungsten Filaments' EfRciency-Photo- rcfrhnn^T^,?o^^"il! t^ Lamps-Overshooting-Helion Filaments-Metallic f^r^^^ Filaments — Osmium Filament Lamps — Tantalum Filament -^^^P^ 749-754 CHAPTER XLIII. VACUUM TUBE LAMPS, FRAME ARC LAMPS, NOTES ON ILLUMINATION AND PHOTOMETRY Vacuum Tube Lamps— The Cooper-Hewitt Mercury Vapor Lamp— ^^AoHo^ ^^f^.;?°'S^''-?^^^^t Arc-Cooper-Hewitt Tube-Visual AcuUy r;v rho.oL^^^ .^""^^T^^. ^".V.^ Lamp— Starting Characteristic— Station- nf^ni?.r^^^^^.^K'^V''""^^^''%"y^°l?^ ^^^^ ^^ Q^^^tz Tube Lamps-Data ^L?e ^^^ ^"i^^ ^"^"^P^J-X^^ Mercury Arc Lamp for Alternating Cur- rents—The MacFarland-Moore Vacuum Tube Lamp— Neon Lamps— ±-, and to get a-, &- must be added to the product. But by the resfular formula the product of the sum and the difference of two numbers is equal to the difference of their squares. A number ending in 5 can be squared or multiplied by itself thus: Multiply the figures next to the 5 on its left by a num- ber one higher, and annex 25 to the product. Thus to square 25 we proceed as follows: 2 is the figure next to 5 on its left; MATHEMATICS, 2^ 3 is the number one higher than 2. 2X3 = 6. Annexing (not adding) 25, we have as the answer 625. To square 165 we mul- tiply 16 by 17, giving 272, and annexing 25 we have 27,225 as the answer. Cancelation. — Cancelation is a process which is rather neg- lected, yet which may be very useful. Suppose we have to divide 1894 by 707. Instead of doing it by long division, we may apply cancelation, thus: 101 270.571 We have divided both numbers by 7, and canceled the original ones. Instead of dividing by 101, we simply diminish 270.571 by 1 per cent, which is done by subtracting from it 1/100 of itself, thus: 270.571 — 2.705 = 267.865, for an approximate result. It is obvious that cancelation is not always of much use. In the above example it is only of value as it enables us to use the percentage method. Often numbers are so intractable that can- celation is quite inapplicable. The essential is that the divisor shall be divisible by some number without giving a remainder. Cancelation always gives a simplification in such cases, but it is often hardly worth while to use it. The limitations of the percentage method must be kept in mind. Often as above the only thingtwhich makes cancelation of value is the applicability of the percentage method. Power of Ten or Exponential Notation. — This adjunct to calculations has become almost indispensable in working with units based on the C. G. S. system. It consists in using some power of ten as a multiple, which may be called the factor. The number multiplied may be called the characteristic. The fol- lowing are the general principles. The power of 10 is shown by an exponent which indicates the number of ciphers in the multiplier. Thus 10- indicates 100; 10' indicates 1,000 and so on. The exponent, if positive, denotes an integral number, as shown in the preceding paragraph. The exponent, if negative, denotes the reciprocal of the indicated power of 10. Thus 10-"-' 1 1 indicates ; 10-^ indicates and so on. 100 1000 30 ELECTRICIANS' HANDY BOOK. The compound numbers based on these are reduced by multi- plication or division to simple expressions. Thus: 3.14X10' = 3.14 3.14 X 10,000,000 = 31,400,000. 3.14 X 10-^ = 10000000 Regard must be paid to the decimal point as is done IOjOUUUoUO here. To add two or more expressions in this notation if the ex- ponents of the factors are alike in all respects, add the character- istics and preserve the same factor. Thus: (51 X 10^) + (54 X 10°) = 105 X 10^ (9.1 X 10-9) + (8.7 X 10-9) = 17.8 X 10-^. To subtract one such expression from another, subtract the characteristics and preserve the same factor. Thus: (54 X 10«) — (51 X 10°)«=3 X 10^ If the factors have different exponents of the same sign the factor or factors of larger exponent must be reduced to the smaller exponent, by factoring. The characteristic of the expres- sion thus treated is multiplied by the odd factor. This gives a new expression whose characteristic is added to the other, and the factor of smaller exponent is preserved for both. Thus: (5 X 10^) + (5 X 10^) = (5 X 10^) + (5 X 100 X 10^) = 505 X 10^ The same applies to subtraction. Thus: (5 X 10^) — (5 X 10^) = (5 X 100 X 10^) — (5 X 10^ = 495 X 10^ If the factors differ in sign, it is generally best to leave the addition or subtraction to be simply expressed. However, by fol- lowing the above rule, it can be done. Thus: Add 5 X 10-2 and 5 X 10^. 5 X 103 = 5 X 10^ X ;^f)-2. (5 y^ ^QB y^ lQ-9^ + (5 X 1 Q-^) = 500005 500005 X 10-2. This may be reduced to a fraction = 100 5000.05. To multiply add the exponents of the factors for a new factor, and multiply the characteristics for a new characteristic. The MATHEMATICS. 31 exponents must be added algebraically: that is, if of different signs the numerically smaller one is subtracted from the other one, and the latter's sign is given the new exponent. Thus: (25 X 10«) X (9 X 10«) = 225 X 10^ (29 X 10-«) X (11 X 10^) =319 X lO-i. (9 X 10^) X (98 X 10-) =882 X 10^°. To divide, subtract algebraically the exponent of the divisor from that of the dividend for the exponent of the new factor, and divide the characteristics one by the other for the new character- istic. Algebraic subtraction is effected by changing the sign of the subtrahend, subtracting the numerically smaller number from the larger, and giving the result the sign of the larger num- ber. (Thus to subtract 7 from 5 proceed thus: 5 — 7 = — 2.) Thus: (25 X 10«) ^ (5 X 10«) = 5 X 10-2 (28 X 10-^) ^ (5 X 10^) =5.6 X 10-n Logarithms. — The use of logarithms can be learned in a few hours. All manuals of algebra give the theory, and the applica- tion with examples is generally given in manuals of trigonometry. The table of logarithms is generally given in the latter manuals, but not in algebras. Logarithms should be taken in the right aspect, as an aid to multiplication and division and extraction of the square root, and as an almost indispensable assistance in extracting higher roots. They assist immensely in arithmetic, and thorough familiarity with them should be acquired. The only point which presents the least difficulty is the charac- teristic — for some obscure reason this is regarded as a sort of obstacle by the beginner. There is even a tendency to omit it altogether in calculations. This tendency is a very bad one. The characteristic should be written out always, because sooner or later cases will arise when its absence will occasion confusion and error. The logarithms of constants are often included in tables of logarithms, and are frequently very useful. A number of tables of logarithms are published in book form. In purchasing one, see that the type and printing are clear. 32 ELECTRICIANS' HANDY BOOK, Angular rieasurement.— A unit circle is a circle whose radius is equal to 1, or whose diameter is equal to 2. Angles are measured by the fractional part of the arc of a circle which they include in their sweep. The arc of an entire circle is divided into 360 parts called degrees, and indicated by a little circle at the top of and following the figures, thus: 45°, 90°, reading "45 degrees," "90 degrees." It will be observed that the angle has no linear measurement, feet or inches for example. The degrees assigned to it express its proportional measure- ment, the whole circle being taken as equal to 360°. Thus 45° are il or i/s of an entire circle, 27° areJI or ^ ^^^ 300 4x) of an entire circle. The length of the circumference of a circle is expressed in terms of its diameter, thus: 7t d, d standing for diameter, and 7t for 3.14159 + . In alternating current formulas, some quantities are used which are what are known as functions of angles. Such are the sine, cosine and tangent. These three are the principal ones employed in alternating current formulas, and are all that will be described here. The cut, Fig. 1, shows a circle. It has two lines drawn across it through the center. Such lines are called diameters. One- half of a diameter measured from the center to the circumfer- ence is called the radius. The angles begin at the right-hand end of the horizontal diameter, and are counted toward the top of the circle, and so all around it against the movement of the hands of a clock. The upper end of the vertical diameter marks the end of an angle of 90°; the left-hand end of the horizontal diameter, an angle of 180°; the lower end of the vertical diame- ter, an angle of 270°; and coming back to the starting point, the right-hand end of the horizontal diameter, an angle of 360°, or one of 0°, according to how it is taken. Radian System of Angular Heasurement.— A radian is the ^ngle measured by the arc of a circle equal in length to the radius. The circumference of a circle of radius 1, which is the unit circle, is 2 ;r, which is equal to 6.2832—. A circle with a radius of 10 inches measures about 62.8 inches around. The circumfer- MATHEMATICS. S3 ence of a circle contains 2 tt radians; a radian is equal to one circumference of a unit circle divided by 2 7t . Radians are shown in Fig. 2; they are the six equal angles which nearly fill the circumference. As the circumference of a circle is equal to 360 degrees, a 360° radian is equal to 360° -^ 2 tt, or =57.3° approximately. 6.2832— When 2 7t appears in a formula, it is generally in the radian system. Fig. ].— Sine, Cosixe and Tangent. Fig. 2.— Radians. Trigonometric Functions. — Fig. 1 is a circle. It is divided into quarters by two diameters, one horizontal and one vertical. The quarters are designated by numbers, and referred to their arcs, which are quadrants. The upper right-hand quadrant is the first quadrant; the upper left-hand quadrant is the second quad- rant; and the lower left-hand quadrant is the third quadrant, and the other is the fourth quadrant. A radius OA prolonged outward determines an angle in the first quadrant. The vertical line from the outer end of the radius to the horizontal diameter is the sine of the angle. This sine is marked AB; the angle is included between the lines OD and O A. 34 ELECTRICIANS' HANDY BOOK. The sine of an angle is always a vertical line, and is always measured up or down, as the case may be. The horizontal line from the outer end of the radius O A to the vertical diameter is the cosine of the angle. The line E A is the cosine of the angle. The tangent of an angle is the vertical line from an extremity of the diameter to the prolongation of the radius marking the angle. For the angle shown, the tangent is the line indicated by the letters D C. The numerical value of the sine divided by that of the cosine gives the numerical value of the tangent. Numerical Values of Circular Functions are expressed in terms of the radius, whicn is taken as 1 except in logarithmic tables, when it is taken as 10'*^. The value when logarithms are dropped is taken again as 1. The value can be applied to a circle of any radius by multiplying it by the radius of the circle in question. Greek Letters.— 7t. This is the Greek letter p. It is best pronounced "pi." If the Continental pronunciation ''pee'' is used, there is danger of confusing it with the English letter p. Sup- pose that a quantity denoted by p is to be multiplied by 7t ; con- fusion would at once ensue if 7t was called ''pee" and not "pi." It indicates the factor by which the diameter of a circle must be multiplied to give the circumference. For approximate calcula- tions its value may be taken as 3 1/7, or what is the same thing, 22/7. If decimals are to be used, 3.1416 or 3.14159 may be used, the latter being accurate enough for almost any purpose. The very usual custom of multiplying the diameter by 3 to get the circumference is so very inaccurate that it should never be used. Take a circle of 37 inches diameter. Multiplied by 3 it gives 111 inches circumference. Multiplied by 3 1/7 it gives 116 2/7 or 116.286 inches circumference. Multiplied by 3.1416 it gives 116.239 inches circumference. Multiplied by 3.14159 it gives 116.2388 inches circumference. Reduced to sevenths, the last two products read between 1/7 and 2/7 for their fractional part. The error in 116 2/7 is only 0.0469 inch, or about 1/20 of an inch. It is evidently un- necessary for every-day work to use the decimal expressions. MATHEMATICS. 35 The exact value of tt has never been calculated. It has been deduced to over a hundred places of decimals. 0. This is the Greek ''th," or theta, a double letter as we would call it in English; it is really an aspirated "f." It is used a great deal in alternating current calculations, to indicate the angle of lag in alternating current work. cp. This is the Greek *'p7i," or phi, an aspirated p; it is used to indicate the angle through which an alternating current wave has advanced from the 0° position. When vector diagrams are used, the measurement begins from the right-hand end of the horizontal diameter. GD. This is the Greek o (long). It is spelled omega and pro- nounced as spelled. It is used to indicate the frequency of an alternating current in radians per second. Let f equal the fre- quency per second of the alternations of a current; then cd = 2 7t f. A single cycle takes 2 tt or 6.2832 radians for its comple- tion. If the numerical value of go in any given case is divided by 6.2832, the quotient will be the number of cycles per second. The product of go by t, or go t, is equivalent to q} in formulas re- lating to alternating current. Useful Constants. — There are certain constants and figures of frequent use which should be memorized. The value of 7€ is one of these. It is approximately 3 1/7, 3/22, or 3.14159. The radius of a circle squared is equal to one-fourth of the square of the diameter. One-fourth of 7t is 0.7854. This is a good figure to remember. The area of a circle is equal to the square of its radius multiplied by tt (3.14159), or to the square of its diameter multiplied 'by 7r/4 (0.7854). Thus a circle of one foot diameter is 0.7854 square foot area; one of two feet diameter is of one foot radius and of 3.14159 square feet area. The factor 0.7854 shows that a circle is approximately 8/10 the area of the square inclosing it. This gives a quick method of approximately finding the volumes of round cisterns and tanks. Suppose a round tank is 12 feet in diameter and 15 feet deep. The area of the inclosing square is 12 X 12 = 144 square feet. 144 X 8/10 = 115.2 the approximate area of the round cis- tern, and 115.2 X 15 = 1728 cubic feet. This is the approximate volume, which can be made quite close to the truth by the per- 36 ELECTRICIANS' HANDY BOOK. centage method. It is evident on inspection that 0.8 exceeds 0.7854 by a little less than 2%. 1728 less 2% is 1728 — 34 r= 1694. The correct answer is a little over 1696. Many other practical factors and quantities may be noted. A speed of one mile an hour is equal to 1.45+ foot or 1 foot 5%+ inches per second. A railway train going one car length per second goes at about 40 miles an hour. One hundred yards in 10 seconds is about 20 miles an hour. The number of 30 foot rails passed over in 20 seconds is the approximate speed in miles per hour. The pull on the draw bar of a car on a level is about 20 pounds per ton per mile an hour. Thus at two miles an hour it is 40 pounds per ton, and so on. A cubic inch of water makes nearly a cubic foot of steam at atmospheric pressure, half this volume at 15 pounds pressure, one-third at 30 pounds pressure, and so on. Water is 816 times heavier than air. A cubic inch of iron weighs nearly one-quarter of a pound; a cubic inch of copper, 0.32 pound; of lead, 0.41 pound. A cubic foot of water weighs about 62% pounds. To reduce kilometers to miles, multiply by 0.6 and add one-thir- tieth. Sixty-two miles an hour is 100 kilometers an hour. 1 kilowatt is equal to a little over 11/3 horse-power (1.3404). 1 B.T.U. (British thermal unit) is equal to 772 foot-pounds. 1 cubic foot of air weighs 537 grains. 1 cubic foot of hydrogen weighs 37 grains. 1 liter of hydrogen (the crith) weighs 0.08961 gramme. Torque. — Torque is force exercised in the rotation of a wheel or similar object, or the force which a rotating wheel or similar object exerts. Thus, in the case of an electric motor its twisting force, or the force with which its shaft is rotated, is its torque. The armature of an active dynamo resists the force which the belt exercises on the belt wheel, and energy or horse-power has to be used to keep it going. This resistance is torque. The strain produced by the belt is driving torque; the resistance offered by the belt-wheel keyed on the armature shaft is the MATHEMATICS. 37 resisting torque, strictly speaking. Of these two terms, the one most used is driving torque only. In a motor the case is re- versed. The armature is drawn around and kept in rotation by the field magnets, and the armature exercises torque, and by means of its torque, and because of it, can drive machinery. In the generator, the belt exercises torque; in the motor, the arma- ture exercises it. If we know the torque and the speed of the machine, we have the actual horse-power. Torque is usually expressed in this country in pounds pull on a one-foot radius, which is that of a 2 foot pulley or belt- wheel. The horse-power exerted by a motor whose speed and torque are known may be calculated by the following formula. In it T indicates torque, H. P. horse-power, r radius of torque, S revo- lutions per minute of the motor shaft. T X r X 6.28 X S H. P. =: 33,000 Suppose the torque exerted by a 4-foot belt-wheel driven by a motor was 10 pounds, and that the motor made 2,000 revolutions per minute. Substituting these figures, the formula becomes: 10 X 2 X 6.28 X 2000 251,200 H. P.= . .— ^ ^1=7.612 33,000 ~ 33,000 actual horse-power. If a machine is rated at a definite horse-power and speed, the torque is calculated by the next formula, which is a transposition of the other. H. P. X 33,000 T=- r X 6.28 X S Suppose a 7% horse-power machine has a speed of 2,000 revolu- tions at full load. To determine the torque on a pulley of 4 feet diameter, which gives rr=i2, we substitute as below: 7% X 33,000 255,750 ^^ ^xX28-^2000 = -25T20- = 1^.18 pounds torque at 2 feet radius. 38 ELECTRICIANS' HANDY BOOK. In these formulas the factor 6.28 is 2 ;r , or the factor by which the radius of a circle must be multiplied to give the circumfer- ence. It is practically accurate to consider the torque of an electric machine identical when run either as dynamo or motor if the speed and current are the same. In many cases it is easy to run a dynamo as a motor. The Prony brake in some of its many forms can be applied, and the torque determined with the simplest possible appliances. The torque developed by the dynamo when run as a motor is taken as that which would be absorbed by it when run as a dynamo. Actual horse-power, or that exerted by a machine, is often S'la. 3.— Prony Brake. called brake horse-power, because it is determined by a Prony brake. The Prony Brake is an apparatus for determining the horse- power of a machine, such as a steam engine, or electric motor, or dynamo. A Prony brake is shown in Fig. 3. A belt pulley is turned by the machine under trial; the pulley is keyed to the shaft M. A strap brake passes around it, armed with wooden shoes. One end of the strap is fastened at D, the other at B'. The latter fastening is adjustable by the screw and hand-wheel S. The arrow indicates the direction of rotation of the wheel. The hand-wheel S is turned until the weight is just held in equi- poise, with the lever between the two stops. A spring balance is often used instead of the weight. The shaft under the con- ditions outlined above rotates with power enough to sustain the weight on the lever or that indicated by the spring balance. Calling the half diameter or radius of the pulley r, and the dis- MATHEMATICS, 30 tance from the center of the shaft to the point of application of the weight L, we have for the turning stress or force, which is torque, of the shaft M: Pull X r ' Torque = L From the torque thus determined and the number of revolu^* tions the horse-power is obtained by the formula below: 6.28 S Horse-power = Torque X 33,000 in which S is the number of revolutions per minute made by the machine. The torque is the force component, the rotation of the shaft is the space component, and the two give energy, and the energy rate is power. The Dynamometer is an appliance which indicates the power a machine at work in exerting, when the speed of the machine is known, indicating directly the force. This force may be exer- cised directly, as when a team of horses is pulling a wagon or when a locomotive engine is pulling a train of cars. A spring balance used as draw bar or coupling link is a dynamometer for such cases. Its reading in pounds multiplied by the speed of the horses or engine in feet per minute, and divided by 33,000, gives the horse-power. If the dynamometer gives the torque or pull of a belt, then the radius of the pulley must be known, and the revolutions per min- ute. Formulas will then give the horse-power. In the illustration. Fig. 4, a transmission dynamometer is shown. It transmits the power of a machine, whence it derives its title, c is a shaft connected by a universal joint c' to the machinery to be driven. The pulley C with inside teeth is keyed to this shaft. It is turned by the pulley B, and B is turned by the pulley A, to whose shaft a a with universal joint the working machine is connected. Noting the directions of rotation indi- cated by the arrows, it will be seen that B driven by A has its axle forced downward. It is acted on with more or less force, ac- cording to the power exercised by the machine. The lever D, on which B is mounted, has a limited range of motion about its ful- 40 ELECTRICIANS' HANDY BOOK. crum at D. This motion is counteracted by the weight P, acting through the lever T on the knife edge e of the lever D. The torque can be taken at any time without interfering with the running of the machine, and without absorbing any of its power. Luminiferous Ether. — ^^This is a theoretical thing whose exist- ence has never been proved. It is assumed to be the cause of the dissemination of light and of the phenomena of electricity. It is best thought of as something like a gas but so much more tenuous that it cannot be detected in any way. It passes through many substances, especially through non-conductors of electricity such as glass. Conductors of electricity are almost impene- trable by it. On this distinc- tion between transparent and opaque bodies, the first not con- ducting electricity and the others l conducting it, is found a basis for SB the theory that light and elec- tricity are closely related. Clerk Maxwell's celebrated electro- magnetic theory of light leads to the same conclusion, and a confirmation for it may be found in the opacity of conduc- tors such as metals and gra- phitic carbon and the transpar- ency of non-conductors such as glass, amber and carbon in the modification known as diamond. This is a general statement, and open to qualifications which it is unnecessary to introduce here. Fig. 4.— The Dynamometer. CHAPTER II. ELECTRIC QUANTITY AND CURRENT. Electric Quantity. — While electricity is about the most in- definable word used in science, we have as a starting point to assume that it is of such a nature as to be susceptible of possess- ing quantity. We have to use the conception of df^finite and definable quantities of electricity without being able to say what we mean by electricity itself. The conception of an electric cur- rent is that of the transfer of quantities of electricity along a wire or conductor, just as in a current of water gallons are transferred through a pipe. An electric current heating the filaments of incandescent lamps, producing the electric arc be- tween carbon terminals, exciting electric magnets and driving powerful motors, is familiar enough. But the idea of quantities, stored up in receptacles, is less so. A quantity of electricity may be stored upon the surface of any insulated body. Coincident with its storage is the storage of another equal quantity of opposite polarity somewhere else. A quantity of electricity cannot be stored or charged upon a surface unless an equal and opposite charge is stored elsewhere. It is something like chemical decomposition. It is impossible to take a quantity of hydrogen from water without producing a corresponding quantity of oxygen, equal thereto in saturating power. Storage of Electric Quantity. — The surface of bodies seems to be the only part concerned in the storage of electricity. The coexistence of two charges and the impossibility of a single charge existing by itself, caused the early investigators to found the two-fluid theory of electricity. Current phenomena are treated more simply by assuming the existence of a single elec- tric fluid. The assumption is therefore made, although rather out of harmony with the phenomena of electric charges. 42 ELECTRICIANS' HANDY BOOK. One of these phenomena is that two oppositely-charged surfaces attract each other, and that their charges tend to combine, form- ing a current while doing so. But the single-fluid versus double- fluid controversy is an academic question; there is certainly no fluid in electricity; and we can speak of a current as of water, or of positive and negative charges as of oxygen and hydrogen in the water molecule ad Wbitum. Condensers. — The typical receptacle for electricity is termed a condenser. It comprises two surfaces adapted to receive and to conduct electricity, insulated from each other. To enable the surfaces to conduct electricity to every part of their area, and to give it up when wanted, they are made of metal. To save space the metal is thin. To separate and keep them insulated from each other, and to modify, owing to a most curious property, their storage capacity, an insulating material is placed between Fig. 5.— The Condenser in Section. them. A sheet of paper as insulator, with a sheet of tinfoil on each side of it, is a condenser. Paper is not procurable of unlimited area, and the same is true of tinfoil. It would also be very inconvenient to have con- densers as big as table-cloths. Acccordingly, to increase the area of the tinfoil it is piled up like the leaves of a book, with paper between the leaves. Every leaf of tinfoil is kept in electric connection with the leaf once removed from it. This brings the tinfoil into two sets, the pieces of each set being in connection with all pieces of its own set and insulated from the other set. The cut, Fig. 5, shows the arrangement in a diagram of its cross section. The dark lines a, a^j and o^ represent one set of sheets of tin- ELECTRIC QUANTITY AND CURRENT, 43 foil, all connected together. The dark lines &, &i, and &2 repre- sent the other set, also connected together. The shaded part intervening represents the dielectric, which may be paper, mica, or glass. In some standard condensers it is simply air, plates of metal being used instead of tinfoil. A and B are the con- ductors, by which it may be charged and discharged. They are twofold, so that one pair can be used for the charge and one for the discharge. One set of sheets receives a positive charge ( + ) when the other receives a negative one ( — ). Fig. 6 shows the way a condenser is built up. It is inclosed in a box with bin'ding posts for the two sets of leaves. Various modifications of connections are applied in practice. Fig. G. -The Condenser. Charging. — If electricity of one kind is poured into or over the surface of one set of leaves of tinfoil, the other electricity must be given some means of accumulating on the other leaves. Therefore, simultaneously with the pouring in of one kind, means must be provided for accumulating another kind. One must be poured over one set of tinfoil, and the other over the other. If a charge is to be given by a galvanic battery, for instance, its opposite terminals, A and B must be connected one to one set, a, a\ a-, the other to the other set, h, &\ &% of tinfoil sheets. In an exceedingly short space of time each set receives its charge. The tinfoil being a conductor, conducts the current everywhere. To discharge the condenser, the oppositely-charged sets of tinfoil are brought into electric contact, the current passes 44 ELECTRICIANS' HANDY BOOK. for an infinitesimal space of time in one direction, and then in diminished intensity in the other, and so beats back and forth like the swinging of a pendulum until the charge is gone, and the opposite electricities have combined. The quantity of elec- tricity which constituted the charge has disappeared. Meaning of Quantity of Electricity. — It would seem that there must be the same quantity of electricity in the condenser after as before the discharge. But a "quantity" of electricity is determinable by and recognized . by its effects. The discharged condenser is perfectly neutral and inert, therefore there is no quantity of electricity in it. Keeping clear of the question of double or single-fluid theory, we may conclude that electric quantity is quite different from hydraulic quantity, which is gallons, liters, or other measure of a fluid. The same is to be said of electric current. It is far different from a current of water. But it is convenient to treat the electric phases of quantity and current as being analogous to quantity and current of water or steam. It is in the actions of water and steam that convenient analogies to electric action are found. Earthing a Condenser. — Another way of charging a condenser is to connect one set of leaves to the earth, or ''earthing" it. The earth is arbitrarily taken as of zero potential. If one kind of electric excitation is imparted to the set of leaves not connected to the earth, the electricity of the same kind is expelled into the earth out of the other set. There is another way of picturing the action, treating the earth as an inexhaustible reservoir of negative electricity, ready to receive negative electricity from one side of a condenser, leaving it positively charged, or to pour in negative electricity, leaving it negatively charged. Capacity of Condensers. — The quantities of electricity which can be stored in condensers are exceedingly small. An incan- descent lamp may use up a coulomb of electricity every second. It would take an enormous condenser to supply it for even a single minute. Such condensers accordingly in practical use are largely employed in the class of electrical work requiring slight currents, such as telegraphy and telephony. The accumulation and instant discharging of quantity following each other in rapid succession play an exceedingly important role in much work in eijEctric quantity and current. 45 modern electricity, where alternating currents accumulaie quan- tity, discharge it, and accumulate it again twenty-five to sixty times in a second. This opens another field for the use of con- densers. The effect of the action is treated of by engineers under the term "capacity." A condenser charged with a quantity of electricity greater or less, as the case may be, can be taken away from its connections and carried about like a pail of water. Electricity could be poured out of it into another condenser, and it could thus estab- lish a current. The distant end of an Atlantic cable might be connected directly to the earth. Then if one set of leaves of a charged condenser were also connected to the earth, a very brief current could be sent through the cable by connecting the other set of leaves to its ungrounded near end. Single Surface Condenser. — A quantity of electricity can be accumulated and held upon any insulated conductor. A piece of tinfoil on the middle of a sheet of glass could be charged with a quantity of electricity. This would at first sight seem precisely the same as pouring water into a receptacle. But the dual element has not disappeared. The charged bit of tinfoil produces an opposite charge on objects around it, on the surface of the experimenter's skin, on the walls of the room, and else- where, there being theoretically no limit to the area affected. The little bit of tinfoil only operates as a container of electrical quantity in conjunction with surrounding objects. It represents one set of leaves of the condenser, the surface of surrounding objects represents the other set of leaves. Unit of Quantity. — A quantity of electricity stored in a con- denser is termed a charge. Poured through a conductor, it pro. duces a current. It is thought of as a measurable thing, and its unit is called the coulomb. A current of one coulomb per sec- ond is called a current of one ampere. One coulomb at a poten- tial of one volt constitutes a unit of energy called the volt- coulomb or joule (pronounced ''joivV'). A joule is equal to nearly one-thousandth of a British thermal unit; 1047 joules have energy enough to heat one pound of water one degree F.; 746 joules would exercise one horse-power for one second. This is not a direct way of estimating the coulomb, because it 46 ELECTRICIANS' HANDY BOOK, is used as a factor of a compound unit, but energy units are so familiar that this method of conceiving of the value of a coulomb is of use. A coulomb of electricity as such is often considered as producing direct results. The most that can be said is that results follow the application of electric energy which vary in direct proportion with the coulombs. A coulomb without asso- ciation with electromotive force can do nothing, and properly speaking cannot be directly measured by its effects, but can be indirectly measured by effects which vary in direct proportion with electric quantity. At Niagara Falls tons of aluminium are produced by electric decomposition of chemical compounds (haloid salts) of alumi- nium. The quantity of metal produced is due to coulombs of electricity passed through the fused mixture containing the aluminium salt or salts. An enormous number of coulombs of electricity are used annually in the production of aluminium. In electro-plating works silver is deposited in greater or less thickness upon tableware and other articles. The quantity of silver deposited depends upon the quantity of electricity used in doing it. One coulomb of electricity deposits 1.134 milli- grammes of metallic silver. It will separate from water about 172 cubic centimeters of a mixture of hydrogen and oxygen gases. This gives a sort of relation between electric quantity and con- crete measures and weights, which makes electric quantity more realizable than it would be without such aids to the imagination. A thunder cloud as one surface, with the earth's surface and the surfaces of all objects thereon as the other surface, can store up quantities of electricity just as a condenser can. A square mile of thunder cloud, at such tension of electromotive force as to be ready to discharge a lightning stroke, need only have a quantity of electricity of seventy coulombs in its charge. Its quantity of electricity would only deposit 80 milligrammes of sil- ver from a plating solution. Coulombs of electricity forced by electromotive force through conductors of properly adjusted resistance produce quantities of heat with accompanying light, of incandescent and arc type. Forced through motors, quantities of mechanical energy are pro- duced, measured by foot-pounds or other unit. In these opera- ELECTRIC QUANTITY AND CURRENT. 47 tions of electric light and power, the energy produced or absorbed is always expressible by a compound unit, such as the foot- pound. These operations are due also to a twofold action of elec- tricity; they are due to potential drop and to quantity com- bined. In the compound units, such as foot-pounds, by which the action of combined potential and quantity is measured, we discern always a potential unit, the foot for instance, and a quantity unit, such as the pound. To them corresponds the com- pound unit of electrical energy spoken of above, the volt-coulomb or joule, 1.356 of which are equal to one foot-pound of mechanical energy. Electric quantity can be measured by things amenable to the simple processes of weighing and measuring. There is danger, on account of this direct proportion existing between electric quantity and the effects of electric energy, that the agency of electromotive force will be overlooked. The coulombs passed through a decomposable solution are di- rectly proportional to the quantity of products of decomposition. But for this decomposition a fixed quantity of electromotive force is required. Therefore a constant value of electromotive force for each case accompanies each decomposition, so the nat- ural tendency is to leave it out of consideration, although the coulomb would be impotent without accompanying voltage or electromotive force. But when heat energy comes into question, the simple ratio disappears, and it is found that heat energy is proportional to volt-coulombs or joules; not to coulombs, but to coulombs raised to the second power. This is the reason why, in referring elec- tric quantity to quantities of physical energy, the joule was used instead of the coulomb on a preceding page. The Storage of Quantity of Electricity involves a factor that applies to the storage of any physical thing, namely, capacity. Capacity is the relative power of storing electricity of a surface or combination of surfaces. Electricity charged upon a surface tends to escape from it and to join that upon the oppositely-charged surface. This tend- ency establishes a potential difference or electromotive force be- tween the two surfaces. 48 ELECTRICIANS' HANDY BOOK. Capacity is defined quantitatively by means of this potential difference. A condenser which will hold one coulomb of elec- tricity at a potential of one volt has a capacity of one farad. It is somewhat as if we should say that a vessel which would hold 5270 grains of air at a pressure of ten atmospheres would have 1728 cubic inches capacity. The weight of the air represents the quantity or the coulombs, the ten atmospheres represent the voltage or the volts, and 1728 cubic inches represent the capacity or the farads. All this is simply an analogy. If the pressure of the air were doubled, the capacity of the vessel would be unchanged, but it would hold twice the quantity of air that it held at the lower pressure. It is manifest that the capa- city of a vessel could not be expressed in grains or other weight of air unless the pressure of the air were specified. A unit of capacity different from the unit of quantity is needed. It is exactly so wuth electric capacity. The potential or elec- tromotive force of its charge must be expressed to define the capacity of a receptacle of electric quantity. This is why differ- ent units are used for capacity and quantity. A measure of a capacity of one gallon holds a quantity of water defined as one gallon, and holds this amount under all circumstances and con- ditions. But an electric measure of fixed capacity, such as a par- ticular condenser, can hold any quantity of electricity until it breaks down and discharges through its dielectrics, puncturing them and destroying its materials of construction, if they are sus- ceptible of injury. Dielectrics. — The substance separating two oppositely-charged conducting surfaces is called the dielectric. It may be any sub- stance which will not conduct the electric current, as otherwise the surfaces w^ould discharge into each other. The nature of the dielectric affects the operation of the condenser, and the effect depends on specific inductive capacity or inductivity. Specific Inductive Capacity or Inductivity. — The nature of the insulating substance or dielectric which separates oppositely- charged surfaces has an effect upon the voltage or potential differ- ence due to a charge of a given quantity. Air and gases are the poorest dielectrics. Sulphur is 3.2 times better than air. Assume two sheets of metal separated by air and brought by a ELECTRIC QUANTITY AND CURRENT. 49 certain charge or quantity of electricity to a potential difference of 3.2 volts. If a layer of sulphur of equal thickness separated them, their potential difference would be only 1 volt. The rela- tive quality of dielectrics in this regard is called Specific Induc- tive Capacity, or Inductivity. The inductivity of some dielectrics is given here. Air, it will be recollected, is 1, and a vacuum about the same. Glass 3.0 to 10.00 Shellac 2.95 to 3.60 Vulcanite 2.50 Turpentine 2.15 to 2.43 Paraffin 1.68 to 2.30 Petroleum 2.04 to 2.42 Beeswax 1.86 Sulphur 3.20 Mica 4.00 to 8.00 The application of these figures is to be seen in the formula for calculating the capacity of a condenser. This formula for microfarads is K = 885 X lO-io X X In this formula a is the area in square centimeters of all the leaves of dielectric between the conducting plates; x is the thick- ness of the dielectric, and k is the inductivity. Examples of Capacity. — The capacity of the earth is only 0.007 farad, or 7,000 microfarads, and that of the sun is 0.076 farad, or 76,000 microfarads. Polarized electrodes immersed in an acid solution have im- mense capacity. Two square inches of platinum electrode im- mersed in dilute sulphuric acid, and polarized a little over 1/50 volt, have a capacity of 175 microfarads. This is the capacity of 80,000,000 square inches of tinfoil or other metal surface separated by % inch of air. If the platinum is more highly polarized, its capacity increases. The polarization is brought about by using them as electrodes for the decomposition of water. Hydrogen adheres to and is occluded by one plate, and oxygen by the other. This establishes a difference of potential between them. The description of Grove's gas battery given elsewhere may be referred to in this connection. riicrofarad. — The farad is too large a unit of capacity for 50 ELECTRICIANS^ ^' HANDY BOOK. ordinary use, so a microfarad, or one one-millionth of a farad, is the standard unit. Current and Rate Units. — The working electrician is so ac- customed to deal with electricity in action, that his mind always turns in that direction. The mechanical engineer deals in many units of energy, such as the erg, foot-pound, and the like; but the electrical engineer instinctively refers to electricity in its effects. A charged condenser does not look a bit different from an uncharged one, though one contains potential electric energy. But an active conductor is surrounded with thermic and other phenomena in the way of force and energy, which make the bringing out of the recognition of its activity by the eye an easy matter. Current intensity is the thing most easily recognized and whose effects are most often witnessed. It is the production of current that is the end and aim of nine-tenths or more of engineering practice. For such reasons as the above the ampere, a unit of rate of quantity transfer, is far more used than the coulomb, a unit of quantity alone. The above shows the origin of a clearly discernible habit of thought among electricians. They, do most of their work with rate units of quantity and of energy. Such units are for rate of quantity, which is current, the ampere; for rate of energy, which is power, the watt. Conductors and Non=Conductors. — The old-time division of substances into conductors and non-conductors of electricity had so much truth in it, that it is preserved to the present time. There is a group of substances that conduct the electric cur- rent well; there is another group that conduct it so badly that they are termed insulators and non-conductors, although every one of them has some conducting power. It is fair to say that between the two extremes thus broadly stated is a field contain- ing comparatively few substances. The majority of substances can be put into one or the other category. Ether Waves Produced by Electricity. — If an electric dis- turbance is produced, the iuminiferous ether is disturbed, waves are produced in it, and the disturbance is propagated through space. For waves to be produced in a medium, it must possess restitutive power. Mechanical waves can be produced in water, ELECTRIC QUANTITY AND CURRENT. 51 because its particles move practically without friction between each other. Any disturbance rectifies itself by the particles working back to their original position and disseminating waves. The absence of intermolecular friction makes restitutive power possible. The force of gravity is the force called on to effect the restitutive action, which restores eventually to their places the particles disturbed by the action which caused the waves. Water is elastic, and without any visible disturbance can propa- gate waves of a totally distinct type, whose production is due entirely to its elasticity and not to its absence of friction or to its weight. Such waves are sound waves. Water conducts sound because its elasticity gives it the restitutive power re- quired for the sound wave. The elasticity of the air makes it also a conductor of sound, and gives it restitutive power for the sound wave. We can hear the hum of an insect high over us in the air, and hardly realize that his minute vocal organs start a series of waves which disturb a mass of air of many tons in weight. The diaphragm of a telephone receiver, acted on by a field due to the irregular current induced by the voice of a distant speaker, is forced into vibrations which reproduce the voice. The elasticity of the iron plate is the restitutive power making possible the starting of sound waves from it as a new center. Action of a Conductor. — If we use the idea of a current in speaking and thinking of electric action, we may picture to ourselves the following representation of the action of a con- ductor. An electric disturbance is produced in ether, and ether waves are set in motion. But just because the ether is restitut- ive, it resists the transfer of anything resembling quantity. Any attempt of quantity to escape from a center through the ether is futile. The elasticity of the ether throws it back on itself. But if a tube were opened through the ether, quantities of electricity could be poured through it, and the choking effect of a restitutive medium being removed, transfer of quantity could take place. This gives us the clue to a useful presentation of the conduction of electric quantity — of the electric current flow- ing through a conductor. 52 ELECTRICIANS' HANDY BOOK. An electric conductor such as a wire of copper, iron, or aluminium, can be pictured as constituting an ether-free cylinder, a tube free from the restitutive ether, and quantities, such as coulombs, of electricity can flow through it. Crude as the above may seem, especially in view of the ion theory, it presents a useful analogy for current transmission by a conductor. Time Required to Produce a Current. — Suppose we had a long tube or pipe through which we began to pump a fluid such as water. It would take some time for the water to reach the end. If a current of electricity is started in a conductor which has some capacity, it takes a measurable time for the current to be appreciable at the further end, and a considerable time before it reaches full strength at the further end. Once it has attained this strength, it can be maintained indefinitely. If the water pipe were inclined a little upward, the water would take a measurable time to reach the end, and would reach it at first as a thin layer, and w^ould require some time to be emitted in full strength. The gradual increase of fiow at the distant end would be still better shown by a pipe which was level, better yet, inclined downward. Let the pipe be inclined downward, and the water would flow under the influence of gravity. It would first trickle in drops or in a relatively small stream from the end, and would only gradually acquire the strength of the entering current. This strength once acquired would be maintained. Production of Current. — The water acts like the electric cur- rent ill the latter case, as its entire mass is acted on by gravity. Every particle is pushed along Individually; it is not merely an end push. A similar action is predicated for an electric con- ductor. The current is pictured as urged through it by action all along the conductor from the surrounding ether. An electric current is not due to a simple end thrust. Current Amperes and Coulombs. — An electric current then is the flow through a conductor of a quantity of electricity caused by electromotive force. As a current is a thing of some dura- tion, frequently of very long periods, we have to define its volume as it passes by us, and say it is of so many coulombs per ELECTRIC QUANTITY AND CURRENT. 53 second, for instance. We can save the enunciation of two words by omitting ''coulombs per second," and saying "amperes" instead. An ampere of current is one coulomb per second. If an ampere flows for one minute, it is the transfer of 60 coulombs; if 60 amperes flow for one second, it is the transfer of 60 coulombs also. The electric current is caused by electromotive force, which is measured by units called volts; it passes through conductors whose relative qualities are generally expressed by stating their relative resistances in units called ohms. We can have a circuit including an electromotive force of ten or any number of volts, and also any number of ohms. Such a circuit may be spoken of from the standpoint of electromotive force or resistance as a ten- volt circuit or a ten-ohm circuit, but neither epithet can be applied to a current. The expression a ten-volt current or a hundred-volt current, once so frequently used, is just as bad a misnomer as such expressions as a ten-ohm or a hundred-ohm current would be. Current Strength or Intensity. — The intensity of a current is measured and dellned by the quantity of electricity it trans- fers in a unit of time. It is the rate of transfer of electric quantity. Its intensity or strength has to be measured in quan- tity-time units, such as coulomb-seconds, which are amperes. The latter word is universally used, as a ten-ampere or twenty- ampere current. The true conception of an ampere has presented such difficulty to many students that it is open to question whether it would not be preferable to use the double unit cou- lomb-second in its place. It is, of course, too late to introduce any such change now. Analogy for the Ampere. — A good analogy for the ampere is the miner's inch. This is a measure of rate of flow of water. It is in universal use in the western mining districts. It is the quantity of water which will pass through an aperture one inch square In a board two inches thick under a head of six inches. The cut, Fig. 7, illustrates the conditions. In one second a miner's inch delivers 0.1937 gallon of water, just as an ampere in one second delivers one coulomb of electricity. The head of water may be taken as representing electromotive 54 ELECTRICIANS' HANDY BOOK. force, and the obstruction offered by the limited size of the hole as representing resistance. Speed of a Current— It will now be evident how absurd is a question often asked: How long will it take for electricity to go through a wire of any given length, such as the Atlantic cable? The first trace of current may go through with the velocity of light, but it will take a measurable time for the current to attain sufficient strength to affect the telegraphic instruments in use on the line. It is not even a auestion of the Fig. T.—The Miner's Inch Analogy of the Ampkr^. velocity of propagation of an electric disturbance— it is a ques- tion of charging a conductor of tangible and perhaps very great capacity. Arrival Curve.— The current's slow growth at the end of a long conductor is indicated by a wave-like curve. In sea cables this arrival curve, as it is called, is rendered more abrupt by the use of condensers. To illustrate how slowly a current may reach its full strength, the Atlantic cable worked directly may be cited. Starting with it uncharged and connecting it as part of a circuit, 108 seconds would be required before the current would attain 9/10 of its full value. In 1/5 second it would attain 1/100 of its full value. Theoretically, an infinite time would be required for attaining the full strength of the original current. ELECTRIC QUANTITY AND CURRENT, 55 This feature of slow growth of current is greatly diminished in extent by the use of condensers, so that the above example is not a practical one. Direction of a Current. Memoria Te^hn'ca. — The idea of a moving of or transferring of electric quantity through a con- ductor implies a direction of the current thus formed. This direction has to be established on conventional grounds. To remember it, we may refer to the galvanic battery for a con- venient memoria technica. In the battery the zinc plate is the active one. The other plate may be pictured to the mind as merely gathering electricity and delivering it to the conductor. The current in the outer portion of a galvanic battery's closed circuit flows from the copper, platinum, or carbon plate to the zinc plate. The letters of the alphabet give the clue, as z, stand- ing for zinc, is the last letter of the alphabet, and the zinc is the last to receive the current. Field of Force and Lines of Force Due to Current. — The ether surrounding a conductor seems to play a part in urging a current through it. In electricity everything goes by recip- rocals, and a current affects the ether which surrounds it. It is thrown into a state of stress, circular lines of force which build up a sort of cylinder around the conductor being formed. Every impulse of electric current that goes through a telegraph wire produces circular lines of force around the wire, somewhat as if it was thickly strung with rings. This occurs for the whole miles of length of the wire. Once produced, the lines of force persist as long as the current lasts. Electromotive Force. — This may be defined as electric pres- sure which under certain conditions causes electric current. It is comparable to the pressure of steam in a boiler, which will force a current of steam through an opening, just as electro- motive force will force a current of electricity through a con- ductor. It is not energy, but appears as a factor of energy in the joule or volt-coulomb, and as a factor of power in the watt or volt-ampere. The practical unit of electromotive force is the volt, and the term voltage is often used as a synonym of electro- motive force, as is also the expression potential difference, drop of potential, or difference of potential. As will be seen later. 56 ELECTRICIANS' HANDY BOOK. there is a distinction to be noted. Electromotive force is often written in abbreviated form as E. M. P. or e. m. f., and is often spoken of by these three letters. Production of Electromotive Force. — It is produced in vari- ous ways. If chemical changes are allowed to take place in obedience to chemical affinity, electric energy is set free, and the e. m. f. constituent of it is produced. Mechanical energy can by the dynamo- or magneto-generator be converted into electrical energy, and electromotive force appears. The economical pro- duction of electric energy, with the inevitable impressment or production of electromotive force, is one great object of the study of the electric engineer. Dynamic and Static Electricity. — Electricity in the mani- festation called a current is treated as dynamic electricity. The current can never exist without the coexistence of electromotive force. Electromotive force can exist w^ithout a current. The latter condition is called static electricity. Electromotive Force and Energy. — It is fair to say that elec- tromotive force is always associated with some form of electric energy. An instance of static electricity is a stick of sealing wax rubbed upon the coat sleeve. This has a very high electro- motive force impressed upon it, and if connected to the earth will produce a current. The electromotive force while in the static condition was a constituent of potential electric energy. This statement is a broad one, but expresses the general condi- tion, and like other broad statements may be open to some modi- fication. One of the most familiar sources of electromotive force is a galvanic battery. On closed circuit this will maintain a current due to electromotive force produced by chemical change. The existence in a battery of chemical combinations or substances whose affinities call for chemical change, shows the presence therein of potential energy. As long as the elements of the battery tend to satisfy their affinities by chemical change, so long will they represent potential energy and will maintain electromotive force. When the battery becomes exhausted, the chemical affinities no longer strive to be satisfied as before, and the electromotive ELECTRIC QUANTIIY AND CURRENT. 57 force disappears simultaneously with the potential energy of the battery. If the battery is on closed circuit, the electromotive force pro- duces a current, and active or kinetic electric energy appears. If the battery is on open circuit, electromotive force is still there as a component of inactive potential electro-chemical energy. Suppose a wire ring cut in one place were moved across the field of an electro-magnet. If this were done in a certain way, electromotive force would be impressed upon the ring. If while so moving its ends were tested, they would cause the reading of a voltmeter to show the presence of voltage on the circuit. The energy element of this combination is purely potential as long as the ring is discontinuous. A voltmeter may, as said above, be used to close the gap, when energy will at once ap- pear, because a current passes through the winding of the volt- meter. This is another example of the association of electro- motive force with potential energy. In the mechanical world the analogous condition obtains. It is hard to conceive of force except in coexistence with po- tential or kinetic energy. A mass of matter, a stone or weight, solicited by gravity represents force, and also potential energy, because if released it will fall and develop kinetic energy. But place it at the center of the earth, and it will no longer tend to fall; it will lose its power to produce kinetic energy, and it will cease to possess weight. Force will disappear because gravity no longer acts upon the body, and simultaneously with the disappearance of force, potential energy will disappear, be- cause the body in its new position can no longer produce kinetic energy. The force centered in the body disappears simultaneous- ly with the potential energy due to that force and to its position with reference to the earth. The comparison excludes cosmic forces: it refers only to terrestrial gravity. Conservation of Electricity. — ^The cause of electromotive force is conveniently referred to the assumption that there are two kinds of electricity, positive and negative. Or if it is de- sired to avoid any revival of the old double fluid versus single fluid controversy, a change in nomenclature will effect it; we 58 ELECTRICIANS' HANDY BOOK. may term the two kinds of electric disturbance positive and negative excitation or charging. A positively-charged body at- tracts a negatively-charged one, and in this attraction is to be sought the cause of current, which cause is electromotive force. Whenever one object is positively charged, an opposite or posi- tive charge is imparted to some other object or objects, which theoretically may be ^even the celestial bodies. This opposite charge is equal in amount, so that a sort of analogy to the doctrine of the conservation of energy, is found in electricity. The algebraic sum of the positive and negative electricities or electrical charges in the universe is equal to zero. This doctrine has been called the Law of the Conservation of Electricity. Electromotive Force and the Static Charge. — The conception of the necessary existence of an opposite charge for every charge of electricity, and of the fact that any object may act in this role, is very important. It tells against the conception of elec- tromotive force as a simple pressure or push, but suggests that it must operate in some way on both extremities of a circuit in opposite senses, or over the whole length of a conductor. Tak- ing an analogy from everyday mechanics, it suggests a bar moved in the direction of its length by a pull at one end and a push at the other end of the bar, given together at the same time. The value of this analogy is to prevent the idea that electromotive force acts only on the end of a conductor pushing electric current through it. Although the action is still the ob- ject of theorizing, it is certain that it is not so simple as that. Electromotive Force in Thunder Clouds. — When a cloud be- comes charged with electricity, the earth becomes charged op- positely. The two tend to combine, and the tendency may be- come so intense under enormously great electromotive force that the opposite electricities combine in a series of currents of inconceivably short duration, and which surge back and forth, also for an infinitesimal space of time, and constitute the light- ning stroke. There the electromotive force may mount into millions of volts, and project a large quantity of electricity through the enormous resistance of air, so as to produce destructive effects. ELECTRIC QUANTITY AND CURRENT. 59 It is no trivial force that splits trees as we see them when light- ning has struck them, especially when we realize that but a small portion of the stroke may have been exerted on the tree, the majority expending itself on reaching the tree through the air. Irregular tubes of melted sand are sometimes found in the earth. These have been formed by the heat of the electric discharge of lightning. A very tangible quantity of heat is needed to effect the melting. When we realize that it is done ia an infinitesimal space of time, it is evident that the rate of heat energy and of electric energy (watts) causing it is very high. The action of electromotive force in the disruptive discharge of electricity, such as that seen in the Leyden jar discharge or in the lightning stroke, is far different from its action in producing an ordinary current such as passes through a wire of an electric circuit. The violent discharge of the jar or of the lightning beats back and forth somewhat like a rebounding ball, but it is the same electromotive force that is operative in producing the minute currents that affect the telephone. The lightning dis- charge, with its oscillations, is comparable to the alternating currents of telephony somewhat as are sound waves in air to light waves in the etlier from the standpoint of frequency. The two are cited as illustrations of the extremes of electromotive force. That of the lightning is almost immeasurable on account of its magnitude, that of the telephonic circuits is the same on account of its minuteness. A lightning stroke a mile in length is calculated to absorb an electromotive force of 5,000,- 000,000 of volts, the telephone current, calculated at about 1/100 of a microampere, requiring an electromotive force of about 1/1,000,000 of a volt for its development. An electromotive force of one volt is a little less than that of a Baniell cell in good order. Electromotive Force the Cause of Current. — The electric current is caused to flow through a conductor by electromotive force. As all conductors possess some resistance, and as a con- stant current* once started moves through each part of the con- ductor with equal intensity, we should anticipate that electro- motive force would be expended in driving the current through each part of the conductor. This is what actually occurs. 60 £jLECTRICIAN8' HANDY BOOK. Drop of Potentaal. — We start at the origin with a definite electromotive force, and it grows less and less as we progress E along the line. Ohm's law (page 18) expressed as R = — tells us that the electromotive force varies with the resistance. Hence if from beginning to end of a conductor a drop of 10 volts is observed, then, for every portion of the conductor of 1/10 its total re- sistance, a drop of 1 volt exists. Analogies of Drop of Potential. — A sim- ple analogy may be taken from a wire. Fig. 8, hanging vertically from a bracket and subjected to twisting at its lower end. To show the action pointers are to be fastened to it at intermediate points of its length, projecting at right angles from the w^ire. As the bottom is twisted, each pointer turns through an arc. The pointer nearest the bottom turns through the longest arc, that nearest the top through the shortest arc, and the intermediate ones through arcs pro- portional in length to their distance from the end. If the degrees through which the pointers move are treated as volts, the drop in volts along a wire conducting a current is illustrated, the twist representing the cur- rent. As the degrees through which the pointers move grow less and less as remoter from the twisted end, so in a conductor the volt- age drops. The current is the same through- out it, and in the twisted wire every part of its length is subjected to an identical twisting strain. Another excellent analogy is shown in Fig. 9. A horizontal pipe conducts water. It has vertical pipes connected to it along its top. The height of the column of water in each of these indi- cates the pressure at that point. It is evident that it will be less Fig. 8.— Torsion Wire Analogy. ELECTRIC QUANTITY AND CURRENT. 61 and less as the outer end of the pipe is reached. The difference of height of any two neighboring water columns indicates the hydraulic drop, an exact analogy of the electromotive force drop. Electromotive Force and Difference of Potential. — There are two terms which are almost synonymous, yet which have a distinction one from the other — electromotive force and differ- ence of potentiaL If a difference of potential is maintained be- tween the ends of a conductor or between any two points on it, a current will pass. The intensity of this current can be de- FiG. 9.— Hydraulic Analogy of Drop of Potential. termined, the resistance of the circuit can be determined, and the product of the two will give the difference of potential. A suit- able instrument of the galvanometer type can be connected to the two points on the circuit, and its reading will give the differ- ence of potential, usually in volts, fractions of or multiples of volts. This is simple enough. A complete electric circuit may next be considered, consisting of a galvanic battery and an outer circuit connecting its terminals. The resistance of the battery is determined, and also that of the outer circuit. On closing the circuit a current passes, and its intensity is determined. On multiplying the sum of these resistances by the current intensity. 62 ELECTRICIANS' HANDY BOOK. we have as before what appears to be a difference of potential. But if we try to determine the difference of potential by an in- strument, such as a voltmeter, we can find no two places to which to connect its terminals, so that it will show the difference of potential we have determined. Its readings are always less. If a number of electromagnetic lines of force are forced to thread themselves through a closed conducting circuit, such as a ring of wire, a current of electricity will pass through it as long as the lines of force increase or diminish in number. The cur- rent will continue to pass as long as any change in their num- ber occurs, and the more rapid the rate of change, the more in- tense will be the current. Multiplying the resistance by the current as before, we get what we might be disposed to term a difference of potential. But on applying our voltmeter, we can find no two points of the circuit between which more than one- half the difference of potential required to account for the current exists. The current is due to the electromotive force. If we could connect a voltmeter to two consecutive points of tne circuit, and force it to indicate the difference' of potential existing between them the long way around, we should find it equal to the electro- m^otive force. But there is no way of doing this. The term difference of potential always indicates the true difference exist- ing between points, which is the minimum one. No hypothetical maximum is allowed for. Electromotive force includes difference of potential as one of its phases and measures. But many cases occur in which it goes beyond difference of potential, and produces a current perhaps twice as great as could be accounted for by simple potential difference. Voltage. — This word is almost a synonym of potential differ- ence, except that it includes the idea of its measurement in volts. Applied to an open circuit, it may be identical with the electromotive force existing in that circuit. CHAPTER III. THE ELECTRIC CIRCUIT. The Electric Circuit. — The existence for any time of a cur- rent of electricity always implies the existence of what is called a circuit. If tw^o surfaces are oppositely charged, they may discharge into each other, but the discharge will last but a minute fraction of a second and will not be a continuous current. A reservation might be made in the case of a circuit actuated by a battery, but the electrolyte of the battery is always treated as a conductor. Constitution of a Circuit. — It consists of a conductor whose ends are connected when in action; when they are disconnected temporarily, it is an open or broken circuit. When completed and connected, so that it forms a re-entrant path for the cur- rent to flow around, it is called a closed circuit. It is called circuit because its ends are to be joined, making a sort of irregular circle, closed loop, or endless path for the current to go through. A straight piece of conducting material, such as a piece of wire or metallic rod, could be used to carry a momentary current or discharge of electricity, but this would not properly be an electric circuit. A lightning rod may offer a perfectly straight path for the discharge of a thunder cloud, and powerful electric currents may surge back and forth through it, currents which would make the metal of the rod fairly explode in a white-hot shower of melted metal, were they not of such inconceivably short dura- tion. This is a conductor only, not a circuit. The galvanic battery, with the conducting wire joining its ends in electrical bonds, gives a continuous, endless path for electrical action. The dynamo with its outer circuit does the 64 ELECTRICIANS' HANDY BOOK. same. The telegraph system or the overhead trolley system, using the earth alone or in part for the return current, is treated as an electric circuit. The earth is taken as representing a con- ductor, although its function may not be strictly that of a con- ductor. Condensers in a Circuit. — A condenser consisting of two con- ducting surfaces, separated by insulating material, operates as an absolute break in the continuity of a circuit. For a continu- ous direct current a condenser in a circuit would open it as effectually as an open switch would. Where short pulses of cur- rent are to be transmitted, condensers may be introduced in the line. This is often done in submarine cable and telegraph prac- tice. These break the circuit for the passage of a consecutive current, but the dots and dashes of the Morse code are better n Fig. 10. - Condensers in a Circuit. transmitted than by a through metallic connection. Such an arrangement, illustrated diagramatically in Fig. 10, is called a circuit. Open and Closed Circuits. — If electric conductivity exists all through the length of the circuit without any break, it is called a closed circuit. A prisoner within it would be closed in by it. To get out he would have to find or make an opening. An elec- tric circuit with such an opening is called an open electric cir- cuit. Once the conception of an electric circuit as a closed ring of conductors is formed, the meaning of open and closed circuit is fixed in the mind. To pull a switch away from its contact point, mechanically speaking, opens the switch. This opens any circuit of which it forms a part, and the circuit becomes an open circuit. If the switch is closed, the circuit becomes a closed circuit. Circuits Without Appliances .—An electric circuit closed and THE ELECTRIC CIRCUIT. 65 with a current passing through it may be composed of a simple conductor without any generator or other appliance in it. A piece of wire with its ends joined, constituting a metallic ring or loop, may become an electric circuit. All that is necessary is to move it across a magnetic field of force, so as to cut lines of force under certain conditions, and a current will go through it, and it will become an electric circuit. Conditions for carrying it out are shown in diagram in the cut, Fig. 11. Fig. 11.— Ring Moving in Field of Force Under Conditions Producing a Current. Appliances and Generator in Circuits. — Current is produced in a circuit by electromotive force impressed upon it, and in very many cases a generator or several, such as dynamos or bat- teries, form part of the circuit. Appliances for utilizing the cur- rent, such as lamps and motors, may also be included. In calcu- lating the resistance of the circuit, all must be taken into account. A galvanic battery may be in circuit with miles of wire in measuring apparatus wound in thousands of convolutions. The battery may include a number of plates of carbon and zinc and half as many separate cups of solution. Or each cup may con- tain two solutions, kept imperfectly apart by porous diaphragms 6G ELECTRICIANS' HANDY BOOK. . or by the difference in specific gravity of the solution. Switches or contact plugs may come in, galvanometers or other apparatus, but the whole, complicated as it may be, constitutes an electric circuit. Electrolytic Conductors. — If we take the case last cited, we see that the current has two kinds of conductors provided for it, one metallic and the other liquid. Through the liquid portion, except perhaps for a very small fraction of the current, no ordi- nary conduction of electricity takes place. As the solution is decomposed, electrical excitation accumulates on the plates and is discharged through the outer circuit by true conduction. Within the battery decomposition of the water group takes place, and electrolytic conduction takes place, something quite distinct from true conduction. An electric circuit may provide true conductors for part of the circuit, and electrolytic conductors for another part. Actions of a Circuit. — A circuit is the seat of three things — • electromotive force, resistance, and current. The current pro- duced in a closed circuit by a given electromotive force is modified by the resistance of the entire circuit, and an identical current exists in all parts of it, whatever the local resistance may be. Although current is not energy, it cannot pass through a conductor except at the expense of energy, and whenever a cur- rent is passing through a conductor, energy is being expended therein. Every part of an active circuit is a seat of energy. This being the case, it follows that in every part of the cir- cuit electric energy disappears and some other form, usually heat energy, is produced in its place. If we take one point of the circuit as our standard of reference or point of departure, as we go from it we should look for a drop of some kind. The direc- tion of an electric current is so very hazy a conception that we cannot prescribe any direction in which a drop should take place. Abandoning a priori deductions, we can go right to the fact. In any portion of an active circuit we shall find an identical current. Between any two points of an active circuit we shall find a difference in potential by using any of the usual measuring instruments. THE ELECTRIC CIRCUIT. 67 When a current is passing through a conductor, the electro- motive force causing it is shown in the existence of a differ- ence of potential. The difference of potential between any two parts of an active circuit is called the drop or fall in potential. We now see how every portion of a circuit carrying a current, which is not energy, is a seat of energy; the drop of potential causing the current is the necessary element. Current multi- plied by potential difference is power or rate of energy, and Fig. 12 —Multiple Aug or Parallel Connection. wherever current exists, a potential difference exists with it. This refers to practical conditions, not to atomic or molecular. Later the conception of the wattless current will be given, and may appear to be somewhat opposed to this statement, but the actual existence of a wattless current is open to discussion. Parallel and Shunt. — "In parallel with," "in shunt with," "in multiple arc," and similar expressions involving these words indicate a division of the conductor into two or more branches v/hich reunite, so that the current is divided among them. 'Branch" applies in the same cases. Fig. 12 shows six appli- ances in parallel, or in multiple arc. Series. — A series connection indicates that one appliance fol- lows another, as shown in Pig. 13. Fig. 13.— Series Connection. Series flultiple. — This indicates a connection in series of such groups of lamps as shown in Fig. 14. Each group has to pass the same current. Multiple Series. — This connection is shown in Fig. 15. It is analogous to multiple-arc connection, and each set of lamps in 68 ELECTRICIANS' HANDY BOOK. series has approximately the same drop of potential if the two main leads are large enough. Fig. 14.— Series Multiple Connection. Fig. ]5.— Multiple Series Connection. Series and Parallel. — The expression three in series and two in parallel indicates that there are a total of 3 X 2 =i 6 appliances arranged in two parallel series of three each, as shown in Fig. 16. Two in series and three in parallel indicates that six appliances ^^— o— o— o— ^ ^^-o— o— o— ^ Fig 16.— Three in Series and Two in Parallel. Fig. 17.— Two in Series and Three in Parallel. are arranged in three parallel series of two cells each, as shown in Fig. 17. This class of expression can be varied indefinitely, as ten in series and five in parallel and the like. Outer Circuit means the portion of a circuit not included in an appliance. Thus a storage battery circuit might include a line of wire, motors, and lamps. The line wire, motors, and lamps THE ELECTRIC CIRCUIT. 69 would be the outer circuit; the full circuit would include them and the battery. Short Circuit. — If from one terminal of a motor a conductor was carried to the other, it would be a shunt for the motor, and if of low resistance compared to the motor, it would **short- circuit" the motor. A conductor of low resistance in parallel with one of high resistance, or in parallel with an appliance ab- sorbing a large drop in potential, is a short circuit for the other conductor or appliance, and is said to short-circuit it. Conductibility, Conductance, and Conductivity. — The prop- erty of conducting electricity is called conductibility. The con- ducting power of any conductor is called its conductance. The specific conducting power, which is the relative power compared with a standard, is termed conductivity. The conductance of a conductor depends on several things. The longer it is, the less will be its conductance; while the thicker or greater in cross section it is, the greater will be its conductance. Anything which lowers the conductivity of a con- ductor affects also its conductance, and in the same way. The conductivity of a conductor is its relative or its specific conduct- ing power as compared with other conductors. It is expressed on the basis of a valuation of the conductivity of the best con- ductor as one hundred. As Ohm's law was originally stated for resistance, the quality of conductance is little used, its reciprocal, which is resistance, being universally used in electrical calculations. It is a pity that this is the case, but units of resistance will always remain in use by the engineer. It has a positive action in the production of light and heat. Without resistance electric lamps and heating effects of the current would be impossible. Resistance. — Resistance is the reciprocal of conductance. It * i is expressed by ; — ■ and is a conception inferior in conductance every way to conductance, but has been so woven into the sci- ence that it will always be used in preference to conductance. No one thinks of a copper wire as an electric resister; a tele- graph line is not laid over miles of country to resist the passage 70 ELECTRICIANS' HANDY BOOK. of electricity. An attempt has been made to create a unit of con- ductance equal to the reciprocal of the ohm. It was proposed by Sir William Thomson (now Lord Kelvin) to give it the rather barbaric name of viJio. In the interest of etymology it was fortunate that it was abandoned, as in the interest of science it is unfortunate. The negative aspect of resistance appears in its definition as the property of an electric conductor by which it opposes the passage of an electric current. Specific resistance is the rela- tive resistance of a material; this should be called resistivity. Resistance is generally used to indicate the resistance of some specific conductor, such as actually in use or liable to be employed in practice. Resistance and Energy. — When a current is passing through an elec- tric circuit, electromotive force has to be expended to drive it through, as the resistance of the circuit op- poses the transmission of current, and the current driven by electro- motive force through a resistance indicates the expenditure of electric energy. The conductor of definite resistance through which the cur- rent is thus forced becomes hot, and this proves that energy has been ex- pended upon it. The energy can only have been obtained through the electric current and electromotive force. Energy results from an electric current passing through a resistance. If different parts of a circuit differ in resistance, the heating effects will be greatest at the points of greatest resistance. Local resistance localizes energy in a circuit. If a conductor through which a current is passing is immersed In a vessel of cold water, it will heat the water. A thermometer whose bulb is in the water will indicate a rise in the tempera- ture. The apparatus (the electric calorimeter) is shown in the cut, Fig. 18. Fig. 18.- -Electric Calori- meter. THE ELECTRIC CIRCUIT, 71 The Ohm. — This is the resistance through which an electro- motive force of one volt will produce a current of one ampere. There has been much difficulty in determining accurately a stan- dard. Mercury at the temperature of melting ice has been the conductor, and the length of a column one square millimeter in cross section, which would give a resistance of one ohm, was determined. Four ohms came into use, of the following designa- tions and length of mercury column: True ohm 106.24 centimeters. B. A. ohm 104.9 centimeters. Board of Trade ohm 106.3 centimeters. Legal ohm 106.0 centimeters. The present standard is the International ohm, the resistance of a column of mercury 106.3 centimeters long at the tempera- ture of melting ice, which mercury weighs 14.4521 grammes. Mercury is of all metals the one most easily purified, and being liquid is unaffected by strain. Internal and External Resistance. — Internal resistance is the resistance of a generator, whether dynamo or battery. Exter- nal resistance is the resistance of the portion of a circuit outside of the generator. Circuit Without Resistance. — Assume that a circuit carries a direct current and has no resistance. It is a purely theoretical conception, and at first sight seems paradoxical. It may be asked what would result were an electromotive force of one volt impressed on the circuit. The first suggestion of a solution would be that an infinite current would result. But an infinite current multiplied by finite electromotive force would give an infinite rate of energy, and this is absurd. The solution lies in a proper appreciation of Ohm's law. In it are linked together three factors — current strength, electromotive force, and resist- ance. Current strength multiplied by electromotive force is taken as representing rate of energy. Resistance is never absent from an electric circuit, and never will or can be. The unit of rate of electric energy made up of current strength and electro- motive force, and called the watt, ceases to be a unit of power unless resistance or its equivalent is opposed to it. A fourth element is omitted from the problem. In the absence of resist- 72 ELECTRICIANS' HANDY BOOK. ance the current would tend to increase indefinitely. As the current increased, it would build up an increasing field of force around the conductor. Energy is required to do this, and by the law of conservation of energy an opposition to the increase of current would result. The formation of a field of force is accom- panied by the development of counter electromotive force, which is electromotive force operating in the reverse direction to the original. Energy is required to increase the strength of a cur- rent in a circuit under these conditions. The current would go on increasing forever, building up an increasing field of force, and energy would be absorbed on the circuit as long as electro- motive force was impressed on the circuit. Electrolytic Conduction. — When two plates of metal or other conductor are immersed in a solution which does not attack them, and are not in contact with each other, if a sufficient potential dif- ference is established between them, a-current may pass. It will pass if the liquid is an electrolyte. An electrolyte is a liquid de- composable by electricity. Even solids are supposed to some extent to be subject to electrolysis. Electrolytic conduction is conduction at the expense of the electrolyte which is decom- posed. Suppose two plates of platinum are immersed in a solution of dilute sulphuric acid. Let the plates be connected to the termi- nals of an electric circuit, and let a difference of potential be established between them. If the difference of potential is less than a volt, a very minute current will pass. Next let the poten- tial difference be increased. Nothing occurs until a certain potential difference is attained, about 1.48 volt, when suddenly a strong evolution of gas occurs from both electrodes and a cur- rent passes, which is many times stronger than the preceding one. It is unnecessary at this place to discuss the ion theory. The old view of electrolysis is still to be considered the practical one. Electrolysis is the separating of a substance into two con- stituents differing from each other in chemical relation. An electrolyte must be a compound substance. By the action of the current it is separated into two unlike substances. It must have such a composition that it can be resolved into two parts. THE ELECTRIC CIRCUIT, ?3 The way the conduction takes place is thus explained: The solution touching one of the electrolytes gives up to it a part of its chemical constituents. The rest combines with the oppo- site constituent of the next layer of solution, displacing its simi- lar constituent, and this takes place all through the liquid until the other electrode is reached. At its surface necessarily there is set free the opposite constituent of the electrolyte. Such- is the old theory, and one which holds its ground with many at the present day. Thus in the case of the acidified water, hydrogen is liberated at one pole, setting free oxygen. This instantly combines with the hydrogen of the next sheet of molecules, setting free its oxygen. The action is repeated until the other electrode is reached, at which oxygen is liberated. Exactly the quantity of hydrogen required by chemical laws to combine with the oxygen is set free. The two gases are liberated in exact chemical rela- tion with each other. If chemically-pure water is used, electrolysis will be greatly reduced. It is probable that with pure water there would be none, but water always contains some impurity, and it is im- possible to perfectly purify it. We are justified, however, in saying that for electrolysis to take place in water, some salt or soluble substance must be present. Water containing a dissolved substance is not the only electro- lyte. Frequently there are substances which when melted by heat become electrolytes. Chlorides and fluorides of the alkaline and other metals are electrolytes when they are melted and are kept in liquid state by heat. Such electrolytes are used in the production of metallic aluminium by the Hall process. From such electrolytes tons of aluminium are precipitated in the works at Niagara Falls and elsewhere. Solutions of metallic salts in water form the electrolyte used for electroplating. The metal is deposited on the article to be plated, and an anode, as it is called, of the metal of the bath is often employed, which is dissolved and keeps up the strength of the solution. CHAPTER IV. OHM'S LAW. Three Elements ia a Circuit.— There are always three things present or to be taken into account in considering the operation of an electric circuit. They are so bound up with its existence as known to us that we cannot eliminate any of them. The first one is current intensity, which is due to the second one, electromotive force, acting against the third one, resistance. They are indicated in formulas by the respective letters C or I for current intensity, E for electromotive force, and R for resistance. Ohm's Lawy following out what has been said in the last few pages, is to the effect that current intensity is equal to electro- motive force divided by resistance. The statement expressed as an algebraical equation becomes: R If the equation be taken in its broadest sense, the exposition of its effect just given covers it. If it be applied to a specific circuit, which therefore is of fixed resistance, it tells us that cur- rent intensity is proportional to electromotive force. If the volt- age in any part of a circuit is doubled, the current will flow with double intensity through that portion. The case may arise where a fixed electromotive force exists and the resistance varies. The equation states that in such a case the current intensity is inversely proportional to the resist- ance. With a fixed electromotive force, doubling the resistance will halve the current intensity, and so on. The statement may be transformed to read thus: the resist- OHM'S LAW. 75 ance is equal to the electromotive force divided by the current. In algebraic form this is expressed as — • -I Following out the same system of interpretation, we deduce the facts that with constant current, the resistance varies with the electromotive force, and that with constant electromotive force the resistance varies inversely with the current, and the current inversely with the resistance. The last case represents the condition of parallel lighting work. By turning on lamps, the resistance of the circuit is low- ered. The plant we may assume to be so organized as to main- tain a constant voltage, therefore we know from Ohm's law that the more lamps we light, thereby reducing the resistance, the more current will be used in inverse proportion to the resist- ance. Finally, Ohm's law may be stated thus: The electromotive force is equal to the resistance multiplied by the current in- tensity, in algebraic form — E = RI. This states that with constant resistance the current intensity varies with the electromotive force, and that with constant cur- rent intensity the electromotive force varies with the resist- ance. Examples of Ohm's Law. — Assume an electroplating bath to be worked at a fixed resistance, and we wish to increase the am- perage of the current passing through it. The voltage must be E increased, because I^— , and wc have assumed that R is in- variable. Assume that a number of lamps are placed in series, and that each one requires the same current. If the number is increased, the resistance of the circuit will be increased. To keep the current constant, the electromotive force must be E increased, because 1= — , and if R is increased, E must also be R E increased, or else the value of the fraction — , and consequently R :6 ELECTRICIANS' HANDY BOOK. the \;^liie of I, will change. The form E ^ R I could be well used here. An example may be given of what may be termed a fallacious case, where Ohm's law seems to fail but does not. Assume a battery of a considerable number of cells connected in series through a circuit of slight resistance. If the number of cells is doubled, and they are kept in series, the electromotive force will be doubled. While only a very slight increase of cur- rent through the circuit will be produced, yet the voltage or electromotive force has been doubled. The fallacy of the deduction that this contradicts Ohm's law lies in the neglect to consider the resistance of the battery. In doubling the number of cells, not only is the electromotive force doubled, but the resistance of the circuit is nearly doubled, so that only a trivial increase of current is produced. Such cases are frequent, and generally as simple as the above. Five Forms of Ohm's Law. — The law can be stated in five forms, three as given — E E Ir.-, R--, E = RI; XV i and the following tw^o — R 1 I 1 j.--^yand---g The first three are those most used; the first one is more used than any of the others. The first group should be memorized if possible. Importance of Ohm's Law. — The consensus of opinion of in- structors in electrical engineering would probably be to the effect that good work has been done if in a three years' course Ohm's law is well instilled in all its bearings into the student's mind. It is infallible and universal; it has no exceptions. Its action may be limited or obscured by other reactions, but it is always in force in electrical circuits. It binds together firmly the three factors of an active electrical circuit. Sir Isaac Newton held when young that there should be no need of studying geometry; that to a properly developed mind it should be obvious. The simplicity of Ohm's law given in the three algebraic forms, with the verbal statement of each and OHM'S LAW. 77 the various interpretations, tell all there is of it. But the stu- dent of electricity cannot exercise himself too much upon it. Reading over these few paragraphs should not be considered equal to the acquirement of Ohm's law. Power. — The product of volts by amperes gives the unit of rate of energy, which is power. From the first and third forms of Ohm's law we get values for I and E respectively — 1=^ and E = RI. Multiplying the first equation by E and the second by I, we have EI =5! and EI = RP U This gives as the expressions for electric energy: ^, R P and E I R The first states that with constant resistance the energy rate or power varies with the square of the electromotive force. The second states that with constant resistance the energy rate or power varies with the square of the current. Other interpreta- tions less useful or at least less used are that with constant electromotive force the energy rate or power varies with the current, inversely with the resistance, and with constant current varies directly with the rQsistance. Examples. — To increase the energy on a circuit operated by a very high resistance generator or battery, the resistance must be lowered. Such a circuit works at approximately constant voltage. To increase the energy on a constant current circuit, the resistance must be increased. The first statement is of merely theoretical value, for the increase of energy will be through the entire circuit, and all that in the battery is of no economic value. Lowering the resistance in this case throws energy into the battery. In the other case, increasing the resist- ance makes the proportion of energy absorbed by the battery or dynamo less. As electric energy is distributed in practice, the law most quoted is to the effect that energy varies with the square of the current. The statement is incomplete unless it states that for it to be true the resistance must be constant. 78 ELECTRICIANS' HANDY BOOK. The great problem which the engineer has to solve is the localization of energy. The energy absorbed in a battery or other generator is lost as far as utility is concerned. The same is to be said of that expended on the transmission line. Constant Current Circuit. — Ohm's law E = R I states that with constant current the electromotive force varies directly with the resistance. If a fixed current is passing through a cir- cuit, the energy rate I E in any part will be increased by in- creasing electromotive force expended on that part. Ohm's law as given above states that lo increase this, the resistance of that part must be increased. But the energy localized by the increase of resistance is heat energy, and such energy is only desired in certain things, such as lamps; in motors, heating is undesirable from several poin^ft of view. It indicates low efficiency, and may do injury. The use of the drop system solves the distribution of energy for all cases on an active circuit, when uncomplicated by special circum- stances. The general law is this: Concentrate the drop of potential where the energy is to be utilized. A motor produces a drop by its counter electromotive force and resistance. The drop due to the first cause is useful, that due to resistance is useless. Energy expended on the resistance is wasted. A typical constant-current circuit is an arc lamp series system. To increase the energy rate on the outer circuit, resistance must be added; the more lamps there are in series, the more energy will be expended. To add resistance so that the energy will be of use, lamps are added in series. To prevent waste of energy oil the line, it is made of size sufficient to give low resistance compared to that usefully contained in the lamps. Constant Potential Circuit. — The constant potential circuit is next to be considered. Let a fixed difference of potential be maintained at the terminals of any apparatus. The formula energy rate or I E = ^ states that with constant electromotive H force the energy rate varies inversely with the resistance. To develop energy in any appliance whose terminals are kept at constant potential, its resistance must be lowered. A typical constant-potential system is a parallel circuit incan- OHM'B LAW. 79 descent lamp system. On this, to increase the energy expended more lamps in parallel are put in operation, thus reducing the resistance. But it is interesting to note that for each number of lamps in operation, this becomes a constant current circuit. Therefore it is subject to the general law that resistance must be concentrated where heat energy is to be utilized. In this case it is in the lamps. The mains and feeders carrying the current to the lamps should be as large as is consistent with the re- cuirements of capitalization. Drop and Fall of Potential indicate the electromotive force expended on any part of a circuit. Thus a 50-volt incandescent lamp has a drop of 50 volts when burning. The terms are synonyms of Potential Difference, Drop may, as in an incan- descent lamp, be due to resistance, or. as in an arc lamp, partly to counter electromotive force. R I Drop and Counter E. M. F.— The first is the fall in potential brought about by resistance. By Ohm's law such drop is expressed by the equation E = R I.. Without a current there is no R I drop; with a given electromotive force the drop varies with the resistance, and is equal to the product of resistance by current strength. This drop is to be distinguished from that produced by counter electromotive force. In charging a storage battery, each cell gives between two and three volts counter electromotive force, and this is almost independent of current strength. This produces a drop which is a counter electromotive force drop. The drop of 110 volts in an incandescent lamp is an R I drop as far as is known; in an arc lamp, the drop is supposed to be a combination of a counter electromotive force and R I drop. To determine the R I drop, the resistance R of the portion of the circuit in which it is to be developed is multiplied by the current strength I ; the result is the R I drop in volts. Thus the R I drop of a 220-ohm lamp passing y^ ampere of current is 220 X i/o = 110 volts. Examples of Power Calculations. — We have seen that the energy exerted by a current through a given resistance is ex- pressed by any of the following expressions: IE=^5!=:RP 80 ELECTRICIANS' HANDY BOOK. The last expression shows that the heating or mechanical equivalent of a current passing through a fixed resistance is proportional to the square of the current. This can be very simply shown by assuming that lamps are to be lighted. Let each lamp be of 100 volts — 200 ohms standard. Such a Tji -i f\r\ lamp will require by Ohm's law I :z= _ _ = 0.5 ampere of it 200 current. If we double the current, we have enough for two lamps in parallel. Two lamps in parallel have half the resistance of one lamp. To get our original resistance, we must put two lamps in series and two in parallel. Double the original current will light these four lamps, giving four times the watts as be- fore. By similar process we will find that to take three times the current without changing the resistance, three lamps in series and three such series in parallel will be required, giving nine times the number lighted by three times the current. Therefore the lamps which can be lighted by currents vary with the squares of the currents at constant resistance. The lighting of a single lamp exacts a definite amperage and voltage. Keeping the amperage constant and varying the resist- ance, it is to be determined how a change in voltage will affect the light given on a portion of the circuit. Here the law of the square does not hold. If we have a 100-volt lamp requiring half an ampere of current and double the voltage, the lamp would give an immensely high illumination, and would burn out in a very short time. If the voltage were doubled, it w^ould be necessary to take care of the increase by putting another lamp in series with the first. The current would remain one-half ampere, but for the double voltage only double the lamps would be lighted. There is no contradiction involved in these two cases. A watt is the product of first powers of electromotive force and current, and the lamps lighted vary with the watts. In the first case, by placing the lamps in parallel the current was increased as many times as there were parallel series of lamps. To keep the resist- ance the same, as many lamps had to be placed in series as were in parallel. This multiplied the voltage by a multiplier expressing the number of lamps in series or in parallel, both OHM'S LAW. 81 being the same. To get the watts expended on the lamps in the first case, the amperes had to be multiplied by the lamps in parallel to get the increased current intensity. The number of lamps in series gave a figure by which the voltage had to be multiplied to give the new voltage. Take the case of three lamps in parallel and three in series, and call the amperes and volts for a single lamp i and e respectively. The watts for a single lamp will then be indicated by e i. There are three lamps in parallel, so the new amperage will be 3 i. But there are also three lamps in series, in order to keep the resistance the same as with one lamp. The voltage therefore for the nine lamps arranged as described is 3 e. The product of the new voltage by the new amperage is — 3eX3t=:9ei, ar nine times the watts required for one lamp. Taking the expression for electric power — if it is interpreted R for fixed resistance, then the power at fixed resistance will vary with the square of the electromotive force. This is the case with the nine lamps. The electromotive force was trebled, the resistance was kept constant, and nine times the watts resulted. To keep the resistance constant, both voltage and amperage had to be increased in precisely similar ratio. There is no con- tradiction involved. Calculation of Resistance of Parallel Circuits. — Suppose three conductors each of 10 ohms resistance are placed in paral- lel. The combined resistance will be one-third that of a single circuit. A bridge three planks wide will be only one-third the obstacle to the passage of a crowd that a bridge one plank wide would be. The combined resistance of the three circuits is ex- pressed by 10/3=: 3.33 ohms. Assume that the three conductors are not of the same resist- ance. Let one be of 5 ohms, another of 3 ohms, and the third of 2 ohms resistance. The combined resistance is found by adding the reciprocals of the resistances and taking the recipro- cal of the sum. The reciprocal of a number is the quotient of 1 divided by the number; the reciprocal of a fraction is the new 82 ELECTRICIANS' HANDY BOOK. fraction having the denominator of the old fraction for numera- tor and the numerator for denominator. The reciprocal of 3 is 1/3; the reciprocal of 3/4 is 4/3. The reciprocals of the resist- ances of the three conductors are 1/5, 1/3, and 1/2, and the sum of these three fractions is 31/30, and the reciprocal of this sum is 30/31 ohm, the resistance of the parallel conductors. Examples of R I Drop Calculations. — The R I drop is equal to the product of the resistance by the current. A 16-candle- power lamp rated to pass i/o ampere of current has a resistance of 220 ohms, and 220 X % == 110 volts, which is the drop. The formula for the drop is E = R I. R I drop varies v^ith the cur- rent for fixed resistances. Take three conductors of 5, 3, and 2 ohms resistance, and a^.sume that a current of 16 amperes is to pass through them. What is the drop? The resistance of the three parallel conduc- tors has been calculated as 30/31 ohms; the current is 16 amperes. The drop is R I == 30/31 X 16 = 480/31 = 15.48 volts = E. By means of the drop the current passing through each one E can be calculated by Ohm's law, I == — . For the 5-ohm conduc- R tor it is 15.48/5 = 3.096 amperes; for the 3-ohm conductor, 15.48/3 = 5.16 amperes; for the 2-ohm conductor, 15.48/2 = 7.74 amperes. As a proof of the correctness of the figures, the three currents thus determined may be added, when they should give a sum of 16 amperes within the limits of the decimal places to which the operation was carried out: 3.096 + 5.16 + 7.74 = 15.996 amperes. Example of Counter Electromotive Force Drop Calcula- tlon. — A drop may be caused by a counter electromotive force. One battery in opposition to another may give the latter. Sup- pose a battery of seventeen Daniell's cells of 1.06 volts each is working against a smaller battery of six similar cells, what is the drop? It is 6 X 1.06 = 6.36 volts, and the working electro- motive force on the system is equal to the difference of the elec- tromotive forces of the two batteries. The first battery has an electromotive force of 17 X 1.06 = 18.02 volts, and 18 02 — 6.36 = 11.66 volts — the net or working electromotive force. Such a drop OHM'S LAW. "•' is usually accompanied by a drop due to resistance. The resist- ance drop varies with the current; the counter electromotive force does not necessarily. Kirchhoffs Laws.-These are extensions of Ohm's law, and are two in number. The first states that if any number of con- ductors meet at a point, and if all the currents flowing to the point are treated as positive, and those flowing away from it are treated as negative, if the potential at the point remains con- stant the algebraic sum of the currents will be zero. The second law states that in a network of conductors forming a closed polygon, with currents flowing through its members, the alge- braic sum of the products of the currents by the resistances for all the conductors is equal to the sum of the electromotive forces. Conductance and Cross-Sectional Area of Conductors. -The conducting power of a conductor for electricity, or its conduct- ance, varies with its cross-sectional area. A wire of one-tenth of an inch cross-sectional area has one-half the conductance of a wire of two-tenths of an inch cross-sectional area. This is true only when the wires are of the same material. Electric conduc- tors are generally of circular section. The areas of two circles of different diameters vary with the squares of the diameters. A wire four one-thousandths of an inch in diameter has sixteen times the cross-sectional area, and consequently sixteen times the conductance or conducting power, and one-sixteenth the re- sistance, of a wire one one-thousandth of an inch in diameter. Circular Hil System —The circular mil system is based oil the considerations stated above. It is a system of stating the size of electrical conductors, based upon the cross-sectional area of a standard circular electric conductor, and has obtained uni- versal acceptance among American engineers. The length of one one-thousandth of an inch is a linear mil, or simply a mil. The area of a circle one one-thousandth of m inch in diameter is one circular mil. The unit of the system is the circular "mil. A wire of copper of commercial purity, one foot long and one circular mil in cross-sectional area, has a resistance of 10.79 ohms at a temperature of 75° F. (24°-C.) This is a wire of one one-thousandth of an inch diameter. 84 ELECTRICIANS' HAXDY BOOK. Application.— »If we know a wire's cross-sectional area ex- pressed in circular mils, we can determine its resistance by sim- ple division. Resistance varies inversely as the cross-sectional area of a conductor. Therefore, if the resistance of a wire of one circular mil cross-sectional area is divided by the circular mils in the cross-sectional area, of another wire of identical length, the quotient will be the resistance of the latter wire. Thus a wire one foot long and nine circular mils in area has one-ninth the resistance of a wire one foot long and one circular mil in area. In ohms the resistance of the larger wire is 10.79/9 = 1.199 ohm. As the cross-sectional areas of wire vary with the squares of their diameters, a wire 3 mils in diameter has nine times the area of a wire one mil in diameter. To determine the cross-sectional area of a wire in circular mils, square the diameter expressed in linear mils or one one- thousandths of an inch. A wire of 1/20 of an inch in diameter is 50/1000 of an inch in diameter. Its cross-sectional area therefore is (50)^ = 2500 cir- cular mils. The resistance of one foot of such wire, if of copper, is 10.79/2500 = 0.00004316 ohm- Area of a Circular flii. — The area of a circular mil is 0.000000785 square inch. As the circular mils in the cross section of a circular wire are equal to the square of its diameter expressed in one-thou- sandths of an inch, the expression ''square of the diameter" may be taken as the synonym of "circular mils," if the diameter is expressed in one-thousandths of an inch or mils. Examples. — Owing to the facts that commercial copper varies greatly in purity, and that very small amounts of impurity affect its conductivity to a considerable degree, there is nothing final about the figure 10.79 ohms given as the resistance of a foot of wire one circular mil in cross-sectional area. Thus Roebling gives 10.51 ohms, at 75° F. (24°— C.) and 10.18 ohms at 60^ F. (15°+ C.) as the resistance of a foot of one circular mil wire. Accepting Roebling's figures, the use of circular mils may be illustrated by some calculations. A wire is 1075 feet long, and is 0.081 inch diameter. What is its resistance? OHM'S LAW. 85 0.081 inch is 81 mils. A wire of 81 mils diameter has a cross- sectional area of (81)- = 6591 circular mils. The resistance of the wire is 1075 X 10.51 ^ 6591 = 18.896 ohms. A wire is 1100 feet long and has a resistance of 4.404 ohms. What is its diameter? Its cross-sectional area is expressed by the formula — 1100 X 10.51 4- 4.404 = 2625.1 circular mills. The diameter in one-thousandths of an inch is equal to the square root of 2625.1. V2625.1 =1:51.233 mils or 0.051233 inch. Wire Gauges. — Various wire gauges are in use. A wire gauge is based upon a series of cross-sectional areas of wires. Each size of such wires is designated by a number. The numbers ordi- narily are consecutive, the lower the number the larger is the wire; thus No. 1 wire is larger than No. 2, and is smaller than No. 0. If the sizes are to be extended beyond 1, the designa- tions are No. 0, No. 00, No. 000, and so on. American Wire Gauge. — The wire gauge generally used in the United States for copper wire is the Brown & Sharpe gauge, usually written **B. & S. wire gauge," or sometimes simply "American wire gauge." At first sight it seems a purely arbitrary scale, but it is not. It is fair to assume that anyone making calculations of resistance of conductors will have the table to refer to. But* there are a few figures which, if remembered, will enable one to operate without the table and yet to express the size of wire in the B. & S. gauge with a close approximation to truth. No. 10 wire is approximately 0.100 in'^h or 100 mils diameter, with a cross-sectional area of 10,000 circular mils. Its resistance per 1000 feet is 1 ohm approximately. The cross-sectional area of a wire of any number in the B. & S. gauge is approximately 1.26 times greater than the one below it. Thus the circular mils of No. 9 wire are equal to those of No. 10 wire, 10,000, multiplied by 1.26 = 12,600 circular mils approximately. The circular mils of No. 8 wire are equal to 12,600X1.26 = 15,876 circular mils. The circular mils of No. 7 wire are equal to 15,876 X 1.26 = 20,003 circular mils. All these are close approximations to truth. This brings out another feature. A wire three numbers away 86 ELECTRICIANS' HANDY BOOK. from another wire has about double its cross-sectional area if of lower number. Thus, taking 10,000 circular mils as the cross- sectional area of No. 10 wire, we have No. 7 wire, three numbers lower. Its cross-sectional area is twice that of No. 10 wire, or 3 0,000X2 = 20,000 circular mils. No. 4 wire is three numbers lower than No. 7. Its cross-sectional area is 20,000 X 2 = 40,000 circular mils. From the above it follows that by going to higher numbers for a difference of three numbers, we must divide by 2. Thus, No. 13 wire is three numbers higher than No. 10. Its cross-sec- tional area is equal to 10 ^ 2, or 500 circular mils. No. 16 wire is three numbers higher than No. 13 v/ire. Its cross-sec- tional area therefore is equal to 5000 -i- 2 = 2500 circular mils. These figures are only approximate, but are well within practi- cal limits. The following give the true cross-sectional areas of numbers differing by 3 and those determined by the approxi- mate method. B. & S. GAUGE. No. of Wire. 000 2. Areas in Cir ul Mr Mils. Tru(\ 168,100 3. Are sin Circular Mils. Approximate. 160,000 4. Approx. -f 4 Per c nt. 166,400 5. Errors in C. 1 4 Per cent. 1.02 1 83,521 80,000 83,200 0.40 4 41,626 40,000 41,600 0.06 7 20,736 20,000 20,800 0.31 10 10,404 10,000 10,400 0.04 +3% 13 5,184 5,000 5,150 0.62 16 2,601 2,500 2,575 1.00 19 1,296 1,250 1,288 0.62 22 640.1 625 644 0.61 25 320.4 312.5 321.9 0.46 28 158.8 156.2 160.8 1.26 31 79.2 78.1 80.4 1.52 34 39.7 39.0 40.2 1.26 The error in the above approximate process it will be seen varies from less than 2 per cent to over 4 per cent. If for wires I OHM'S LAW. 87 larger than No. 10, 4 per cent is added to the approximate sizes,, and if for wires smaller than No. 10, 3 per cent is added, the results will be well within working limits. This is done in the fourth column of the table, and it will be seen that the error is generally less than 1 per cent. For No. 000 wire it is 1.022 per cent; for No. 10 wire, it is 0.04 per cent; for No. 34 wire, it is 1.26 per cent. To find the size of intermediate numbers, multiply the circular mils of wire of any number by 1.26 to get the circular mils of the next larger wire. Thus, the size of No. 3 wire is obtained approximately by multiplying the circular mils of No. 4 by 1.26. 41,600 X 1.26 =z 52,416 circular mils, the size of No. 3 wire. Multiply the circular mils of wire of any number by 1.60 ta get the circular mils of the second next wire. Thus, the size of No. 2 wire is obtained approximately by multiplying the cir- cular mils of No. 4 by 1.60. 41,600 X 1.60 = 66,560 circular mils, the size of No. 2 wire. The sizes of No. 3 and No. 2 wire given in the table are 52,634 and 66,373 circular mils. The degree of approximation is very good. Especially is this true when it is remembered that no two samples of copper have the same conductivity, and that the temperature variation is considerable in copper. The figure 1.26 is the cube root of 2. The figure 1.60 is 1.26 X 1.26. CHAPTER V. . ELECTRO-CHEMISTRY. The Basis of Electro=Chemistry. — When one coulomb or' electricity passes through water, it liberates 0.0105 milligramme of hydrogen. The chemical equivalent of hydrogen is 1, that of oxygen 16. In one molecule of water there are 2 atoms of hydrogen and 1 atom of oxygen, or by weight 16 parts of oxygen and 2 parts of hydrogen, a total of 18 parts by weight. The 0.0105 milligramme of hydrogen was derived from a certain quantity of water, which bears the same proportion to 0.0105 that the chemical equivalent of the water molecule, which we have seen is 18, bears to the sum of the equivalents of hydrogen In its molecule, which are 2. This gives us the proportion: 2 : 18 : : 0.0105 : x = 0.0945 milligramme water. Hydrogen Liberated by tlie Coulomb. — This tells us that one coulomb of electricity decomposes 0.0945 milligramme of water. Not only is hydrogen set free, but oxygen also. The oxygen can be got by a similar proportion, based on the proportion or' hydrogen to oxygen in the water molecule, which as we have seen is 2 to 16. 2 : 16 : : 0.0105 : x = 0.0840 milligramme oxygen. This result could have been more simply reached by subtract- ing the hydrogen liberated from the water decomposed. 0.0945 — 0.0105 =: 0.0840 milligramme oxygen. Every element is liberated from a compound in strict propor- tion to the coulombs which pass through it by electrolytic con- duction. Proportion of Hydrogen to Oxygen. — We have seen that one coulomb liberates different quantities of oxygen and hydrogen. ELECTRO-CHEMISTRy . 8j> Hydrogen has a chemical equivalent 1/16 that of oxygen. Yet a coulomb liberates only eight times as much instead of sixteen times as much. This is because oxygen is a dyad, which means an element of double combining values, which is its valency,, and hydrogen is one of single combining value, a monad. To get the relative amount of oxygen liberated for a given amount of hydrogen, we might have divided its chemical equiva- lent by 2, the figure of its valency, and put our second proportion directly thus: 1:8:: 0.0105 : x = 0.0840 milligramme oxygen. Atomic Weights and Cliemical ^Equivalents. —In any chem- istry will be found a table of atomic weights and sometimes of chemical equivalents. If the table is properly arranged, each element's valency will be stated. If this information is omitted from the table, it can be found in the text of the book. To find how much of any element will be separated from its combination by a CQulomb of electricity, divide its atomic weight by its valency, and multiply by 0.0105. Electro»Clieniical Equivalents. — Numbers thus obtained are called electro-chemical equivalents. Suppose the electro-chemical equivalent of nickel is required. The atomic weight of nickel is 58.8, its valency is 2, or it is a dyad or is a bivalent. These are three ways of expressing the same fact. We divide its atomic weight by its valency, and multiply the result by the electro-€hemical equivalent of hydrogen: (58.8 ^ 2) X 0.0105 = 0.3087 milligramme which is the electro-chemical equivalent of nickel, or the quan- tity which one coulomb can separate from a solution. If gold is in question, we find its atomic weight to be 197 and its valency 3. Its electro-chemical equivalent is given by the equation : (197 -f- 3) X 0.0105 = 0.6894 milligramme which is the electro-chemical equivalent of gold. If an electro-plater is paying for his electricity by the ampere- second, or, what is the same, by the coulomb, it is of importance for him to know how much his plating costs him in electric current. This he finds out from the electro-chemical equivalent 90 ELECTRICIANS' HANDY BOOK. of the metal he is depositing and the total weight which he deposits. Current Strength and Chemical Decomposition. — The elec- tro-chemical equivalent of silver is 1.134 milligramme. If the strength of a current given by a battery is to be determined, it may be passed through a solution of silver for a known number of seconds. The silver which it separates is weighed, the weight is divided by the seconds of time during which it passed and by the electro-chemical equivalent of silver. The result is the strength of the current in amperes. Silver Voltameter. — The apparatus outlined above is one of the classics of electricity, and is known as the silver voltameter. Summary* —The statements just given may be conveniently summarized. The weight z of an element set free by one coulomb of elec- tricity, calling atomic weight A W and valency V 1, is given by the equation: z = ^^ X 0.0105 milligramme. A current of intensity I will deposit a weight P of an element per second according to the equation: P = ;^ I — 4l^ X 0.0105 milligramme. One ampere hour is equal to 3600 coulombs; it will therefore liberate 3600 X 0.0105 = 37.8 milligrammes of hydrogen, and ^^- X 37.8 of any other element. This is stated very simply. Those who are interested in electro-chemistry should study up the theory of chemical equa- tions and of chemical arithmetic (stoichiometry) also. Example. — A chemical equation may be now written out, and the electro-chemical equivalents calculated. A bath of copper sulphate has a current of 9 amperes passed through it for 35 minutes; how much copper and sulphuric acid will be produced? The chemical formula for copper sulphate is CUSO4, for sul- phuric acid H.SO4, for water HoO. The decomposition is ex- pressed by the chemical equation: CUSO4 + H,0 = Cu -f H,SO, + 0. ELECTRO-CHEMISTRY. 31 The nascent oxygen would usually be caused to attack an anode, but for our purposes we will assume that it escapes. The atomic weights needed are the following: Copper, Cu, 63; sulphur, S, 32; oxygen, O, 16; hydrogen, H, 1. Copper is a dyad. The copper precipitated by a coulomb is — -^ X 0.0105 = 0.3307 milligramme. 35 ampere minutes is equal to 60 X 35 = 2100 coulombs. The 35 ampere minutes will precipitate 0.3307 X 2100 = 694.47 milli- grammes of copper. The molecular weight of the sulphuric acid is obtained by adding together the atomic weights of its con- stituents. These weights are 2 -f 32 + 64 = 98; and for every 63 parts of copper precipitated, 98 parts of sulphuric acid are set free. This gives for the total sulphuric acid the proportion: 63 : 98 : : 694.47 : x = 1080 milligrammes sulphuric acid. Electromotive Force in Chemical Decomposition. — Electro- motive force does not enter into these calculations. It requires a definite amount of electromotive force to break up each chemi- cal combination. For some of them it varies exceedingly little; if it varied more, it could be used as a basis for methods of chemical separation in analysis. Energy in Chemical Decomposition. — As electromotive force is required to break up an electric combination; and as the quantity decomposed or broken up varies with the coulombs, the energy expended varies with both these factors, and the energy rate or power with the volt-amperes or watts. Watts multiplied by seconds give volt-coulombs, and watt-seconds or volt-coulombs multiplied by 10,193.7 gives gramme-centimeters of energy. Calling coulombs Q and electromotive force E, we have for the energy expended in a chemical decomposition ex- pressed in mechanical units of weight and height: Q E X 10,193.7 gramme-centimeters and QE X 0.101937 = kilogramme-meters. The energy expended in decomposition is here expressed in pure mechanical units. The weight of substance decomposed by Q coulombs from what we have seen is Q z. The energy expended may also be expressed in heat units, say in grammes of water heated 1° Centigrade, or calories, some- 92 ELECTRICIANS' HANDY BOOK. times called small calories. The mechanical equivalent of heat is 0.424 kilogramme-meter per calorie. The weight of the sub- stance Q z multiplied by the calories H t corresponding thereto and multiplied by 0.424 will give kilogramme-meters of energy, the expression being 0.424 Q z Ht. This has the same value as the other expression; they can be equated thus: 0.424 Q z U t = 0.101937 Q E, which reduced so as to give the value of E by successive steps is — E = ^-"^^^ z Ut = 4.16 z B.t. 0. L01937 The above equation gives the value of E, or the electromotive force required to decompose a compound. We must know ;:;, which is in grammes 0.0000105 (= 0.0105 milligramme) and H t, which is the heat of combination or the thermo-chemical equiva- lent of the compound dealt with. The quantities of heat expressed in thermal units are termed thermo-chemical equivalents. They have been determined for a great many chemical combinations, and are expressed in gramme-degrees C. or kilogramme-degrees C. The above formula has been deduced for gramme-degrees. It is merely a question of where the decimal point shall be. If the equation is to hold for kilogramme-degrees, it must be shifted to represent one thousand times the quantity; 4160 must be substituted for 4.16 in the expression. Voltage Calculations. — Suppose it is asked what voltage will be required to decompose water. Consulting the table, we find that one gramme of hydrogen in burning, i. e., in forming water, produces 34,450 gramme-degree calories. The electro-chemical equivalent of hydrogen is 0.0000105 gramme. Substituting in the equation we have: E z= 4.16 X 0.0000105 X 34,450 = 1.5047 volts. Thermo-chemical equivalents are expressed in two ways. One is the heat liberated by the combination of a gramme of the substance. The other is the heat liberated by grammes equal in number to the chemical equivalent of the substance. Some- times one and sometimes the other is given in tables. If the ELECTRO-CHEMISTRY. 03 first is used in the calculation, the form of equation given for hydrogen is adhered to. E =: 4.16 X electro-chem. equiv. X thermo-chem. equiv. If the latter is used, the factor 0.0000105 has to be retained. 4.16 X 0.0000105 = 0.0000437. This factor may be kept as a constant, and we have for the second form of thermo-chemical equivalents : E == 0.0000437 X thermo-chem. equiv. Both give precisely the same result, but the second is the more usual and far more convenient form. An interesting point oc- curs in electro-plating. As each portion of metal is deposited, a definite quantity of energy is expended on its separation from the solvent. But for each such quantity of metal deposited an identical quantity is dissolved from the anode, with production of energy. One energy is equal to the other, so that theoretic- ally all the energy required is that needed to overcome the re- sistance of the solution. In practical operation there is always a loss besides this. Where a metal is precipitated and none dissolved, the energy to decompose the salt goes to the expense account. Many battery calculations have been made to determine the voltage given by different combinations. The zinc-copper-copper sulphate couple (Daniell's battery) is thus calculated for its voltage. Zinc is dissolved, forming sulphate; this sets energy free or develops energy. Copper sulphate is decomposed, ab- sorbing energy. Zinc combining with oxygen gives out 43,200 calories, and the oxide combining with sulphuric acid gives out 11,700 calories, a total of 54,900 calories. These figures are for the gramme equivalent of zinc, which is a number of grammes equal to its atomic weight 65.2 divided by its valency 2, or 65.2/2 = 32.6 grammes. The total calories of energy developed in calories are 43,200 + 11,700 = 54,900. The total calories of energy absorbed by the copper separated from the sulphate are 19,200 + 9200 = 28,400 calories. The net calories developed in the combination are 54,900 — 28,400 = 26,500. For the electromotive force we have: E = 0.0000437 X 26,500 — 1.15 volts. CHAPTER VI. PRIMARY BATTERIES. The Primary Battery Cell. — The simplest types of primary battery come under the category of single-fluid cells. A piece of copper and one of zinc, if placed in contact with each other and immersed in a saline or acid solution, will generate a cur- rent of electricity due to the impressing of electromotive force upon the circuit formed by these things — the saline or acid solu- tion, the copper, and the zinc. The circuit may be looked upon as a triangle — one side liquid, one copper, and one zinc. If a cartridge shell contains some dilute acid, and a wire or rod of zinc is immersed in it, but not allowed to touch the cop- per, a galvanic battery is formed. Attach wires to the zinc and to the copper. Connect one to a plate buried in the earth and the other to a telegraphic instrument, and messages can be sent by it over many miles of wire. There is some claim that a battery made out of a percussion cap has sent an electric im- pulse across the Atlantic Ocean. Three Constituent Parts. — In a cell there are three princip::! things as noted above. One is a liquid, the electrolyte, w^hic-i will be decomposed, through attacking chemically a substance, almost always a zinc plate, when an electrical current is passing through it. The second element is the zinc plate or some equivalent solid or liquid material which the solution can at- tack. The other is a material which the solution cannot attack. The two materials last mentioned must be conductors of electri- city. Simple Batteries. — A glass tumbler of dilute sulphuric acid with a plate of zinc and one of copper (carbon or platinum ani PRIMARY BATTERIES. 9.3 some other metals will do) dipping into it and not touching each other constitutes a simple battery. If the zinc is pure, no action will take place until the metals are connected electrically by touching each other or by a conductor such as a copper wire. When such connection takes place, a current will flow and the zinc will be attacked. The amount of zinc attacked will be in exact proportion to the coulombs of electricity produced. The plates of metal conduct the current by regular electric conduction. The liquid as such has hard- ly any true conducting power. A mere trace of conductivity can be found in it, by the production of very trifling cur- rents, practically negligible. But under the influence of the electric current, the liquid is decomposed. In its decomposition it virtually becomes a conductor, and is said to conduct electrolytically. The solution is called an electrolyte, which word means ''decomposed by electricity." Such a bat- tery is shown in Fig. 19, in which the zinc plate is marked Zn, the copper plate Cu, and the direction of the current is indicated by arrows. The current, it will be observed, always flows from plus ( + ) points. Nomenclature* — The general nomencla- ture of the parts of the cell is rather con- fusing, but it is hopeless for anyone to attempt to simplify it, because positive and negative are applied in diametrically opposite senses to the plates, and cathode, anode, electrode, plate, and other terms are embalmed in the literature of the science. In reading an author whose subject is at all understood, the reader will have no trouble in appreci- ating the particular terminology he uses. The simpler termin- ology is generally the better. Any kind of vvemoria tecJinica or artificial memory may be used to keep clear the distinction be- tween positive and negative. Negative and Positive Plates. — Writers in the English, lan- FiG. 19.— Simple Bat- tery. 06 ELECTRICIANS' HANDY BOOK. giiage usually call the plate corresponding to the copper plate of the simple battery described, which is the one unacted on, the negative plate. This is because it is not dissolved or at- tacked. The zinc plate, which is attacked, is then called the positive plate. The direction of current in the outer circuit, oh the telegraph line or other conductor, is taken as from the negative plate to the positive plate. This may be remembered by picturing the unattacked plate as an inert collector of elec- tricity, which it pours out upon the circuit in the form of current. The above terminology is simple and readily remembered. An excellent memoria technica is that the current starts from a plate the initial letter of whose name is generally c (carbon or copper) and goes through the outer circuit to a plate of initial s: (zinc). The current starts from the letter which comes earlier in the alphabet. While solid plates are almost invariably used, a liquid amal- gam of zinc may represent the positive plate, and liquid mercury might be used to represent the negative plate. There is a bat- tery in which the first-described arrangement exists. There is one term which may be advantageously used; it is ^'electrode" for plate. Thus we can broaden the assertion above by saying that a battery may have indifferently solid or liquid ^'electrodes." Cell, Couple, and Pair. — A battery of only two plates is called a cell, a couple, or a pair. An aggregation of cells be- comes a battery. It is a case of the greater including the less. The word pile is often applied to a battery. Properly, this term should be restricted to the real literal voltaic pile described on pages 97 and 98. Exhaustion and Polarization. — When the solution is weak- ened by dissolving the positive electrode, the battery is said to be exhausted. When the negative plate loses effect from accumu- lation of hydrogen, the battery is said to be polarized. The dis- tinction between exhaustion and polarization should be followed closely. Local Action and Amalgamation. — The essential thing in batteries is to avoid what is called local action in the 2inc. Chemically-pure zinc is not attacked by dilute sulphuric acid PRIMARY BATTERIES, . 97 such as is used in batteries. But commercial zinc contains enough impurity to cause local action. This means that it forms a lot of little voltaic couples, and accordingly dissolves in weak acid. To prevent this, the zinc in all acid solution batteries is amalgamated with mercury. The mercury is rubbed over the clean surface of the zinc along with some dilute sulphuric acid. A strip of galvanized iron is an excellent rubber for amalgamat- ing. It will pick up mercury as a soldering iron will pick up solder. Newly-amalgamated zinc shines like silver, but soon loses this luster. It is exceedingly brittle. As a trace of zinc injures mercury, the plates should be amalgamated with a few drops of mercury only. Dipping them into mercury is unnecessary and the mercury thus abused has to be purified before it can be used for other purposes. The first genuine primary battery dates back to the Italian physicist Volta in 1800. This is a good starting point for the description of batteries. Volta's Battery .^This construction goes back over a century. It is adapted for a series of cells in series arrangement, which was called a "crown of cups," or couronne des tosses. A plate of zinc is soldered to a plate of copper at one end, so that the two form a sort of V or U. A number of these are made. Be- sides these double plates, two single plates, one of zinc and one of copper, are provided for the ends. Cups one greater in num- ber than the pairs of soldered plates are partly filled with weak sulphuric acid. The soldered plates are put in, each pair in two cups, zinc in one and copper in the other, each cup receiving the zinc plate o^ one pair and the copper plate of the other pair. After all the soldered pairs of plates are disposed of, the end cups will each have one (a) a zinc plate and the other (&) a copper plate in it. The whole is then completed by put- ting into one cup (a) a copper single plate and into the other cup (&) a zinc plate. Care must be taken that the plates do not touch. The electromotive force of the zinc-copper couple is less than one volt. It is of only historic interest. Volta's Pile, or the Galvanic Pile. — A series of disks of copper and zinc are cut out of sheet metal. They may be some 98 ELECTRICIANS' HANDY BOOK. inches in diameter. Half as many disks of bibulous pasteboard or of cloth are cut out. These must be about a quarter of an inch less in diameter than the metal plates. The pasteboard or cloth disks are moistened with acid. Any excess must be squeezed out. A piece of heavy glass is a good basis for the erection of the pile. This is laid on a table or elsewhere, and a disk, which we will assume to be of copper, is placed upon it. The pile may be started with zinc. On this is placed a disk of pasteboard or cloth. It must not be too wet. Next comes a disk of zinc, then one of copper, then pasteboard or cloth, and Fig. 20.— The Galvanic Pile. Fig. 21.— The Wollaston Battery. BO on until fifty or one hundred plates have been used. The exciting solution may be water and sal ammoniac or a mixture of water and 1/20 its weight of sulphuric acid. The disks of cloth or pasteboard must not be so saturated as to permit tl:e weight of the plates to squeeze out tli^ acid. No acid must get upon the edges of the plates. Ears may be left on some of the disks, certainly upon the end ones, for attaching the wire of the circuit. If ears are provided on some of the intermediate plates, various voltages may be taken from it. An improvement is to solder the zinc and copper plates of each pair together, either over the entire face or accurately around the entire edge. The galvanic pile is mainly of historic interest. The cut, Fig PRIMARY BATTERIES, 99 20, shows a double column or pile. The zinc plates are marked z, the copper plates A, the cloth u, A bar c c connects them at top. It will be observed that the order of copper and zinc is reversed in the two columns. This keeps them in the same rela- tion to the current. The terminals dip into two vessels of salted or acidulated water, b h, which can be used as terminals, by dipping other plates at the end of wires therein. Wollaston's Battery. — This is a copper-zinc combination. The copper plate is bent into a U shape, and the zinc plate lies within its bends. In the cut, Fig. 21, C C are the copper plates, Z Z the zinc plates, B B the cups. Blocks of wood separate the plates at the bottom. Their connections are screwed to a wooden bar at the top. This with the wooden blocks keeps them fixed in position. The cut, Fig. 22, shows the plates with their terminals O 0'. Hare's Calorimeter. — This is another historical battery devised by Offershaus in 1821 and modified by Hare in 1824. It is shown in the cuts, Fig. 23. A sheet of copper and one of zinc are wound into a spiral. Pasteboard strips are used to keep them from touching each other. They are dipped into a vessel of acid or sal ammo- niac solution when to be used. "Wire ter- minals are soldered to the zinc and copper respectively. Notched standards are pro- vided, to carry the weight of the plates and to keep them out of the solution if desired. The zinc terminals are marked Zn and the copper ones Cu. Zamboni's Pile. — A glass tube one-half to one inch in diameter is coated on the inside with sealing wax. It is filled with disks of silver paper coated on th.. back with powdered manganese dioxide rubbed up with thinned mucilage. The disks must dip in easily, so as not to get manganese dioxide on the inner surface of the tube. A pile of one thousand such pairs gives enough electromotive force to deflect a straw suspended by a silk filament. A pair of Zamboni's piles, each of some two thou- sand pieces of paper, are sometimes arranc^d to attract and dis- charge alternately a strip of gold leaf suspended electrometer Fig. 22.— Plates of WOLLASTON 'S BATTEKYo lOO ELECTKICIANS' HANDY BOOK. fashion by a filament. It is said that such a pendulum will oscillate for years. riodern Batteries. — We now come to batteries which are more than historical. Some go back to early days, but are still used or have been used in modern days. These necessarily must, if excited by acid, be protected against polarization. A general division into two classes may be made, single-fluid and double- fluid cells. We shall first consider single-fluid cells. Smee's Battery.— Zinc and platin- ized platinum or platinized silver in weak sulphuric acid. Modifled by Pat- terson, who substituted platinized iron for the silver plate; by Grove, who Cw+ Fig. 23.— Hare's Bat- tery. (Deflagrator.) Fig. 24.— Smee's Bat- tery. substituted platinized wire gauze; by De St. Amstell, who sub- stituted platinized tulle. Smee's battery has long been a promi- nent battery. The platinizing is not a simple plating with plati- num, but platinum black, which is a very flnely divided form of the metal, is deposited upon the surface by electro-deposition- This form of platinum cannot be covered, by hydrogen; smooth PRIMARY BATTERIES. 101 platinum can be covered virtually by the gas, so as to polarize the battery. As usually mounted, a bar of wood separates the plates, whose upper edges are secured or clamped to it. The exciting fluid is sulphuric acid diluted with water. It may range from one-seventh to one-sixteenth its volume of acid, according to the requirements. The platinum plate before use is dipped every day into a solution of ferric chloride. This oxidizes any reduced deposit or occluded hydrogen. Silver-plated lead plati- nized is substituted sometimes for the regular negative electrode. Fig. 25.— Smee's Battery— Tyer's Form. E. M. F., 0.42 to 0.47. This construction is shown in Fig. 24. Another form is shown in Fig. 25. Mercury is poured into each cell, and bits of zinc are dropped into it from time to time. Thus scraps of zinc can be used instead of a plate. A ball of zinc, cast on the end of a wire, which wire must be insulated, dips into the mercury. Walker (1859) used platinized carbon instead of platinized metal. E. M. F., 0.66. Iron Negative Plates, — Sturgeon (1840) used cast iron; Miin- nich (1849) amalgamated iron; Callan (1845) cast iron in form of a shallow vessel, constituting at once the recipient for solution and the negative electrode. Hughes (1880) used zinc-hydrogen- ated iron, acidulated water; polarization only one-fifth that of Smee's battery. E.M.F., 0.56. Aluminum Negative Plate. — Helot (1855) zinc-aluminium- 102 ELECTRICIANS^ 8' HANDY BOOK, dilute sulphuric acid. The aluminium negative plate is dipped in strong hydrochloric acid for a few minutes to make it less easily polarized. Grove's Battery (1838). — Zinc amalgam-platinum in a solution of sulphuric acid with nitric acid as a depolarizer. This has seen long service as one of the leading batteries of the world. The cut, Fig. 26, shows one construction. The platinum goes in a porous cup, V. This contains nitric acid, 1.33 sp. gr. Outside the porous cups is the zinc, Z, and the whole is contained in a battery jar charged with sulphuric acid. As it produces current Fig. 26.— Grove's Battery. Fig. 27.— Grove's Battery. the nitric acid is reduced, and corrosive and poisonous fumes of nitrogen oxides are evolved. It is credited with an electromotive force of 1.9 volts. Generally, less than this is to be looked for. Another form of Grove's battery, in which a bent platinum plate is used to increase the area and diminish the resistance, is shown in Fig. 27. Various modifications have been tried. Oxalic acid (Royer) has been used instead of nitric. The result was the production of formic acid with evolution of hydrogen. A saturated solution of ferric chloride with a little nitric acid has been recommended. As long ago as 1840 (Hawkins) and 1841 (Olfers) the platinum was replaced by iron in concentrated nitric acid. Iron is not attacked by this acid when concentrated. The acid if diluted PRIMARY BATTERIES. 103 attacks it, so this formed an objection to its use, as the acid soon becomes dilute by use of the battery. Uelsmann proposed to substitute silicon iron for platinum in the Grove cell. The E. M. F. varied with the concentration of the nitric acid. Buff (1857) proposed aluminium in place of the platinum. Grove himself in 1839 had tried wood charcoal and retort carbon as substitutes for platinum. It is said that he thought that in the scientific world platinum only would be considered **truly in harmony with science." Callan (1847) substituted platinized lead for the platinum, and replaced the nitric acid by a mixture of 4 parts concentrated sulphuric acid, 2 parts nitric acid, and 2 parts saturated solution of potassium nitrate. Carbon Negative Flates. — Early in the last century Gautherot found that wood charcoal would act as a negative electrode in Volta's battery. The early inventors in this line were Leuch- tenberg (1845), Fabre de Lagrange (1852), J. Walker (platinized carbon) (1859). Tommasi (1881) used zinc-graphite in dilute sulphuric acid. The graphite is heated to redness, and cooled in a current of car- bon dioxide or nitrogen. E. M. F. at first, 1.37 volts; falling rapidly to 1 volt, and after a few hours to 0.83 volt. Moving Electrodes. — A number of batteries have the plates moved in and out of the solution, so as to depolarize the negative plate. Becquerel (1852) was an early investigator in this line. Erckmann made his plates disk-shaped, mounted them on an axle, and rotated them. About half their depth was immersed in the acid. Brushes rubbed against the plates as they rotated. Maiche (1864) devised a copper- or carbon-iron couple with very dilute nitric acid (one per cent). The copper or carbon electrode was disk-shaped and rotated as in Erckmann's battery, with about one-third its area immersed. Skene and Kuhmaier used a zinc copper cell with dilute sulphuric acid; the copper is moved in and out of the liquid by clockwork. Bunsen's Battery (1842).— Amalgamated zinc carbon in dilute sulphuric acid mixed wuth fuming nitric acid as depolarizer. Bunsen improved on Grove's cell by substituting relatively cheap 104 ELECTRICIANS' HANDY BOOK. carbon for platinum. E. M. F., 1.89 volts. He placed the zinc in the porous cup and the carbon in the exterior vessel. Archereau reversed the relation of the plates, and put the carbon in the porous cups and the zinc outside. The battery has been elabo- rately tested by Meylan (1886) with the following results: Exciting liquid, sulphuric acid consisting of equal volumes of 60° Beaume sulphuric acid and water. Depolarizer, nitric acid of 36° Beaume. Zinc, active surface 116.25 square inches. Ex- ternal resistance, 1.27 ohms. Internal resistance, 0.04 ohm, falling to 0.035 ohm and rising to 0.12 ohm. Electromotive force. Current. Energy. Starting 1.93 volts. After 15 minutes closed ciccuit 1.87 '' 1 .42 amperes. ** 24 hours *' *' 1.77 '' 1.33 " 56 watt hours. *' 30 ** *' '' 1.73 '* l.zi '' 70 A modern Bunsen battery is shown in Fig. 28, the carbon and zinc plates being indicated by C and Z, and positive and negative by the regular signs + and — . The Bunsen and Grove cells have the advantage that no salts are used in their solutions, so there is no trouble with crystalli- zation in the carbon or platinum compartment. The evolution of nitrous fumes corroding the battery connections and exacting special ventilation is a very bad feature. Modifications of Bunsen's Battery. — Numerous attempts have been made to modify it. A few only will be mentioned here. Liais and Fleury (1852) made a carbon cup act at once as the porous cup and as the negative electrode. The nitric acid was poured into its interior. Miergles (1868) and Faure (1880) pro- posed a stoppered carbon bottle to hold the nitric acid and act as negative electrode. Boettger (1868) used carbon soaked in nitric acid as negati^^e electrode. To suppress the fumes of the depo- larizer Balsamo advised maintaining a layer of turpentine on top of the nitric acid. This suppresses a great deal of the emana- tion and suffers change in its own composition. Archereau sug- gested the use of tin scrap contained in a vessel inverted over the cell to combine with the nitrogen oxides. Rousse used a layer of oleic acid to absorb the fumes. Thann used as depolarizer a mixture of 500 grammes nitric acid to 60 grammes chloro- chromic acia (Cr, O, CI,). The latter is obtained by acting on PRIMARY BATTERIES. 105 10 parts potassium bichromate and 17 parts of sodium chloride with 30 parts concentrated sulphuric acid. It absorbs the nitro- gen oxide fumes. Qibbs' Battery.— Prof. Wolcott Gibbs in 1878 produced a truly philosophical modification of the depolarizing liquid in the Bun- sen battery. He used as depolarizer nitric acid of 1.4 sp. gr. saturated with ammonium nitrate. This solution gives off innoc- uous nitrogen, as it is reduced by the nascent hydrogen. Poggendorff's Battery. — This is one of the leading batteries. Fig. 28.— Bunsen's Battery. It resembles closely the Bunsen battery, and dates back to 1842, the same year that Bunsen's battery is referred to. It is a zinc- carbon couple with porous cup and as an exciting liquid salt solu- tion or dilute sulphuric acid. Its depolarizer is a solution of potassium bichromate. The latter oxidizes the nascent hydrogen, so as to prevent depolarization, and gives off no gas or fumes. By using dilute sulphuric acid as the excitant, and a strong solution of potassium bichromate in dilute sulphuric acid as the depolarizer, an electromotive force of 2 to 2.2 volts is obtained. The next cut, Fig. 28, serves as a representation of the Poggen- 106 ELECTRICIANS' HANDY BOOK. dorff battery. There is no material difference of construction between them, and Poggendorff' s battery is often called Bunsen's battery. Sometimes the Poggendorff battery is made up without the porous cup, as in the Grenet battery described further on. The zinc and carbon are immersed in the same solution. A mixture of sulphuric acid, water, and potassium bichromate in solution forms at once the exciting and depolarizing solution. This solution acts upon the zinc disadvantageously. For all purposes, where a powerful battery is needed and where constancy is less desirable than power in small compass, some form of the Poggendorff or bichromate couple is very available. The disadvantage of the omission of the porous cup is the dissolving of the zinc on open circuit, even if amalgam- ated. The combined depolarizing and exciting fluid attacks the zincs under the above circumstances. This difficulty is met by withdrawing the zincs from the solution when the battery is not in use. A great variety of this class of battery differing in mechanical details has been devised. Modifications of Poggen dorff 's Bat- tery. — Some representative batteries of the Poggendorff porous cup type are the following: Fuller's flercury- Bichromate Battery is shown in Fig. 29. The zinc electrode in the shape of a cone or pyramid is cast around the lower end of a copper wire, which must be insulated. It rests on its base in the bottom of the porous cup. It is in height but a fraction of the height of the cup. Mercury is poured in to the porous cup, so as to lie in contact with the zinc to keep it amalgamated. The carbon electrode is in the outer vessel. In the illustration, Fig. 29, Z indicates the zinc electrode. The porous cup receives the acid solution; the depolarizing solution is in the outer vessel. In starting, no acid need be put into the Fig. 29 - -Fuller's Bat- tery. PRIMARY BATTERIES. 107 porous cup. Enough will soon find its way in from the depolarizer to start the cell to working. It is claimed that on an ordinary working circuit this battery can be run for six months without renewal. The internal resist- ance can be varied, as in any other porous cell battery, by using porous cups of varying thickness and porosity — 1/{> ohm to 4 ohms are given as ranges of resistance of the commercial cell. It has been used in England extensively for telegraphic service. Camacho Cascade Battery. — This battery provides for the constant renewal of the bichromate depolarizer. The cut. Fig. 30, shows a series of cells arranged on steps. The depolarizing solu- tion is caused to flow slowly from the upper vessel into the porous cup of the upper cell. Thence by a pipe it flows from the bottopa Fig. 80.— Camacho*s Cascade Battery. of this porous cup into the porous cup of the next lower cell. This flow goes through as many cells as desired. Baudet Siphon Battery. — In this construction the regular por- ous cup construction is used. Siphons with india-rubber starting bulbs connect the outer cups of contiguous cells. The zinc plates are contained in the porous cups. The depolarizing fluid, when all the siphons are charged, will siphon from one cup to the next as long as a difference of level obtains between the liquid in the first cell and the outlet of the siphon connected to the last. Depolarizing solution is slowly admitted to the first cell, and 108 ELECTRICIANS' HANDY BOOK. siphons along the row to the end one. The effective level of the outlet siphon of the last cell can be adjusted by a trap, which also keeps the siphon charged. Radiguet Battery.— In this battery the zincs are in the porous cup. The porous cup forms one division of a double vessel, some- what heart-shaped in contour, whose other division Is glazed. The combined glazed and porous cup oscillates about a journal. When tilted in one direction, the porous division descends into the main battery cell, and the acid runs from the glazed division Into the porous one. The zinc plate is fixed in position in the porous cup division. When the combined cell is tilted in the other direction, the porous cell division is withdrawn from the main battery cell, and the acid runs out o:^ it into the glazed division. The effect of this is that when the battery is not in use, the zinc is out of contact with acid, and the acid solution is in a separate impervious receptacle away from the depolarizing solution. The latter is in the main cell with the carbon. A single motign of the lever or handle turns the porous cup down with the zinc in it, into the depolarizing solution, and the acid simul- taneously flows in and surrounds the zinc. Other modifications of the Poggendorff cell show the dip battery principle applied to the zinc plates — the carbons being left immersed. The porous cell being only an imperfect expe- dient, this withdrawal of the zincs leaves the solutions to intermingle by diffusion through the pores of the porous cup, so this withdrawal of the zincs is only a partial solution of the problem. Grenet's Battery.— In this battery, shown in Fig. 31, the zinc plate, Z, is drawn out of the solution by the handle a when the battery is not in use. This cell is variously constructed on the general lines shown. Dip Batteries. — Many bichromate batteries are mounted so as to have all their plates withdrawn from the solution. All the plates are attached to a bar by which they are all raised simul- FiG. 31.-Grenet's Battery. PRIMARY BATTERIES. 109 taneously from the liquid. A sort of windlass is often mounted on a frame to effect the lifting. Partz's Battery utilizes the different specific gravity of the liquids. The carbon lies horizontally on the bottom; the zinc, also horizontal, is suspended above it half way up the jar. It is first charged with a solution of magnesium sulphate 1:4, or ammonium chloride 1:5, or some similar salt. Five per cent to ten per cent of hydrochloric acid may be added to reduce the resistance, but it exerts local action upon the zinc. A solution of sulphuric acid and chromic acid is poured in through a glass tube, which reaches to the bottom of the vessel. This depolariz- ing solution of high specific gravity lies under the other solution, floating it up and covering the carbon. As the depolarizer is exhausted, more is added through the tube. This battery is credited with over 2 volts E. M. F. Depolarizing flixtures and Exciting Solutions in Batteries of the Poggendorff Type. — The Bunsen battery carrying out the principle of Grove, but substituting carbon for platinum, opened the possibility of new depolarizing solutions. Many such, which would attack platinum, are available for carbon. Poggendorff's substitution of chromic acid for nitric acid did away with nitrous fumes. A number of solutions for carbon-porous cup batteries have been tried, and many are of interest. D'Arsonval (1881). — A depolarizing mixture of 1 volume nitric iicid, 1 volume sulphuric acid, and 4 volumes water saturated With copper sulphate was employed by him. Ruhmkorff (1867) and Dupre (1885). — Carrying out a sugges- tion due to the earlier scientist, Dupre used as polarizing solution a mixture of water 600 parts, sodium nitrate 510 parts, potassium bichromate 60 parts, and sulphuric acid 720 parts. The potas- sium bichromate absorbs the nitrogen oxides. Mauri. — His depolarizer consisted of potassium chlorate 50 parts, potassium nitrate 25 parts, mercuric chloride 4 parts, iodine 5 parts. Koosen (1873). — His depolarizer was based on the use of potas- sium permanganate. Two solutions are described: a, potassium permanganate 300 parts, sulphuric acid 100 parts; &, potassium permanganate 100 parts, sulphuric acid 250 parts. Water enough 110 ELECTRICIANS' HANDY BOOK. to dissolve the potassium salt is used. It must be mixed with great care, the acid being added little by little to the aqueous solution of permanganate. E. M. F., 2 to 1.7 volts. Lacombe. — Saturated solution of potassium chlorate and ferric sulphate or chloride, to which is most carefully added sulphuric acid. Potassium permanganate may be substituted for the chlorate. Duchemin. — Picric acid solution mixed with sulphuric acid was employed as a depolarizer. It is reduced to picramic acid. riixtureof Sulphuric and Nitric Acids.— Many depolarizing mix^ tures were made by mixing these two acids. The idea was to have the water combine with the sulphuric acid, so as to give a stronger nitric acid to do the depolarizing. Potassium Bichromate 5oiutions. — Formulas. — Poggendorff. — Potassium bichromate 100 parts, water 1000 parts, sulphuric acid 50 parts. Delaurier. — Potassium bichromate 18.4 parts, water 200 parts, sulphuric acid 42.8 parts. Chutaux. — Potassium bichromate 100 parts, mercury bisul- phate 100 parts, water 1000 parts, sulphuric acid 66° (B.) 50 parts. Dronier's Salt. — A mixture of one-third potassium bichromate and two-thirds potassium bisulphate. It is dissolved in water just before use. Tissandier. — Potassium bichromate 16 parts, water 100 parts, sulphuric acid 37 parts. Finely-pulverized bichromate is used. It is dissolved as far as it will in the water heated to about 100° F. The acid is then added, and the mixture shaken until all dis- solves. Kookogey. — Potassium bichromate 227 parts, boiling water 1134 pnrts, sulphuric acid added while water is at boiling temperature 1588 parts. It is allowed to cool, and the liquid is decanted from the crystalline residue which forms on cooling. Trouve's. — Water 80 parts, pulverized potassium bichromate 12 parts, concentrated sulphuric acid 36 parts; all parts by weight. The pulverized potassium bichromate is added to the water, and the acid is added slowly with constant stirri'ng. As much as 25 parts potassium bichromate may be added to 100 parts of water. PRIMARY BATTERIES. Ill The heating produced by the acid and water dissolves nearly all the potassium salt. Use cold. The Daniell Battery.— The Daniell battery is the type most used probably of all primary batteries. It is of low voltage, a little over one volt, and of high resistance, several ohms in all ordinary sizes. Its great constancy and cheapness of its first cost and of its solution have made it the telegrapher's battery par excellence. It is being replaced by caustic potash and other batteries to some extent. The typical cell contains a porous cup for the zinc. It is filled with water. A copper plate is placed outside the porous cup, vir- tually surrounding it. A pocket or receptacle for copper sulphate crystals is provided near the top of the copper plate, and is often made out of the same copper as the plate. Sometimes to start it off some salt, sodium sulphate, or zinc sulphate is added to the water. Daniell produced the cell in 1836. Tommasi gives the invention to Becquerel in 1829. Walker in 1830 made a similar couple, using animal membrane for a diaphragm instead of unglazed porcelain for the porous cup. The action of the cell is this: The copper sulphate dissolves. Its sulphuric acid attacks the zinc, its copper is deposited as metal on the copper plate. The fluids move by or move through the porous cup. Under the action of a current, on closed circuit, the level of the copper sulphate solution rises. The action of this battery is subject to the defects of all porous- cup batteries. The solutions mix through the diaphragm so that the depolarizing solution comes into contact with the zinc. This is very injurious because the metallic copper precipitates on the zinc. This is done at the expense of the zinc, which is dissolved, constituting a source of expense. The dissolving of the zinc increases the specific gravity of the solution, which has to be weakened sooner than would be the case without this wasteful dissolving. The zincs have to be scraped occasionally, to free them from the copper. Both the latter features of wrong action involve extra labor. The electromotive force varies slightly according to the salts present in the zinc compartment and with the presence or absence 112 rrLECTRICIANS' HAXDY BOOK. of free acid. The great constancy of this battery has made it in the past a favorite for testing purposes as a standard of electro- motive force. Scientific investigators have made many investiga- tions of its reactions and determinations of its electromotive force. The latter varies from 1.160 to 1.03 volts. 1.07 volts is usually taken as the electromotive force. The chemical reactions involved are put thus: For the vessel containing the zinc plate: H, SO, + Zn =r ZnSO, + 2H. For the vessel containing the copper plate: 2H + Cu SO, — H,SO, + Cu. The electromotive force is but slightly affected by heat. If the surface of the copper is oxidized, its voltage is slightly increased by light. Dilution of the solutions is almost without effect on its voltage. The quality of the metals in the electrodes, whether rolled or rough, crystalline or not, makes very little difference in the voltage. The resistance of the Daniell cell is said to depend more upon the area of the copper than on that of the zinc. Amalgamation of the zincs is not favored by all investigators. Different mate- rials for the diaphragms have been tested, and have naturally been found to have no influence on the electromotive force. riodifications of Danieli's Battery. — These are not so numer- ous as might be anticipated, unless we include the gravity cells. Varley proposed to surround the porous cup with a layer of zinc oxide. This will decompose any copper sulphate which works its way through the walls of the porous jar. Copper oxide will be precipitated, and zinc sulphate will be formed. One great annoy- ance is the deposition of metallic copper on the porous cup's exterior. Borseul wound a spiral copper wire around the cup with a spiral plate at its lower end, the middle of the wire at- tached to the copper plate. The wire was supposed to catch all the copper as it precipitated. Parelle and Veritee, in their balloon or flask battery, place the copper sulphate in a glass flask with narrow neck, as shown in Fig. 32. It is filled with water, and inverted neck downward into the porous cup. It supplies copper sulphate for a long time. PRIMARY BATTERIES. 113 Fig. 33.— Balloon or Flask Battery. The solution in the outer jar must be weakened from time to time; otherwise, the battery takes care of the solution au- tomatically. Trouve's blotting-paper bat- tery, shown in Fig. 33, con- tains a copper and a zinc plate- marked Cu and Zn. The space between is filled with disks of blotting paper. The lower sheets of paper to one-half the total number are soaked in copper sulphate solution and allowed to dry. An insul- ated copper wire runs down through a central hole to the copper plate and is soldered thereto. Another wire is con- nected to the zinc plate. When the battery is to be used, water is poured upon the disks until it shows at the edges; they are pressed together and placed in the jar. This battery will give a small current for months. Eisenlohr (1849) used a sodium or potassium bitartrate in the zinc division. Buff used liquid zinc amalgam. An insulated wire runs down through the solu- tion into the mercury. Gaiffe's cell is a com- bination gravity and porous cup cell. The zinc is in the shape of a cylinder, and is suspended from the edge of the jar near its top. The porous jar is glazed or treated so as to be im- FiG. 33.— Trouve's Blotting-Paper Battery. 114 ELECTRICIANS' HANDY BOOK. pervious for its lower half. It contains the copper plate, and a wire extends from the copper plate up over the edge of the porous cup and down to the bottom, where it is carried around the lower part of the half porous cup in a circle. Any copper sulphate in the porous jar as it works its way through descends on account of its specific gravity to the bottom of the outer jar. When the circuit is closed, this cop- per sulphate is the first decomposed, and the copper ring acts as an electrode. When this part of the solution is exhausted, the copper in the porous jar becomes the negative electrode. Then the cell works like a regular Daniell's battery. This construction favors the preservation of the zinc from local action or attack by the copper sulphate solution. D'Arsonval (1881) has, by using caustic soda solution in the zinc compartment, brought up the voltage to 1.5 volts. Reynier reached the same voltage, using a seamless bag of parchment paper for the porous cup, a 30 per cent solution of caustic soda for the zinc compartment, and sodium bisulphate or sulphuric acid in the copper sulphate solution. He used other mixtures, whose complication tends to exclude them from every day use. Sand Type of Daniell's Battery. — Several cells have been de- vised in which a layer of sand replaces the porous cup. Minotto (1863) uses sand, D'Arsonval uses animal black or bone black, Coronat uses sawdust. There are other modifications. Gravity Battery. — This term is almost restricted to one type of cell, the copper-zinc-copper sulphate couple. It is based on the exact reactions of the Daniell cell, but has no porous cup, relying entirely on the various specific gravities of the constituent liquids to keep them separated. The construction of the modern gravity battery is cheap, because the porous jar is dispensed with. The original gravity battery dates back to 1859, when it was pro- duced by Meidinger. There is apt to be a little uncertainty about the originators of fundamental things in the world of practical science, but this Inventor going back nearly fifty years has given his name to the gravity battery, and the title adheres to it still. Meidinger's Battery (i859). — The cup was contracted in diam- eter at about one-third of its height, so as to form a shoulder, on PRIMARY BATTERIES, 115 which a cylinder of zinc rested. A smaller cup rested on the bottom of the main cup, and contained the copper electrode. This cup held strong copper sulphate solution, whose high specific gravity operated to prevent it rising and attacking the zinc when on open circuit. A glass tube with a hole in its bottom was arranged to keep up the strength of the copper-sulphate solution. A flask such as shown in Fig. 32 is sometimes applied to this battery. The next steps in the development of this cell were in the direction of simplification, and in modern cells there are often only three parts, two electrodes and the jar. A copper electrode which rests on the bottom of the jar, a zinc electrode of approxi- mate disk shape supported in a horizontal position near the top, and the battery jar are the three parts. To charge it, the copper electrode is put into the jar, resting on its bottom, and crystals of copper sulphate are introduced to a depth of two inches or more. It is then carefully filled with water to within an inch of the top. The solution of copper sulphate is of higher specific gravity than water, and stays at the bottom more or less com- pletely, especially if the battery is in use. But if the battery is little used and remains on open circuit, most of the time the copper sulphate solution rises and acts upon the zinc, attacking it, depositing metallic copper upon it, and impairing rapidly the condition and efficiency of the battery. The zinc dissolves, forming zinc sulphate, whether the battery is working or not. In the first case, the zinc should and must dissolve; in the second case, when the battery is not working, the solution is due to local action and is a defect. The inevitable formation of zinc sulphate acts to increase the specific gravity of the overlying solution, and to diminish the characteristic gravity feature of the cell. Accordingly, from time to time some of the zinc sulphate solution must be withdrawn and its place supplied by water. This dilution with water and the occasional addition of copper sulphate, called in the telegrapher's vernacular "blue- stone," should be all the attention the battery requires. If left much on open circuit, additional attention is called for — the occa- sional scraping of the zincs to free them from deposited copper. The cut. Fig. 34, shows Lockwood's construction of the gravity 116 ELECTRICIANS' HANDY BOOK. cell. A spiral wire connected to the copper plate in the bottom of the jar lies above the copper-sulphate crystals, and is designed to prevent the copper-sulphate solution rising and attacking the zinc. It acts by decomposing the solution. There are many other varieties. Modification of the Gravity Cell.— Thomson's battery, start- ing with saturated solutions of both copper sulphate and zinc sulphate, has the latter underlying the former, as it is of higher spe- cific gravity. The zinc is in the bottom, the copper near the top of the cell. Cardarelli in 1883 is credited with the same idea. Cu- pric chloride has been used as a substitute for the copper sulphate. On open circuit the copper is at- Fja. 34. -The Gravity Battery. (Lockwood's.) Fig. 35.— D'Infrevtlle's Wa ste- LESS Zincs for Gravity Batteries. tacked, reducing the cupric chloride to cuprous chloride. On closed circuit this reaction does not take place. Delaney in- closed the zinc in a paper envelope, and the copper sulphate in a strawboard box. The zinc is but little subject in this battery to local action. D'Infreville's wasteless zincs provide for the attach- ment of partly expended zincs to the bottom of the new one. In ordinary practice nearly half the zinc is wasted, as the plates get so corroded as to require replacing. In this system such plates are attached below the old one, their stem, which is slightly conical, being forced up into a hole in the center of the other zinc, as shown in the sectional diagram, Fig. 35. The half-dissolved PRIMARY BATTERIES. 117 old plates are thus used up. The cut shows a partly-expended one below a new one. Sir William Thomson's gravity battery, shown in Fig. 36, consists of a shallow tray on whose bottom rests a sheet of copper. Copper-sulphate solution covers the copper plate. Four wooden rods rest on the copper, and carry a grating of zinc con- tained in a parchment paper tray or box. The resistance, owing to the large surfaces and their nearness, is low compared to the ordinary Daniell or gravity battery. Thomson's battery in a Fig. 30 —Thomson's Battery. measure comes between the two, as the parchment paper dia- phragm and the specific gravity of the copper-sulphate solution each play a part in preventing local action. The trays are piled one on top of the other. Caustic Alkali Batteries. — Many batteries have been based on the action of caustic alkali on zinc. It dissolves the metal much as an acid does, and brings about polarization of the nega- tive electrode unless some means are taken to overcome it. Black oxide of copper, cupric oxide, is the favorite depolarizer in this class of battery; so much so, that the name "oxide of copper bat- tery" is often applied to the class. 118 ELECTRICIANS' HANDY BOOK. Lalande and Chaperon (1881).— These inventors have done much to bring the caustic alkali-oxide of copper couple into prominence. Amalgamated zinc copper in a 30 per cent solution of caustic alkali with copper oxide as depolarizer is the combina- tion. The alkali acts on the zinc, and the nascent hydrogen re- duces the copper oxide to the metallic form. The electromotive force may be as high as 0.98 volt. With electrodes 4 inches square and 2 inches apart, the resistance is 0.25 ohm. In one form an iron battery jar is used, which forms the negative elec- trode, taking the place of copper. As soon as a portion of the oxide of copper becomes reduced, the latter may operate as a cop- per electrode to some extent. In one form, Fig. 37, the battery jar and negative electrode are Fig. 37.— Lalande's Trough Battery. represented by an iron tray, A. A layer of oxide of copper, B, is spread over its bottom. Insulating blocks, L, carry an amal- gamated zinc plate, D, which rests horizontally upon them. The caustic alkali used as excitant is B = (a little more than) 14,000. The magnetizing force was now reduced, when the left- hand curve gives the effects of residual magnetism. On the re- duction when H = B = (a ^i^^le more than) 10,000, and this value of B== 10,000 is the residual magnetism. To reduce B to Fig. 82.— Hysterecis Curves. a demagnetizing force of H == (about) — 23 is needed. On fur- ther applying minus values of H» opposite magnetism is in- duced in the steel until H = — 90 a value of about —14,000 is reached for B, If now H is brought back to zero, B = (^ 1^^^^® more than) — 10,000, just as before the positive values of B, rnd this again is permanent magnetism of opposite polarity to the preceding. As before, B becomes zero when H has the same numerical value as before, but of opposite sign, or H= (about) 20 when B = 0. On increasing H' ^^^ value B= (a lit^^^ "^ore than) 1400 is reached when H ^^^ ^^^ ^^^ value of 90. The curves give an open figure; they inclose an area, and the THE FIELD OF FORCE, 187 ■whole resembles an Indicator diagram. Like the latter, it repre- sents a cycle which could be repeated indefinitely. Loss of Energy Due to Hysteresis. — The area is proportional to the energy converted by hysteresis into useless heat. The loss of y^ J steresis affects the operations of much electro- magnetic machinery and of alternating-current transformers. Hysteretic Constant.— A very simple formula for the loss has been produced by C. P. Steinmetz. Calling Ti the loss measured in ergs due to hysteresis per cubic centimeter of iron and for a single cycle, the formula reads as an equation: The Greek letter rj (eta) is a constant called the hysteretic constant. The equation holds good for a frequency of cycles of alternation up to 200 per second. This is twice that of stan- dard alternating current systems. Remembering that 10^ ergs are equal to one watt or volt-ampere, we can at once see just what waste of energy there may be occasioned by hysteresis in any case. The hysteretic constants for various qualities of iron are given in the table. Very soft iron wire 0.002 Most ordinary sheet iron 0.004 Soft annealed cast steel 0.008 Cast iron 0.016 Hardened cast steel 0.025 To get the loss in watts from the above, it is simply necessary to substitute the proper coefficient for ^ in the equation and divide by IQ-'^, or what is the same thing, to multiply by 10-^ Suppose the material used had the «oefficient 0.003. The watts loss would then be equal to 0.003 X IQ-'' X B^'^ X ^- The number of cycles indicated by n has to be introduced, because Steinmetz's original equation refers to a single cycle only. When a magnetizing force is applied without change to a piece of iron, its magnetization increases sometimes for half an hour or more, sometimes to the amount of several per cent of the magnetization. This is termed viscous hysteresis by Ewing, its discoverer, and sometimes it is termed magnetic creeping. CHAPTER IX. MAGNETS. The Electro=Magnet.— If a bar of iron is inserted in the axis of a coil of wire through which a current is passing, it will become magnetized and will attract iron. If free to move, one end, and always the same end, will point toward the north pole of the earth; not directly in that direction, except over a limited area of the earth's surface. Turning back to page 170, we see in Fi,g. 76 the diagram of a straight electro-magnet. The letter N indicates the north-seeking end of the pole, the letter S the south-seeking end. They are generally called the north and south poles of the magnet. Tractive Force of the Elect ro= Mag net.— A piece of iron by presenting a good path for the lines of force in the'vicinity of an excited electro-magnet virtually concentrates a number of them within itself. Other things being equal, a line of force tends to become as shore as possible, acting something like an India-rubber band. Hence the lines extending from magnet face to armature tend to become as short as possible, and this tendency pulls the armature toward the magnet, just as if a multitude of India- rubber bands connected the two. Spreading of Lines of Force. — In air lines of force spread apart, which might seem to contradict the above. But the lines not only tend to shorten their paths, but do not easily change direction. A line starts out straight from the surfaces of a mag- net, and curves gradually toward the other surfaces. This ten- dency to start straight (normally) from a surface gives the lines of force a feather-like contour. illustrating Lines of Force About a flagnet. — The cut, Fig. 83, shov:s tlio direction of lines of force about the north and south poles Gi a magnet, as shown by iron filings on a card or MAGNETS. 189 slip of paper. All these effects shown by iron filings may be made to give permanent records by using a piece of blue print paper, The paper is placed over a somewhat obscure place. r such as employed by draughtsmen, the poles in a horizontal position in The filings are dusted on tho paper, which may be tapped or shaken a little. It is exposed to strong daylight or sunlight with the filings in place, and is then soaked in water, the filings first being removed. Very interest- ing prints can be made in this way. Spiral EIectro=nagnet. — If an active conductor is surrounded by a spiral of iron, as shown in Fig. 84, the spiral will become magnetized and will become a magnet, with poles at N and S. this connection. U=Shaped Electro=Magnets. — The horseshoe or U-shaped elec- tro-magnet is a type which has been very extensively used. The core represents -a portion of a circle, three sides of a rectangle or some similar form, and generally two coils of wire are wound r Fig. 8J.— Magnf.tic Lines of Force Shown by Filings. Fig. 78 may be referred to in Fig. 84.— SpiRAii Electro-Magntt. upon two of its sides. The sides are called legs or limbs, the connecting portion of the core is the yoke. A typical magnet, such as used in telegraph instruments, is shown in Fig. 85. An- other wound with coned coils of wire is shown in Fig. 86. The wire is wound in opposite directions on the two legs of U- shaped magnets, as indicated in Fig. 87, in which arrows are 190 ELECTRICIANS' HANDY BOOK. used to show the direction of the current around the core, whose poles, marked N and S, are supposed to face the observer. A powerful form is that proposed by Silvanus P. Thompson and 1EM_, Fig. 85 —Typical Instrument Fig. 86.— Electro-Mac net with Coned Electro-Magnet. Con s. shown in Fig. 88. -A thick, short magnetic circuit is provided by the core of this shape. The magnetic circle, Fig. 89, is very similar, and shows how a U-shaped magnet can be excited by a single coil. This form is made for lecture purposes about three-quarters inch thick, bent >^^r:^^ ^ — ^ Fig. 87.— Winding of a U-Shaped Electro-Magnet. Fig. 88.— S. p. Thompson's Electro-Magnet. into half circles of about two inches, internal diameter. It is exceedingly powerful, presenting a path of high permeance for the lines of force. Joule's electro-magnet. Fig. 90, is a very old form, and one which has given very high tractive power. It was one of a num- MAGNET8. 191 Fig. 89.— Magnetic Circlr. ber of forms of electro-mag- net devised by J. P. Joule in the first half of the last cen- tury. The volt-coulomb or joule is named in honor of this distinguished scientist. The hinged electro-magnet. Fig. 91, needs no armature. When a current is sent through its coils, the two legs swing together and their ends touch each other. ^ example of a U-shaped magnet with a single coil is seen in Fig. 92. This type is called by the Germans a limp- ing magnet, which S. P. Thompson renders club-foot. A pivoted armature is pro- vided for these particular magnets. Annular Chambered flagnet.— A number of electro-magnets, whose exciting coils are contained in annular chambers or grooves have been devised. One used for lecture experiments is shown in Fig. 93. It may be- called the electro-mag- netic Magdeburg hemi- speres. The magnet and armature are indi- cated by a a and are identical. The section of one. A, is shown with the exciting coiL C. The iron-jacketed electro-magnet. Fig. 94, is practically one part of the above device, and is intended to attract a flat armature. A practical application of this type is shown in the electro- Fig. 90.— Joule's Eiectro-Magnlt. 192 ELECTRICIANS' HANDY BOOK. magnetic clutch, Fig. 95. Brushes B B bear upon insulated rings C C on the hub of a band wheel, which is free to rotate on a shaft. Current entering by the brushes excites the annular coil, which magnetizes the band wheel and draws it against the disk A A. The latter is keyed to the shaft and rotates with it. When ml I I CTV. V ^^^ ^^^^ ^^^ ^^^^ hand wheel are x'^^^^^^^^S^^^^^ drawn together, the wheel has >r^^^^^^//^%. to turn with the shaft. Electro = flagnetic Tractive Power.— A pair of wheels may be drawn together by a coil, as is shown in Fig. 96, thus one wheel being caused to grip or press against another, so as to turn it. The arrangement shown is of very limited application, and owing to the poor magnetic circuit, is far from efficient. A better arrangement is shown in Fig. 97, where a current of electricity passed through a coil car- FiG. 91.— Hinged Electro-Magnet. Fig. 93.—" Club-Foot " or Limping Electro- Magnets. ried by a car wheel increases its traction on a rail. The coil is annular and lies in the groove around the wheel. The current enters by brushes, as in the clutch just illustrated. Multipolar flagnets are shown in two examples — Joule's "zig- zag," Fig. 98, and Roberts', Fig. 99, electro-magnets. These, in the light of what has been said, explain themselves. In them the MAGNETS, 195 usual letters N and S indicate north and south poles, and the arrows indicate the direction of the current. Various Armatures, — Cam mechanism due to Robert Houdin, the famous French magician, is shown in Fig. 100. E is the electro-magnet attracting its armature a. The cam A acts upon B. By varying the shapes of the faces of the cams, all sorts of results in the motion of the distant rod can be reached. The armature shown in Fig. 101 is attracted upward from the posi- tion shown in the dotted lines when the magnet is excited. It also presses against the drum, which is part of the core, and rotates it so as to turn the gear wheel on the further end of the shaft. A spiral spring may pull upon the short arm to draw the armature back. This operates like a ratchet and pawl I ".'^^ Fig. 93.— Annul. ' r Chamber- LD ElLCTRO-MAGXET. Fig. 94.— Iron-Jacketed ElECTRO-^I AG NET. mechanism, as it only operates to turn upon its up-stroke. In this magnet the core must be free to turn in the coil. In Fig. 102 is shown another magnet with rotating core. A is the rotating core, turned in one way by the pull upon the armature projecting from its lower end. The arm D is of brass, C is of iron. The core B is fixed. Other pivoted armatures are shown in the cuts. Figs. 103 and 104. 194 ELECTRICIANS' HANDY BOOK. The Natural Magnet is a mineral consisting of a combination of iron and oxygen, whose composition is indicated by the chem- ical formula, Fe^O^. The mineral is called magnetite, and is char- acterized by being attracted by the magnet just as iron is, only not so powerfully. Some samples of magnetite do more than this, as they attract iron themselves. Such are natural magnets, known to the ancients as the lodestone. The attractiveness foi iron is localized in each piece, being at a maximum at certain kz/v/y/yyyyy/. Fig. 95.- Electro-Magnetic Clutch. Fig. 9G.— Electro- Magnetic Drive. points. These points act upon the compass needle, each repelling one end of it and attracting the other end. If the mineral were suspended by a delicate enough pivoting or suspension, one of the attracting points on it would seek the north pole. The Permanent Magnet is a piece of steel which has been charged with magnetism, and which retains it. It attracts iron, its ends doing so most strongly; tends to point north and south, the same end always tending to the same pole; and thus de- termines what are generally called its north and south poles. Sometimes they are called the north-seeking and south-seeking poles. MAGXETS. 195 Action of riagnet Poles on Each Other- — The north poles of two magnets tend to repel each other, and the south poles repel each other exactly the same. A north pole of one magnet attracts the south pole of another. Like repels like, and unlike attracts unlike. Magnets repel each other just as much as they attract each other. Making Magnets by Single Touch.— One process of making a magnet is shown in Fig. 105. A bar of steel lying on a table is stroked from center to end with one pole of a permanent magnet, 'the arrow showing the motion. The stroking magnet is returned through the air to the center of the steel bar, and a second stroke Cio. 97.— Electro-Mag- netic Car Wheel. Fig. 98.— Joule's Zigzag Electro- Magx et. Fig. 99.— Roberts' Electro-Magnet. is given. This is repeated a number of times, and then the same operation is gone through with the other pole of the magnet on the other half of the bar. The end of the bar stroked with the north pole of the magnet will be a south pole and vice versa, as indicated by the letters N, N and S, S in the cut. This process is called single touch. The stroking may be done for both halves, vrith two magnets simultaneously, as described above for one. The north pole of one magnet and the south pole of the other are brought together or nearly so on the center of the bar, and simul- taneously moved out along it, are swept back to the center through the air, and the stroking is repeated. A little bit of wood may be placed across the center of the bar to keep the magnets from touching each other at the beginning of the stroke.. 196 ELECTRICIANS' HANDY BOOK. Making flagnets by Double Touch.— For this the opposite poles of two magnets are brought close together, separated by a slip of wood or pasteboard, the magnets being inclined at an angle of over 90° to each other, like a V with very wide angle. Fig. 106. The apex is placed on the center of the bar, and is moved ten to twenty times slowly back and forth over the whole length of the bar. In both single and double touch the effect is increased by rest- ing the ends of the bar to be magnetized upon the opposite poles FiQ. 100.— Cam Mechanism for Electro- Magnets. of two other magnets. The poles must be the same as those of the magnet with which the stroking of the end in question is done. flaking U-ShapedMajcnets.— This type of magnet is universal- ly called a horseshoe magnet. A bar of iron of this shape may be magnetized by stroking with another horseshoe magnet, from near the bend to the ends, or from ends to the bend. As for straight magnets, the magnet must be returned through the air. A piece of iron should be laid across the ends during the process. An excellent way of magnetizing U-shaped bars used for volt- meter magnets is to place the ends of the bar against the tvv^o MAGNETS. 197 poles of a powerful electro-magnet. Each end touches its own pole, and the adherence is strong. The operative now rocks it back and forth a number of times as it adheres to the electro- magnet, thus slightly jarring it and causing it to become per- manently magnetized. Magnetizing by Coil and ElectrO'VTagnet.— A compactly-wound coil of wire was proposed by Elias of Haarlem for making mag- nets. Through such a coil a current was passed, and the coil was moved from end to end of the bar to be magnetized. The coil may be slid thus over a U-shaped bar while its ends are in Fig. 101.— Calombet's Electro- magnetic Pawl. Fig 102.— Waterhouse's Pivoted Armature. contact with a powerful electro-magnet. A successive turning on and off of the current of the electro-magnet is used sometimes. Another suggestion was to apply the steel bars while red hot to the poles of an electro-magnet, and to pour cold water on them while there. Steel for Magnets. — Tungsten steel is considered the best ma- terial for permanent magnets. Hopkinson gives the analysis of such a steel: Iron 95.371 Carbon 0.511 Manganese 0.625 Silicon 0.021 Phosphorus 0.028 Tungsten 3.444 19S ELECTRICIAX8' HANDY BOOK. Chrome steel containing 0.687 carbon and 1.195 chromium and no tungsten also gave Hopkinson good results. Preservation of flagnets. — Jarring should be scrupulously avoided. The armature of a magnet should not be allowed to come against the magnet violently. It should be gently put into place. Jerking the armature off does no harm unless a positive jar or clicking is produced. A horseshoe magnet should have its armature in place when it is put away, and bar magnets should be in pairs, with poles in reverse direction and connected by short bars or armatures. Examples of Permanent flagnets —A compound U-shaped magnet is shown in Fig. 107. The body is made of thin bars Fig. 103.— Oscillating Fig. 1G4.— Siemens^s Pivoted Armature. Armature. supposed to be magnetized separately, and then fastened to- gether. An iron armature a with a hole serves to show its lift- ing power. Weights are attached by means of the hole. The above would often be termed a horseshoe magnet. A true horse- shoe magnet is shown in Fig. 108. There the poles are very close together. Such a magnet can be used for magnetization by double touch, on account of the proximity of the poles. An iron bar with a wheel of lead or brass mounted on its center and placed across the legs of a magnet, as shown in Fig. 109, will if it is inclined roll down the magnet around the poles and up the under side of it, actuated by the momentum of the little flywheel. In the next cut, Fig. 110, little bars of iron with disks at the ends are placed together, as at A. On bringing a magnet above them, they become similarly magnetized, and as MAGNETS. 199 they lie with north pole to north pole and south pole to south pole, they are driven apart by mutual repulsion, indicated by Polarized and Hagnetized.— When magnetism is spoken of, ¥iG. 105 -Making a Magnet by Single Touch. these words are synonyms. A polarized piece of steel is a mag- netized one. A polarized relay in telegraphy is one whose action depends upon a permanently magnetized armature. Constancy of riagnetism.— For instruments such as volt- meters, the critical thing is to have magnets of great constancy. Fig. 106.— Making a Magnet by Double Touch. To secure these, they must not be too strongly saturated, as such a procedure produces magnets which lose readily part of their strength. riutual Action of Currents —Two parallel conductors through which currents are passing attract each other if the currents are flowing in the same direction. If one current is flowing in one direction and the other current in the reverse direction, the conductors repel each other- 200 ELECTRICIANS' HANDY BOOK. Ampere's Theory of riagnetism.— Based on the above facts and on the construction of the electro-magnet, the celebrated Ampere's theory of magnetism has been formulated. It ac- counts for the mutual attraction and repulsion of magnets, and for their tendency to, place themselves in the magnetic meridian and to have one end seek the north pole. A current of electricity is assumed to circulate around each molecule of a magnet. The cut, Fig. Ill, shows the theory. It Fig. 107.— Compound U-Shaped Magnet. Fig. 108.— Horses Magnet. Fig. 109.— Magnet with Flywheel Armature. Will be seen that the effect is as if a single current circulated around the outside of the magnet. The parts of the currents ad- jacent to each other in the interior counteract each other, and the outside currents virtually coalesce into one. This is the con- ception of a magnet according to Ampere's theory. It will be seen that the current denoted by the outside ar» rows corresponds to the current through the windings of an electro-magnet. If the observer faces the north pole, the Amper- ean current, as it is called, will circulate in direction opposite to the motion of the hands of a watch. If we face the south pole. MAGNETS. 201 the current will coincide in direction with the motion of the hands of a watch. Memoria Technica.— A watch indicates seconds, and could prop- erly have the letter S marked upon its glass. It would then rep- resent the south pole of a magnet, its hands in their motion giv- ing the direction of the Amperean currents. The watch has been used before to fix on the mind the relation between an elec- tric current and its lines of force. The "S" may be taken as the symbol for ''seconds" and "south pole." Taking the face of a watch as indicating the south pole of a Fig. 110.— Rolling Armatures. Fig. 111.— Ampere's Theory of Magnetism. magnet, it tells us how the lines of force go. As the watch tells us that time flies from us, it tells us that at the south pole the lines of force fly from us. They issue from the north pole and return to the south pole through the outer circuit. Ampere's Theory of Terrestrial Magnetism.— A magnet points north and south, approximately, the same pole always pointing north. By Ampere's theory this is accounted for by supposing the earth to be a great magnet, and to be encircled by currents flowing around it, approximately parallel to the equator. If currents of like direction attract each other, then if placed at an angle with each other they will tend to coincide in direc- tion. Currents tend to become parallel with each other, and to 202 ELECTRICIANS' HANDY BOOK. coincide in direction also. If two conductors are free to rotate, and currents are passed through them, they will tend to rotate like a compass needle until parallel with one another, with the current flowing in the same direction in each. The theoretical ampere currents of the earth force the ampere currents which are supposed to encircle a magnet into parallel- ism and similar direction, and thus cause the compass needle to point to the north. Attraction and Repulsion of flagnetic Poles.— In Fig. 112 are shown two pairs of magnets. One pair has its north pole facing the south pole of its neighbor. The arrowheads indicate the direction cf the Amperean currents. The currents in both poles Fig. 112.— Ampere's Theory Explaining Attractive and Repulsive Forces Between Magnets. correspond in direction, and as currents of like direction attract each other, the north pole of the magnet S N and the south pole of the magnet Si Ni attract each other. This refers to the upper pair. In the low^er pair one magnet has been turned end for end. The Amperean currents are now opposite in direction, and the north poles of the magnets SN and Ni Si repel each other. If the south poles were brought together, repulsion would also exist, because the Amperean currents would again be opposite in direction. Action of a Current on the flagnet. — A compass needle in the vicinity of an electric current is acted on by it, and tends to place itself at right angles thereto. It never can unless the current is at right angles to the magnetic meridian, but the I MAGNETS. 203 tendency is present. Thus a current deflects a compass needle, if the compass is held near the conductor, unless the conductor is at right angles to the magnetic meridian, or lies nearly or quite east and west. Remembering that the magnetic needle of the compass is sup- posed to have Ampere currents circulating around it in planes at right angles to its axis, this directive tendency of the compass needle will be recognized as an effort of the Ampere currents to place themselves in parallelism with the current in the conduc- tor. If held above the conductor, the needle will be deflected in one direction; if held below, it will be deflected in the other. Ampere has devised a rule for remembering the ways in which a magnetic needle will be acted on by a current in a conductor near to it. Ampere's Rule. — If a man were swimming with the current in the conductor and had his face turned toward the magnetic needle, its north pole would be deflected toward his left hand. This means that if the needle was above the conductor, he would have to be on his back to face the needle; if it was below, he would have to be on his face. Hence the needle will turn in reverse ways according to whether it is above or below the con- ductor. If the direction of the Ampere currents be formulated in the mind, it will be seen that the above deflection of the magnets simply brings them in parallelism with and coincident in direc- tion as regards their nearest portion with the current in the conductor. A coil of wire traversed by a current represents a magnet. In physical experimenting, such coils called solenoids are used to illustrate the Ampere law. They will, when passing a current, tend to point toward the magnetic pole; their unlike poles will repel each other; and they will act exactly as magnets do. If an inert bar of iron is surrounded by a conductor carrying a cur- rent, Ampere's law will be exemplified, and we will have an elec- tro-magnet. An electro-magnet is a bar of iron around which a current of electricity is caused to flow, so as to represent the Ampere mag- netizing currents of the permanent magnet. As we can make 204 ELECTRICIANS' HANDY BOOK. the artificial currents very strong, and give them as many turns around the iron (called a core) as we wish, an electro-magnet can be made very strong, many times stronger than the best perma- nent magnet of equal weight. Right-handed Screw Law. — The relation of north and south pole to the current circulating around a magnet core is ex- pressed by the right-handed screw law. It is to this effect: A right-handed screw, such as a corkscrew. Fig. 113, placed so as to coincide with the axis of the magnet and turned in the di- rection of the current, will move toward the north pole of the magnet. The arrows and polar letters N and S in the cut indicate the relations. This is merely another state- ment of the watch law. Assuming the arrows to indicate the di- rection of current circulating around an iron bar S N, it will be seen that if the end N were pointed at the reader, the current would be against the motion of the hands of a watch. The end pointing thus should be and is the north pole. If the corkscrew were turned in the reverse direction, its motion would indicate a current in the oppo- site direction to that shown by the arrows. If the lower end were pointed at the reader as before, the current would coincide in direction with that of the hands of a clock or watch, and the pole would be a south pole. Again, imagine a corkscrew pointed at the face and turned. If turned right-handedly, it would advance if the screw had a grip on anything. Its direction of turning would give the polarity of the lines of force due to a current moving in the direction of the observer. The reverse also holds. Both these statements express the watch-face rule for lines of force due to currents. Fig. 113.— Corkscrew Analogy of the Magnet. CHAPTER X. INDUCTION. Electro = Magnetic Induction. — If an electric conductor lies in a field of force, it may be in the vicinity of a magnet pole, it will be unaffected by the field, as far as any electromotive force in it is concerned. If the conductor is moved so as to cut the lines of force, or if the magnet is moved while the conductor is sta- tionary, which brings about the same result of cutting lines of force, electromotive force will be impressed upon it. There are many variations in the relations of conductors and fields of force which have the effect of impressing electromotive force upon such conductors, and producing currents in them if they form or are part of a closed circuit. In general terms the inductive effects summarized above involve attraction or repulsion between pole and conductor. Tlireading, Interlinking, and Cutting Lines of Force.— There are two general ways of taking cognizance of the action of a field on a moving conductor. It may be referred to cutting of lines of force by the conductor, or to changing the number of lines of force which pass through the space included in the electric circuit. The latter may be looked upon as a ring, or irregular circle-like lead of wire. The passing of lines of force through this circle of wire is often called threading or interlinking of lines of force. The latter expression is correct because lines of force form closed circuits of their own. Induction. — When an electric conductor forming part of a cir- cuit is swept through a field of force an electromotive force is impressed upon it. If the ends of the conductor were connected to a proper instrument, such as a voltmeter, the electromotive force would affect its index, and it would be evident that electromotive force actually existed. The cutting of lines of force by an electric 206 ELECTRICIAXS' HAXDY BOOK. conductor represents the impressing of force upon or transfer- ring of force to the conductor. The term force as last used applies to electromotive force. If the proper conditions are estab- lished the electromotive force impressed on the conductor by the field of force will produce a current. If these conditions do not exist no current will be produced. Thus there are two varieties of induction. In the one case energy in the form of volt-coulombs, or other electromotive force-quantity unit, is developed, and by the law of the conservation of energy the motion of the conduc- tor through the field of force is resisted, so that energy has to be expended upon it to move it across the lines of force. In the other case no current is produced and no energy is required to move an open-circuit conductor through the field. ^- FiG. 114.— Ring Moving in Field of Force Without Cutting Lines of Forciu. Conditions for Inducing Electric Energy. — The conditions for thus producing current are two. The conductor must form part of a closed circuit, and the number of lines of force passing through the loop or opening of the circuit must vary in number; or a portion of the circuit must cut lines of force. In most cases of dynamo generators both the latter conditions exist at once. As the armature conductors cut lines of force they vary the num- ber of lines of force interlinked with the circuit. Examples of Interlinking. — Assume a uniform field of force and let a ring of conducting material be moved in it. The cuts, Figs. 114 to 117, illustrate several conditions, the motion of the ring being indicated by the arrows. INDUCION, 207 In the case illustrated by Fig. 114 the ring is swept through the field of force but cuts no line of force as its motion is parallel to them. Therefore no electromotive force is impressed upon it. In the case shown in the next cut, Fig. 115, lines of force are |, cut, therefore electromotive force is impressed; but as the num- ber of lines of force embraced in the ring is unchanging, no current is produced. Each half of the ring has electromotive force of the same polarity impressed on it and the two oppose each other, so that no current results. In Fig. 116 the ring is swung around so that it not only cuts lines of force, but the number of lines embraced by it is constantly varying, hence E <- <- Fig. 115.— Ring Moving in Field op Force Cutting Lines of Forcr Without Change of Interlinked Lines. electromotive force and current both result. In the next cut, Fig. 117, the ring is swept in a straight line through a non- uniform field of force. It not only cuts lines of force, but the number passing through it varies constantly. Electromotive force and current both are produced. In the first two cases no power is expended on moving the ring through the field; in the last two power is so expended. Motionless Conductor in a Field of Force of Varying Den- sity. — Where a ring or convolution of wire or other conductor is placed in a magnetic field, lines of force will pass through it, if its plane of position is at an angle to the general direction of the lines of force. Lines of force would be said to thread through 208 ELECTRICIANS' HANDY BOOK. it, but would have no effect whatever upon it. We have seen that a current would flow through it, actuated by electromotive force, ^- ^- < ff \ ^--^^ ■ < 1- ^v < ^'-^ ^ Fig. lie.—RiNG Moving in Uniform Field of Force Under Conditions Producing a Current. if the wire were moved so as to vary the number of lines of force embraced by the circuit. Suppose the wire or conductor to be Fig. 117.— Ring Moving in Field of Force Under Conditions Producing a Current. kept motionless and the density of the field of force to vary. This would cause the lines of force embraced by the circuit to vary INDUCTION. 209 in number. Electromotive force and current would be produced in the conductor exactly as if it were moved. Energy Relations. — Energy would be absorbed whether the field of force was increased or diminished in density under the above conditions. The presence of the closed circuit would be the cause of such expenditure. It would by counter-electromotive force resist any change of field density which would produce energy in its conductor, and exact the expenditure of additional energy. Fields of Force in Practice. — In practical engineering fields of force are produced by magnets, which are generally electro- magnets. They vary in the number of their poles, but follow pretty closely some general rules. The poles are nearly always of even number; for every north pole there is a south pole; the north and south poles are placed in alternation with each other. Fields of force may be moved past conductors or past coils form- ing parts of circuits; or the conductors and coils may be moved past them; or the relations of field to conductors or coils may be kept changing, as in inductor generators. In all such cases elec- tromotive force is impressed on the circuits. The conductors or coils which are thus treated form part of armatures, and consti- tute the active portions of the armature windings. The effect of the processes is to cause the number of lines of force interlinked with the circuit to vary. A variation of 10^ lines of force per second produces an electromotive force of one volt. Direction of Current Induced by Cutting Lines of Force. — If the north pole of a horizontally placed magnet face the ob- server the lines of force will come out of it toward him, will curve around and pass through the space surrounding the pole away from it to the south pole. If a perpendicular conductor is swept from left to right across the north pole an electromotive force will be induced in it, tending to produce in it a current from above downward. Let a letter N be marked upon the pole. Rule lines upon the end parallel to the oblique stroke of the N. Cut a narrow slit in a card and holding it with the slit vertical move it to right or to left. The lines will appear through the slit like a series of dots, and will appear to move up or down — up for a motion to the left, down for a motion to the rigl:t. Their ,210 ELECTRICIANS' HANDY BOOK. apparent motions indicate the direction of currents induced in a vertical conductor moved across the north pole, to left or to right. For the south pole the directions are the reverse. The cut, Fig. 118, illustrates the principle. In it the south poles are diagonally shaded in the opposite sense to the north pole. The same process of using a slotted card will show the direction of currents in a conductor swept across them. In the cut the arrows a b and c d indicate the direction of cur- rent induced by motion in the direction of the dotted arrows. With motion in the other direction the currents would have the reverse directions. Two Systems of Induction. — Electro-magnetic induction can be referred to two causes. One cause is the cutting of lines of force Fig. 118.— Directions of Ixduced Currents. by conductors. This generally, as far as effective, has the result of changing the number of lines of force threading the circuit. The other way is to directly change the number of lines of force threading the circuit, without reference to cutting them by con- ductors. The first cause is represented by the conductors on a drum armature, one of which is indicated in diagram in Fig. 119; the second is represented by a type of generator in which coils whose planes are parallel to those of the field magnet coils are swept past the poles. Fig. 120. In the first case the electromotive force is at a maximum for any conductor when it is directly opposite the pole; in the second case it is at a maximum when the arma- ture coil is midway between two poles. Generator Without Motion. — We are accustomed to think of IXDUCTIOX, 211 a dynamo or electric generator as a machine in rapid motion when active, and inert when at rest. But from what has been said it follows that it would be perfectly practicable to have a Fig. 119.— Closed Loop in a Bipolar Field. dynamo without any moving parts if some way could be devised to change rapidly the intensity of the current passing through its field. A close analogue to such a dynamo is the alternating- current transformer. In it is a field of rapidly varying density with a conductor placed so as to have the lines of force pass through it. The changes in number of lines of force passing through it develop pulses of electromotive force so that an alter- nating current is produced. Examples of Induction. — If a tele- graph wire or trolley wire were carrying a steady current and were set swinging by the wind, it would carry with it its field of force which would swing back and forth some- thing like a huge cable. Theoretical- ly there would be no limit to its di- ameter, but its intenser field would be within a limited radius from the conductor as an axis. This field swinging back and forth would sweep its lines of force across any Fig. 120.— Bobbin Field and Disk Armature Coils. :il2 ELECTRICIAN IS' HANDY BOOK. contiguous conductor, and if the ends of the latter were con- nected to each other or to the ground, would cause currents of brief duration to flow back and forth through it. Thus the trol- ley wire's, every swing, however slight, produces some current through the rails below it. The varying currents in a telegraph wire as Morse signals are sent through it make varying fields of force and set up pulses of electromotive force with consequent currents in contiguous con- ductors. The interference of telegraph signals with telephone circuits is a very familiar instance. Telephone Receiver a Dynamo. — The telephone receiver can be used as a transmitter and originally was used as such. The minute vibrations of the thin plate which is supported close to the pole of a permanent magnet make the lines of force move about a little so that the coil of wire surrounding the end of the magnet is cut by varying numbers of lines of force. These varia- tions induce electromotive force and currents, which reproduce in a distant telephone receiver the sounds of the voice. The telephone receiver used thus is really a dynamo. In actual tele- phone practice the microphone is used as a transmitter and induc- tion does not play any direct part in its functions. Laws of Induction. — There are several laws affecting electro- magnetic induction w^hich may be given here. Faraday's Law is based on his discoveries in the induction of currents by the cutting of lines of force. It is given in the follow- ing words: ''When a conductor is moved in a magnetic field so as to cut the lines of force, there is an electromotive force impressed on the conductor, in a direction at right angles to the direction of the motion, and at right angles also to the direction of the lines of force." Fleming's Rule for remembering this law and the connection between the three factors motion, magnetism, and induced elec- tromotive force is this: Hold the thumb and first finger of the right hand at right angles to each other. Let the forefinger rep- resent the lines of force and point in their direction. Then the hand will represent the north pole of a magnet. The thumb will represent the direction of movement of a conductor. The INDUCTION. 213 latter is represented by the middle finger pointing at right angles to the other two. Then moving the conductor in the direction of the thumb, an electromotive force in the direction in which the finger points will be impressed upon it. The words ''current induced" may be substituted for "electro- motive force impressed." When direction is attributed to electro- motive force it refers to the direction of current which such elec- tromotive force would produce. The cut, Fig. 121, shows the relations of the three factors as described. Ampere's Rule Adapted to Induction. — Suppose a man swim- ming along a conductor with his back to the north pole of a magnet whence lines of force is- sue. Then if he and the conduc- tor together be moved toward his right hand, the induced current will flow in the direction in which he is swimming. When the movement is not at right angles to the lines of force a certain proportion of the move- ment can be found which will be at right angles and this represents the effective portion of the movement. The object of the adop- tion of the idea of perpendicular movement is for the sake of simplicity. Clerk riaxwell's Rule. — If a magnet is in the presence of an active circuit which therefore produces a field of force, each por- tion of the circuit acts upon the magnet in such a direction as would cause the magnet, were it free to move, to take up the posi- tion in which the greatest possible number of its lines of force would be embraced by the circuit. Lenz's Law is a most convenient statement of the relations be- tween the motions of magnetic poles and currents induced by their lines of force. While such relations can be worked out from a lower basis, the summarization known as Lenz's law will be found an admirable tool to work with. It is all-essential to Fig. 121.— Fleming's Rule. 214 ELECTRICIANS' HANDY BOOK. understand that it is in strict accord with Ampere's law. It is generalized by Daniell, author of ''DanielTs Physics," thus: "When- ever a closed circuit capable of bearing an electric current lies wholly or in part in a magnetic or electro-magnetic field of force, any disturbance in the intensity of the field of force will induce a current in the circuit; and the direction of the induced current is determined by the rule that the new current will increase the already existing resistances (not electrical but mechanical resist- ance), or develop new resistance to that disturbance of the field which is the cause of induction." The law is more briefly expressed thus: "When a conductor is moving in a nagnetic field a current is induced in the conductor in such a direction as by its mechanical action to oppose the motion. It is also divided into two divisions, one for a generator read- ing thus: The induced current is always such that by virtue of its electro-magnetic effect it tends to stop the motion that gen- erated it. In accordance with this statement a dynamo requires more energy to be expended on it when it is generating current than when idle, because the passage of the current increases all elec- tro-magnetic effects and also the Lenz effect of resistance to motion generating the effects. For motors the converse division of the law is put thus: The motion produced in a motor by the passage of an electric current is always such that by virtue of the electro-magnetic inductions which it sets up it tends to stop the current. This division covers the case of counter electromotive force. Examples of the Application of Lenz's Law. — Synchronous alternating-current motors are addicted to varying in speed, going at one instant faster than the generator and at the next instant going slower. This action, disastrous to all regularity, is some- times called hunting. It is checked by inserting coils of wire or other conductors in the field magnet pole faces. Currents are induced in these by change of velocity of the rotations. By Lenz's law such induced currents tend to stop the objectionable hunting motion which produced them. A direct-current motor not doing anything and running with- INDUCTION. 215 out any mechanical resistance ought, it would seem, to run at an indefinite and almost unlimited speed. As the motor turns it in- duces electro-magnetic effects, which are greater the faster it re- volves. By Lenz's law these effects are such as to oppose its mo- tion. As they increase with its speed, opposition to its motion increases with its speed. It cannot, therefore, exceed a certain rate. This is another way of stating the action of counter electro- motive force upon a motor. Complaint is sometimes made that electric cars go too fast in cities. Their speed could be easily limited by constructing their motors so as to have any desired limit of velocity of rotation. Two conductors carrying current in the same direction attract each other; in opposite directions, repel each other. If one wire is carrying no current but has its ends connected and the current through the other is increased, an opposite current will be in- duced in the second wire. The wires repel each other when cur- rent is induced, and in the opposite case, when induction is dimin- ished, attract each other. Lenz's law fails to touch this case be- cause there is no mechanical motion. But let a steady current pass through one wire and let the other closed circuit, which includes the second wire, be moved closer to it, and the current induced will resist the motion. It will be a current in the oppo- site direction. If separated the induction will resist the separa- tion, and the currents will be similar in direction. Lenz's law is best taken with its due limitations — that it only applies to the relations of electro-magnetic induction to mechani- cal motion causing it or produced by it. It is not a good practice to try to stretch it to cover induction where there is no mechani- cal motion. Foucault or Eddy Currents. — If a conductor should be so moved in a field of force that the number of lines of force pass- ing through it at an angle with its direction of motion vary, a current will be produced within it. This current will circle around in its mass, will absorb energy and expend it in heating the metal. Such currents are called Foucault or eddy currents. A set of infinitely thin conductors with ends unconnected moved through a field, if insulated from each other, would require the 216 ELECTRICIAXS' HANDY BOOK, expenditure of no energy, on the assumption that being of infinite thinness, no electric circuit can exist in them. If their ends were connected by a conductor under the jonditions already specified as to variation in density of field, then a current would flow and energy would be absorbed. If a heavy solid conductor were sub- stituted for the infinitely thin ones, while local currents would be established in it, there would be no through current all in one direction caused to pervade it. If its ends were connected by a conductor such a through current would be established. The local currents in the mass of the conductor are Foucault or eddy currents. Variations in Impressed Electromotive Force, — The conduc- tor which cuts the lines of force forms part of a circuit, and in Fig. 123.— Closed Loop in a Bipolar Field. cutting the lines of force either increases or diminishes the lines of force threading or interlinked with the circuit. The conductor indicated in the diagram, Fig. 118, starting at the left of the pole, cuts lines at a comparatively slow rate. This is because the lines are not so densely placed as directly opposite the pole. Hence a relatively small electromotive force is impressed upon it at the distant point. It cuts more and more lines of force in a second as it approaches the pole, thereby changing the number of interlinked lines with greater and greater rapidity, so that the electromotive force, and consequently the current, is strongest when the conductor is opposite the pole. It then, for like rea- sons, diminishes as the conductor recedes to one side of the pole in its motion. Such a conductor is shown in Fig. 122, reproduced INDUCTION. 217 from a preceding page. There is another case the opposite of this. The conductor described is a part of a circuit the plane of whose moving portion is in line with the axis of the magnet when the armature conductor is opposite 'the center of the pole. Suppose the current is to be induced in a flat coil swept across the pole, and that the coil is perpendicular to the magnet ax;s when the coil faces the pole. Such coils are shown in Fig. 120 Such a coil will be interlinked with the greatest number of lines of force when opposite the pole, but its change rate will at that point be the lowest. The current induced by the impressed electromotive force will be least at this point. This condition obtains in many alternating-current generators. The maximum electromotive force is induced when the armature coils are midway between two poles; the electromotive force is zero when the coils are opposite the poles and in the densest field. Direction of Current Induced in Coils. — The direction of the current induced in a coil is determined by Lenz's law. If it ap- proaches a north pole the currents induced will oppose its ap- proach; they will therefore be the reverse of the Amperean cur- rents or of the currents in the magnetic coils. As the coil recedes the currents will reverse also by Lenz's law so as to oppose the motion, and will coincide in direction with those of the magnet poles. Fig. 120 and several other cuts illustrate the principle. The poles of a magnetic field are shown facing the observer; the direc- tion of the induced currents is shown by the curved arrows. The coil in which current is being induced is moving in the direction indicated by the arrow. Arrowheads are marked on the coils to show the direction of the currents induced. When directly oppo- site the pole there will be no change in the number of lines of force passing through the coils, and no currents will be induced in them. CHAPTER XL DIRECT-CURRENT GENERATORS ANJ> MOTORS. Dynamo « Electric Generators.— Electric energy is now almost universally produced on the large scale by dynamo-generators including the following parts: A strong magnetic field or fields are produced by one or more electro-magnets. The magnetic circuits include the core of the electro-magnets and a mass of iron between or near their poles which constitutes the armature core. Coils of insulated copper wire are wound upon the armature cores. By mechanically changing the relations of armature windings and fields of force electromotive force is impressed upon the circuit and a current results. The product of electromotive for^e and current is elec- tric power. Mechanical energy is required to operate the mechan- ism for changing field and armature coil relations. This energy is absorbed by the machine, and electric energy is produced in its stead. The easiest way to understand the dynamo, as it is often termed, is to follow up the construction from the simpler to the more complicated types. Interchangeability of Dynamo and Motor. — The interchange- ability of dynamo and motor stands as the subject of one of the greatest discoveries in electric engineering. Electric motors had been constructed for many years before it was definitely decided that the same machine could receive electric energy and convert it into mechanical energy, thereby constituting itself a motor, or could be operated by a steam engine, water turbine, or other prime motor, receive mechanical energy, and convert it into elec- tric energy. It then is a generator or dynamo. As dynamos are calculated to give a definite electromotive force and current, and as motors are calculated to absorb a definite DIRECT-CURRENT GENERATORS AND MOIORS. 219 electromotive force and current, the calculations for motor and dynamo are on the same lines. Varieties of Dynamos. — There are two grand divisions of dy- namos; one is for the production of the direct current, which is a current of unchanging direction; the other is for the production of the alternating current, which is a current reversing its direc- tion periodically, in practice from twenty times upward a second. Although a current which changes in direction may be consid- ered as an aggregation of different currents of opposite direction, this aggregation is always called an alternating current, and is treated as a variety of single current. The principal constituent parts of a dynamo are the field, con- sisting of core and winding; the armature, consisting also of core and windings; the collecting rings or commutator and brushes. The field and armature vary in construction, their windings vary in system, and from these variations many varieties of dynamos are derived. Elementary Idea of an Alternating-Current Dynamo. — As- sume a bipolar (two-pole) field which in the cut, Fig. 122, is indi- cated by two magnet poles facing each other and marked N and S. Let a simple rectangle of wire such as shown be rotated about an axis, a h, in such a field. As one side sweeps across the north pole the other sweeps across the south pole, and electro- motive foBce of opposite polarity is impressed on the two sides of the rectangle, so that a current is produced through it. This current is strongest when the cutting conductors are passing the poles, sinks to zero or nothing when the plane of the rectangle is at right angles to the lines of force extending from pole to pole, and reverses in direction as this point is passed. During half the revolution the current flows in one direction, and dur- ing the other half in the other. This constitutes a dynamo. Collecting or Slip Rings. — In this dynamo the current is con- fined to the rectangle, which is supposed to be a continuous con- ductor insulated from the axle. Suppose it to be cut at one end at the axle, and let the ends be connected each to its own ring fastened around the axle and insulated from it. The rings are also to be insulated from each other. If the rectangle is rotated, electromotive force of alternating polarity will be impressed 220 ELECTRICIANS' HANDY BOOK. upon it, but as it is an open circuit, no current will be produced. Let a spring bear against each ring, and let a wire of greater or less length connect the springs. The circuit is thus closed, and currents first in one direction and then in the other flow through the whirling conductors and the wire. The rings are Fig. 133.— Use of CoiiLECTiNG or Slip Rings. called collecting rings. The currents are treated as one and are called an alternating current. The arrangement is shown in Fig. 123. Brushes. — The springs which bear upon the collecting rings or commutators are called brushes. Often instead of springs, blocks of carbon, pressed by- springs, are used. The brushes must be insul- ated from the frame of the machine. Elementary Idea of a Direct=Current Dy= namo. — In the next cut, Fig. 124, the rec- tangle is shown with its ends connected to segments of rings, each one as nearly as possible 180° or a half circumference in extent. They are insulated from each other and insulated from the shaft and attached to it. Springs at opposite sides press against them. The segments constitute a commutator, whose section is shown in Fig. 125. Let the rec- tangle in Fig. 124 with its two-part commutator be rotated in the Fig. 134.— Rectangle Cohnected to commtitator, DIRECT-CURRENT GENERATORS AND MOTORS. 221 two-pole field. Let tlie springs which are insulated from each other be connected by a wire. As the rectangle passes the points where no current is generated, the springs pass from one com- mutator division to the other. As current goes in one .direc- tion through the rectangle, it is delivered to springs in one sense. As the current reverses in the rectangle, it is delivered to the springs in the other sense, because as the current changes, the springs change their contacts with the commutator segments. Hence the wire connecting the springs receives currents varying from zero intensity up to a maximum, but always in the same direction. Increasing the Electromotive Force by Increasing the Turns. — ^The electromotive force is propoj;tional to the lines of force cut per second by the whirling conductors. It may be increased Fig. 135.— End View of Two- PiECE Commu- tator. Fig. 126.— Double Rectangle Connected TO Two- Part Commutator. by increasing the turns in the rectangle. In Fig. 126 they are shown doubled (the dotted line is the axis of rotation) and they may be increased any number of times. The turns are insulated from each other and are continuous. Doubling the turns doubles the electromotive force, and so on. Increasing the Electromotive Force by Adding an Armature Core. — The field may be made denser by filling the space between the poles as completely as possible with a mass of iron. This is done by providing a cylindrical iron core, which almost fills the gap between them, and winding the wires on that. The denser field giving more lines of force, it follows that more lines of force are cut per second, and that a higher voltage results. Armature and Core. — The iron cylinder with the wire wind- ings constitutes an armature. The iron cylinder alone is the 222 ELECTRICIAXS' HAXDY BOOK, armature core. The wires are the windings of the armature. Each convolution of the wire is called a turn. Field Poles. — The early magnetos, dynamos, and motors were based on the horseshoe or U-shaped magnet as a producer of the field of force. Where a single magnet was used, this constituted a two-pole or bipolar construction. Recent practice favors the use of more than two field poles, or of multi-polar dynamos. In these as a rule each pair of poles in- duces two parallel currents, and in typical w^inding there is a brush for each pole, and the brushes are spaced at equal angles around the commutator. As a general rule, the number of poles is even; there are as ^many north poles as there are south poles. The poles alternate with each other, a north pole coming next to a south pole. Fig. 127 shows a sec- tion of a four-pole field with arma- ture core, the lines of force being in- dicated by arrows. Dynamos and motors can there- fore be classified from their number of field poles as bipolar, four-pole, six-pole dynamos. Two general di- visions are bipolar dynamos and multipolar dynamos, the latter in- cluding all except bipolar ones. Open=Coil Armatures. — The ele- mentary armatures described up to this are open-coil armatures. They may have any number of coils and any number of turns in each coil. Open-coil armatures are used in practice principally on the Brush and Thomson-Houston dynamos. In them they are greatly developed from anything shown here. They are used in great number on testing and signaling magnetos. The name open coil is given to them because no closed circuit can exist in their windings; the outer circuit has to be connected to the brushes to give a closed circuit. Spindle or H Armature. — The sj^-indle or H armature had in early days a considerable vogue. It is now definitely abandoned Fig. 127.— MULTIPOLA.R Field AND Armaturk Core with Magnetic Circuits Indicated DIRECT-CURRENT GENERATORS AND MOTORS. 22S in favor of better constructions, except for very minor uses. It is a single-division drum armature. The contour of the core is that of a cylinder with two grooves running lengthwise of its surface and diametrically opposite to each other. The cross sec- tion of such an armature represents a sort of letter H, whence one of its names was derived. It was a very distinctive armature with Werner Siem^ens in his early machines. It had a two-bar commutator and was an open coil. It was a poor form, as it had low permeance and inevitably gave a highly pulsatory current, as it only admitted of two di* visions in the commutator. The cut. Fig. 128, shows this arma- ture. Closed'Coil Direct=Carrent Armature, —This is a type whose windings are so connected as to form a closed circuit. This is irrespective of the brushes. The great majority of machines have this type of armature. Fig. 128.— Siemens's Spindle or H Armature. The characteristic current distribution in a closed-coil direct- current armature involves parallel currents in its windings. In a two-pole dynamo the current in one half of the windings, is parallel to that in the other. In a four-pole dynamo there are four divisions, each with its own current in parallel with that in the next division. The same principle applies to any num- ber of poles. The collecting brushes are in typical constructions equal in number to the poles. While the above is true for most dynamos, the windings of the armature can be greatly modified, so as to bring about differ- ent current distributions. The above are characteristic, and represent the usual practice. It follows that the currents in a closed-coil direct-current arma- ture never go through the winding consecutively. The brushes. 224 ELECTRICIANS' HANDY BOOK. are placed on the commutator at points where parallel currents meet. When the outer circuit is open, the electro-motive forces Induced meet at these points and neutralize each other, so that no current is induced in the windings, although they are in closed circuit. Cutting Lines of Force Without Change in Number of Inter- linking Lines can also produce a current. The circular con- ductors which have just been illustrated as having no current produced in them, although they cut lines of force, have electro- motive force impressed upon them. But the electromotive force can be located in halves of the ring separated by a diameter, in a general way perpendicular to their direction of motion. The electromotive force in one half is of similar polarity to that in the other half, so that they oppose each other and no current is produced. Electromotive force cannot exist without the possibility of a current being produced by it. A current by some mechanical arrangement could be taken from the extremities of the diameter of the ring without any change in number of interlinking lines of force. A machine in which this is done is called a homopolar, uni- polar, or acyclic generator, and .is described very briefly else- where. It has not gone into very extensive use, although it probably has a future. The ordinary generator produces its effects by so cutting lines of force that the number interlinking the circuit changes as they are cut by the conductor. Without such change no current would be produced in ordinary machines. The point is mentioned here to fix the fact that the cutting of lines of force is what pro- duces electromotive force, and that the variation in number of interlinking lines is something which happens in ordinary gen- erators when the lines of force are cut. la homopolar dynamos the above variation does not occur. The cause of impressment of electromotive force is the cutting of lines of force, not the variation in interlinking lines of force. CHAPTER XII, DIRECT-CURRENT ARMATURE WINDING/ Armatures. — Tlie function of an armature is to support con- ductors forming part of an electric circuit, which are to be sub- jected to the action of a field of force whose relation to the wind- ings constantly varies. With fixed relations of the field to the armature conductors there would be no current induced in ma- chines of the usual type as here described. The relations are varied, so as to induce current by the sweep- ing of conductors and field poles past each other. This is univer- sally done by having the poles and armature coils arranged on a circle; and by rotation of either armature or field, or by rotation of a series of "inductors" of soft iron past the poles, the desired varying of relations is brought about. Armatures are wound in many ways. For direct-current work the closed-coil drum armature is much used. It is the success(>r of two historical inventions, the Pacinotti disk armature and the Gramme ring armature. In both of these the closed-coil feature appeared, which characterizes most modern dynamo and motor armatures. The Pacinotti Armature. — ^The modern armature is with a few exceptions wound on principles exemplified by the famous Pacinotti (pronounced Pacheenot-tee) armature of 1864. These principles require the winding to be consecutive from beginning to end, so that the windings form one closed circuit. Such a winding is characterized as re-entrant when in the winding the last end falls into place, so as to be in line with the first end. The winding is carried out as symmetrically as possible, and at symmetrical points it is connected to divisions of the com- mutator. Pacinotti described his armature in 1864; it constituted part of 226 ELECTRICIAXS' HAXDY BOOK a motor. It was a disk-shaped armature mounted horizontally and wound with a continuous winding of wire and with sixteen connections from the windings to sixteen insulated bars on the axis of the disk, which bars constituted a commutator. Under the disk was a circle of iron polarized by an electro-magnet underneath it. and whose legs rose vertically to opposite extremities of a diam- eter of the iron ring. This produced two opposite poles on the ring. There was here embodied the salient features of the modern dynamo, its continuous winding and commutator with numer- ous divisions, each connected to the winding. A toothed iron rin^ with wooden pegs or projecting pieces of boxwood formed the core on which the armature windings were made. Pacinotti had no idea that by turning the armature mechanic- ally a satisfactory current could be produced. His motor con- tained the elements of a dynamo unknown to himself. The Gramme Ring, named from the inventor Gramme, w^as de- scribed in the Comptes Rendus (Paris) in 1871 and 1872. It wa3 patented in 1870. It is a type of armature which acquired an immense vogue, and became in a sense one of the scientific glories of France. It is not much used in this country, where the drum armature is generally adopted. The Brush and Thomson-Houston open-coil dynamos are the principal American machines using it. But abroad many ring-armature machines are built. Gramme's original ring core was made smooth and of circular cross section and was entirely overwound with wire. riodern Types of Closed=Coil Armatures. — The modern arma- tures may be grouped into four classes: Ring armatures, drum armatures, pole armatures, and disk armatures. The^ring arma- tures are based on the Gramme ring. The drum armature i3 sometimes taken as being derived from a Gramme ring by filling up the central opening with iron. The pole armature recalls an early type, and finds one of its great applications in alternating- current machines. The disk armature dates back to one of Pacinotti's machines. In modern American practice the ring and disk armatures are not much used. Armatures for direct current vary from those designed for al- DIRECT-CURRENT ARMATURE WINDING. 27 Fig. 129.— Two-Part Gramme Ring Armature. ternating current. The latter will be described by themselves. The ring armature may be taken as to a certain extent the pro- totype of the drum armature, and will be first treated. ^ The two-pole or bipolar dynamo is a little simpler than the multipolar one, and will be the starting point for the description of arma- ture windings. The Gramme ring is a ring of soft iron on which the armature coils are wound. Fig. 129 shows in diagram a ring, supposed to rotate about an axis perpendicular to the pa- per in the two-pole field. The dotted lines show the course of the lines of force. The ac- tive parts of the armature windings are those on the outside of the ring. It may be made more complicated and efficient by doubling the parts, as in Fig. 130. B^ and B2 are the conductors from the brushes; at A and C no current is induced; at D and E the maximum is induced. The arrows give the course of the currents, and show how they meet in opposite commutator sections. The windings operate in parallel of two To make a real working armature, a large number of windings are requisite, and the commutator is divided into many sub- divisions. Such a ring '!V / // s k N Fig. 130.— Four-Part (Irammt? Ring Armature Showing Course of Current. is indicated in diagram in Fig. 131. Commutator Connections of Ring Armature. — In practice a commutator is mounted on the shaft, and wires are led from the windings of the ring to it, and the brushes bear against the 22S ELECTRICIANS' HANDY BOOK. commutator and take current from it. The wires are led from symmetrically or evenly spaced portions of the winding, and Fig. 131.— Gramme Eing Armature Showing Commutator Connections. each one is connected to its own commutator segment or leaf. The commutator is used from mechanical considerations, other- wise the current could be taken from the conductors on the out- side of the ring, as indicated in the diagram. Fig. 132. Cores of Ring Armatures were originally made of iron wire wound into a circle of any desired thickness. Present prac- tice makes them of thin ring- shaped laminations. A closed ring is somewhat troublesome to wind, so ring cores are often made in two semi-circular halves, over which the coils al- ready made up can be thrust, after which the two halves are bolted together, so as to form a ring. Permeance of the Ring Core. — One of the objections to the ring core for two-pole machines is its low permeance as compared with a drum core. Every effort may be made to reduce the central opening of the ring, yet it is Fig. 132.— Relation of Gramme Ring to Commutator Brushes. UlttJ^JUT-VUKKEJST ARMATURE WINDING. 229 not easy to imagine a ring armature core of as high permeance for a bipolar field as a drum armature core would be. The cross section of the core varies greatly with different con- structors. The modern appreciation of the laws of the magnetic circuit has led to the production of ring cores of good^ per- meance. Some of the older rings were thin, and consequently of low permeance. For multipolar machines a ring armature may have about as good permeance as that of a drum armature. This is on account of the course taken by the lines of force, which go into the core and out of it within a small portion of its circum- ference. In a multipolar machine they follow a U-shaped path from pole to pole, largely through the outer layers of the core. In the bipolar machine the lines of force have to go from one side of the ring to the other. This develops the bad permeance ot the ring armature to the highest degree, as the lines of force following the curved path of the core have a longer distance to travel than that followed by them in going directly across a drum- armature core. In plain ring winding there is a brush for each pole, w^ich brushes normally collect current from points of the commutator nearly symmetrically located between the poles, so that a like difference of potential exists between each pair of contiguous brushes. By special winding one pair of brushes can be made to answer for a multipolar machine. This has the objection that it leads to unsymmetrical positions of the brushes, with conse- quent uneven voltage between the brushes appertaining to the two sides of the commutator. Idle Wire. — In the ring armature the wire on the outside of the ring is the active portion. All on the inside is idle as far as the impressment of electro-motive force is concerned. This is on^ of the objections to this type, an objection which is of some mo- ment, as low resistance in the armature is an element of effi- ciency. Current in a Ring Armature. — The course of the currents in a ring armature in a bipolar field is shown in Fig. 133, in which the arrowheads show the direction. The currents meet at two points at the extremities of a diameter which is at right angles to the diameter determined by the axes of the poles. This axis is 230 ELECTRICIANS' HANDY BOOK. marked A B. The arrowheads indicate the current in the eight coils. The impressment of electromotive force which causes the current is principally in the coils 2, 3, 6 and 7 of the arma- ture in its present position, but the armature is supposed to be turning, so that the coils acted on are constantly dif{:erent ones. The brushes remain fixed in position, indicated by the line A B. The arrows under N and S are supposed to be one in front of and the other behind the core, and indicate the current which will be impressed by the induction of the poles. The curved arrow denotes the direction of rotation of the core. Open - Wound Fcur * Part Ring Armature. — In Fig. 134 Bi Ba indicate the brushes which are taking current from the horizontally-placed pair of coils. These are in the position in which the highest degree of electromo- tive force is impressed upon them; and during the period of such active impressment, the brushes receive current from them. After a rotation of about half the arc of the commutator division, the coil is open-circuited, and the other one has its circuit closed, as the brushes come in contact with the commutator sec- tions connected to it. This illustrates the principle of the open- coil armature. It apparently fails to utilize half the ring sur- face, but in any case one-half of this surface represents the locus of impressment of by far the greater part of the electromotive forcp. The arrows have the usual significance. Mounting of a Ring Armature. — The diagram showing the relations of a ring armature to its field in a series-wound dyna- mo is shown in Fig. 135. The field magnet i- wound to give consequent poles, N N and S S, above and below the vertical diam- eter of the ring R. The current is taken from the opposite sides Fig. 133.— Currents in the Gramme King Armature. DIRECT-CURRENT ARMATURE WINDING. 231 Fig. 134.— Open-Coil. Ring At'.m attire. of the commutator and goes through the field coils. The axle A B of the armature is journaled in the magnet yokes. Multipolar Ring Armature. — The ring armature can be used in a multipolar field with- out change. All that is necessary is to have more hrush*es than two, so as to take the current off from several parts of the winding. If the currents in a Gramme ring are traced, it will be found that neutral points are es- tablished equal in number to the poles of the field. If the winding and com- mutator connections are symmetrical, the neutral points will lie midway between the radii which go through the axes of the poles. The current is easily traced by following the rule given on page 210, and treating the parts of the wire outside the ring as conductors corresponding to the arrows of the diagram on the same page. It is in any case obvious that if the current is induced in one direction in front of the north pole, it will be induced in the other di- rection in front of th« south pole. The cur- rents therefore meet midway between the poles, and are to be taken thence by a brush. This brings a brush midway between each pair of poles, so that they aggregate one brush for each pole. The development of a four-pole ring winding is shown in Fig. 136. The Drum Armature,— It has been noted that the wire on Fig. 135.— Diagram of "Windtng of a Gramme IUng bi.HiES Machine, 232 ELECTRICIANS' HANDY BOOK. the outside of a ring armature is the active part. The large opening of the ring decreases its permeance. If the opening were filled with iron and the idle wire suppressed, one improve- ment would result — the lowering of reluctance or increasing of permeance, and in some cases there would also be brought about a reduction of resistance. If a reduction of resistance occurs, it is due to the reduction in length of the wire. This reduction is to be looked for in the transition from a thick ring core with small central aperture to the drum core — not in the transition F:o. 133 —Development op a Four-Pole Ring Armature. from the old-style thin-bodied ring core with large central aper- p ture. In the ring armature shown in Fig. 131 imagine the center open- ing filled with iron and the inner wires removed. Other leads must be carried across the two ends, so as to bring the whole quantity of wire into one consecutive coil, and from symmetrical- ly-located points on the windings leads must be carried to a com- mutator. This gives a drum armature. A drum armature is unlike a ring armature in one respect. If wound for a bipolar field, it will not operate properly in a four- pole, or other multipolar machine. In a six-pole field it v/ould give some result; in a four-pole field, none. DIRECT-CURRENT ARMATURE WINDING. 233 Action of the Drum Armature. — A conductor on the periphery of a drum armature swept across a field pole has impressed upon it electro-motive force the reverse of that impressed upon one swept past the opposite pole. If the current induced flows to the com- mutator end in one conductor, it will flow away from it in the opposite conductor. It follows that to obtain a continuous cur- rent such conductors should be connected to each other. Then the current as the circuit is completed will flow in one direction through one active wire, then across the end of the core in th^ connecting wire, in the reverse direction in the other active wire, and across the other end. If the wires correspond in angular distance to two opposite poles, this course of current will be given by the impressed electromotive force. This is the reason why wires opposite a north pole must be connected to wires op- posite a south pole. If a wire were connected to one directly opposite on both ends, there would simply be a series of short-circuited conductors, ag- gregating as many complete short circuits as there are leaves or bars in the commutator. To secure a continuous winding, the conductors exactly op- posite are not directly connected. Direct connection is made as described between conductors nearly but not quite opposite to each other. With every cross or end connection a step in ad- vance (wave winding) or a step in retardation followed by a longer one in advance (lap winding) is made. The final result is the same in either case; a uniform progress around the cylin- der is made by the windings, somewhat similar to a spiral. Drum Armature Windings.— To form an idea as easily as pos- sible of the essential features of the drum winding, an example may be given of a winding with .very few conductors. The winding of a drum armature may be divided into three classes or parts. The first are straight lines of wire or con- ductors which cut the lines of force. These lie upon the cylin- drical surface of the core, parallel with its axis. If we stop here, we have simply a lot of short straight pieces of insulated w^ire occupying the places of the elements of a cylin- der. They are of- the same length as the core. The commutator end of the armature may be termed its front end. The straight 234 ELECTRICIAXS' IIA2^^DY BOOK. conductors now must have their ends connected across the front and rear ends of the core. The rule which must be followed for closed-coil winding is that each wire must be connected to one at an angular distance from i. corresponding approximately to the interval between the nearest north and south poles. For two-pole fields such as arc now^ being described this distance is approximately 180°. At the back of the armature, wires run across its surface connecting the conductors on the periphery of the drum or cylindrical core, subject to the rule just stated. These operate with the front connections to connect all the wind- ing into one continuous circuit. At the front each of the wires lying on the cylindrical surface is connected to an armature bar. To the same bar is connected a wire connected to an opposita conductor, which is approximately 180° distant from the first one in a two-pole field. If it were a four-pole field, the angular distance would be approximately 90°. Simple System of Armature Winding. — In a simple type of bi- polar winding the rear end of a wire might connect with one which would be one wire out of perfect opposition. In front it would connect to a commutator bar. The same commutator bar would then connect to another wire nearly opposite, which would be near to the one with the rear connection. This if repeated would join all the wires into one continuous lead. The number of commutator bars would be equal to one-half of the peripheral surface wires. Eight=Conductor Drum Armature. — Suppose there were eight surface conductors or wires. Starting with any desired wire, let them be numbered consecutively. As the most natural way, we may start with wire 1. In front it is connected to a commutator bar, which we may designate as a, From this commutator bar the second connection runs to wire 4. This is one less than half th3 wires. The rear end of wire 4 is connected across the rear of the core to wire 7. Now returning to the front or commuta- tor end, wire 7 is connected to commutator bar d, and the second connection from commutator bar d goes to the front end of wire 2. Counting forward, this is one less than half the number of wires. The rear end of wire 2 connects with the rear end of wire 5. The front end of wire 5 connects through the commu- DIRECT-CURRENT ARMATURE WINDING, 235 tator bar c to wire 8, also an interval of one less than half the wires. The rear end of wire 8 connects to the rear end of wire 3, and the front end of wire 3 through the commutator bar h con- nects ^ith wire 6. The rear end of wire 6 connects with the rear ena of wire 1. This closes the circuit. A winding table for the above is given here. 1 a 4 indicates that wire 1 connects through a to wire 4. 7 d 2 that wire 4 connects to wire 7, that wire 7 connects 5 c 8 through d to wire 2, and so on as explained. 3 & 6 above. The letters a, &, c, and d denote bars of the com- mutator. Twwlve=Conduct or Bipolar Armature. — ^The winding of this is shown in the diagram, Fig. 137. The dotted lines indicate the wires crossing the distant end of the core, the full lines those crossing its front. Again a departure of one wire from 180° angular distance is adopted. Wire 1 connects in front with wire 8 and on the rear with wire 6. Wire 1 to wire 7 would be 180° angular distance, whence it is evident that the winding is based on a departure from 180° of an interval of one wire. The commutator sections should be six in number, and should connect to the centers of the set of conductors which cross its end of the core. Sixteen=Conductor Bipolar Armature. — The winding is sho^n on the basis of a departure of one wire from 180° in Pig. 138. The neutral line is shown at right angles to the polar axis of the field. On part of the circles representing the end view of the active conductors crosses are marked. These indicate that the current in those wires goes away from the observer. The circles with central points indicate that in the conductors they repre- Fig. 137.— Twelve-Co NDur tor Bipolar Drum Armature Winding. 236 ELECTRICIAN^ ;^' HANDY BOOK. sent, the current is coming toward the observer. To remember this system, the points may be taken as indicating the points of arrows flying toward the observer and the crosses as indicat- ing the feathers of arrows flying away from the observer.^ This system of indication is often used in textbooks. + 'B Fig. 138.— Sixteen-Condenser "Ripol-ar Drum Armature AVinding. Winding Tables.— The winding tables for these three armatures, omitting commutator bar letters, are as follows: Eight-Conductor. Twelve-Conductor. Sixteen-Conductor. a 4 d 2 c 8 b 6 1 8 3 10 5 12 7 2 9 4 .1 6 1 8 15 6 13 4 11 2 9 16 7 14 5 12 3 10 In the twelve-conductor winding, wire 6 connects with wire 1, and in the sixteen-conductor winding wire 10 connects with wire DIRECT-CURRENT ARMATURE WINDING. 237 1, thus making the windings re-entrant. The windings may be studied out on the cuts, when the full significance of the wind- ing tables will be apparent. Windings for Multipolar Fields, — In bipolar winding, every- taiug in the way of the spacing of conductors is referred to 180°, oi to one half of a circumference. The cross connections over the ends of the drum core connect conductors separated a little rcore or a little less than 180° from each other. The angular distance 180° is the distance from center of pole face to center of pole face. When a drum armature is wound for a multipolar field, the angular distance between adjoining north and south poles is substituted for the 180° of bipolar winding. Suppose that there are eighteen conductors on the cylindrical surface of the core. This gives four and a half conductors to each pole if there are four poles in the field. The quarter circumfer- ence is the controlling factor. Conductor 1 in bipolar wind- ing might connect to number 12 or 14; in four-pole winding it may connect to number 6, number 6 to number 11, and so on, going five conductors at each connection. Eighteen = Conductor Four=Pole Armature. — This four-pole winding with eighteen conductors is illustrated in Fig. 139 in diagram as hitherto, and in Fig. 140 a circular development of the identical winding is given. The dark spots near the center of the latter diagram indicate the points where the brushes take the current. The outer lines forming the points of the star are the connections crossing the rear end of the core, and corre- spond to the dotted lines of Fig. 139. The short straight lines running from inner circle to outer circle represent the straight Fig. 139.— Eighteen-Conductor Four- I OLE Drum Armatuue, Wave Winding. 238 ELECTRICIANS' HANDY BOOK, conductors on the periphery of the drum. The cylindrical surface Fig. 140.— Circular Bevelopment of Armature Winding. of the drum is represented by the annular area between the two circles. The lines within the inner circle represent the con- nections at the front or commu- tator end of the core. Circular Developments are used a great deal to illustrate armature windings. The points outside the rings have no real existence as shown. They mere- ly indicate the center of the cross connections over the hsad of the armature core. Commutator Connections are shown in Fig. 141, a fourteen- section armature winding with seven commutator divisions. Wave and Lap Winding.— There are two divisions or classes of Fio. 141.— Commutator Connec- tions IN Fourteen-Conductor Drum Armature. DIRECT-CURREXT ARMATURE WINDING, 239 winding i'or drum armatures, named as above. In the first a uniform progression is obtained in the winding; in the second a retrograde step of a definite number of conductors is followed by a forward step of a larger number. Thus in wave winding each step is progressive; in lap wave winding the sum of every two steps is progressive. The development of these windings most obviously shows the origin of their names. An example of the development of v^ave and lap winding for an armature with eighteen periDheral conductors will be shown. Wave Winding. — The peripheral or active conductors are rep- Fig. 142.— Development of Etghteex-Conductor Wave "Winding. resented in Fig. 142 as vertical lines and numbered from 1 to 18. The cross connections at one end of the drum core are rep- resented by the lines of V-shaped connections above the vertical lines; the cross connections at the other end of the core are represented by the V lines below the vertical lines. If the reader will follow the course of the wires with a pointeii|Of any kind, he will see that there is a wave-like progress. The winding is exactly what is shown in Figs. 139 and 140. Lap Winding;.— The next cut, Fig. 143, shows the same arma- ture with eighteen conductors as before, buf with lap winding. Thus on the top of the diagram a wire starts from conductor 1 and goes to the right to conductor 6, which is five conductorb 240 ELECTRICIAXS' HAXDY BOOK. in progress. From conductor 6, instead of going forward the winding goes baclv on itself, or to the left to conductor 3; from conductor 3 the lead goes forward to conductor 8; tlien back to conductor 5, and so on. This ends by the winding from conduc- tor 17 going forward to conductor 2 and back to conductor 1. This ends the winding and leaves it re-entrant. Thus the wind- ings form a series of laps, going forward five sections and back- ward three sections, gaining two divisions for each two steps. Development of Commutator Connections. — The commutator bars are shown in the development as little rectangles, and they Fig. U3.— Development of Eighteen- Conductor Lap Winding. are indicated by small letters. There is one bar for each pair of nearly opposite conductors, and in the development they are shown connected to the angles, either above or below the dia- gram. These angles in the development simply represent the centerU of the end windings, w^hich go across the ends of the drum; they do not necessarily represent any angle or bend in the wire. Development of Field Poles. — These are represented back of the diagram, and each one is marked N or S according to its kind, whether nortli cr south pole. Development of Current Induced. — This is determined for DIRECT-CURRENT ARM A' U RE WINDING. 241 ilrum Vv^indings by the rule given on page 210, and the field poles in the diagram are shaded diagonally in accordance with that, rule. If the conductors are carried from left to right, those in the range oi the north pole will have downward currents in- duced in them, when the outer circuit is closed; those in the range of the south pole will have upward currents induced. Arrowheads are drawn to follow out this induction, and where the currents meet on the commutator, the brushes take off cur- rent to the outer circuit. A twenty-four conductor four-pole lap winding is shown in Fig. 144. Straight Developments. —The cuts. Figs. 136, 142 and 143, show a system of development much used in illustrating armature wind- ings. It is defective be- cause it has disconnected ends. If the paper were bent into a cylinder, these disconnected ends would come together, and the winding would form a closed circuit or be re- entrant. As drawn, this connection has to be as- sumed, just as the circular contour has to be assumed. Winding a Drum Arnfkture. — The drum armature winding in course of completion is shown in perspective diagram in Fig. 145. When completed a wire will pass around the cylindrical core in one continuous circuit. From symmetrical points leads are connected to the commutator bars. When such an armature rotates in a two-pole field of force, it will impress electromotive force upon a circuit connected to fixed brushes, two in number, bearing against the commutator surface at points 180° distant from each other and at approximately right angles to the diame- IiG. 144.— Twenty-four -Conduct OR, PouR-PoLs Drum Armature, Lap Winding. 242 ELECTRICIAXS' HANDY BOOK. ter connecting the center of the poles of the field magnet. The system may be followed out on the leads connected to the com- mutator bars, cL e, and f. The rest is incomplete, but the con- tinuity of the winding is shown in the part mentioned. Some- Fto. 145.— BnijM Armature tn Process of Winding, times wooden peg» are driven into slots in the core to keep the winding in place while being put on, as shown in Fig. 146. Another diagram illustrating drum armature winding is given in Fig. 147. The heavy black line represents one turn of the armature winding, fastened at one end to its proper commuta- FiG. 146.— Operation of Winding a Drum Armature. tor division. From the same division a second turn starts, and going around the drum connects to the next commutator division. The two diagrams illustrate the general lines on which drum armatures are wound. DIRECT-CURRENT ARMATURE WINDING. 243 General Considerations in Laying Out Drum ArmaLure Windings for direct-current generation admit of no final descrip- tion, because such windings can be executed in many different ways. A simple method of doing it, which follows the lines of what has been already described, is the following: The number of poles in„the field must be known. Usually these are of even number and in pairs of north and south poles, the two alternat- ing with each other. The number of layers of wire to be carried by the core is to be settled, and finally the number of commuta- tor bars. The controlling factor in settling the last factor is the total voltage. This divided by the number of bars gives the voltage between adjacent bars. The lower this is kept, the let'* ^IX 4 /i Fig. 147.— Conductors on a Drum Armature. danger will there be of sparking or arcing on the commutator. Another point to be kept in mind is that an increase of armature divisions, other things being equal, produces a more even electro- motive force and current. Single Layer Winding for Bipolar Field.— If the winding oi tbe armature is based on single active conductors, there must be twice as many of them as there are bars in the commutator. But for each such conductor any number of leads of wire may be substituted. The windings, whether in one or several layers, must be divided into twice as many sections as there are divi- sions in the commutator. By section, as will be seen later, is meant a group of wires lying side by side on the armature periphery. Each such division forms a portion of a continuous 244 ELECTRICIAN^ S' HANDY BOOK, coil wound around the core along its periphery and over its ends. Each such coil will leave two ends. These are connected each to its own commutator bar. Suppose that there are to be thirty-two divisions in the com- mutator; there must then be at the least sixty-four active con- ductors on the cylindrical surface of the core. There may be substituted for a pair of single conductors connected across the ends of the core a coil of any number of wires, whose free ends are treated as are the ends ot a single conductor in the simple case of sixty-four conductors. Suppose that there are two poles in the field. Then 180° is the controlling factor. A circle is drawn to represent the end view or cross section of the cylindrical armature core. Around this circle represent- ing the core section sixty-four points or little circles are drawn, evenly spaced from one another. These represent the end view of the sixty-four conductors or groups of conductors. The points or little circles are numbered consecutively. Starting rrom circle number 1, a full line is drawn across the targe circle to circle number 32 or 34. Either one of these is one removed from the 180° position, which latter is held by conductor number 33. Suppose number 32 has been selected. From it a dotted line is drawn to number 63. This is also one less than 180°, being two points distant from point 1, and removed one point from 180'\ Then from 63 draw a full line to a point removed by two points from point 32 and removed by one point from 180°. This is point 30. The same process is kept up until the line drawn from point 34 to point 1 closes the circuit, and makes the winding re-entrant. Double- Layer Winding for Bipolar Field. — Suppose that there are sixty-four conductors as before, but arranged in two superimposed layers. The circumference of the circle is divided into thirty-two parts, and sixty-four points are distributed around it in two concentric circles, each containing thirty-two points. The inner circle of points is numbered from .1 to 32, and the outer circle of points from 33 to 64. Starting from num- ber 1, a full line is drawn from it to number 16, and a dotted line from number 16 to number 31, and this is continued until all but one of the inner circle is connected and number 18 is DIRECT-CURRENT ARMATURE WINDING. 245 reached by a full line drawn from number 3. The inner circle of conductors could now be closed and made re-entrant by con- necting number 18 to number 1. This is the only open portion left. But this would leave out the outer layer of conductors. Accordingly, number 18 is connected by a dotted line to number 33 on the outer layer. A full line connects number 33 to number 48, a dotted line connects number 48 to number 63, and eventually all is closed and made re-entrant by connecting number 50 to number 1 of the inner layer. The object of drawing some lines dotted and others full is simply to distinguish between the ends of the core. The dotted lines cross one end, the full lines cross the other. Commutator Connections. — Every cross connecting wire on one end of the core must be connected to a commutator bar. Taking the full lines for the crossings on the commutator end, each of these must connect to a commutator bar. If Figs. 143, 144. and others are referred to, the"^ connection to commutator bars wHi be found indicated in them. The windings of the inner layer connect to alternate commutator bars, 16 in number; the windings of the outer layer connect to the remaining alternate bars. riultipolar Windings. — These may be laid out by the method given for bipolar windings, except that the controlling angular distance is 90° instead of 180°. Suppose a thirty-two-section armature is to be connected for a four-pole field. The conductors are drawn as dots or little circles around a circle as before and numbered. Starting from number 1, it is connected to a con- ductor one less or one more than required for 90°, say to num- ber 8, by a full line. Number 8 by a dotted line is connected to number 15, number 15 by a full line to number 22, and thus the process is kept up until a dotted line from number 26 to number. 1 closes the armature and makes it re-entrant. This is a single-layer wave winding. Suppose that as before we had two layers, each of thirty-two conductors. Then when number 26 v/as reached on the inner layer, precisely as above, a dotted line would connect it to number 33, a full line would connect number 33 to number 40, and so on until number 58 would be reached by a full line from number 51; then a dotted line f^om number 246 ELECTRICIANS' HAXDY BOOK. 58 to number 1 would close the winding and leave it re-entrant. The commutator connections are made substantially as described above. Multipolar Lap Windings. — The last three examples progress evenly, and are therefore wave windings. To make them lap windings, conductor number 1 should connect with a conductor more than 90° distant, and this last conductor should go back in its connections toward number 1. Thus, taking a thirty-two- section four-pole winding, numter 1 may connect by full line to number 10, this by dotted line to number 3, this by full line to number 12, and so on until the armature is closed by a doited line from number 8 to number 1. This system makes the wind- ing a lap winding whose net progression is two conductors in- stead of seven, as in the wave winding just described. Variations on the above are innumerable. The controlling angular distance has here been taken as one conductor more or less than 180° or 90°. But other distances can be taken. The absolutely essential feature is that conductors directly con- nected must be acted on simultaneously by opposite poles. Nomenclature for Drum Armature Windings. — A single turn of conductor comprising two peripheral conductors and the con- nections across the end of the core of a drum armature may have it"^ front ends connected to two adjacent commutator bars. A coil of many turns of wire may occupy the same place, and have its front ends connected to two adjacent commutator bars. Either of these portions of a winding are called ''elements." The active portions of an element lie on the cylindrical or periph- eral portion of the core, one for one pole and the other for the othc: pole, and are called ''sections." In connecting one "sec- tion" to another, so as to form an "element," a definite number of sections are bridged over or are caused to intervene be- tween the sections of an element. A sixty-four-conductor wind- ing under this nomenclature is a sixty-four-section or a thirty- two-element winding. In the sixty-four-section winding de- scribed on page 244 the distance from number 1 to number 32, which is a bridging of 32 — 1 or 31 sections, is called the "spacing," and it is a spacing of 31 sections. General Formulas. — For a bipolar winding we start from one DIRECT-CURRENT ARMATURE WINDING. 247 of the ends of section 1. The cross wire is taken across the end of the core to a section a little more or a little less than 180° removed from it. If there are four poles in the field for 180°, there must be substituted 90°, if six poles 60° must be substi- tuted, and so on. These controlling angles are equal to the quotient given by 360° divided by the number of poles in the field. V Taking a sixteen-section two-pole winding, it would have in- cluded 180°, had the cross wire gone from conductor 1 to con- ductor 9. Therefore, the wire may be taken to conductor 8 or conductor 10, one being 157°, the other 202° distant in angular measurement from conductor 1. Bipolar Winding Formula.— Denoting the total number of con- ductors on the cylindrical surface of the armature by Z. and the number in one element by &, Z/b is equal to the number of elements in the winding; and the number of sections, being two in each element, is equal to _ _ and is denoted by s. b Let the spacing be denoted by y in the cases cited above. The general expression for spacing is y= — ± a, in which a is any number compatible with tho requirements of re-entrant winding and the production of series connection through the winding. For J=L in the last formula we may substitute _?_ because 2 5 = -_ , and therefore A rr^, and the formula becomes: b 2 b V = ~± a. bipolar Winding by Formula. — To put a continuous re-entrant winding on a bipolar drum armature on these lines, s must be prime to y. If s and y have a common factor, the armature winding will have parallel re-entrant coils equal in number to this factor. Thus assume s -:= 16, giving a sixteen-section winding with 248 ELECTRICIANS' HANDY BOOK. eight armature bars. Let p represent the number of pairs of poles in the field. Let a = 1, and if the winding is bipolar, it follows that p = 1. _ _ = half the sections or 8. In the I) equation y = '^('V/- ^^ By substituting for p and a their values, each being equal to one, and for - - its value, or 8, we have: b 2/ = 8 ± 1 = 7 and 9. The one value of s which is 16 is prime to either of thes^^ values of y, so this winding will be re-entrant in one continuous coil. The values s = 16 and i/ = 6 or 10, which would result from making a =: ± 2, are not prime to each other, because they have a common factor 2. It follows therefore that this spacing would give two re-entrant windings, parallel to each other. Multipolar Winding by Formula.— There is no difference in general principles between bipolar and multipolar windings. Taking the four-pole winding described on page 245, the angular distance between sections is 78%°, where in the sixteen-section bipolar winding the distance was 157 or 202%°, The general formula just deduced can be mad^ to apply to a winding for a multipolar field. Denote the number of pairs of poles, which is half the number of single poles, by p. Denote the spacing by y. The value of y is then given by the formula: [e armature with thirty-tw( Ld letting a equal 1, we hav( 1 / no\ y - ^~r , '") -h 1 = 7 or 9. Suppose a four-pole armature with thirty-two sections. Then p =r 2 and 5 = 32, and letting a equal 1, we have for the spacing: The equation: 1 / 6' ^ 8 may be transformed to read: s=i2py ± a. DIRECT-CURRENT ARMA'^^URE WINDING. 249 This formula may be used to deduce the number of sections. Assume that a four-pole winding is to have about sixteen sec- tions; the spacing y will be about 16/4 or 4; and p, which is the number of pairs of poles, is 2. Substituting these values ^^ the last equation, it becomes: s=r2 (2X 4) ± 1 = 15 or 17. This method is only of interest for windings of many sections. For ordinary purposes a simpler plan is to take a number of sections divisible by the number of poles. Then select for the spacing a number one or two greater or less than one-quarter of the sections, remembering that it must be prime to the total number of sections, and without a common factor if for series winding. Thus assume a four-pole armature; 24 is divisible by 4, and will answer for the total number of sections, s. We have then s = 24, p = 2, and y = "^^ ± a. We may try for a the a X 2 24 values 1 and 2. The formula then becomes y =■ -^ — - ± 1 or ± 2 = 4, 5, 7, or 8. Of these, 5 and 7 are prime to 24, having no common factor. A spacing of 5 or 7 on a thirty-two four-pole winding will be in series and re-entrant. This v^^inding with a spacing of 7 is described on page 245. Lap Winding.— In bipolar and multipolar lap winding we have a net value for y. We have to go forward a distance equal r.pproximately to the distance between the contiguous pole cen- ters, and then to go back a lesser distance, leaving a net spacing equal to the difference between the two distances. This net spacing is equal to the algebraic sum or arithmetical difference of the two. It is an object to have the commutator bars even in number. To do this the number of sections must be divisible by 4. Thus, a fourteen-section or eighteen-section winding would give a commutator of seven leaves or nine leaves. CHAPTER XIII. THE DIRECT-CURRENT GENERATOR. The Magneto ( enerator, — This is a generator in which the field is produced by one or more permanent magnets. Fig. 14S shows a bipolar generator in dia- gram. Very large machines have been constructed with permanent magnets for the field. The cut, Fig. 149, shows the De Meritens machine, used for the production of the arc light, and the relation of field poles to armature coils is shown in the small diagram on the left, Fig. 150. The Modern Multipolar Dyna= mo has its yokes contained in and forming parts of a species of frame of iron. This is a circle or poly- gon. From its inner periphery cores, one for each pole, project toward the center like incomplete radii. The ends of the cores cut to the periphery of a smaller circl? form or define the armature cham- ber or tube. A drum or pole armature ro- tates in the space between the poles. One brush for each pole is typical. From these brushes one or more circuits may be supplied. The position of the drum, as it is acted on by the radial pulls of the symmetrically placed Fig. 148.— Bipolar Magneto Generator. THE DIRECT-CURRENT GENERATOR. 251 cores, tends to hold a central position, which is correct. The virtually circular frame subjected to radial pull alone is exceedingly strong, and the magnetic pull cannot deform it. There is no question of the material of the foun- dation, for nothing more than a magnet yoke, and perhaps not even that, comes in contact with the foundation. To the mechanic's eye the symmetry of the multipolar machine is attractive. The project- ing pole pieces are short and thick, so as to mini- mize leakage of lines of fiq. 149. _deMeritens Magneto Generator. forco. The rotating part of the machine, the commutator and the brushes, are at a distance from the floor, and less liable to pick up dirt than in the old type of machine. The poles may be of any num- ber, limited only by practical considerations. For direct cur- rent work relatively heavy cur- rents as a rule are generated, so that the necessity of using thick wire tends to limit the number of poles. The construction is symmetri- cal, and the field sections may be made on the interchangeable plan, even if some special planing is needed to bring them smooth- ly into place in setting up a machine. But the construction is so strong that there is never any need of replacing field sections. Fig 15 \— Relation of Armature Coils and Field. 252 ELECTRICIANS' HANDY BOOK. Advantages of flultipolar Construction.— The old type of two-pole dynamo wiili parallel magnet legs has been abandoned generally for the multipolar type. The objections to the bipolar type are as follows: To take the drag of the heavy armature off the bearings, the armature end of the magnet has to be placed downward. The magnetic pull tends to lift the armature from the bearings and make it run easily and prevent wear of the under journal-box. But the placing of the armature end downward makes it impos- sible to use an iron base for the machine. Such would short- circuit the lines of force, and would thereby weaken the field of force in which the armature rotates. The long magnet legs give much magnetic leakage, as shown in Fig. 81, page 184, thus further weakening the field. Never- theless, good results were reached with the old-time bipolar dynamos. Field Winding of Dynamos. — The general principle upon which the field magnets of dynamos and motors are wound is ex- ceedingly simple, and is what ^as been described under electro- magnets. Each pole piece has in the typical and almost univer- sal class of machines to be of opposite polarity to its neighbor. The windings, if directly on the pole pieces, follow the rule of electro-magnet winding, so that the current around the south poles follows the direction of motion of the hands of a watch if the pole is facing the observer, and the reverse holds for the north pole. The windings in a bipolar magnet if on the legs compare exactly with those of an ordinary electro-magnet. The wire crosses from the front of one leg to the rear of the other, so as to give one north and one south pole. A single winding on the yoke connecting the legs is sometimes used for both poles. On multipolar machines with windings on the poles, the same rule is followed as for bipolar windings, the wire crossing from front to rear of the pole pieces adjacent to each other. Series Winding.— The simplest or most natural conception of a dynamo is the series-wound dynamo. In it the terminals of the armature are connected as follows: One is connected to one end of the field winding. The other is connected to the end of the THE DIRECT-CURRENT GENERATOR. 253 outer circuit. The other end of the outer circuit is connected to the remaining field terminal. The cut, Fig. 151, shows a diagram of a bipolar series dynamo, and Fig. 152 shows the same in con- ventional diagram. This type of connection is of almost historical interest. It -s impossible not to recognize in it the foundation of the modern dynamo. The self-exciting dynamo, relatively small in size, had no difficulty in replacing the old magnetos. It must not be for- gotten that powerful magnetos were constructed in old times, and were used for lighthouse il- lumination. Fig. 151.— Series-Wound Dynamo. Fig. 153 —Conventional Kepresenta TiON OF A Series- Wound Dynamo. But when the self-excited dynamo appeared on the scene, with a field enormously intensified over that of the old magneto, a veritable revolution was made. The modern engineer often winds his fields in parallel with the outer circuit, or has them wound with two coils part in parallel and part in series. He may use a small independent machine to excite the field, which also is an old idea, so that the principal machine has only its armature coils traversed by the current it produces. Yet the self-exciting series-wound dynamo must be regarded as one of the parents in the already long line of ancestors. The winding is seen to be adapted to produce opposite polarity of adjacent poles. 254 ELECTRICIAXS' HAXDY BOOK. Action of Series Winding. — The action of series winding brings about several conditions. The armature can generate no elec- troraotive force until the field is excited by the current this electromotive force produces. Therefore to start it everything must be done to favor the production of current. The dynamo is best started on very low external resistance, and the armature may have to be speeded up. To facilitate starting, it is impor- tant to have good permeance, or to have a good magnetic circuit. The polarity of the machine is fixed by the polarity of the field magnet poles. As shown in the cut, the direction of the cur- rent is indicated by the arrows. But if for any reason the dyna- mo began self-excitation or building-up with the north and south poles reversed, the current would fiow in the other direction. This happens not infrequently with series-wound machines. For elec- tric light this may or may not be of importance; for charging storage batteries or electro-plating it is imperative that no change occur. If the resistance of the outer circuit is increased, the electro- motive force diminishes. If the same resistance is diminished, the electromotive force increases. These two effects are due to the effect of resistance on the total current which passes through the field magnet coils. If used for constant-current lighting, the addition of a lamp will cut down the electromotive force exactly when it is most needed. If used on parallel-circuit lighting, each new lamp light- ed will cut down the external resistance, strengthen the field, and increase the electromotive force. This involves danger of burn- ing out the lamps. Series winding therefore has its defects, and the tendency is to adopt other windings. Shunt Winding.— A shunt-wound dynamo is one whose field magnets are wound in parallel wUh the outer circuit. The ter- minals of the armature winding, which are the brushes, are con- nected each to two wires. Ov(^ is a terminal of the outer circuit, the other a terminal of the field-magnet winding. The cut, Fig. 153, shows a binolar dvnamo shunt-wound. Fi^. 154 shows a conventional renresentation of the same. It will bo seen that the potently 1 differpnce or voltage expended in the- THE DIRECT-CURRENT GENERATOR, 255 field magnet and outer circuit are identical. The energy expended on the field magnet is totally lost as far as any economic effect is concerned. It is of importance to keep its value as low as possible. The volts are fixed and beyond control. The only way of reducing the watts of energy expended in the field is to re- duce the amperes. Accordingly, the winding of a shunt dynamo is of fine wire and of many turns. This causes it to carry only a small proportion of the total cur- rent. The watts absorbed by it, as the volts are relatively con- stant, is directly proportional to Fig. 153.- Shunt Wound. Fig. 154.— Conventional Fepresent^ TiON of Shunt Winding. the amperes of current which pass through it. By this way of winding the field coils the proportion of energy expended on their excitation is kept as low as in the series-wound machines. Action of Shunt Winding. — The action of a shunt-wound dy- namo is the reverse of that of a series-w^ound one. If the resistance of the outer circuit is increased, the field mag- net receives more current, and the voltage at the armature ter- minals increases. The effect is that produced in the series ma- chine by short-circuiting. If a shunt-wound machine is supplying lamps operated in par- allel, the resistance of its outer circuit will be decreased as more and mors lamps are operated. This causes less current to be shunted into the field, and the voltage falls. ^5t) ELECTRICIANS' HANDY BOOK. The effect of taking current from the field reduces its mag- netization. This in its turn reduces the electromotive force generated by the armature. This reduction comes in as a third step, and again cuts down the field current. Nevertheless, some shunt dynamos with low-resistance armatures regulate them- selves fairly well within a reasonable limit of action. If the resistance of the outer circuit is raised, the intensity of magnetization is increased, as more current is shunted around the field. A shunt-wound dynamo may supply a constant-current system of lamps very well. This is the system where the lamps are in series. If new lamps are added to the series, the resistance of the outer circuit is increased, more current is shunted through the field coils, and the electromotive force and voltage of the outer circuit increase. This is in the direction of meeting tho greater demand for potential. The series machine, because of its connection, must have the full current pass through its windings. This current cannot be changed. The current passing through the field windings in the shunt machine can be varied. This may be done by placing a variable resistance in circuit with the field windings. By in- creasing this the field is weakened and vice versa. Compound Winding.— A dynamo consisting of a combination of the series and shunt machines is called a compound-wound dynamo. The field magnet is encircled by two windings. One is a pro- longation of the outer circuit, exactly as in the series dynamo. The other is a finer wire circuit, in parallel with the outer cir- cuit, exactly as in the shunt-wound dynamo. Of the compound-wound machines, there are two variations shown in the diagrammatic cuts, Pigs. 155, 156, 157 and 158. Short=Shunt Compound Winding.— The first variation, Figs. 155 and 157, is the short-shunt machine. The shunt field circuit is connected directly to the brush terminals. The outer circuit, with the series field circuit in series with it, is connected to the same terminals. The shunt field coil is in parallel with the line, con- taining outer circuit and series magnetizing or field coils. Long'Shunt Compound Winding.— The second variation, Figs. THE DIRECT-CURRENT GENERATOR. 257 156 and 158, is the long-shunt machine. In this only one ter- minal of the shunt coil is connected directly to a brush terminal. The other end of the shunt coil connects to the outer circuit be- yond the outer end of the series field coil. In this connection the shunt coil is in series with the armature and outer circuit and in parallel with the series coil. Action of 5hort=Shunt and Long=Shunt Windings — There is not much difference in the action of these two kinds of windings. Fig. 155.— Compound- Wound Dynamo, Short-Shunt. Fig. 156.— Compound-Wound Dynamo, Long-Shunt. In the short-shunt winding an identical current goes through the shunt as long as the same voltage is maintained at the armature terminals or brushes, because the shunt coil takes its current from those terminals. In the long-shunt winding there is a slight variation in the voltage of the shunt coil, with constant voltage at the brushes, if there are variations in the current in the outer circuit. Self => Regulation of Compound-Wound Dynamos,— If a com- pound-wound dynamo is supplying a circuit at constant poten- tial, it may be almost self-regulating. Suppose the resistance of the outer circuit to be diminished. This sends more cur- 258 ELECTRICIANS' HANDY BOOK, Fig. 157.— Coxventionali Representa- tion OF Short-Shunt Dynamo. rent through the series coil, and thereby acts to increase the intensity of the field. But the reduction of resistance in the outer circuit reduces the current in the shunt winding. This action goes to reduce the intensity of the field. By giving proper propor- tions to the two exciting coils, the intensity of the field can be kept prac- tically constant as the resistance of the outer circuit is increased or diminished. The arma- ture being kept at a con- stant speed of rotation in a constant field of force by the engine or other source of mechanical power, impresses on the circuit the identical elec- tromotive force. As its resistance and that of the series field coil is constant, the voltage at the terminals remains constant. This applies to an ac- curately arranged wind- ing. Whether the result is reached by calculation or by trial, it can be attained very closely. At high or low cur- rent strength there is apt to be a comparatively slight change in voltage. Characteristic Curves.— On page 283 are given characteristic curves of series-wound and shunt dynamos. If it is realized that the characteristic of a compound-wound machine may be almost a horizontal line, its self-regulating powers will be seen. This appears from Figs. 176, 177 and 178. YiG. 15S. -Conventional Representation OF Long-Shunt Dynamo. THE DIRECT-CURRENT GENERATOR. 25^ Over=Compounding. — The result of such even action as de- scribed above is the maintenance of constant voltage at the termin- als of the machine. In electrical work all sorts of conditions may have to be met. A very usual one is that on a circuit a constant voltage is required, not at the generating plant, but in the heart of the district, perhaps miles away. In an over-compounded dynamo the series coil is given so many turns in proportion to the turns in the high resistance shunt coil that its influence overbalances that of the shunt coil. The effect of over-compounding is to cause the voltage at the terminals of the machine to rise with increase of current. The proportional increase of voltage with increase of current can be accurately regulated by the relative sizes of the coils. It is only necessary to follow what has been said of the series dynamo, and to regard the compound-wound machine as a series dynamo greatly reduced in its characteristic action. Over-compounding enables a constant voltage to be maintained in any point of a district. The resistance of the mains between the dynamo in the station and the given point in the district is known. The drop in voltage due to that resistance varies with the current. The over-compounding of the machine can be regu- lated to give the same increase in voltage with the increase of current, and thus the voltage at any desired point in the district can be kept constant, following Ohm's law. Example of Compound Winding Calculation.— Suppose the re- sistance of a single lead of the mains to be 0.01 ohm. Then that of the two leads is 0.02 ohm. Suppose a maximum current of 500 amperes is needed. The drop due to the specified resistance and current is obtained by Ohm's law: RI = E or 0.2 X 500 =: 10.0 volts. This of course is an extreme case. But the dynamo by over- compounding c?.n be made to vary its voltage at the terminals in this or any other desired proportion to the current. With the resistance given above, and the variation in voltage for the cur- rent as calculated above, which variation is at the terminals, a constant voltage would persist at the outer end of the leads. 260 ELECTRICIAXS' HAXDY BOOK. Excitation of Field Coils in Compound Dynamos, — The series field coils of dynamos can only be excited ty the working cur- rent or by a portion of it. If tlie machine is compound-wound, the series coils are taken care of by the machine. The shunt coils may receive their current from various sources. To make the machine self-regulating, it would seem that the shunt coil should be fed from the machine proper. This practice makes the dynamo self-contained. Two other systems of shunt-coil excitation are used. In one system the terminals of the shunt coil are connected to the leads or bus-bars of the main circuit; in the other, a separate source of current is used for the shunt coils. When several dynamos are operated, and constant potential is maintained in the circuit at all times, a new element in the magnetization of the field is in- troduced because the magnetization, as far as the shunt coil is concerned, in this arrangement is independent of the speed of the dynamo. The excitation becomes zero when a self-exciting dynamo stops. Effect of Independert Excitation of Shunt Coil. — If the ter- minals of tlie shunt coils are connected to an always active outer circuit, to station bus-bars for instance, the shunt coil excites the field as long as the connection is kept closed. As the dynamo runs slower the field excitation diminishes, but with less rapidity than before, and is never reduced to zero until the bus-bar or main circuit connection is broken. It is a case of under-com- pounding. The great advantage of it is that it makes it possible to excite the field before starting a dynamo. The field before the armature begins to rotate is not only excited, but the correct polarity is established. The instant the dynamo begins to work, electromotive force is impressed upon the armature coils, and there is no difficulty in bringing the voltage up to that of the main circuit. Disconnecting or Opening the Shunt Coll. — The capacity of the shunt coil is considerable. It cannot with safety be dis- connected by a simple opening of a switch. A bank of lamps is generally mounted in series with it. The field break switch is placed between the lamps and the main circuit. When it is opened, the resistance of the lamps prevents undue sparking. THE DIRECT-CURRENT GENERATOR. 261 ^>^ * ^' Separate Excitation of Shunt Coil. — The shunt coil may also be excited by an independent-; source of electric energy. This may be a storage battery or an exciting dynamo. The separate excitation brings about a particular result. The exciting ma- chine will be run at a constant voltage, so that the current passed through a separately excited shunt coil can be absolutely constant. The inevitable variations in voltage on the outer circuit bring about some variation in current in shunt coils fed from the bus-bars, which variation may be slight, but it exists. Otherwise, the result of separate excitation is not to be distinguished from outer circuit or bus-bar exci- tation. It gives another dynamo or storage bat- tery to be looked after. Exciting Series Coils from Main Circuit.— A very obvious way of ex- citing the field coils of a compound dynamo is to send current through its series coils from the main circuit. This is done by closing two switches, one connecting a terminal of the field series-coil with one lead of the main circuit, and the other connecting the other end of the same coil with an equalizing bar or by special connection with the other main-circuit lead. This leaves the armature for the moment out of circuit. The dynamo can then be started and brought up to the proper potential. The armature has electromotive force impressed on it at once, and excites the shunt coil. Thus it is brought with cer- tainty into action, and the polarity is fixed from the start. Separately = Excited Generators, — The separately-excited dyna- mo closely approaches the magneto in its action, as the strength Fig. 159. V. y -Separately-Excited Dynamo. 262 ELECTRICIANIS' HA^'DY BOOK, of the field does not directly depend upon the current generated. The diagram, Fig. 159a, shows the connections for the separ- ately-excited machine. The field-magnet coils are entirely separ- ated from the commutator connections. A current passing through the coils produces a field of definite and irreversible polariiy. The armature rotates in the field, and impresses electromotive force of definite polarity on the circuit. The current which excites the field magnet may be derived from a small dynamo, or from any source desired. Action of the Separately-Excited Dynamo. — This arrangement has several advantages. The absolute irreversibility of polarity may be a very valuable feature. Thus, when storage batteries are being charged with a self- excited machine, tlie polarity sometimes becomes reversed. In such a case, if there is any charge in the battery, it discharges through the dynamo, and the latter be- comes a motor, and the charge is wasted and lost. A similar trouble occurs in electroplating. But with separately-excit- ed machines this class of trouble is impossible. If its voltage is Insufficient to fully charge a battery, it will at any rate not act as a motor and discharge what may be in the battery. The electro- plater is certain that with a separately-excited dynamo his articles in the plating bath will receive the desired deposit and will not strip and lose what they have received. Regulation of Separately -Excited Dynamos and Magnetos.— There are three general factors of regulation of magnetos awi separately-excited generators. The speed of rotation of the armature may be altered. This changes the lines of force cut in a given period. 10^ lines of force cut per second, it will be remembered, gives one volt. The brushes may be pushed forward on the commutator. This introduces demagnetizing turns in proportion to the advance of FiQ. 1591.— Rheostat f^r "Regulating Separately-Excited Dynamo. THE DIRECT-CURRENT GENERATOR. 263 the brushes. This is described in Chapter XIV. Thus the mag- netic circuit, although produced by separate excitation, can be reduced by self-regulation. Another way is to change the normal magnetic flux through the armature by outside means. An old device with magnetos was to provide a movable piece of iron, which could be moved toward or away from the poles of the magnet. This as it approached the poles shunted off more and more of the lines of force from Fig. 160." -Separate-Circuit Dynamo. Fig. 161.— Separately and Self- Exciting Series Dynamo. the armature, weakening the field and reducing the electromo- tive force. In separately-excited machines the current passing through the field-magnet coils can be weakened by the introduc- tion of resistance into the exciting circuit, or by any other means. A rheostat-like arrangement can be introdmced to cut out some of the coils of the field, as shown in Fig. 159a, in whieh R denotes the rheostat. The Separate- Circuit Dynamo has either two separate arma- tures in the field space, or has two sets of coils. Whichever it is, one armature or coil set is used to excite the field, the other to supply the current to the circuit. Fig. 160 shows a 264 ELECTRICIANS' HAXDY BOOK. diagram of such a dynamo with two commutators, from one of which the field current, and from the other of which the field magnet current is taken. Separately and Self=Excited Dynamo.— The diagram, Fig. 161, shows this machine, in which a current from an outside source passes through one field coil, and the main current of the dynamo passes through a second field coil. Multipolar Dynamo Connections. — To avoid complication and to give diagrams readily understood, only two-pole machines have been illustrated in this chapter. But everything which has been shown for such machines applies to multipolar machines. The conventional diagrams may be used for multipolar machines ex- actly as employed in this chapter. The few turns of wire indi- cated may refer to the winding of any number of poles. Conventional Representations of Machines. — The tendency of engineers is to simplify their diagrams as much as possible and to indicate a machine with numerous poles by a few lines only, as if it were of the simplest construction. Except for small machines the bipolar construction may be considered to be de- finitely abandoned, as is explained elsewhere. The brushes are conventionally drawn as if they were set tangentially. This is done for a purpose, as it serves to indicate the direction of rota- tion of a machine. Often this is not essential as far as the draw- ing is concerned, and the brushes may be shown as radial brushes or lines, as in Fig. 159a. Later the representation of an alternating-current machine will be spoken of, and it will be seen that the distinction between the direct-current and alternating-current machine depends upon the representation of the brushes. In these conventional figures those remain in use which are the simplest and most effective as regards freedom from misunderstanding. CHAPTER XIV. ARMATURE REACTIONS. Armature Polarity Due to Its Windings. — The armature of a direct-current dynamo, by the polarity it acquires from its wind- ings, modifies the course of the lines of force. If the dynamo is idle or on open circuit, no current passes through the armature, and any lines of force which may exist go straight across from field magnet poles to the armature core. But the current which goes through the windings of a dynamo- or motor-armature operates to produce north and south poles in it. These are situated at points about equally distant from the poles of the field. In a bipolar machine the line connecting the north and south poles of the field is approximately or exactly at right angles to the line connecting the north and south poles of the armature core. Action of Field Poles on Armature Core.— The field poles tend to induce polarity in the parts of the core nearest to them, the north pole inducing south polarity and the south pole north polarity. The effect of the combination of polarities, one due to the field poles' induction and the other to the induction of the armature windings, is to give resultant poles to the armature at intermediate points. The lines of force emerge from one pole of the field, go obliquely to the opposite resultant pole of the armature, obliquely through the armature to its opposite corresponding pole, and thence obliquely to the adjacent pole of the field. Field Distortion. — This armature reaction introduces mani- festly an element of complexity into the subject. It is no longer a simple set of straight lines of force which constitute the field, but an S-shaped volume of polarized ether, constituting a dis- torted field of force. 266 ELECTRICIANS' HANDY BOOK, Armature Reaction Diagrams.— The reaction of the magnet- ized armature core is easily understood from an inspection of the diagrams. The diagram,, Fig. 162, shows an idle armature lying between two pole pieces of an active field magnet. The wires are indicated by circles. Those with crosses show the current going away from the observer, those with dots the current approaching; those with neither show idle wire. The iron core of the arma- "^X5^ 0000 Fig. 162.— Unexcited Armature in Excited Field. CXTXl nN ^^-' nmn uuuu -' s ^^ Fio. 163.— Excited Armature in Unexcited Field. ture has induced in it two poles opposite those of the field mag net, and the general course of the lines of force is indicated bj dotted lines. The diagram, Fig. 163, shows an excited armature between the poles of an idle field. There the poles in the armature lie at ARMATURE REACTIONS. 267 right angles to the field pole pieces, and the same conventional signs are used for the currents in the wires. The arniature poles in this figure are at N and S. In Fig. 164 both field and armature windings are supposed to bo passing current. It will be seen that there are four poles, two N poles and two S poles, each pair at right angles to the other. The S pole of the field tends to establish an N pole in the arma- ture core opposite to itself. The N pole of the field tends to establish an opposite and corresponding S pole in the armature core. The windings of the armature tend to produce their own poles on the vertical line as shown. The resultant poles in the armature lie between the two pairs. The resultant N pole lies s Fig. 164 —Excited Armature in Excited Field. in the right-hand upper quadrant; the resultant S pole in the left-hand lower quadrant. Varying Densities of Field.— Not only are the poles of the armature core thus displaced out of symmetry. The lines of force are densest in distribution between opposite-named poles of the field and armature core. They are crowded together toward the horns or ends of the pole pieces that lie in the direc- tion of the motion of the armature. They are thinned out at the other horns. All this is shown in the cut. This crowding to- gether of the lines of force is due to a reaction between the core poles and the field poles. This reaction in the case shown in the figure tends to displace the S field pole upward and the N field pole downward again in the direction of rotation. Were there no displacement of poles, the poles of the armature 268 ELECTRICIANS' HANDY BOOK. should lie upon a line at right angles with the line connecting N and S poles of the field. In the three figures these poles would lie on the vertical line. But owing to the armature reaction, the brushes have to be shifted in a dynamo in the direction of the rotation. Their line of position is now oblique to the line con- necting the centers of the field magnet poles. Neutral Points. — The points connected to the brushes are termed the neutral points. These normally lie at the ends of the oblique diameter described in the last paragraph. It is perfectly evident that the armature reaction m.ay vary under different conditions of load. Especially is this to be looked for in shunt or compound wound dynamos. The neutral points vary according to the relative intensities of magnetization of field magnets and armature cores. Brush Adjustment.— To meet this variation of positions of the neutral points, the brushes are mounted on a rocker so as to be movable back and forth. They are rigidly connected with each other, so as to always be at opposite extremities of a diame- ter, but by turning their mounting or "rocker" back and forth, their position can be made to coincide with that of the neutral point. Demagnetizing Turns, — The brushes are advanced from the ends of the symmetrical line of the armature through a distance which may be stated in terms of an angle of so many degrees. If we go back from each brush against the direction of rotation a distance equal to twice this angle, we get a space called the demagnetizing belt, and the turns of wire comprised in this belt are called demagnetizing turns. In the cuts, Figs. 165 and 166, the condition is shown, n n' is the line connecting the neutral points; a h and c d cover the demagnetizing turns. The same conventional signs show the direction of current. It is obvious that the demagnetizing belt is working against the field-magnet turns, and reducing the intensity of the field of force. The brushes should be kept as near the symmetrical points as possi- ble. The arrows in Fig. 165 show the general direction of the armature currents. Reduction of Field Density*— Referring to the same figures, the turns outside the demagnetizing belt tend to diminish slightly ARMATURE REACTIONS. 269 the intensity of field. Tliis is by their action in crowding together the lines of force at the advanced horns of the field magnet poles. This reduces the permeability of the iron at that point, and hence reduces the field density or intensity. Demagnetizing turns are entirely distinct from the armature reactions described on the preceding pages. The turns in the de- magnetizing belt are in direction of current the reverse of those in the field magnet. Action of the Demagnetizing Turns.— The action of the de- FiGS. 165 AND 166.— Neutral. Line and Demagnetizing Turns or Armatuhe. magnetizing turns is to weaken the field. The armature core is a part of the magnetic circuit, and w^hatever affects the lines of force which go through it affects the whole circuit. The demag- netizing coils have no action except when a current is going through them, and their action varies with the intensity of the current. It is simply a matter of ampere turns working in opposition to the ampere turns of the field. The electromotive force of dynamos naturally rises as the speed increases. But most series-wound machines reach a maximum, and then tend to fall off in electromotive force. It is due in great part to the advance of the brushes under increased load. This throws more turns of wire into the demagnetizing turns, and thereby increases the counter or back ampere turns. The distortion of the lines of force has also something to do with this. 270 ELECTRICIANS' HANDY BOOK. Dead Tarns. — It follows from the above and from some other reactions which may be included, that the increase in electro- motive force is not strictly proportional to the speed. Thus, if the electromotive force were to be increased ten per cent, it might be necessary to increase the revolutions twelve per cent. The extra revolutions of the armature required, above the propor- tion of the voltage gained, are called "dead turns." Spurious Resistance. — Self-induction in a conductor does two things. It resists the starting of a new current through a con- ductor, and tends to pro- long the passage of an ex- isting current if anything occurs to diminish it. The latter action is what pro- duces the spark on open- ing the circuit of a spark coil. Consulting Fig. 167, it will be seen that one coil of an armature is shown short-circuited by one of the brushes. As the armature rotates, coil after coil is thus short-circuited. T and U have just been short-circuited, and W and X will next be. The self-induction acts to send a current through the coil for perhaps only a portion of the exceedingly brief period when it is short-circuited. This is a loss of energy. As it passes on, a new current has to be started, and is resisted by its self- induction. This is another loss of energy. These actions increase in degree with the speed, and reduce the current. They act like resistance, and the term spurious resistance is applied to them. Spurious resistance increases with the speed of rotation. Anything which will reduce the inductance of the armature will reduce spurious resistance. The fewer turns of wire in the armature, the less will the inductance and the spurious resistance be. But in designing dynamos and motors, spurious resistance is rather a minor consideration. Fig. 167.— Short-Circuiting of an Armature Coil, by a Brush. ARMATURE REACTIONS. 271 Eddy or Foucault Currents.. — If electromotive force is im- pressed upon the windings of an armature, it follows that it will also be impressed upon all metal parts of it. The core is not only not exempt, but its periphery is subject to nearly as strong induction as the coils themselves. Accordingly, currents whose direction is determined by the regular laws of induction are pro- duced in them. These currents existing within the mass of the metal are termed e^idy currents or Foucault currents. The por- tions of the metal nearest the surface are most impressed with electromotive force; the outer portions cut more lines of force per revolution than do the inner portions of the core. The elec- tromotive force impressed on one side of the core is of opposite polarity to that impressed on the other. Differential action of electromotive force on a conductor will set up currents according to Ohm's law. These make themselves known by the heat which they produce in the metal in which they are generated. A copper or iron wheel rotated rapidly in a strong electric field becomes hot from the generation in it of eddy currents. Eddy Currents in Armature Cores. — Energy is expended on the production of these currents, which is totally lost as far as any useful effect is concerned. There is no available way of sup- pressing them in armature cores except by a somewhat crude method. The core is built up of a quantity of pieces of thin iron, insulated from each other, and set at right angles to the direction in which the impressed electromotive force tends to produce a current. It is called a laminated core. A cylindrical armature core is accordingly made of a pile of circular disks of thin iron. Between them are placed layers of some insulating matec-ial, such as paper, and thus the possible path of the currents is so much shortened that they amount in the aggregate to comparatively little. The object is thoroughly attained by making the disks as thin as possible and insulating them well. Eddy Currents In Core Disks. — Eddy currents are established in core disks, though of relatively little importance. The cut. Fig. 168, shows how such currents act in laminations. The thick- ness of the lamination is greatly exaggerated in the cut. 272 ELECTRICIANS' HANDY BOOK. Eddy Currents in Pole Pieces.— Any alteration in the distribu- tion of the lines of force in the field will cause eddy currents in the contiguous masses of metal. Thus, eddy currents may be produced in the pole pieces of the field magnet if the iron core of the armature is not perfectly cylindrical. Some types of arma- ture have projecting teeth of iron, the disks being out of contour to give projections. These teeth, as they sweep past the smooth circle of the pole pieces, virtually carry an intense little field of force of their own with them, and thus start eddy currents in the pole pieces. Every eddy current rep- resents joules of energy, and has to be accounted for in the power. End Leakage of Lines of Force in Armature.— Lines of force may leak around into the flat ends of the armature core. These will be enough to start currents flowing through these disks. If the armature core extends out beyond the pole pieces, this source of eddy currents is disposed of. Eddy Currents in Conductors. — If the conductors or windings of an armature are very thick, eddy currents may be produced in them. This makes them carry several currents, one running counter in part of its course to the regular current. All eddy currents, representing loss of power or waste of en- ergy, must be suppressed as far as possible. The worst ones which can be produced are core currents, and these are minim- ized by laminating the core. Fig. 168.— Eddy Turrents in Armature Laminations. CHAPTER XV. CHARACTERISTIC CURVES. CO o > Characteristic Curves. — The action of dynamos under various conditions is represented by what are known as characteristic curves. These are diagrams constructed on the usual basis of lines at right angles to 100 r 1 1 \ \ \ \ 1 each other. The vertical line may be divided into parts representing volts — generally the difference of potential between the ter- minals of the dynamo. The horizontal line may be di- vided into parts represent- ing amperes. The vertical and horizontal scales may also be given other mean- ings. Hopkinson, who in 1879 first proposed this way of representing the action of a dynamo, used the values of the total electromotive force for the vertical line. Fig. 169 is an example of a characteristic curve, E, of a series-wound dyna- mo. It shows the electro- motive force for small currents increasing much more rapidly than it does when the current increases. It is evident that with enough current taken' out of the machine, the electromotive force would remain practically constant. .^E ^ ^ ^ / / / / 7 20 RE VS. 1 { IE ME ^IS / / / l/ Pig. io 20 30 40 50 60 70 AMPERES 16».— Charactkristic Curve of a Series-Woujsd Dynamo. -274 ELECTRICIANS' HANDY BOOK. The variations in current are produced with constant speed of rotation by changing the external resistance. The internal re- sistance remaining constant, and the field excitation varying with the current intensity, are two of the controlling factors which produce the curve. Horse=Power Lines. — Seven hundred and forty-six watts or volt-amperes are an electrical horse-power. Points on a charac- llso o > 40 1 1 \ \ \ \ \ \ S ; \ y V135- \ \ \ V^ \ \ \ W \ \ 1 \ \ \ \ \ '% ^N 1 \ ^! \ <% -% ^^- \ -?. ''? N "^. 1 \ h. ^^ -.^ ^' ^t "~^-, ■* ^^-. ^ ->» 1 10 20 30 40 50 60 AMPERES Fig. 170.— Hobsb-Power Lines. teristic curve diagram can be determined where the product of the volts and amperes is equal to 746. These points joined give a one-horse-power curve. Other points are where the product of the volts and amperes is equal to 746 X 2 or 1492 watts, which are «qual to two horse-powers. These points connected give a two- horse-power curve. The process is carried out for other horse- powers, as shown in Fig. 170. The characteristic curve in connection with this set of horse- CHARACTERISTIC CURVES. 275 100 90 80 70 60 o > J 50 40 power curves tells two things. It indicates the relation of amperes to volts in a specific machine corresponding to the various horse-powers. Types of Characteristic Curves.— There are different types of characteristic curves. The one just spoken of is referred to the electromotive force of the dynamo, and is called a total charac- teristic curve. Another type of curve is referred to the potential difference ex- isting between the ter- minals of the dynamo. This potential is easily de- termined by a voltmeter. Such a curve is called an external characteristic curve, or sometimes the terminal- potential curve. Drooping Characteris- tic—In the cut, Fig. 171, the internal characteristic curve, e, corresponding to the external one, E, is shown in dotted lines. Be- yond a certain point, about 27 amperes, the voltage be- gins to decrease in value. A curve of such a shape is called a drooping curve. Sometimes characteristics droop much more than this. An advantage attaches to this drooping. It indicates that, should the machine be short-circuited while running, the electro- motive force will not increase. A machine with drooping charac- teristic is advantageous for constant-current arc lighting. The straight line J is the characteristic of the armature. The curve e is derived from E by subtracting the ordinates of J from those of E, and drawing the curve e through the points thus determined. 10 __ ^E ^ / / \ Z"^' .^^^ // 720 F ,EV. ^••--. "•^6? / f ( // // X ^,'' J // ^y'' -^' // ^ ' t"-- ^^' 10 20 50 60 70 30 40 AMPERES Fig. 171.— ToTAii and External Char- acteristic Curves of a Series Dynamo. 276 ELECTRICIANS' HANDY BOOK, Interpretation of Characteristic Curves.— The resistance of a working dynamo and its circuit is made up of various compo- nents. The total characteristic gives the equivalent in ohmic resistance of true and spurious resistance under different condi- tions. By Ohm's law resistance is equal to electromotive force -pi divided by current, or R = _ 1 In the characteristic curve diagram, E is given by the vertical distance (ordinate), I by the horizontal distance (abscissa). Dividing one by the other, we get resistance in ohms. Or draw- ing a line from the lower left-hand corner (origin) to any point on the curve, the tangent of the angle that line makes with the horizontal line will be proportional to the resistance at the point in question. This gives another basis for interpretation. If resistance is increased, the line making with the horizontal an angle whose tangent is equal to the resistance will swing back to the left, so as to increase the angle. This increasing of the angle wi]l increase the tangent, which is as it should be, because the tangent represents resistance, which by the conditions assumed also in- creases. Such a line is called sometimes the vector line of watts. As it approaches the vertical the tangent increases, and if the value of the current intensity, or of I, is kept constant, the value of the electromotive force or of E will increase. When the tangent of the angle becomes infinitely great, then an infinite electromotive force will be required to maintain a finite current. Data for External Characteristic Curves.— When the volt- meter and ammeter have been used to determire the relations cf current and electromotive force in a dynamo under different out- puts, the data are obtained for characteristic curves. Here an important distinction is to be drawn. The voltmeter gives the potential difference existing on the outer circuit onlv. The amperes given by the ammeter are those which pass through the whole circuit. Data for Total Characteristic Curves — This is a distinction which oft?n has to be made. Amperes are identical over all parts of a circuit. Volts vary in proportion to the relative resistances of the parts of the circuit affected. If, having obtained the data CHARACTERISTIC CURVES. 277 for an external characteristic, the electromotive force of the whole circuit can be substituted for the potential difference given by the voltmeter, the new data will give the total characteristic. The resistance of the dynamo being known, Ohm's law gives the electromotive force of the system. Suppose a current of 50 amperes is taken on the external characteristic, the dotted line curve e, Fig. 171; this gives 60 volts. By Ohm's law the external resistance is equal to _or^ — = 1-2 ohms. The internal resist- i 50 ance of the particular dynamo tested was 0.6 ohm. The total resistance of the circuit was therefore 1.2 + 0.6 =: 1.8 ohms. The electromotive force of the whole circuit at 50 amperes of current is now deduced by Ohm's law: E = R I = 1.8 X 50 = 90 volts. Drawing Characteristic Curves. — In the ways above described a number of points on a characteristic curve are found, and from these the curve is constructed. For each value of current strength a value of voltage as given by the voltmeter for an external characteristic is given, and a value of electromotive force calcu- lated as above for the total characteristic is also given. These points are marked upon a sheet of paper. The current measurements are taken on horizontal lines, the voltage and electromotive force measurements on vertical lines. The curve is then drawn through these points. A thin flexible strip of wood called a spline is a simple appliance for such purposes. A more efficient instrument is the flexible ruler. Its construction is based on the use of a bar of lead. This is bent to any desired curve, and holds its shape. The splines spring back when re- leased. The easier characteristic to get is the external. It has to precede the total characteristic. Its data are absolute and useful. The data of the total calculated as described leave armature reactions out of account. When a characteristic is given and no statement is made that it is a total characteristic, it may be taken as an external one. It is bad practice not to state whether a charf?cteristic is external or total. Internal Characteristic— It is obvious that there is an inter- nal characteristic. This is based on unvarying resistance. It is therefore a straight line. 278 ELECTRICIANS' HANDY BOOK, \sQ may prove this by returning to the radius vector of watts. If resistance is constant, the tangents of all radius vectors of watts must be constant. This is equivalent to saying that all such radius vectors must coincide in direction. They will all be represented therefore by parts of one straight line. They will vary among themselves Y loor 90 80 70 60 13 50 o > only in length measured from the lower left-hand corner (origin). Terminology of Ana- lytical Geometry.— The word given in parenthe- ses is one of the terms used in analytical geom- etry. The vertical line on the left is the axis of Y or of ordinates; the lower horizontal line is the axis of X or of abscis- sas; horizontal lines are abscissas; vertical liues are ordinates; the inter- section of the axes of X and Y is the origin. Line of Ohms.— The diagram hitherto has not directly shown ohms. It is divided into squares. A diagonal to the lower left-hand square drawn from the origin will be an angle of 45° with the horizontal, and its tangent will be 1. This is taken as 1 ohm. The vertical line or ordinate through the right-hand end of this square is taken as the line of ohms. A radius vector of watts drawn to any part of the curve will have its tangent given by the part of this vertical line cut off. The value of this tangent will give the ohms resistance. In the diagram, Fig. 172, the resist- ance at the point B is 2 ohms, at A 4 ohms, and so on. The line of ohms is erected on the point 10 of the base line. 30- 20 10 L- E ^ ^ / c_.... i/ / I" / CO I o_ b. O z -i / "1\ / ! / / 1 / i / / / \ / /I 1 // 1 r d: IF 10 .20 30 40 AMPERES 50 60 70 Fig. 172.— Line of Ohms in Character- istic Curve Diagram. CHARACTERISTIC CURVES. 279 The line from O to B is the radius vector of watts. It intersects. the line of ohms at a distance 2 from the base, taking the side ol: a square as unity. This shows that at the point B the resistance is 2 ohms. As B corresponds to 90 volts and 45 amperes, by F. 90 volts ^ , , . ^ Ohm's law R = — • = or 2 ohms, which corresponds J- 45 amperes with the diagram. On the same diagram the resistance line, or line of ohms, can be used for either internal or external characteristic. The inter- section of the internal radius vector of watts with it will give the constant internal resistance. General Notes on Characteristic Curves. — The changes of resistance are effected by the manipulation of the outer circuit by the observer. Resi-stance is thrown in and out as desired, in order to get the different points of the characteristic. To give a characteristic any raeanin,g, one of the factors must be kept constant. This is always th'e revolutions per minute. But a charactei'istic could be based on fixed resistance, and the changes in current and voltage could be brought about by varying the speed. Then the tangents of the radius vectors of the watts curve being equal as denoting the resistance, the curve would merge into a straight line. Fixed current might be the basis. Then the characteristic would become a simple vertical line, its position would give the amperage. Its length would give the voltage as long as the resistance was unchanged. If the resistance was increased, the radius vector would cut at higher up, and the voltage would be given by the place of intersection. Fixed voltage as a basis would give a horizontal line, whose length would give the amperage. The intersection of the radius vector with this line determines the amperage corresponding to any desired resistance. This makes it clear why characteristics are given with fixed speed of rotation. The straight lines do not give the peculiarities of a dynamo as fully as do the curves. By changing the speed, we virtually change one dynamo into another. Critical Current. — A characteristic curve of a series dynamo starts as a nearly straight line. At its beginning its radius 280 ELECTRICIANS' HANDY BOOK. vectors are virtually one except in length. At first doubling tlie voltage doubles the current approximately. But after a while the curve bends to the right. On examining Fig. 170, it will be seen that for a given increase in voltage, the amperage will increase more rapidly than before. The point where this change is noticeable is a sort of critical point. The current corresponding to it is called the critical current. It is obvious that there is nothing accurate about it. The critical current is the minimum current required to excite the field. With insufficient ampere turns, a field magnet will not produce an adequate intensity of field. Therefore with a series dynamo too high an external resistance, cutting down the current, will weaken the field. This weakening may be enough to arrest the dynamo in its functions, and cause it to give hardly any electromotive force. It may easily prevent it starting into action from inaction. A series dynamo must be started on low external resistance, and the resistance must n^ver be so high as to cut the current down below the critical value. The electromotive force given by a dynamo increases with the speed. The resistance may accordingly be increased as the speed increases without affecting the current. Therefore there is no critical speed or critical resistance for a series dynamo, except In the most general sense on short circuit. Shunt=Wound Dynamo ^Characteristics — There are three pos^ sible characteristics of this type of machine. Pne is the total characteristic, which includes the armature current and the electromotive force. The armature current is equal to the sum of the currents passing through the field winding and shunt winding. The second is the external characteristic. This is based on the voltage between terminals and the total current of the outer circuit. This current is a part only of the armature current. The third is the so-called internal characteristic. This is based on the same voltage as for the second case and on the amperes in the shunt or field magnet windings. Possibly some ingenious person might evolve a fourth and a fifth characteristic, taking the armature into cases two and three. In practice the external and internal characteristics are most used. These are the second and third of the preceding list. CHARACTERISTIC CURVE8. 281 Critical Point of Shunt= Wound Dynamo. — The external char- acteristic of a shunt-wound dynamo is given in Fig. 173. It begins at the top at P. On open circuit all the current is shunted into the field, and the voltage between terminals reaches its max- imum. The resistance of the outer circuit when open is infinite. The outer circuit is now closed through a very high resistance. This shunts a certain amount of current from the field coil, and weakens the field so as to reduce the voltage. The resist- ance is gradually lowered, shunting more and more cur- rent from the field as more passes through the external cir- cuit, until a sort of critical point is reached. This point is where a reduction of external resistance begins to rob the field of so much current that the electromotive force falls more rapidly than before. At this point the watts are at a maximum. At last the curve, at the 35-volt-32-ampere point in Fig. 173, reaches a point of instability, and with very little change of resistance runs down to a zero value. The horse-power curves are interesting in their relations to the characteristic. The ordi- nate or vertical next to the left-hand axis of ordinates can be used as an ohm line. A straight edge will then give the resistance for each point on the curve. At P it is infinite, because the tan- gent of an angle of 90° is infinite. The volts at P were obtained on. open circuit, which is infinite in resistance. In the case shown in Fig. 173 the critical current may be called 32 amperes. It is not critical to the same extent as in a Y P 60 \ \ \ \ \ ^^^ V ^^ \ \ \^ 50 \ \ \ \'-0 \^ \ \\ 40 o > 30 \ \ \ \ \ \ )~ \y y^ ^ 20 10 / / AMPFRF*^ Fig. 173.— External, Characteristic OP A Shu ^T- Wound Dynamo WITH Horse-Power Lines. 282 jjJLECTRICIANS' HANDY BOOK. series dynamo. But the long, almost straight descent of the curve toward tJie origin gives a critical factor. This is external resistance. With insufficient external resistance the electromo- tive force falls to zero. With infinite resistance it rises to a maximum. At an intermediate resistance, which in this particu- lar case is about one ohm, the power is ready to rise rapidly or fall rapidly for a slight change in resistance. If the outer resist- ance is increased, the power rises on account of an increase in Fig. 174— Outer Cihcuit and Total Current Characteristic in A Shunt Dynamo. Fig. 175.— Total Characteristic OF fcnuNT Dynamo. voltage. If the resistance decreases, the power falls by an almost equal decrease of voltage and current. Roughly speaking, the characteristic shows one horse-power with a resistance of 1 ohm and a resistance of 6 ohms. With a resistance of about 1% ohms it shows the maximum power, nearly 2 horse-power. These variations in resistance are in the external circuit. Total Current Characteristic in Shunt Dynamo.— The total current is that which flows through the armature, and which is the sum of currents in the field coil and in the external circuit. The characteristic of the external circuit as just deduced is the CHARACTERISTIC CURVES. 283 basis. The new one is drawn by adding to the abscissas or to the horizontal lines of the diagram, lengths representing the current which at each given voltage will flow through the field. In the cut, Fig. 174, the inner of the two curves is the charac- teristic of the outer circuit. The radius vector is drawn at an angle giving a tangent equal to the armature resistance. A straight line is the characteristic when the resistance is con- stant. Therefore, the distance e s is the amperes of current which would exist in the armature at the voltage corresponding thereto. This and the corresponding lengths are then added each to the corresponding abscissa of the external curve, as at m n SERIES MACHINE Figs. 176 and 177.- Ohm-Volt Curve of Series and Shunt Dtnamos. and i i^ and the new curve is drawn through the points thus determined. The outer curve is the total current characteristic thus con^ructed. Total Characteristic of Shunt Dynamo.— So far the curves have been based on potential difference at the terminals. To get the characteristic based on total electromotive force and total current, we start with the curve of total current e, Fig. 175. The radius vector J gives the armature characteristic. Take a point p, Fig. 175, on the curve of total current. This is on an ordinate denoting p e amperes. The voltage of the armature at this current is the length a x. This added to p x, the voltage at the terminals, gives the electromotive force q x at the amperage O X. In this way points are found on a curre which give the 284 ELECTRICIANS' HANDY BOOK. relations between the electromotive force and total current, just as described for other cases. Ohiii=Volt Curves.— Curves can be laid out with the resistance of the circuit as one of the ele- YlB __H D ments. In parallel lighting service the resistance of the outer Circuit increases as lamps are extinguished, and decreases as they are lighted. The ohm-volt curve is espe- cially adapted for expressing the conditions of such service. Two such curves are shown on a small scale in Figs. 176 and 177, one for a series dy- namo, one for a shunt dyna- mo. For a compound dyna- mo the corresponding curve is a combination of the two. In Fig. 178 the curves are su- perimposed. It will be seen that if combined, the result will be a straight voltage line. This is the condition desired for parallel circuit lighting and supply, where the voltage should be constant under all changes of resistance. Thus the sum of E G and E F is equal to H E. This line repre- sents the sum of the voltages of the curves at the resistance de- noted by E. The same summation at other points would give other points representing the sum of the voltages of the two , curves at various resistances. The locus of these points, or the place where they would be found, would be a line approximately straight and horizontal; it is the line B D of the cut. The line B D is the combination of the other two curves and indicates their combined action. This action is that of giving identical voltage at varying resistances. OHMS Fig. 178.— Combined Series and Shunt Ohm-Yolt Curves for a Compound- Wound Machine. CHAPTER XVL THE DIRECT-CURRENT MOTOR. Direct' Current Electric Molor and Torque. — The direct-cur- rent electric motor is a machine driven by the direct current which is generated in any desired way, which current is forced through it by electromotive force. As motors are constructed in modern engineering practice, the driving of the motor causes the armature to rotate. This it does with greater or less force, and the force developed is torque. Torque is a twisting or turning force. The armature driven around with torque or twisting force is connected to machines, so as to do useful work. Reversibility of Dynamo and Motor. — It is only a few years ago that the doctrine of the conservation of energy was definitely formulated and accepted as the cornerstone of natural science. In nothing is it better illustrated than in the reversibility of the dynamo and motor. If such a machine is turned by ipechanical power, resistance will be encountered if the circuit is closed, and mechanical energy will be absorbed. On the circuit, by the heat- ing of the wire and other means, the presence of electrical energy can be discerned. Energy is conserved. The mechanical energy expended in driving the machine has not disappeared; it has been converted into electrical energy. In the motor exactly the opposite action takes place. Electri- cal energy is absorbed by the motor, and mechanical energy is given off by it. It is another example of the conservation of energy. The same machine can act in one or the other role. In engineering practice an electric machine often automatically changes from motor to dynamo, or the reverse. Sometimes this action is a cause of serious trouble if not detected in time. Generator and Motor Connected. — If we have two direct- current machines connected by two leads, so as to form a com- 286 ELECTRICIAN^ S' HANDY BOOK. plete circuit, and both are turning, each one in turning will gener- ate electromotive force. Referring now to the cut, Fig. 179, there are shown two such machines connected, and both armatures are supposed to rotate in the direction indicated by the upper arrow. The polarity of the electromotive force due to rotation tends upward from the lower brush to the upper. This tendency is indicated by the arrows on the end of the armature pointing upward. The condition of things shown in the cut is the operation of a generator and motor on one circuit; both are direct-current and GENERATOR MOTOR LINE Fig. ITO.—CcNNECTiON of Generator and Motor. bipolar. The bipolar type is selected for the sake of simplicity. V>^hat is to be noted applies to all direct-current machines. Counter Electromotive Force. — The left-hand machine is the generator sending current over the line and through the coils of its armature. The latter begins to revolve. As it does so, it generates in accordance with Lenz's law electromotive force, which operates to resist its rotation. This it does by opposing the electromotive force on the line, thereby cutting down the driving current. Such an opposing electromotive force is called counter electromotive force, and is indicated by the arrows on the end of the armature. This example illustrates a broad principle underlying the operation of direct-current machines primarily. If in a direct- THE DIRECT-CURRENT MOTOR, 287 current machine the electromotive force and current are in harmony with each other, working in the same direction, the machine is a generator. If the current forced through the ma- chine and the electromotive force due to its rotation oppose each other, the machine is a motor. The above applies to alternating-current synchronous motors operated by single-phase current, as will be seen later. Action of Counter Electromotive Force. — Counter electromo- tive force tends to reduce the speed of rotation of direct-current motors. It does this without any direct waste of power. A motor may work with highest efficiency at a certain rate of speed. If the counter electromotive force reduced it below this speed, its efliciency would be reduced, but there would be no direct relation necessarily between the counter electromotive force and the reduc- tion in efficiency. Counter electromotive force is not a hurtful resistance. Counter electromotive force tends to prevent a direct-current motor from going too fast. The faster its armature rotates, the greater will be the counter electromotive force produced, and the less will be the torque of the machine. The torque v/ill diminish, because the current will diminish as the counter electromotive force increases. As the armature turns against mechanical re- sistance of various kinds, friction, air resistance, and sometimes a load such as that due to machinery driven by it, the reduction of torque reduces the speed of rotation. Relation of Speeds of Generator and flotor Connected. — If the two machines are identical, and if the motor turned without any friction or resistance of any kind, the greatest speed the motor could attain would be equal to that of the generator. Identical armatures rotating at identical speed in identical fields of force generate the same electromotive force. In the generator and motor these w^ould be opposed to each other; and when the motor turned at the same spo^d as the generator, no current would pass. In the condition supposed, the armatures would rotate synchronously, and no mechanical energy would be gener- ated in the generator or expended in the motor. Such syn- chronism could not possibly exist, as no armature could rotate without experiencing some resistance. 288 ELECTRICIAyS' HANDY BOOK, The slower a motor runs, other things being equal, the greater will be the current passing through it, and the greater will be the net electromotive force producing the current. The volt- amperes will be greater therefore as a motor runs more slowly, and slow running is due to increased mechanical resistance. The volt-amperes represent energy absorbed; the mechanical resist- ance overcome represents energy developed. As is manifestly proper, they increase and diminish together. Counter Electromotive Force and the Armature. — The arma- ture of a working motor is ordinarily of such low resistance that the current which would pass through it at the potential of the working circuit would heat it so much as to injure it. As the armature rotates it has counter electromotive force impressed upon it, which acts like resistance, and reduces the current passing through it. Counter electromotive force protects the armature from burning out. Reduction of current in the armature re- duces torque, so that the turning force of the armattire is re- duced as its speed of rotation increases. Thus a slowly-turning armature takes more current and exerts higher torque than a rapidly-rotating one. To protect it from burning out a rheostat is generally used to start it, so that it begins rotating with a re- duced current, only receiving the full electromotive force of the circuit when it is turning fast enough to protect itself by comnter electromotive force. CHAPTER XVII. OPEN-COIL AND HOMOPOLAR GENERATORS— SIZE AND OUTPUT OF GENERATORS. Open=CoiI Armature Winding.— The windings of the armature of a direct-current dynamo need not be re-entrant. In the one- coil armature they are disconnected at the ends, as is seen in Figs. 123 and 126, page 220. The old two-pole magnetos and dy- namos with single-coil armature of the H type, Fig. 128, page 223, all operated on this principle. The Gramme ring, introducing into the larger field of engineering practice the older Pacinotti prin- ciple of closed-coil winding, was hai'.ed as an immense advance. Yet to the present day the open-coil winding is used on some of the most successful dynamos. The Brush Dynamo uses open-circuit windings. The diagram, Fig. 180, shows the principle of the armature winding. The coils are carried on an iron ring-shaped core, which is a variety of the Gramme-ring core. The coils may be of any even number; for each pair of coils opposite to each other there is what amounts to one two-part commutator. Returning to the diagram, Fig. 180, the outer end of each armature coil is connected to a commutator bar or segment, as indicated in plane development. The inner ends of each pair of opposite coils are connected across the armature to each other. When a pair of coils are in the neighborhood of the vertical line, or in the position of coils C, a, the maximum number of lines of force are passing through them. For the instant the path of these lines of force coincides with the path followed by the coil, so that there is no change in the number passing through the coil for that instant. Hence the coil is inactive and is out of circuit, no brush contacts being made with its commutator seg- ments. From the coils A^ A„ which on account of armature reaction 290 ELECTRICIANS' HANDY BOOK, are in the position of best action, current is tal^en by the brushes P Q. The arrows give the direction of the current. From brush Q the current goes to brush R. The coils Bi B, have left the position of best action, and hence have less electromotive force impressed upon them than have the coils Ai A2. The same applies to the coils D^ D, as regards electromotive force, for converse reason, that these coils D^ Do are approaching but have not reached the position of best action. The brush Q, taking the Fig. 180. -Development op the Brush Dynamo Winding. current from A,, delivers it to the B and D coils, four in number, in parallel of two, by the brush R. It divides between them, re- unites at brush S, goes through the field F M, and outer circuit back to P. The object of dividing the current between coils B\ B, and D^ Do in parallel is in a sort of accordance with Ohm's law. The electromotive force in these coils being below the maximum, the resistance is also reduced by connecting them in parallel. OPEN-COIL AND HOMOPOLAR GENERATORS. 291 Each pair of coils in a Brush machine can be pictured as rep- resenting an open-coil independent armature of two coils. Thus an eight-coil machine is in a sense equivalent to four independent machines, caused to co-operate in producing a pulsating direct current. Brush Dynamo Construction.— In Fig. 180a is shown a Brush dynamo with its upper field section removed and its armature hoisted up out of its bearings. It shows the ring armature with grouped windings. On the shaft are rings of larger diameter Fig. 180a.— Brush Dynamo. than that of the shaft which carries them. These are oiling rings When the armature is in place, these hang with their lower sec^ tion immersed in oil. As the shaft rotates they turn also, travel- ing around it and carrying up oil to it so as to continuously feed the bearings with oil. This is termed ring oiling. The Thomson^ Houston Armature is wound on the open-circuit principle. The armature contains a group of three coils, or thrcf^ sets of coils and a three-part commutator. The coils are wound on a hollow spheroidal frame, and the resulting armature is nearly spherical in shape. The coils are all wound in the same r3nse, right or left-handed. Three ends of the coil windings are connected together; the other three ends go to three commu- 292 ELECTRICIANS' HANDY BOOK, tator divisions. In the newer machines a ring armature has been used. The diagram, Fig. 181, shows the connections. Three ends ol: the coils are connected together at D, as shown; three go to the commutator segments, A, B, and C. The small arrows show the direction of the current. The coils consist of numerous turns of wire; the diagram shows each as a single turn for the sake of simplicity. The next diagram, Fig. 182, shows the coils as three radii, A, B, and C, connecting three commutator segments. Each radius represents a great num- ber of turns of conductor on the core. FP are four brushes; LL indicate lamps on the outer cir- cuit. Arrows show the direction of the current, and a curved ar- row shows the direction of rota- tion of the armature. The dotted lines mn show the position of the neutral line of the field. Coil B in the diagram is in th( neutral position, and is cut out. The positions of coils A, B, and C may be referred to the figures on a clock face. B is at 10 o'clock, A is at 2 o'clock, and C is at 6 o'clock. In this po- sition current goes from C to A. When the armature . turns so as to bring B a little further on, it connects with the brush F', and this coil B through the brush F' and the coil C through the brush P' work in parallel with each other and in series with A. Next C moves on toward 4 o'clock, and is cut out, leaving A and B working in series with each other. As A passes on toward 10 o'clock, just before it parts company witJi the brush P, the brush F comes in contact with C, and A and C are in parallel with each other and in series with B. In practice the angular distance between the brush ends P' and F' and the brush ends P and F is about 60° respectively, and this keeps the parallel pair of coils in parallel with each other for some considerable part of the rotation. Fig. 181.— Thomson-Houston Armature Winding. OPEN-COIL AND HOMOPOLAR GENERATORS. 293 The dynamo is automatically regulated by moving the brushes F F' backward or away from P P', and shifting P P' forward when more current is needed and vice versa. If the angular dis- tance between the brush ends P and F and the brush ends P' and F' is 60°, there will be no period when all the section coils are not doing something, two being always in parallel with each other. By bringing the brushes closer together the current is dimin- ished by increasing the period of time during w^hich one of the coils is cut out and inactive. Homorolar, Acyclic or Unipolar Dynamo. ^ — We have seen that a closed ring sw^ept through a uniform field of force has no current induced in it, although electromotive force is impressed on it. Electromotive force is impressed upon its two halves of Fig. 182.— Diagram of Circuits of Thomson-Houston Dynamo. the same polarity in each, so that they counteract each other. If to opposite ends of the horizontal diameter of the ring, as shown in the drawing, Fig. 115, the ends of a conductor were connected, current would go through it. A simple conductor representing the vertical diameter of the circle could take its place, and nat- urally would. A current is thus produced by sweeping a con- ductor through a uniform field of force, and without varying the number of lines of force which are interlinked with the circuit. A generator constructed on this basis is named as above. Up to the present time, comparatively few have been made. Various ways of producing the field can be used. Let cylin- drical north and south poles of a dynamo face each other. The axis 01 the armature corresponding with the center of the cylin- 294 ELECTRICIANS' HANDY BOOK. drical field, radial conductors swept through it will have elec- tromotive force impressed upon them, and if one brush connects with the inner and one with the outer end of a radial con- ductor, current can be taken from the brushes. The conductors cannot be connected as in ordinary dynamos, but the brushes must take current from opposite ends of the conductor. Any number can be put in the armature, and are connected to col- lecting rings, so as to carry out the principle described. Two north poles may be placed within two south poles, thus making an annular or ring-shaped field. A conductor swcpl through this field, and lying parallel with the axis of the field, will have electromotive force impressed upon it, and a current can be taken from brushes connecting with its ends. Any num- ber of conductors can be placed in the field, so as to form a hollow cylinder. They cannot be connected, as in a drum arma- ture, or one will counteract the other. The current has to be taken from their ends. Owing to the absence of a commutator, this type of machine presents great advantages as a generator of direct current. Its names are derived from the feature that the active conductors move through a field of uniform or unvarying density. Relation of Size and Output of Dynamos.— Considerable dis- cussion has been given to the question of the relation of the sizes of identically-shaped electric current generators to their respective outputs. If a dynamo is reproduced in all its relative proportions, but of double the linear dimensions, what will be the relative power output of the two? Manufacturer's and Thompson's Rules — One rough rule is to treat the output as varying with the weight. This is a manu- facturer's rule, a mere approximation to accuracy. It is express- ed in mathematical terms by saying that the capacities of iden- tically proportioned dynamos vary with the cube of the linear dimensions. Prof. Silvanus P. Thompson has given the fifth power of the linear dimensions as the correct figure. If one dynamo was twice as large in linear dimensions as an- other, it would, according to the "manufacturer's rule," have eight times the capacity of the smaller one; according to Prof. Thompson's rule, it would have thirty-two times the capacity. * OPEX-COIL AND HOMOPOLAR GENERATORS. 295 This is a considerable discrepancy. Yet if we investigate it on a purely proportional basis, the discrepancy may be still greater. Deduction of Thompson's Factor.— Assume a dynamo of double the linear dimensions. The length of the magnetic circuit will be doubled. For identical value of magnetization or Q twice the ampere turns will be needed. The wire on the field will have twice the diameter on the large dynamo of that of the wire on the small one. This will give it the same number of turns, but four times the capacity for current. Therefore the intensity of field or B will be multiplied by four on account of the increased current, and divided by two on account of the greater length of magnetic circuit, giving half the permeance per unit of cross section. This is a net increase of field intensity to twice that existing in the smaller dynamo. The field will have four times the cross-sectional area. There- fore being of twice the intensity, the lines of force in it will be 4 X 2 = 8, or eight times as many as in the smaller dynamo. The armature will have twice the linear dimensions of the smaller one, and therefore can carry twice the turns of wire per layer and twice as many layers as the smaller one carried. This gives four times as many convolutions. If it rotates at the same number of revolutions per minute as does the smaller armature, it will cut 8 X 4 = 32, or thirty-two times as many lines of force as did the armature one-half its size in linear dimen- sions. This gives thirty-two times the voltage. This coefficient 32 is the fifth power of 2, or 2^ = 32. If we stopped here, we should have Prof. S. P. Thompson's figure. The output of a dynamo is governed by the voltage it can produce, and by the amperage it can carry. The size of the armature wire con- trols the amperage. It must not be subjected to so heavy a cur- rent as to get overheated. If it is assumed to be of the same size in the larger as in the smaller dynamo, the larger one will give the same current at thirty-two times the voltage, which gives thirty-two times the capacity in watts. The rule of the fifth power thus deduced is not rigorously ac- curate, because sources of loss vary with the sizes of dynamos, and tend to favor the output of the larger sizes. In the above calculation a doubling of intensity of field mag- 296 ELECTRICIANS' HANDY BOOK. netization is assumed. This is not to be looked for in pranL-tice. With equal excitation, the output would be reduced to 16, or 2^ times that of the smaller. The Sixth=Power Rule.— If we assume the air gap to be in- creased in depth or thickness, and the same field intensity to be maintained, we can get a still higher result. Assume the field of equal intensity to be of four times the area. Assume that the same voltage is to be impressed on the circuit. Then one-fourth the windings are required on the arma- ture, as there are four times as many lines of force in the field. To put one-fourth the windings on an armature of twice the circumference, the wire should have eight times the diameter of the wire on the smaller armature. But a wire of eight times the diameter of another one can carry 8- = 64, or sixty-four times the current, giving an output sixty-four times as great as that of the generator which is half its size in linear dimensions. Both of these deductions are on the basis of an equal number of revolutions of the armature per minute. It would be nearer truth to take an identical peripheral velocity. This would reduce 2^ or 32 to 2* or 16, and 2' or 64 to 2^ or 32. The latter figure allows for an increase of the thickness of the air and copper gap in the larger machine to eight times what it was in the smaller one. This is certainly excessive. Calling n the relative linear size of the larger dynamo, the authorities give the student his choice of the following powers of n to express the increased output of the n times larger dy- namo: Prescott n- Mascart and Joubert n- Hopkinson n^ Rechniewski n^ Manufacturer's Rule, a little over n^ Ayrton n^*' Frolich n* Deprez n^ Thompson, S. P n^ All things considered, the fourth power or rJ is a safe figure to take. CHAPTER XVIII. GENERATOR AND MOTOR CONSTRUCTION. Disks for Smooth=Surface Armature Cores. — To prevent the production of eddy currents of serious intensity, armature cores are built up of thin sheets or disks of iron. Disks are cut to give the cross section of the drum, and are laid up with sheets Figs. 183 to 186.— Smooth-Contour Core Disks. of paper intervening to form a cylindrical pile. The disks are often cut with a large hole in the center. They are often fastened together with bolts, w^hich run through the pile of iron and paper from end to end. The bolts are insulated by tubes of in- sulating material. The cylindrical core thus produced may be 297 298 j:.LECTRICIANS' HANDY BOOK. keyed directly to the main shaft, or is carried by spiders. The cuts. Figs. 183 to 188, show various examples of core disks with Figs. 187 and 188.- Smooth-Contour Core Disks with Spiders. smooth contour. Sometimes the core is built up of segments of disks, as shown in Fig. 189. Disks for Orooved Armature Cores.—A general practice in drum armatures is to have a series of longitudinal grooves in the cylindrical core surface to receive the conductors. Disks for these are cut out with peripheral notches, as shown in Figs. 190 to 192. Fig. 189.— Segmental Cobz Disk Construction. However carefully the piling up of such disks is done, the notches or grooves are not to be relied on as being perfectly true and smooth. Smoothness is essential to avoid cutting the insulation. Accordingly, each core thus built up is often placed in a filing machine, where the long grooves are filed out one by one until they are of exact size and smooth. This constitutes the armature core, which is keyed to the armature shaft of the machine. In Fig. 193 is shown such a core. GENERATOR AND MOTOR CONSTRUCTION. 29 :i Formed Coils. — The material of the winding differs lor differ- ent machines. It may be composed of round wire, of square wire, or of flat bars. In some worlds each coil is shaped on a form to Figs/ 190 to 193.— Disks for Grooved Armature Cores. the exact contour of the place where it is to lie on the armature. In the case of a heavy bar of copper, this exacts some rather elab- orate bending. Sometimes an iron jig is used for this purpose, sometimes it is done largely by hand process. In any case, each Fig 193 —Laminated u. vm i^i«j\iA'iuitE Core on Shaft. coil or element appears as an Irregular rectangle. Such coils aro termed formed coils. American practice favors the use of formed coils, whenever it is possible to employ them. They are shown in Figs. 194 to 196. Wire Wining— Tha older way of winding armatures was the simple process ol' hand-winding vvith insulated wire. This is 3g Field. GENERATOR AND MOTOR CONSTRUCTION. 305 the conductors on the periphery of the drum or ring as the case may be, were it not for mechanical considerations. Position of Commutators. — The commutator is keyed on the \J/,r Figs. 203 and 303a.— Disk Armature Induction. armature shaft. It must be accurately in center with it, and its surface must conform to the requirements stated. The end disks or spiders which hold it to the shaft are seen in the cuts, and the insulation between them and the commutator bars is ^M_ Fig. 201.— Section of Field and Armature of Disk Dynamo. Fig. 205.— AR>f\ TUBE Coils of Disk Dynamo. shown in the cuts. In Figs. 199, 200, and 201, page 302, are shown different arrangements of commutators and armature. Brushes and Brush Holders. — The name brush is applied 306 ELECTRICIANS' HANDY BOOK, to the conductor, generally very unlike a brush, which bears against a moving surface, also a conductor, so as to make an electric contact. The moving part may be a simple insulated ring upon a rotating shaft, it may even be a shaft, or is a com- mutator with conducting segments insulated from the shaft and from each other. Fig. 2^Q —Section op Commutator and Brufh Mou' ting. In direct-current dynamos such as are being now described, the brushes beir against the cylindrical surface of the commutator. Tangential Brushes. — The first brushes were springs of metal placed tangentially, and pressing against the ring or commuta- tor. These were succeeded by compound brushes made of a number of pieces of copper secured one on top of the other and GENERATOR AND MOTOR CONSTRUCTION. 307 beveled and trimmed accurately square at the end, so as to line with the commutator divisions. Wire gauze is a constituent of some brushes, and carbon bearing directly or radially on the Fig. 207.— Commutator Constritction. Fig. 208.— Section op Commutator and Brush Mounting. surface of the commutator is now the generally accepted type of brush. It presents advantages not possessed by other brushes. Trimming Metal Brushes.-— Metal brushes must fit the com- mutator circle with their hollow beveled ends, and must be 308 ELECTRICIANS' HANDY BOOK, trimmed perfectly square. Gauges are provided for the pur- pose. A sharp cold chisel may be used, or a shears may suffice. A little touching up with a file may be needed. If a file is used, no filings must be left on the metal, as they might stick between the commutator segments and short-circuit the bars. Radial Brushes. — The construction of a radial car- bon brush and brush holder is shown in the cut, Fig. 210. A block of carbon is held in a socket. The block must move freely in the socket FiOo 209.— Section of Commutator AND Brush Mounting. Fig. 210.— Radial Brush and HOLDEU. and must be free from side shake. It is pushed downward { by a flat spring, and rests upon the commutator surface. The lower surface of the carbon is shaped to fit the commutator. The spring is so long that its pressure is sensibly even for dif- ferent lengths of carbon blocks. The carbon blocks constantly wear, so this feature of even pressure, whether they are long or short, which is equivalent to old or new, is a valuable one. Another brush mounting is shown in Fig. 211. Here the brush is fastened in a socket by a clamping screw, and a spiral pres- sure spring forces the brush against the commutator. The block is held by strips of hard copper, which act like springs and conduct the current. GENERATOR AND MOTOR CONSTRUCTION. 309 An important advantage in radial brushes is that if an arma- ture accidentally in starting or stopping, or while out of action, should be turned backward, radial brushes will be uninjured, while tangential or inclined ones will probably get caught in the commutator and be bent back and injured, so that they will re- quire straightening and trimming before the dynamo can be started again. Position of Opposite Brushes. — Brushes on opposite sides should not be set so as to bear upon exactly the same zone of CLAMPING SCREW Fig. 211. -Radial. Brush and Holder. the commutator. They should be staggered a little, so as to bear upon the whole surface of the commutator as near as may be. Brush Rigging. — In multipolar machines, the brushes are some- times carried in sets on two insulated rings, which are carried on a metallic ring. For the latter a seat is turned in the frame of the machine. In Fig. 212 is illustrated the ''brush rigging" as it is called of an eight-pole machine, and the next cut shows a set of the brushes. The mounting of a brush of one of these sets is shown in Fig. 211. The two rings are called the posi- tive and negative bus rings. A commutator is alwavs apt to wear a little uneven. The brush holders are made as light as possible, so as to follow^ any irregu- 310 ELECTRICIANS' HANDY BOOK. larities of shape of the commutator. A properly-treated commu- tator with brushes of good quality will wear very evenly. Relation of D pth of Air Gap to Sparking.— To avoid spark- ing, equal induction should be exerted by each pole. This is in- sured by equal permeance of the magnetic circuit. An unequal depth of air gap is the most potent disturbing cause. Some irregularity often results from wearing of the bearings, which throws the armature out of Fig. 312.— Brush Ftggtng for a MUl.TIPOi.AH DYxNAMO. Fig. 215.— Brt-s^ks of Same Brush 1{IGG1^'G. center, and diminishes the air gaps on one side and increases them on the other. To minimize the trouble resulting from this displacement, it is a standard practice to make the air gap rather deep. A small displacement of the armature shaft under these circumstances has less disturbing effect than it would with very small air ga^s. A reduction of a thirty-second of an inch would not be of mnch effect in a half-inch air gap, while the same re- duction would reduce a one-sixtcenth-inch air gap fifty per cent, and brine Rbo^if very irregular or one-sided 1^ '^notion. Field riagn t for Multipolar Dynamos — Almost all large dy- namos are now made of the multipolar tyr^e. "^^-o ^^oles are often short cylinders carried on the inner surface of q v,(ir,vy polygonal GENERATOR AND MOTOR CONSTRUCTION. 311 or circular iron frame as shown in Figs. 214, 215, and 216. The poles project radially toward the center. The frame constitutes the field yoke. It is often made in two parts joined on the hori- zontal diameter, and the poles lie on the diagonals of the circle as shown. The effect of this is that the upper half of the field can Fig. 214.— Foitr-Polb Dynamo Frame WITH Section op Winding ON One Pole. Fig. 315.— Multipolar Dynamo Frame Showing Brackets for Carrying Brush 1 igging. be lifted off without touching the armature surface, a very essen- tial requirement, as any rubbing of pole face against the surface of the windings might injure the insulation. In some machines the pole pieces are made of steel and cast- welded into the frame. This process is executed by imbedding the pole pieces in the mold, letting their ends project into the space where the metal for the semi-circles of the field frame is to run. The melted cast iron makes a perfect joint with the steel. Cast-iron pole pieces are used upon some of the machines. Formed coils are often used on poles as shown in Fig. 217. Sometimes they are wound with flat conductors bent edgewise as shown in Fig. 217a. To illustrate what has preceded, the cut, Fig. 218, may be re- 312 ELECTRICIAXiS' HANDY BOOK. ferred to. It shows the principal parts of a four-pole dynamo, the rear post, which carries the armature bearing, being omitted to avoid confusion. The ring system of oil feeding is used, and Fig. 216.— I' OLE AND Pole Winding of Multipolar Dynamo. Fig. 217.— Pole Winding WITH Formed Coils. Fig. 217a.— Edgewise Bent Field Winding and Pole ring oilers are shown under the left end of the edgewise armature shaft. The loose rings when in place rest upon the revolving shaft bearing and turn with it, picking up oil and feeding it to the shaft. GENERATOR AND MOTOR CONSTRUCTION, zu 314- ELECTRICIANS S' HANDY BOOK. Laminated Field Magnets.— Sometimes the field is made up of laminations cut as shown in Figs. 219, 220, and 221. The windings lie in the inner notches, and are so placed as to establish alter- nate poles all around the circle. The plates are insulated with paper. Sectional Laminated Field flagnets.— The laminated field <&^0^^^&^^ Figs 219, 220 and 221.— Laminated Fields. magnets on large dynamos are often made in sections, as shown in Figs. 222 and 223. These pieces are bolted together so as to form a circle. It will be noticed that the lower one has perfor- ations, and not notches. The windings are passed through these holes, and a very solid construction is the result. GENERATOR AND MOTOR CONSTRUCTION. 315 Such fields as those shown in Figs. 219 to 223 are most used in alternating-current generators. Details of Multipolar Field Windings.— The windings of the field poles are often made on tin spools with deep flanges, the surface being most thoroughly insulated by mica, paper, and cloth. One by one these are wound full of wire, which is shel- FiGS. 223 AND 223.— SECTiONAii Laminations for Field Magnets. lacked and, baked, so as to produce a solid mass, ready to be slipped upon the pole. A laminated pole piece placed over the end of each cylindrical pole holds the winding in place. It projects on both sides of it, and is hollowed out accurately to the circle of the armature. The space between it and the armature constitutes the air gap. The field castings have to be carefully machined before being set up. The bases and joint faces are planed off, the sections are bolted together, and the inner faces of the poles are then turned off. CHAPTER XIX. THE ALTERNATING CURRENT. Alternating Electromotive Force. — Although the term alter- nating current is universally used, alternating electromotive force is its producer, and precedes it often. The true original i? generally alternating electromotive force. An alternating electromotive force is one which alternates in polarity, regularly or in obedience to some law. It starts at a value of zero, rises to a maximum of one polarity, descends to a value of zero again, changing in direction reaches a maximum of opposite polarity, whence it returns to zero again. These alterna- tions are repeated over and over again. Cycle, Wave, and Frequency. — The changes described in the last sentence constitute a cycle or' wave of alternating electro- motive force. If for the word "polarity" we substitute the word "direction," the sentence will describe a cycle or wave of alternat- ing current. When a wave of alternating current or of alternating electromotive force reaches its point of greatest value, the whole circuit is affected; when it reaches zero value, the whple circuit is at zero. The expression wave must be clearly understood. It does not mean that waves run over the circuit, but it does mean that the whole circuit passes simultaneously through the values of the cycle, ranging from zero to a maximum in one direction, back through zero to a maximum in the other direction, and then back to zero, completing a wave. The number of waves per second is called the frequency of the alternating electromotive force or current. Electromotive Force rnd Current Curve. — We often gain in our conceptions of things by analogies. It is evident that the action of such electromotive force is analogous to that of a wave. A line representing the cross section of a wave would be a con- 316 THE ALTERXATING CURREXT. 317 venient way of picturing to the mind the action of alternating electromotive force. It is so good a representation that it is universally used. Such a line is shown in the cut, Fig. 224. The horizontal line is what the geomstrician would call the locus (place) of zero values. The vertical lines are drawn of length corresponding to the distance of points of the curve from the horizontal or base line. They are called ordinates of the curve. The curve comprises one full wave or cycle. The portions of the curve below the horizontal line represent the electromotive force of one polarity; the portions above the line that of the other polarity. The lengths of the ordinates rep- resent the strengths of the electromotive force. The shape of 270^ Fig. 324.— Sixe Curve of Generating Circle. the curve represents the w^ay the electromotive force increases and diminishes. If the proper conditions prevailed, an alternating current of exactly identical form would be produced by the alternating electromotive force. A similar curve would represent the cur- rent, its ordinates would indicate by their length the intensities of the current. Their position above or below the zero line would indicate the direction of the current constantly changing; for part of the cycle in one direction and for part of the cycle in the reverse direction. The current is the factor generally referred to in practice; alternating electromotive force is its invariable cause and con- comitant. An alternating-current generator is also an alternat- ing electromotive force generator, and an alternating-current sys- 318 ELECTRICIANS' HANDY BOOK. tern is an alternating electromotive force system. It is thus through all the branches of alternating-current work. But the expression "alternating current" is always used, except where there is special reason for using the other expression, "alternat- ing electromotive force." Production of Alternating Electromotive Force and Current. — Alternating electromotive force is impressed upon a circuit by means of a special type of electric machine called an alternating current dynamo, alternating current generator, or alternator. In practice the armature contains one, two, or three separate wind- ings. The terminals of the windings are connected to various numbers of leads constituting ihe outer circuit. The armature is so wound that its entire winding is simultaneously impressed with the same electromotive force at the same instant, but this electromotive force varies though the values of the wave or cycle just described. In turning through the arc represented by the quotient of 360° divided by half the number of poles in the field of the alternator, one cycle or wave of alternating electromotive force is impressed upon each of the armature windings and cir- cuit. The above is strictly true for typical alternators, but is sub- ject to modification for some constructions. Length of a Wave. — A wave curve is divided into or measured by 360 parts. This refers its length directly to degree measure- ment, as there are 360 degrees in a circle. Such system har- monizes with the fact that alternating electromotive force is always generated by the circular motion of an armature or field, and that a single wave is produced by a complete rotation of the armature or field through 360 degrees, or by a rotation through an integral portion of 360 degrees. Fcrm of Alternating E. M. F. and Current. — The form of the curve represents the nature of the variations of the electro- motive force or current, as the case may be. The form of a cur- rent is the nature of its variations, and is graphically represented by its curve. The same applies to electromotive force. An al- ternating current may increase slowly or suddenly, by an even curve or a series of jumps, and may return to zero value in var- ious ways also. If it is to be represented by a curve, the irregu- THE ALTERXATIXG CURRENT. 31^ larities of the current will be shown in its form. Hence the form of the current is spoken of, and expresses the consecutive varia- tions in intensity. In Figs. 225 and 226 waves of various forms are shown. The dotted lines represent in each case separate waves of exactly the same length. The two dotted lines com- bined produce as a resultant in each case the heavy black line, which gives a third wave form. Length of Wave and Frequency. — Waves of the same fre- quency are those which are produced the same number of times in a given period of time. The curves representing them are // \ fT A / 6 V-'-. •' / V--'' V---' ^«^ J V. Vy fj ^ ^^:rN ^ '^ -^ Figs. 325 and 2t6.— Waves of Different Forms and Resultant of Two Waves. drawn of the same length per wave if the waves are of the same frequency. In the diagrams of waves in this section of this book, the waves on each diagram are of the same frequency, and are therefore drawn of the same length. Cause of the Form of Alternating; Electromotive Force and Current. — The weaves of current may vary in forn;, and the form depends upon the construction of the generator, with particular reference to the shape of the field-magnet pole pieces and of the armature core. The distribution of lines of force through the field may be varied indefinitely by changing the shape of the poles, even if the armature core is left unaltered. The current varies in intensity as the rate of change of the number of lines of force interlinked with the circuit varies. This rate of change de- 320 ELECTRICIANS' HANDY BOOK. pends on the distribution of lines of force in the fi^ld. Hence any form, within reasonable limits, can be given to the alternat- ing electromotive force and current waves. The form universally striven after is known as the sine curve. Alternating Electromotive Force and Current Curves. — When an alternating electromotive force is producing a current, separ- ate curves may be drawn for each. One zero or base line is used for both curves. The portions of the electromotive force curve above the line represent the portions of the electromotive force producing current represented by the portions of the current Fig. 227.— Current and Electromotive Force Curves. curve above the line. The same applies to the curves below the line. The length of a wave of current must be rigorously equal to that of a wave of electromotive force producing it. The height may be different, and as a matter of convenience usually is so drawn. Fig. 227 shows two such wave curves. The current curve may cross the zero line at the same points where the electromotive force current does, or may not, according to the conditions of the circuit. The single zero line or base line is the locus or place where current and electromotive force have zero value. When the current curve crosses this line, it indicates that the current at that instant ceases to exist. It is the point when it reverses THE ALTERNATING CURRENT. 321 its direction, dropping to zero in one direction and starting in the other direction from zero. The same applies to alternating electromotive force and its polarity. Drawing the Elejtromotive Force and Current Curves.— A straight horizontal line is drawn, and a portion of it is taken to represent the period of time taken by the system to produce one vvave or cycle. For several reasons it is most convenient to divide this into periods of time, each equal to 1/360 of the period of a 3ycle or some integral divisor of 360, such 1/36 or 1/16. Suppose the line is divided into sixteen parts; at each a perpendicular is drawn. These are called ordinates of the curve. If electromotive force is the subject, each ordinate is drawn of length proportional to the voltage at the period it represents. For one polarity the ordinates are laid off upward, for the other polarity downward. Through their ends a line is drawn, and this represents the form of the cycle. The same method can be used for alternating-cur- rent curves, the lines being drawn of length proportional to the amperage of the current at different periods of the cycle. The method is illustrated in Fig. 224. Degree System. — The portion of the zero line containing or subtending a complete wave, consisting of one positive and one negative portion, is by the degree system divided into the 360 di- visions as described, each division representing one angular degree of rotation of a two-pole armature in a two-pole magnetic field. This armature may be only hypothetical. If the armature is four- pole and rotates in a four-pole field, each degree on the zero line of the current curve or electromotive force curve will repre- sent one-half of a degree of its rotation; if it is a six-pole con- struction, one-third of a degree will be represented, and so on. Thus the ordinate at any point by the degree system can be re- ferred to some point in the rotation of the rotor of the alter- nator. This indicates a reason for referring the divisions of the zero line to degree measurement. The form of the alternating current almost universally used in represented by the sine curve, and to make the construction of the curve intelligible and easy, the degree system is essential. The Sine Curve. — The sine curve or curve of sines, as it is also called, is shown in Fig. 224, and is based on the following 322 ELECTRICIANS' HANDY BOOK. principles: On a horizontal base line are erected perpendiculars. The line is divided into 360 parts, representing degrees of a circle. Each perpendicular is laid off equal in length to the sine of the angle expressed by the degree mark on which it stands. Up to the 180th division the perpendiculars are on one side of the base line; for the rest of the 360 divisions they are below it. The curve can also be drawn by the use of what is called the gener- ating circle. Generating Circle. — At one end of the horizontal zero line a circle, as in Fig. 224, is drawn, with its center on the line. Sines of various angles are drawn, and from the end of each sine a horizontal line is drawn, under or over the zero line, according to the position of the sines. The intersection of each horizontal line with a perpendicular to the base line erected on it at a point corresponding to the degrees of the arc of the sine gives a point on the sine curve. The arcs begin at the right-hand end of the horizontal diameter of the circle. Half of the sines are above and half below the base line, and the complete circumference of the circle gives the sines to determine the ordinates for one full cycle or wave. In the cut 22i/^° is the angle used in the construction. The curve is drawn through the ends of the lines. The com- plete circumference of the circle gives one full wave. An ellipse or various other figures or closed curves could be substituted for the circle, when other forms of waves would be produced. In practice the circle only is used, as the tendency of engineering is in the direction of the production of sine curve currents. By reference to the generating circle, the sine curve may be thus simply described. Imagine the periphery of a circle straight- ened out so as to become a straight line. Mark upon it the de- grees of the circle. Erect perpendiculars on each degree mark, for the first 180° above the line, for the rest below it, and make each line equal in length to the sine of the angle indicated by its degree mark. The curve is drawn through the ends of these lines. The construction is shown in the cut. Fig. 224. The left-hand quadrant of the generating circle is divided into angles of 22i/>°, THE ALTERNATING CURRENT, '^23 or one-sixteenth of a circumference. For each angle sines are drawn, such as M P. On the base line divisions corresponding to the angles are laid off, and ordinates erected on them. Each sine determines the length of the ordinate corresponding to its angle. Thus, the sine MP of 45° determines the length of the ordinate M P erected on the second or 45° of the sixteen divisions of the horizontal base line. Interpretation of the Generating Circle.— It is not necessary to draw a sine curve to represent the form of this universal type of alternating-current cycle. The cycle can be represented by the circle alone. Thus, the line drawn from the horizontal diameter, perpendicular to it, and intersecting the circle at any point, gives the value of the electromotive force or current at the instant represented by that point on the circle. This line at 90° has a maximum value. This means that when a period is one-quarter advanced from its beginning, the electromotive force or the cur- rent, as the case may be, has its maximum value. The same applies to the 270° of the circle. Values below the diameter ar^ of polarity or direction opposite to that of values above it. At. 180° the value of the perpendicular is zero, and after that it be- gins to increase in the opposite direction. This means that when the cycle is one-half completed, the electromotive force or cur- rent, as the case may be, has a value of zero, and immediately begins to increase, but in the opposite polarity or direction. Rate of Change. — At the top or bottom of the loops the curve for two consecutive points is horizontal, where for an infinitely small period it moves parallel to the base line. Hence, if it rep- resents current strength, the current for this instant will not chane-e. Where the curve crosses the base line there is a place where the curve between two consecutive points is most steeply inclined to the horizontal line. At this place it approaches the vertical direction nearer than elsewhere, and at this place the electromotive force or current, whichever the curve represents, is changing most rapidly in value. It has here its highest rate of change. Graphic Representation of Rate of Change. — If a radius sweeps through the arc of a quadrant, the successive sines will indicate the successive values of the alternating current it may 324 ELECTRICIANS' HANDY BOOK, be taken as representing. As it starts from 0°, the sines will in- crease most rapidly in length and their rate of change will be greatest. As it approaches 90°, the sines will increase least rap- idly in length for a given angular change, or their rate of change will be least. Let a second radius at 90° from the first be as- sumed to move around with the first one, always remaining per- pendicular to it. The sines of the angles fixed by the positions of this second radius give the rate of change of current or elec- tromotive force corresponding to the sine curve of the other or first radius, In Fig. 228 the two radii are shown, one marked R. of Ch. being the radius vector of rate of change. As a sine curve can be drawn from the first radius, a second one can be drawn from the other; the second sine curve will be a rate-of- change curve. The rate of change at any given point of the first curve will be pro- portional to the ordinate of the second curve at this point. The two curves will occupy fixed positions with reference to each other. They are said to be in quad- rature with each other, as will be explained later. Pig. 232 may be referred to as showing two curves generated by radius vectors at right angles to each other. Radius Vector and Resultant.— A line may be drawn from a common center or point called the origin, and may have two qualities. One is its length. This may be greater or less; and as the line is to be taken as representing a quantity, the length of the line must be proportional to the quantity it represents. This proportional size is determined by reference to some other line, otherwise the question of length would not come into con- sideration. The other quality is its angular position with refer- FiG. 238.— Radius Vector or Rate of Change. THE ALTERNATIXG CURRENT. 325 Fig. 239.- Vectors and Resultant. ence to any other quantity, which quantity is taken as indicated by another line drawn from the origin as before. The angular position is the number of degrees included be- tween the two lines. Thus, suppose two quantities represented by two lines, Fig. 229, O B and O C, drawn from a common center at an angle with each other and one longer than the other. They would indicate two quantities out of phase with each other and one greater than the other. They would form two sides of a parallelogram, whose diagonal O N from the origin O outward would be their resultant, and would represent the combined effect of both quantities. Vector Diagram of a Sine Curve.— From the center of a generating circle as shown in Fig. 230, radius vectors are drawn at equal angular distances around the cir- cle. Each radius has marked upon it a point equal in distance from the center to the length of the sine of its angle. In the diagram, O Q is laid off on OP equal in length to the sine of 60°. The same is done for the other radius vectors, and a curve is drawn. For one-half of the wave this is a circle of half the diameter of the generating circle. Two circles one above and one below the zero line represent the full wave of a sine curve. Phase, Lag, and Lead. — The length of a wave of impressed electromotive force is exactly equal to that of the wave of ctir- FiG 230.— Vector Ptagram of Sine Curve. 326 ELECTRICIANS' HAXDY BOOK. rent it produces. The waves also correspond in form, although in order to secure clearness in diagrams one set are often diawn of lower height than that of the other set. The two sets of waves may be in identical positions. This means that as each wave of electromotive force is impressed on the circuit, a cor- responding wave of current accompanies it. At the instant when the electromotive force is greatest, the current would be greatest. Such waves are said to be in phase with each other, and are shown in Fig. 231. In this and the following wave diagrams, the figures 1, 2, ... . with the vertical lines enable the relations of the curves tc be seen. If the waves of alternating electromotive force take a certain Fig. 231.— Waves in Phase. Fig. 232.— Waves in Quadrature. time to produce the current waves, so that the current wave reaches its highest intensity after the electromotive force wave producing it has begun to diminish, the current weaves are said to lag. Such a condition is shown in diagram in Fig. 232 by the one set of waves lagging behind the other, crossing the zero line later than the others. The reverse may hold. One set of weaves may reach their height before the other set reaches it. They are said to lead, w^hich condition Fig. 232 also serves to illustrate. Such condi- tion is shown in diagram by the one set of waves reaching their highest points while the other set are still rising. Angle of Lead and Lag.— We have seen that angular measure- ment can be applied to waves, and that the length of a wave is THE ALTERNATING CURRENT. 327 360°. A fraction of the length of a wave is expressed in de- grees. Half a wave is 180°, one-quarter of a wave is 90°, and so on. The difference in period between two waves is expressed therefore in degrees also. One set may lag 40° behind the other; as 40° is 1/9 of 360°, a lag of 40° means a lag of 1/9 of a wave length. The same applies to all lags and leads. The expression in degrees is called the angle of lag or of lead as the case may be. It is designated by 0. Quadrature and Opposition.— If the angle of lag or of lead be- tween two sets of waves is 90°, which is the quarter of a circle, the waves are said to be in quadrature with each other. The waves of Fig. 232 are in quadrature. If the angle of lag or of lead is 180°, the waves are said to be in opposition, as shown in Fig. 233. Basis of Lag and Lead.— The waves of alternating elec- tromotive force are the usual basis — the current waves are said to lag or lead. But if the current leads the electromotive force lags, and sometimes the current is employed as basis. A diagram of sine curves of two sets of waves, out of phase with each other, can be interpreted In two ways — as showing one set lagging or the other set leading. Average Values. — The average intensity of a sine current is equal to the average of the sines. As half are positive and half are negative in value, the average value of the current is zero. This is more simply put if the current is simply thought of as alternating, half one way and half the other, and therefore as neutralizing itself. This is begging the question. The thing that concerns the electrician is the practical value of the alternating current. It is a component of energy under proper conditions both coming and going. The average value of the ordinates of a sine wave gives the average value of the thing it represents. It can be expressed Fig. 233.— Waves in Opposition. 328 ELECTRICIANS' HANDY BOOK. as a fraction of the longest of the vertical lines. By geometrical process it is proved to be equal to 0.63633 of the maximum ordi- nate, the longest of the vertical lines — it is nearly two-thirds of the line indicating the height of the crest. This figure is of little practical use, because what the engineer is concerned with is the power on his circuit. In the case of direct current, the product of electromotive force by amperes gives the power. Direct-cur- rent power plants operate either at constant potential or con- stant current, while the electromotive force and current are con- stantly varying in alternating-current distribution. The formulas for energy rate or power I E, P R, and _ are given in a preceding It section of this book. The two latter can be used for alternating current work, to get two of the factors for the power of the circuit — the effective electromotive force and effective current. Effective Values. — The square of the value of a current is pro- portional to the watts it can produce in a conductor of definite resistance. The watt or volt-ampere is expressed by E I and by I- R and the latter expression shows that with a fixed resistance the watts produced by a current vary with the square of the current. * The effective value of an alternating current is expressed as the intensity of a direct current which would with the same re- sistance develop the same number of watts. This current would produce the same quantity of heat with the same resistance. Doubling the current with the same resistance would give four times the heat. The heating effect of an alternating current at any instant is proportional to the square of its intensity at that instant. The average of the squares of the values of the current through a half cycle is proportional to the average heating effect of the current. The square root of this average value is the effective value of the current. Calculation of Effective Values. — Taking PR and _ as ex- R pressions for the rate of energy in a circuit, if the average value of the squares of current or of electromotive force be taken, it will give the expressions for average power. If the square THE ALTERXATING CURRENT. 329 root of average P or average E- be extracted, it will give what is known as the effective values of current and electromotive force. The effective value of an alternating quantity is defined as the square root of the mean square of the ordinates of the sine curve. The effective value is thus calculated. Both the sines and cosines of the respective arcs of a quad- rant vary exactly in the same ratio, but oppositely or comple- mentarily disposed on all parts of a quadrant. We have the re- lation sin-& -h cos^ G =: 1. if the radius is equal to unity. As the sine and cosine vary in the identical ratio, the average sine- is equal to the average cosine-. Therefore from these considera- tions we have: Average sin- Q + average cos- 9=1 Average sin^ 6 = average cos^ disregarding the signs and only concerning ourselves with the numerical value. .*. 2 X average sin.^ ^ = 1 Average sin^ = ^2- Average sin G =^/1P = 0.7071. V /2 - 1.41 ^ This gives us the factor used in obtaining the effective value of the thing shown by the sine curve. The effective value is then equal to the maximum value multiplied by 0.7071. The maximum value in the unitary circle is 1, and the formula is therefore ap- plicable to other cases by substituting for 1 the value of the ^in 90° X radius = radius. Form Factor, — The quotient obtained by dividing the effective value by the mean value varies with the form of the curve. For a 7 '7 sine curve the value is zzzl.ll. This value is called the form - (38t) factor. This factor is of interest as giving the relative heating powers of alternating and direct currents. The alternating current will have about eleven per cent more heating power than will the direct current, which is of the same average strength. If an alternating current voltmeter is placed upon a circuit in which the volts range from +100 to — 100, it will read 70.7 volts, although the arithmetical average, irrespective of + or — sign, is really 63.7 volts. If a direct electromotive force were 330 ELECTRICIANS' HANDY BOOK. to act upon the same instrument, it would have to be of 70.7 volts value, to give the same reading. If an alternating sine current ammeter reads 100 amperes, it means that the current fluctuates from -i-141.4 to — 141.4 am- peres, but produces the same heating effect as if it were a 100- ampere direct current. An interesting point to be made here is that if a generator is wound with two windings for alternating currents, its effective electromotive force will be 1.11 times higher than if operated as a direct-current dynamo, by commutator and proper connec- tions. If w^ound with one open circuit of wire with two end con- nections to the collecting rings, its electromotive force will be 2.22 times higher. The tendency of an electromotive force on a circuit to cause the piercing of the insulation depends on the maximum voltage. This voltage must be taken cognizance of for this and similar effects. If on a Siemens dynamometer the current given by an alternator was found to be any given number of amperes, the maximum cur- rent would be found by dividing the reading by 0.707, or by multi- plying by 1.41+. Formulas for Effective Values. — The values can be expressed as vulgar fractions thus: I max Effective current =r i=r E max Effective electromotive forces _^ It will be noticed that in cases where the electromotive force or current intensity has been determined by an apparatus, it is always the effective values that are given by the readings of the instrument. As these are the working values, the coefficient 0.707 is of comparatively little use in practical working. Power Factor. — The rate of energy or the power developed in &. circuit by a direct current is expressed by E I. This will not answer for the alternating sine current. The periodic change in its values renders a constant necessary, with which the prod- uct of the effective current intensity by the effective electro- THE ALTERXATING CURRENT: 331 \ motive force is multiplied to give the average power. This is a practical quantity; it is the cosine of the angle of lag, or cos (p, E max J max Average E I or power = cos 9. Call effective current I, and effective electromotive force E. The equations for the effective values of current and electromotive force on page 330 give us the values of E max and I max, thus: E max E= or E max=:E. =0 the power is at its greatest, and is equal to the product of the effective values of electromotive force and current intensity. This case is when there is no lag of cur- rent, or when the current and electromotive force are in phase. Qualities of a Circuit. — There are three qualities of a circuit which affect the action of alternating current and electromotive force upon it — resistance, inductance, and capacity. Each one has two effects — one effect upon the current, the other effect upon the phase relations of electromotive force and current. Resistance acts to reduce alternating current, just as it does in the case of a direct current, in accordance with Ohm's law. This action upon an alternating current is greatest when the 332 ELECTRICIAXS' HAXDY BOOK. current is greatest, as at the top of the wave, and is without effect when the current is zero, represented by the sine curve cross- ing the zero line; in other words, it has most effect when it has most material to work upon. Its second action is to tend to bring alternating current and alternating electromotive force into phase with each other. Reactance.. — The rate of alternating current which can pass through a circuit is modified by the resistance, inductance, and capacity in the circuit. The effects of each on the current can be indicated by ohms; the retarding effect of inductance and the reverse effect of capacity can be expressed in ohms, just as if they were resistance. The ohmic values of capacity and inductance are called react- ances — capacity reactance and induction reactance. If the reactances and resistance of a circuit are known, its effect upon an alternating current is determined by Ohm's law subjected to certain modifications. Inductance. — ^^The relation of lines of force to current is in- duction. Current produced in a circuit by changes in the lines of force interlinked in it is induced current, and electromotive force so produced is said to be impressed on the circuit. The rela- tions of lines of force produced by current in a circuit to that circuit constitute inductance or self-induction. As self-induc- tion is due to changes in current intensity, it is an all-important thing in alternating current practice, where the current varies in intensity many times in a second. Inductance in an alternating-current circuit acts to diminish or to absorb the electromotive force. It operates most when the current is at its zero value, because that is when the rate of change is greatest. This period is 90° removed from the period of maximum current, so that inductance acts in quadrature with resistance. Its second action is to cause the waves of current to lag behind those of the electromotive force. Inductance and the Henry.— Energy Is required to create a field of force, but not to maintain it. Conductors are related in their properties to the field of force established about them by a current, or in other words vary in their property of estab- lishing fields of force under given current changes, which consti- THE ALTERXATIXG CURRENT. 333 tutes this property called self-induction or inductance, and which is measured by a unit called the henry. If a conductor is so constituted that a rate of change of cur- rent of one ampere per second in it requires the expenditure of one volt, it has an inductance of one henry. The same thing may be stated otherwise. If the inductance of a circuit is such that a current increasing one ampere per second produces in it a counter electromotive force of one volt, the circuit has an inductance of one henry. The henry is sometimes called the coefficient of inductance or of self-induction. It is generally indicated by the letter L. If the current in an active circuit decreases, the lines of force d.' ninish and their potential energy becomes kinetic, and electro- motive force increasing, the normal current is induced on the circuit. Electromotive Force in an Alternating=Current Circuit.— The electromotive force in an alternating-current circuit con- taining: inductance is partly expended in producing changes in the current. The electromotive force expended on increasing the- current intensity varies with the rate of change. As the current increases, so also does the field density increase, and this in- crease of field density is what absorbs the energy indicated by the electromotive force multiplied by the current change. If the circumstances are such that the field diminishes in density, in so doing it generates electromotive force of the polarity cor- responding to that producing the current to which the lines of force are due. Couviter Electromotive Force.— This is the hypothetical elec- tromotive force opposed in polarity to the original impressed electromotive force, and due to inductance. It can only exist when the current is increasing in value. In an alternating-cur- rent circuit it appears when the current is increasing, and has the highest value when the current has the highest rate of change, which is when the current is passing from its period of zero value. It resists the action of impressed electromotive force, which produces an increasing current — that is to say, resists the current in the first and third quarters of a wave. 334 ELECTRICIA2\S' HANDY BOOK. Forward Electromotive Force.— This is the hypothetical elec- tromotive force of the same polarity as the original, and is also due to inductance. It can only exist when the current is dimin- ishing in value, and has its highest value when the current is approaching zero value. It strengthens the action of impressed electromotive force which produces a diminishing current; it tends to increase the current in the second and fourth quarters of a wave when the impressed electromotive force would reduce it. Counter and Forward Electromotive Force in an Alternate ing=Current Circuit.— From what has been said in the last two paragraphs, it will be seen that induced electromotive force in an alternating-current circuit, whether it be forward or counter electromotive force, opposes the action of the impressed electro- motive force. When the latter is rising in value, its action is opposed by counter electromotive force; when it is falling in value, its action is opposed by forward electromotive force. Hence inductance generates for an alternating current what is virtually counter electromotive force for all its phases. Turns of a Circuit and Inductance. — Assume a turn or con- volution of a wire constituting a part of an electric circuit. If an ampere of current is passed through it, it constitutes an ampere turn. Let a current starting from zero value, and increasing to a definite value in one second, be passed through it. The lines of force of the field called into being will exercise induc- tance and produce a certain amount of counter electromotive force, which may be called e. Assume that a second convolution 01 wire is added, so that the current has to go through two turns. Instead of one. As this gives double the ampere turns, twice as many lines of force will be called into existence during the second of growth of a current equal in all respects to the one assumed. Each turn of wire will therefore be impressed with counter elec- tromotive force equal to 2e, because twice the lines of force of the first case act upon it. But there are two turns, each acted on by counter electromotive force of 2e. The total counter electro- motive force is therefore ie. This gives the law: The inductance of a circuit is proportional to the square of the number of its turns, if a constant rate of increase of current is maintained in it. THE ALTERNATING CURRENT. 335 Reactance of Inductance.— A circuit opposes to the passage of an electric current, whether such current be constant or varying, a resistance. This is measured by the practical unit, ohm, and in alternating current topics is often called for precision's sake ohmic resistance. This is the resistance of Ohm's law, indicated in formulas by R, and is independent of the electromotive force and current. Self-induction, which when a current is increasing in strength manifests itself by counter electromotive force, increases w^ith the current and therefore with the electromotive force. Counter electromotive force is a variable quantity in a circuit of fixed inductance. From Ohm's law expressed as I = — L we see that in a circuit of constant resistance the electromotive force must vary directly as the current. Therefore, as induced electromotive force varies directly with the current change, we can deduce an expression which Avill express it as a constant resistance, into which ex- pression current will not enter as a factor. Then in the expres- sion for the entire obstruction offered to an alternating or other type of varying current, we shall have two additive constituents. One is ohmic resistance, independent of current strength; the other is an ohmic equivalent of inductance, also independent of current strength. Ohm's law can be expressed as R ==__!. The inductance of a 1 circuit multiplied by the current change, which other things be- ing equal varies with the ultimate current strength, is equal to the counter electromotive force. As these two factors increase and diminish together, counter electromotive force divided by ■p current strength is a constant quantity. By Ohm's law _ = R or resistance. Hence we can express the effect of counter elec- tromotive force on a varying current, which calls it into ex- istence, by a constant resistance equivalent thereto in its action on such varying current. This resistance can be expressed in ohms, and is called re- actance. 336 ELECTRICIANS' HA^WY BOOK. By Ohm's law R ===_. Therefore, R =_!lzl_ where n is taken i 7li as any multiple whatever. But by the law of self-induction a rate of change in a current will produce a definite counter electromotive force in a specific circuit. If such rate of change be multiplied by a factor, which may be indicated by n, the electromotive force in- duced by it will also be increased in precisely the same ratio, or n times. Calling the rate of increase of current I or nl as the case may be, the induced electromotive force will be E and nE respec- tively, and . — = - — __ =z R. Reactance is therefore expressible in ohms. ^ '"■*■ Ohmic Equivalent of Reactance of Inductance —This numer- ical quantity depends on two factors. One is the inductance in henries of the circuit, and the other is the frequency of the alter- nations of the current. Calling inductance in henries L and fre- quency /, we have the expression for the value of inductance re- actance, A, in ohms: A =: 2 Ttfl^ or 6.28318 fh The numerical factor 6.28318 is 2 tt. Inductance Reactance in Subdivided Conductor. — The induc- tance of a copper wire varies very little for variations in its di- ameter. In round numbers a wire of 167,000 circular mils cross section has 80 per cent of the inductance of one of 42,000 circular mils and 70 per cent of the inductance of one of 6,500 circular mils. The resistance in these three wires would be approximately in the ratio of 4 : 16 : 100, the inductance as 70 : 80 : 100. The resistance varying inversely with the circular mils, increases in a much greater ratio than the inductance, and the discrepancy increases in more rapid ratio as the conductors are reduced in size. With sufficient subdivision the ohmic resistance would in- cea=e in so rapid a ratio, that the inductance could be taken as constant without any considerable error. Assume that inductance is unchanged by reducing the size of the conductor, and that we have a conductor of 1 ohm resistance, and at the given frequency of alternation possessing inductance wnose effect is a reactance of 5 ohms. Assume that a current of 1 ampere is to be maintained. The grnphic solution is first given. THE ALTERNATING CURRENT. 337 The perpendicular line. Fig. 234, is divided for ohms of resistance, the horizontal line for reactance, which is the ohmic equivalent of self-induction. The diagonal A indicates the impedance. Suppose we substitute four wires of the same aggregate section, then each wire will have a resistance of 4 ohms, and by our assumption the same inductance and consequent ohmic equivalent. The impedance of a single con- ductor will be shown by the line F. But this single con- ductor carries only one-fourth the current, or one-quarter am- 1 2 3 4 5 Fig. J234.— Reactance in subdivided Conductor. pere, because there are four of them. Its length represents the total impedance of one of the new lines, and evidently is not four times as long as A, but is but a small fraction greater. There- fore one of the new lines with an Impedance of 6.04 ohms (for 6.04 is the length of R) has only one-fourth the current to carry that the original thick wire of impedance 5.1 ohms had to carry. The voltage drop in the thick line is by Ohm's law; 5.1 X 1 = 5.1 volts. The voltage drop in one of the thin lines is: 6.04 X 14 = 1.5 volts. But as the four thin lines are in parallel, there will be the same drop in each, or the subdivided main will carry the same current as the thick solid one at less than one-third the drop in potential. The assumption made that self-inductance is the same for all 338 ELECTRICIAXS' HANDY BOOK. wires is incorrect, but the increase is so slow that the principle is correctly illustrated. If accurately calculated, the result will be a little less favorable to the bubdivided line. Capacity.— This is the third quality which may exist in an alternating-current circuit. It is the last of the three qualities spoken of on page 331. Its action upon alternating current is the reverse of that of inductance, as it reduces resistance and gives lead to the current. It is indicated by such diagrams as Fig. 235; inductance by such as Fig. 236; non-inductive resistance by such as Fig. 237. Reactance of Capacity — If a condenser is connected in a cir- FiG. 235.— Symbol, of Capacitt. -6666666^)6^^ Fig. 236.— Symbol of Inductance. Pig. 237.— STMBOii of Non-Inductive Resistance. cuit, it will open or break the circuit as far as a direct current is concerned. No current would pass, and the circuit would be blocked as effectually as if the wire were cut. But the circuit with a condenser in it is a closed circuit for an alternating cur- rent. Electricity may be said to be poured into it at one period and out of it at another, so that the alternating action is kept up as if it were a closed circuit. Just as resistance and inductance have each a twofold effect in an alternating current circuit, one upon the current intensity and the other on the phase relation of alternating current and electromotive force, so has capacity. Capacity Increases current or reduces the resistance, or increases the conductivity of a cir- THE ALTERNATING CURRENT. 339 cuit for alternating currents, and acts to give the current a lead over the electromotive force. Its action is exactly the reverse of that of inductance. Infinite inductance would reduce an alter- nating current to zero, while increase of capacity would diminish the reactance of a circuit so that an alternating current in it would be of increased strength. Ohmlc Equivalent of Reactance of Capacity. — It is best to use the farad as the unit of capacity in the reactance formula. Capacity appears in the formula as the denominator of a fraction, so that capacity reactance would become zero if capacity became infinite. The formula is in form the reciprocal of the induc- tance reactance formula, page 336, with farads, indicated by K, substituted for henries. Calling farads of capacity K, and capacity reactance B, we have: 1 1 R = , or '^ 7t ji^ 6.2b318/K This gives the value of capacity reactance in ohms for a cur- rent of frequency f with a capacity of K farads in its circuit. Impedance indicates the impeding effect exercised upon an al- ternating current by the combined ohmic resistance and reactances of the circuit through which it passes. A circuit always contains resistance, and may contain capacity and inductance. If it con- tains two or three of these qualities, the ohmic resistance it of- fers to the passage of an alternating current is made up of the combined effect of resistance and reactance; the latter may be of one or of both kinds. The combined effect is not due to simple addition because induction reactance and capacity reactance are opposed to each other, and each is in quadrature with resistance. Calling resistance R, inductance reactance A, and capacity re- actance B, we have as the value of impedance: Impedance = V R' + (A — B)- If there is no inductance reactance, then A = 0, and the above by regular algebraic process reduces to VR- + B-. If there is no capacity reactance, then B = 0, and it reduces to VR"-*'4-~Ar Electric Resonance. — This term is applied to the condition that obtains in a circuit when the inductance reactance, expressed 340 ELECTRICIANS' HANDY BOOK. by 2;rfL, and the capacity reactance, expressed by ^ are 2 TtfK equal to each other. The formula for the impedance of a circuit containing resistance inductance and capacity is: Impedance ==^H' + (2 tt / L — y^~^) If 2 ;r f L r^- then the formula reduces to: Impedance = ^/I^' or R- Electrical resonance in other words causes an alternating-cur- rent circuit to act as if it had only true ohmic resistance. But its capacity and inductance have not been annihilated, but only put into opposition with each other, and this brings about re- sonance. By Ohm's law we have E =: R I. For the inductance of a cir- cuit the Ohmic equivalent of reactance must be substituted in the above formula. Then the value of I is determined by Ohm's law as if there were neither inductance nor reactance, and with that value of I, and substituting for R the ohmic value of induc- tance reactance or capacity reactance, the value of E for the in- ductance element and capacity element of the system are reached. Suppose that in a system fed by alternating current there is a condenser of 50 microfarads capacity (0.00005 farad) and that there is an inductance of 0.050 henry. Take the frequency at 100 and the effective E.M.F. as 100 volts. The inductance react- ance is 2 ;r X 100 X 0.050 =: 31.4; the capacity reactance is ^ =r. — ^ = 31.8. Take the resistance at 2 2 ;r X 100 X (^.00005 U > yi4 ohms. Then for the total impedance we have: Impedance ^2^-{- (31.4 — 31.8)' which is practically 2 ohms. By Ohm's law a current will flow through such a circuit expressed by: — or = 50 amperes. R 2 Neither of the reactances has been annihilated; they simply counteract each other's effects, but each acts individually the same as ever. The inductance reactance remains at 32 ohms nearly. Through it a current of 50 ohms has to pass. Therefore by Ohm's THE ALTERNATING CURRENT. 341 law, E = I R, we have E = 50 X 32 or 1600 vo-lts as the electro- motive force between the terminals of the coil embodying the inductance of the system. For the reactance an identical figure is obtained. Thus by resonance an original 100-volt electromotive force can generate in parts of a circuit a voltage many times greater. Fig. 238 shows the diagram of a portion of a circuit containing inductance and capacity. All the figures in the above calculation are approximate, deci- mals being omitted or restricted. The equation 2 7t f L=: i may be considered as expressing the condition of resonancy. It follows that if f and L are known, the value of K which gives resonancy can be calculated, and that if / and K are known, the corresponding value of L can be calcu- lated. This is done by the ordinary operations of algebra. The equation tells that if L is large K must be small, and vice versa, in order to bring about resonance. In a circuit in which electrical resonance exists the entire cir- cuit is not affected by it, but only the portions containing induct- ance and capacity. The circuit as a whole passes the current subject to Ohm's law, while the portion containing inductance and the other portion containing capacity work in concert with each other, and if in tune, as it is sometimes expressed, are the seats of high electromotive force. Damage to apparatus some- times ensues from this. While resonance eliminates the effects of the inductance and the capacity upon the circuit taken as a whole, it leaves each one unaffected in its action. The inductance still has its value in henries, the capacity still has its value in farads, and they retain their individual characteristics and power of reaction. Causes of Lag and Lead. — The effect of inductance in a circuit is to cause alternating current to lag behind the impressed alter- nating electromotive force which produces it. The lag, if the circuit possessed neither capacity nor resistance, would attain its highest possible value, which is 90° or quadrature. Induct- ance is the cause of lag. The effect of resistance in a circuit is to cause alternating cur- rent to tend to be in phase with the impressed electromotive 342 ELECTRICIANS' HANDY BOOK, force. If a circuit possessed neither capacity nor inductance, the impressed electromotive force and current would be in perfect phase with each other. The effect of capacity in a circuit is to cause the current to lead impressed alternating electromotive force. This lead, if the circuit possessed neither inductance nor resistance, would attain its highest possible value, which, as in the case of lag above cited, is 90°, or quadrature. Capacity is the cause of lead. It follows that resistance acts in quadrature with inductance and capacity, and that capacity acts in direct opposition to inductance. Summation of Alterna'ing Quantities. — The combined effect of quantities acting additively in alternating manner, so that their alternations may be represented by a sine curve, cannot ^MM> Fig. 238.— Capacity and Ixductance in a Circuit. always be expressed by simple addition. Suppose a sine wave one inch high represents the action of a certain alternating current. Next suppose that a second current is poured into the line, coinciding in phase, intensity, and form with the first. A wave of twice the height would result. The combined effects of the two currents could be expressed numerically by adding them together. It will be understood that whatever is said of current hei-e applies also to electromotive force. Current and electromotive force alternate in exactly the same manner, and either can have its action represented by a curve of the sine type. Suppose now that the currents differed 180° in phase, as shown in Fig. 233. One would be positive in its alternation when the other was negative, one would exactly counteract the other, and the result would be zero. Suppose that the phases of the two currents differ 90° in phase. In some parts of their 'Jiycles they THE ALTERNATING CURRENT. 34a co-operate, in others they resist each other, and a more compli- cated curve is the result. The values of two sine curves can be added together by draw- ing them and constructing a new curve. Its height at any point is determined by adding algebraically the heights of the original curves at that point. Distances below the base line are treated as negative. Fig. 239 shows another method. A F and A F' are the generating circles of two sine waves whose phase difference is the angle between A B and A B'. The two vectors are compounded as in Fig. 239, and with the new vector A B" a new generating circle is produced, from which the new curve is generated. The curve x.^ Fig. 239.— Summation of Sine Curves. I is generated from circle F B D E, curve II from circle F' . . . , and the curve III is produced by adding I and II or by directly generating it from the generating circle B"... Composition of Resistance, Inductance, and Capacity. — Every circuit possesses these three qualities. Their combined effect may be found by a simple diagram, although where accuracy is required, mathematical calculations are essential. Let a horizontal line be drawn starting at an origin and of length to represent the ohmic resistance of a circuit. The react- ance of inductance will be represented by a line at right angles to it, because the two are in quadrature. Draw the inductance line vertical and rising from the origin, and of length to repre- sent inductance reactance. The reactance of capacity is 180° removed from that of inductance, and hence is also in quadrature 344 ELECTRICIANS' HANDY BOOK. with resistance. The line representing it will start from the origin and extend vertically downward. There is nothing absolute -about the position of these lines, except that they must be related to each other as shown. Each line or radius vector must be drawn of length to give the relative value in ohms of the reactance it represents. In the diagram, Fig. 239a, the line O R represents resistance, O I represents inductance, and K capacity. Draw I D parallel to R and equal to it in length. The diagonal from O to D or O D is the resultant of combined effect of resistance and induct- FiG. 239a.— COMPOSTTTON OF "Resistance, Inductance, and Capacity. ance reactance. Then from D draw a line D P parallel to O K and equal to it in length. The diagonal O F will give a line ex- pressing the combined effect of the two reactances and of resist- ance. Multiplication of Alternating Quantities. — The multiplica- tion of alternating quantities has to be done to find the power of a circuit in watts, because the latter unit is a product of electro- motive force by current, a volt by an ampere. The term volt- ampere signifies a watt. If an alternating electromotive force is multiplied by an alter- nating current, a product differing altogether in form and value from the additive or compounded result is obtained. This THE ALTERNATING CURRENT. 345 product is a curve of power, of volt-amperes or of watts. Its amplitudes indicate quantities of watts at the different periods. Power Curves.— The diagrams, Figs. 240 and 241, show each a curve of sines of electromotive force in full line, and one of cur- rent in dots. Multiplying their amplitudes together, new ampli- tudes are obtained, which give the curve drawn in dot and dash, which is the volt-ampere curve, curve of watts or power curve. When the current curve or the electromotive force curve crosses r\ /'~\ Figs. 240 and 241.— Multiplication of Alternating Quantities. the base line, its amplitude is zero. Therefore the amplitude of the power curve at this point must also be zero, because the product of a finite quantity, in this case the amplitude of the other curve at the same point, multiplied by zero is equal to zero. As the electromotive force is, in the case shown in the cut, out of phase with the current, for each cycle or period there are four zero factors. This brings the power curve twice as often to the base line as either of the original curves. It has twice the alternations of either of them. The system receives energy from the alternator during the time the power curve is above the base line. It receives energy in varying amounts whose measure is the amplitude of the curve at that point. It returns energy to the alternator when below 346 ELECTRICIANS' HANDY BOOK. the line, measured as before for any instant by the amplitude of the curve at that point. The second of these diagrams. Fig. 241, shows the electro- motive force curve and current curve in quadrature witli each other. This condition would be brought about by presence of inductance and absence of resistance in the circuit. The power curve in this case is half above and half below the line. It has twice the alternations of either of the other curves, just as before. It indicates the return of exactly as much energy as is received. In such a case no energy is expended on the line. It is the case of the wattless current. The portions of the curve above the line represent power received; those below represent power returned. The two are equal, and as they are opposed in action, the com- bined result is zero. Referring again to Fig. 227, the current and electromotive force curves are divided by vertical lines crossing the zero line at the intersections of the curves. Within the space I, E. M. F. ordi- nates above the line, which are positive, are multiplied by current ordinates below it, which are negative. The result is negative by algebra. Therefore the new curve for this division is below the line and negative. Within the space II, positive E. M. F. ordinates are multiplied by positive current ordinates, giving a positive curve or one above the line. Within the space III, positive cur- rent ordinates are multiplied by negative E. M. F. ordinates, bringing the combined curve below the line. Within the space IV, both current and E. M. F. ordinates are negative. But by algebra negative multiplied by negative gives a positive quantity; therefore the combined curve is above the line here. This result is shown in Fig. 240. If E. M. F. and current curves are in phase with each other, all the multiplications are either of positive by positive or negative by negative, so that the new power curve is all on the positive side of the zero line. Two=Phase Current. — If two electromotive forces invariably in quadrature with each other are simultaneously produced by a generator, the currents produced may be distributed over four conductors, a pair for each current. The combination is called a two-phase current. It is illustrated in diagram. Fig. 242. The THE ALTERNATING CURRENT. 347 full line A. . . is one current, the dotted line B. . . is the other; the 00° distances are marked on the diagram. Three-Phase Current. — If three electromotive forces invari- A .B aI "^*\ /' ^ 90---->r<----90-y^-->i ^ ■ X> 1 / X ,\ / I X ' \ X / , \ / 1 / • ' \ 1 / 1 /'' 1 \ . ■ N / i' V ' ^ \ 1 J '\ ' \ 1 / 1 ^ 1 / / \ / 1 / \ 1 X , / ' \ ' \ X / ' / 1 / 1^* *- 90 ->i^--.90\----> JK. / y .^^ V ^ ^^••J ^^ "*•. * A B Fig. 242.— Two- Phase Current. ably 120° apart in phase are impressed on a six-wire circuit, what is called a three-phase current results. By special connections this current can be distributed by means of three or of four A -?. c A "^*V, ^^^ *'"T*'*N \ # y^* '\, ^^ \ / 1 \/r ^ /^ > V i A 1 X / i \i / \ / \ / i \ / ;'- / \ / ■*- \ 1 / ^ hV / \ / / --1-201 1 \ V' -— 120°|-- /- -1-20- — V \ \ 1 > /\ vj/ /\ *»■-* / Fig. 243.— Th REE-Phase Current. vrires. The diagram, Fig. 243, illustrates it. The full line A..., dotted line B..., and dot and dash lin^ C... are curves of the three currents which really make up the so-called three-phase current. CHAPTER XX. ALTERNATING CURRENT GENERATORS. Generation of Alternating Current. — Alternating current is generated in dynamo-electric generators, which represent one of the simplest or fundamental cases of im^pressment of electro- motive force. The direct current dynamo is a step in the direc- tion of complication, as the alternating current dynamo with its simple collecting rings taking the current as it is generated is the simplest of all mechanical generators. For this reason some authors treat of alternating generators first and then of direct cur- rent generators. Some even make a discussion of the alternating current lead up to the direct current. Single=Phase Armature.— If a coil of wire with disconnected ends is rotated in a magnetic field with its axis of rotation sym- metrically placed as regards the lines of force, it will have im- pressed upon it at each revolution two electromotive forces of op- posite polarity. Its position may be to a considerable degree un- symmetrical as regards the field of force, yet the same will be true. Electromotive force to be utilized must produce a current. To utilize these pulsations, the ends of the open-circuit coil must be connected by an outer circuit. Upon the shaft carrying the coil are secured two copper rings insulated from one another and from the shaft. One terminal of the rotating coil is connected to one ring, and the other terminal to the other ring. A pair of brushes bear against the rings, one brush for each ring, and to these brushes the terminals of the outer circuit are connected. One terminal is connected to each brush, and the brushes are insulated from the frame of the machine. An iron core is placed within the coil, as in the direct-current dynamo, so that an arma- ture is constituted. The electromotive force produced by the rotation of the coil is 348 ALTERNATING CURRENT GENERATORS, 349 proportionate to the rate of change of the number of lines of force interlinked with the circuit by means of the coil. The elec- tromotive force passes from a maximum of one polarity to a value of zero, then to a maximum of the other polarity, back to zero, and then to its original polarity. This varying electro- motive force tends to produce upon the closed circuit a current varying in like manner as regards intensity, and reversing in direction as the polarity of the electromotive force changes. The reversing in direction of current must occur exactly as often as the reversing in polarity of the electromotive force, but lag or Fig. 244.— Element aby One-Phase Alteknator. Fig. 245.— Multipolar Statok. lead will generally operate to prevent the two being simultan- eous, as they would be if there were no inductance or capacity in the circuit. The diagram of such a dynamo is given in Fig. 244. It is the simplest possible representation of an alternating-current gener- ator. The object and function of an alternating-current generator are to impress alternating electromotive force upon a circuit. What disposition is made of that electromotive force, whether it is made to produce a corresponding alternating current or not, and if it procV- s one how near it is to be to that which should be exacted I - n's law — all these are questions outside of the operation of namo, except as regards its unvarying factors, such as cai:r iductance, and ohmic resistance. 350 ELECTRICIANS' HANDY BOOK. Such an armature would produce a single alternating current on a closed circuit, and such a current is called a single-phase current. flultipolar Construction. — The alternating current as used in modern engineering practice must be of high frequency. A com- plete cycle would, with the bipolar construction just described, require a complete revolution. To give high frequency the arma- ture would have to be rotated at very high speed. If the poles are increased in number, a single rotation will give more cycles; in typical constructions one cycle is given per revolution for each pair of poles. If the field contains four poles, there will be two cycles per rotation; if it contains six poles, there will be three cycles, and so on. By increasing the number of poles a given frequency is obtained with fewer rotations of the armature per second. For this reason alternate-current generators generally have a number of poles, bipolar construction not being much used. Generators with more than two field poles are called multi- polar generators, and a multipolar stator is shown in Fig. 245. Gi o iping of Windings — The windings in alternators are gen- erally referable to groupings. The active conductors may gener- ally be assigned to groups, each group of approximately the width occupied by the face of a field pole, and there being a group of conductors for each pole. In a general way multipolar construc- tion by filling the circle of the armature with pole faces tends to make the distribution of conductors on the armature periphery even, but grouping can always be traced out for them. Principle of Alternate- Current Armature Winding. — This principle is that all active portions of the winding of an individ- ual armature winding must coincide in action at each instant, all co-operating to produce the same effect on the circuit. The direct-current armature winding with its commutator works in parallel of two for each pair of poles, while the alternating-cur- rent armature winding operates in series whatever is the number of field poles. The active conductors of an alternate current winding must be so joined that at any instant an electromotive force of uniform polarity shall be impressed upon all of them. Drum Armature Connection^. — Suppose any number of pairs of north and south field poles arranged symmetrically around a ( ALTERNATING CURRENT GENERATORS. 351 IS) dnim-armatur-e core, mounted in bearings. Let the cylindrical surface of the core have conductors Insulated from each other secured to it. If rotated in the multipolar field, each conductor will have alternating impulses of electromotive force impressed upon it, as many in one revolution as there are poles in the field, and changing in polarization or ''direction", in one revolution also as many times as the number of poles. Elementary Four=Pole Single=Pha5e Armature. — In Fig. 246 a core is shown upon whose surface four conductors are placed. If rotated in a four-pole field, electromotive force will be im- pressed upon them of opposite polarity for every alternate con- ductor. The ends of these conductors are to be connected by wires or other conductors extending across the front and rear ends of the armature core. The ar- rowheads indicate the polar- ity of the electromotive force, or the direction of current which it tends to produce in each conductor, at the Instant indicated. To have these currents coincide in direction for the entire winding, the front end of one conductor must be con- nected to the front end of a conductor one pole re- m.oved from it; the rear end of this one connected to the rear end of a conductor one pole re- moved from it in the same direction, and the same system is car- ried out all around the circle. This brings the two ends of the windings out together. One end is connected to one collecting ring, the other is connected to the second collecting ring. Such a winding will give a single-phase alternating current. The action may be described as a zigzag action. The terminals of the wind- ing are subjected to the accumulated electromotive force im- pressed on the active conductors by the four poles of the field. Single-Phase Wave and Lap Winding.— An example of wave winding is shown in Fig. 247 in development. It will be seen Fig. 246. - Elementary One-Phase Drum Armature Wound foii A Pour-Pole Field. 352 ELECTRICIANS' HANDY BOOK. / / tliat the electromotive force impressed on each active conductor of the armature co-operates to produce a current in one direction ALTERNATING CURRENT GENERATORS, 353 all through the windings. The active wire^ are spaced in accord- ance with the distance from pole to pole. In direct-current wind- ing the spacing is usually a little more or a little less than this distance. The next cut. Fig. 248, shows in development a single-phase lap winding. The spacing is regulated as in wave winding by the distance from pole to pole, and a uniform impressing of electro- motive force on all the active conductors is produced. If we trace the course of the conductors in two successive loops in the direct-current lap winding, we shall find our course a series of left-handed or right-handed turns as the case may be, Fig. 249.— Analysis of Direct Current Lap Winding, Fig. 250.— Analysis of Alternating Current Lap Winding. but either left-handed or right-handed all the way around. If we start in the lines of Fig. 143 at the left hand and follow the line beginning at the left, we shall progress in a sort of spiral toward the right hand, always in the same sense. In this particu- lar case it will be against the movement of the hands of a watch. If the alternate current lap winding. Fig. 248, is traced out through its loops, we shall progress with the hands of a clock in one loop and against the hands of a clock in the next loop all the way around. The courses followed can be roughly shown, as in Figs. 249 and 250. Ring Winding for Alternating Current. — In Fig. 251 is shown how a Gramme ring armature can be made to give an alternating 354 ELECTRICIANS' HAXDY BOOK. current. For eacb. pole of the field a single lead is taken from equidistant parts of the windings. Every second connection is taken to one collecting ring, and the others to the other collecting ring. Such an armature rotated in a multipolar fi-eld whose number of evenly-spaced poles is equal to the ring connections will develop a single-phase alternating current. The connec- tions operate to divide the w^indings into groups, one for each pol?. ConventiDna! Representation of Collecting Rings. — In Fig. Fig 2r)l.— Gramme Ring Connected iOR LlNGLE-PHxiSE ALTERNATI^Q CUIiKE^T. Fig. 253.— Bipolar Singi.e-Phase Alternating Curhent Pole Geneiiator. 251 we see the conventional way of representing collecting rings in diagrams. Two circles are drawn from the same center. One is of greater diameter than the o'her, and each represents a ring. These rings are r-eally of identical size, but are conven- tionally represented as of different size, in order to distinguish betvv^een them. Pole Single- Phase Armature. — Armatures for alternating cur- rents are sometim-es of the projecting-pole type. Foles project radially from them, and are wound in the same sense as the poles of the field. A direct current passed through the windings of such an armature would cause one projecting pole to be of ALTERNATING CURRENT GENERATORS. 355 north polarity and the next one to be of south polarity. One consecutive winding goes around all the poles in succession, and the induced single-phase current is taken from its t-erminals. In Fig. 252 direct current from an outside source may pass through the windings of the poles attached to the frame. The armature rotated in the field thus formed delivers current to the circuit connected to the collecting rings. The reverse may be carried out. Direct current may be supplied to the central poles by connections to the brushes. If the central part is rotated, al- ternating current can be taken from the windings of the frame poles. This construc- tion would give one cycle per revolution. In Fig. 253 the same system is indicated for four poles. This construction would give two cycles per revolution. Rotor and Stator.— In both the examples shown in the diagram, Figs. 252 and 253, electromotive force could be impressed on either the sta- tionary or rotary member of the machine. Whichever part it is impressed on is the arm- ature. The approved terminology for alternators calls the part v/hich turns the rotor, whether it is a revolving field or armature; and calls the part which does not turn the stator, whether It is a stationary armature or field. Yet the distinction of field and armature remains. The alternator can always show two parts, one the field through which a direct current is passed, the other the armature on whose windings alternating electromotive force is impressed. Either may be rotor. The general rule for single-phase windings is that armature and field are interchangeable. If a direct current is supplied to the field, electromotive force will be impressed on the arma- ture as the rotor turns. Again, if the armature be supplied Tig. 2">3.- Four-Pole Single-Phasij Alternating Current Generator. 356 j^'LECTRICIANS' HANDY BOOK. with direct current, the fi-eld will have electromotive force im- pressed upon its windings. There is nothing practical in this, because the armature and field are generally wound with widely- different sizes and lengths of wires. One is wound to have good excitation from the source of direct current. The other is wound to give alternating electromotive force of the desired numb-er of volts, under the effects of the field. It is obvious that the wind- ings are apt to be widely different. Inductor Alternator. — The principles of this type of alternator are shown in Fig. 254. The stator is both armature and field. The full line indicates the field winding through which a direct Fig. 254.— Inductor Alternator. current is maintained. It is wound around every second pole, so as to excite north and south polarity in them alternately. Thus, taking a north pole as a starting point, the one next to it would be without polarity, because the field windings do not go around it. The next pole would be a south pole, owing to the direction of the -winding. This succession is kept up all the way around. A segment only is shown in the cut. The full line indicates the field winding, and the field poles are marked N and S. The neu- tral poles between the field poles are marked A A, and the arma- ture winding on them is shown by the dotted line. It is wound in the reverse sense on neighboring poles. The rotor carries heavy masses of soft iron B B, called induc- tors, each one wide enough to cover two poles on the stator and the interval between them. As the rotor turns, it changes the ALTERNATING CURRENT GENEKj^TORS. 357 polarity of the neutral poles. Thus in the position shown, tne left-hand neutral pole, acted on by the inductor extending from it to the south pole on its right, is polarized with north polarity. The right-hand inductor polarizes the right-hand neutral pole above it with south polarity. When the rotor turns through an angular distance equal to one pole face and one pole interval, the opposite polarities are imparted to the neutral poles. In each revolution of the rotor the armature poles vary in polarity as many times as there are poles' in the stator. The great advantage of this type of machine is that the wind- ings are stationary. A ma- chine with rapidly-rotating rotor carrying windings with it is not so solid a construc- tion as one in which the care- fully insulated vv^indings are on a motionless part of the machine. Disk Windings.— This kind of armature has been used extensively in Europe, but not very extensively in this country. In preceding pages ^05. Figs. 203 and ^u5, disk f^^. 255.-Windino of Bisk Arma- dynamos have been shown, turb for Sixgle-Phase md in Fig. 255 the coils and A. C. quRKc.NTs. collecting ring connections of a disk armature are shown. The arrows indicate the directon of the current. This direction changes during a rotation six times, because the armature is wound for a six-pole field. The disk armature does not need an iron core; it is so thin that the lines of force readily strike across it from pole to poiP. Two=Phase Winding.— Suppose it was desired to send out two independent currents of equal periodicity, but differing as re- gards the phases of the electromotive force producing them, one electromotive force to be 90° behind the other. Two inde- pendent machines could be mechanically coupled. This would have to be so effected that the proper phase relation would obtain, 358 ELECTRICIANS' HAXDY BOOK. which would involve setting the armatures so that one would Lo one halt of a pole interval behind the other. The currents could be distributed on four lines of wire, two to each machine. The phase relation existing between the electromotive forces on the two circuits being invariable, the result would be called a two- phase current. An easier way to produce it is to have a second independent winding on the same armature. A single machine then produces the two-phase current. The cut, Fig. 256, shows the principle. The conductors A A Fig. 256.— Two-Phase Winding. are parts of a continuous conductor that goes all around the armature in every second groove. The windings B B do the same. The ends of each winding go to their own pair of collecting rings, of which there are four. The diagram shows rotor and stator as straight; in reality, each one is circular, and one lies within the other in the regular way. In the position shown, the windings A A are being acted on, and the current in them is indicated by the dot and cross symbols — the dots indicating current coming toward the observer, and the crosses indicating current going away from him. The conductors B B in the position shown have no electromotive force impressed on them; their sine curve is crossing the zero line. Either part shown can be stator. Usually it would be the arma- ture. Three=Phase Winding. — What has been said of the two-phase ALTERNATING CURRENT GENERATORS. 359 winding may be repeated with slight variation of three-phase winding as shown in Fig. 257. The three windings are designated by A, B, and C, and as no winding is in a neutral position, dots and crosses are put on all. The three represent three indepen- dent windings, and may deliver current to six collecting rings, a pair for each winding. Corresponding parts of adjacent windings are distant from each other two-thirds of a pole interval or one-third of two poles, which for a bipolar machine would be 120°. This fixes the condition 6 .A / N p. Z 5 _ik 1 N ) ( s \ ( ) I ) ( > z^> 1 Fig. 257.— Thref-Phase Winding. that the currents induced shall be 120° different in phase from each other. Six-Wira Connection of Three=Phase Alternator Winding.— The windings of a three-phase alternator may be variously con- nected. They may be treated as if they were windings of three separate machines, in which case two conductors would be as- signed to each of the three outer circuits which they could sup- ply. This would give a total of six conductors to be led through the district. Almost always other systems are used, which enable the distribution to be effected with three or four wires. A six- ring collector system is shown in Fig. 258. Y or Star Connection.— This connection requires four wires to distribute the power from a three-phase alternator — three active and one neutral wires. The latter passes current when the bal- ance is disturbed, exactly like the neutral wire in the three-wire 360 ELECTRICIANS' HANDY BOOK. system of parallel distribution. The connections are made in the machine and on the outer circuit. The three windings of a three-phase alternator can be taken as beginning at three adjacent points on the armature. From these points collector-ring connections would be made were the six-wire system in use. For the Y system three of these ends, symmetrically distributed with reference to each other, are con- nected together, and one lead is taken from them through the district, which is the neutral lead. From each of the other ends of the three windings a lead is taken, thus giving a total of four leads. In the utilizing of the four mains, each lamp or other appli- ance is connected from one of the active wires to the neutral wire. The balance is kept as true as possible by taking the same amount of power from each act- ive lead. If exactly the same amount is taken, the neutral wire carries no current. The development of a Y wind- ing is shown in Fig. 259. There are three windings, A, B, and C. The A winding begins at A^ and ends at A^; the B winding begins at Bi and ends at B^; the C winding begins at Bi and ends at B^. The six ends which might be connected to six independent line wires are A, B, C, A^, B^, Ei,nd C4. For the Y connection each second end is connected to the neutral wire. In the development these alternate ends are A^, C,, and B4. The remaining ends A^, C4, B^ are connected each to one of the active wires. If the course of the current is examined by the rule given on page 210, and carried out in Fig. 259 by the ar- rowheads,. it will be seen that a strong downward current in Ai is balanced by weaker upward currents in B^ and Ci. The relative strength of the currents is due to the strength of the field through which they are moving, and A^ is evidently in a stronger field than either B4 or Ci. A strong current goes upward to the Fig. 258.— Six-Ring Collector FOR Alternator. ^ALTERNATING CURRENT GENERATORS. 361 upper line, which line indicates a collector ring with brush A, while weaker currents go down from the other collector rings. Delta or flesh Connection.— Taking the three windings as be- fore, a first and last end can be assigned to each. Thus in tne Fig. ^59.— Development of T Connections. Y connection it may be taken that the three interconnected ends joined to the neutral wire are first ends, and the three ends with separate conductors are the last ends. For delta connection the three windings are joined in series. The last end of one winding is joined to the first end of the winding next to it in phase (120° c c Figs. 230, 261, and 262.— Y, Delta, and Combination Connections. removed in phase). The last end of this second winding is joined to the first end of the third winding, and the last end of the third winding is joined to the first end of the first wind- ing. Prom each junction of first and last ends a wire is led through the district which is to be supplied, a total of three wires, there being no neutral wire. 362 ELECTRICIAXS' HAXDY BOOK, Line Connections. — The appliances on the line may be con- nected from wire to wire, so as to maintain the delta distribution over the working circuit, or the Y system may be used with a delta system at its junction, so as to dispense with a neutral wire. The Y connection is shown in Fig. 260, the delta connection in Fig. 261, and the combined Y and delta connection in Fig. 262. Neutral Wire in the Y 5yslem. — This wire is usually treated as a part of the system requisite to its operation. It can be sup- pressed if the appliances on the three divisions are evenly bal- anced, the case being precisely analogous to that of the neutral wire in the three-wire system of distribution, except that it is a case of one neutral wire for three active wires and not of one neutral for two active wires. It is obviously impossible to secure such distribution in ordinary practice, so that naturally the fourth wire has come to be regarded as a necessary part of the connections. An interesting illustration of the properties of the Y connection has been made by causing three carbons to take the place of the three limbs of the Y and producing an arc at the junction, the carbons being drawn apart, as to cause a triple arc to strike. It was maintained without any neutral wire. Another experiment was the lighting of a triple-filament lamp, three lead- ing-in wires connecting to filaments joined at their ends Y fashion. This lamp was ignited without any return wire. CHAPTER XXI. ALTERNATING CURRENT MOTORS. The Induction flotor, — This is a motor whose action depends upon the induction of electro-magnetic polarity in an armature wound with a re-entrant coil, or with a coil whose members are connected in parallel. The coil must not be an open one. The alternating current in the field induces currents in the windings, which induced currents produce polarity in the core of the arma- ture. The polarity of the armature being due only to induction, gives its name to the motor. The Rotary Field.— The produc- tion of the rotary field is the prin- cipal reason for the generation of polyphase currents. By means of this invisible transferring of mag- netic polarity around a circle, one principal type of the alternating-cur- rent motor is operated. The cut. Fig. 263, shows four coils of wire. Let the coils B B receive an alternating current, while the coils A A receive another cur- rent in quadrature with the first. The result will be that when the current in B B is at its maximum, the current A A., will be of zero intensity. Then as the current in B.. decreases, that in A... will increase. When the B current is at its maximum, north and south magnet poles will be established on a horizontal axis passing through the center of the B coils. The A coils when active will establish poles on an axis perpendicular thereto. Poles at intermediate points will be established when current is pass- ing through all four coils. The result of the arrangement is that 363 B B riG.;563.— KoTARY Field Coils. 364 ELECTRICIANS' HANDY BOOK, a north and south pole are kept traveling around the circle by the action of the alternating currents in quadrature with each other. Such currents constitute a two-phase alternating current. The change of one current from one maximum to the other takes place perhaps one hundred times in a second. Hence the resultant poles of the field whirl around it w^ith great rapidity. The first Niagara alternators give a two-phase current with twen- ty-five periods in a second. These produce a rotating field that has fifteen hundred rotations per minute. Three such coils of wire with a three-phase current would give a rotary field. riagnetic Needle in a Rotating Field. — A compass needle piv- oted in the rotating field with its axis of suspension coincid- ing with the axis of rotation of the field would whirl around v/ith the speed of the field once it was started. Such an arrangement would not be an induction motor. An induction motor is one in which the rotating field in- duces currents in the arma- ture, and under the com- bined effect of the field and armature excitation the arma- ture revolves. Armature in a Rotary Field.— If instead of a magnetic needle a cylindrical laminated armature core wound with a re-entrant coil as shown in Fig. 264 is mounted on bearings in the field, it will rotate. This it will do because the alternating cur- rents will induce currents in its wires. This they do directly by their rotary field of force. This whirls around, and thus its lines of force are cut by the windings of the armature core, which cutting induces a current in them, producing north and south poles in the core. The core with its windings is mounted in journals and rotates as did the magnetized needle, but with a very important distinction. To establish in the core the polarity Fig. 264.— Two- Phase Rotating Field AND Armature. ALTERNATING CURRENT MOTORS, 365 described above, lines of force have to be cut by its windings. Therefore it drops behind in its revolutions, and turns from one to five per cent, ten per cent in small motors at full load, slower than does the rotary field. If it by any means was made to synchronize with the field, it would have no induced polarity such as described, and no pull or torque would be exerted upon it. Therefore it constantly falls behind. The aanount of this falling behind is called its slip. The generation of a three-phase current and the operation by it of an induction motor are shown in diagram in Fig. 265. By R GENERATOR Fig. 265.— Three-Phase Generator and Inductiox "Motor. following the figures it will be seen that the stator of the motor receives the identical currents induced in the stator of the gen- erator; but the poles, of the generator stator travel around it. Consequently, a rotary field is produced in the stator of the motor. Three=Phase Induct'on Motor.— The diagram, Fig. 266, repre- sents a four-pole three-phase generator driving such a motor. The generator has twelve armature coils, three sets marked ABC for each field pole, giving a three-phase current. They are con- nected in Y combination. The left-hand diagram represents the generator. The field is the rotor. The motor, also with twelve 366 ELECTRICIANS' HANDY BOOK, coils, marked as in the motor, and Y-connected, is indicated by the right-hand diagram. The motor and generator are connected by three wires, a, h and c. The fourth wire is omitted because it would have no load to carry. The capital letters on the arma- ture of the g^-nerator enable the course of the windings to be followed. The three-phase current produces a rotary field as the two- phase current does on the same general principle. The lag of the currents behind one another acts to cause the poles resulting from the combined action of the coils to rotate around the field. These poles may be the resultant of two or of three windings; ihey are never due to one only in the three-phase motor. Ftg. 360.— Four-Pole Three-Phase Generator and Inducttok Motor. Induction Motors. — Motors constructed on the above principle are called induction motors. One of the most striking features about them is the fact that the coils on the armature, which is the rotor, are self-contained, have their terminals connected so that the winding is purely re-entrant, and have no outside connection whatever. A General Electric Company induction motor is shown In Fig. 267. Rotary and Revolving Field. — What has been described is the rotary field. In it the rotary action is purely electrical, there is no rotation of any part of the mechanism. A revolving field is another thing — it is a field which turns around an axis like a wheel. It is often used in alternating-current generators. There ALTERNATING CURRENT MOTORS, 367 is danger of confusion in the use of these two terms, and the meaning of each should be grasped, so as to keep the distinction between them. By a simple modification of mechanical structure, a rotary field may be mounted on journals and the armature may be fixed. In such a case the field becomes the rotor, and is really a combined rotary and revolving field. Starting Torque, — Polyphase-current induction motors have a starting torque, which single-phase synchronous motors are desti- FiG. 36T.— Induction Motor wrrn SQijTRREii Cage Armature, tute of. This feature has made polyphase currents the favorite type of alternating currents. Squirrel Cage Armature. — This is a favorite type of armature used on induction motors. It consists of a laminated core, with straight conductors of copper lying in longitudinal grooves or holes as close to its surface as possible. The ends are connected to two rings of copper. The windings thus provided have been aptly compared to a squirrel cage, and the name has been defin- itely adopted for them. A simple form is shown in Fig. 268. Starting Resistances are used to develop starting torque. It 368 ELECTRICIANS' HANDY BOOK. is proved in the analytical discussion of the induction motor that at starting the torque is proportional to the rotor resistance. Resistances are provided for changing the resistance of the rotor windings. In the General Electric Company's form L motor, a wound armature is used with distinct circuits instead of a squir- rel-cage armature. The terminals of the circuits come out in the center of the armature, and are connected to each other through resistance grids. The grids have contact points, and a shoe worked from outside the motor by a lever slides back and forth, so as to cut resistance in or out as required. Many arrangements of starting resistance have been used by different makers. Starting Compensator.— This is also used in starting the in- FiG. SJ68.— Squirrel Cage Armature. duction motor. It is a transformer containing a single coil which takes full line voltage. It has one or more taps, and by connect- ing the motor to one of the taps a reduced voltage is obtained for starting. When the motor reaches nearly full speed, it is thrown from the tap directly into the circuit, so as to get full voltage. The change is effected by a switch working in oil situ- ated in the base of the compensator. The motor to be used with this apparatus has the simple squirrel-cage armature, as there is no change of armature resistance to be brought about, and V. is of simpler construction. It is applicable where the motor is not obliged to start with full load, and where there is no objection to the use of a large starting current. ALTERNATING CURRENT MOTORS. 369 Lenz's Law and the Induction Motor. — Lenz's law applies to this motor. The rotary field as its poles move induces currents in the armature opposing the motion of the fields. This motion while not mechanical has exactly the effect of a mechanical move- ment of the poles. Currents opposing the motion, with their action increased by the iron of the core, cause the armature to rotate exactly in accordance with the law. Construction of Induction Motors.— Laminated cores for field and armature are much used, such as have already been illus- trated previously. The windings of the armature, if of the squir- rel-cage type, are not necessarily insulated from the frame. The motor with starting resistance may give a starting torque 50 per cent greater than the full load running torque with about the same excess of current. They are made of high horse-power as well as of smaller power. The Synchronous flotor. — If two single-phase alternators are connected together in one circuit, one may be driven by power so as to impress alternating electromotive force upon the line with accompanying current. The other alternator receiving cur- rent from the line if it is once started into motion so as to cor- respond with the alternations of the other, will continue moving and be a motor. For each alternation of current an identical alternation is involved in its operation; the two machines work- ing together harmonize exactly in the time of their alternations, and are said to be in synchronism. The motor machine is a synchronous motor. When current and electromotive force generated by its motion harmonize in direction, an electric machine of the dynamo type in which such condition exists is a generator. In other words, such condition can only exist in a system to which power is ap- plied — can only exist in a dynamo whose armature is turned by power. If mechanical energy is expended on an alternator, electromotive force and current harmonizing with each other will be the result. The generation of electromotive force opposed to the current received by a dynamo indicates that that dynamo is a motor — is giving out no electrical energy, but is absorbing it and is giving out mechanical energy. Therefore, if an alternator has its cur- 370 ELECTRICIANS' HANDY BOOK. rent opposed to the electromotive force its motion generates, xt becomes under proper conditions a motor. The synchronous one-phase motor is based on these principles. Condition of Operation. — The cut, Fig. 269, shows two curves, one of electromotive force, and one of current. The current lags. The current and electromotive force oppose each other in the section marked I, are together in II, opposed in III, and to- gether in IV. An alternator producing a current and electro- motive force of these relations would during the periods I and III give out mechanical energy and absorb electric energy and Fig. 269.— Current and Electromotive Forck Curves. be a motor. During the periods II and IV it would absorb mech- anical energy and give out electric energy and be a generator. If the lag was 90°, or in quadrature, the periods I and III would be equal in all ways to II and IV, and the machine as far as its electrical functions were concerned would absorb no mech- anical energy and give out no electrical energy. It would be in a wattless condition. If the lag exceeded 90°, periods I and III would be larger than II and IV, and the machine would absorb electric energy and give off mechanical energy and would become a synchronous motor. ALTERNATING CURRENT MOTORS. 371 There need be no structural difference between the generator of a single-phase alternating current and the synchronous motor driven by it. Whether a machine is one or the other is a ques- tion of phase relation of volts and amperes. If the two identical machines of the single-phase alternating-current type are con- nected electrically, and one is rotated by power as a generator and the other by any means is caused to rotate at the same speed, the latter becomes a motor, and will thereafter rotate at the identical speed of the generator and be driven by it as a synchronous motor. In the generator the electromotive force and current will be nearly in phase with each other. In the Fig. 270.— Single-Phase Generator and Synchronous Motor. motor counter electromotive force will be generated, and will almost exactly oppose the current. In the generator the curves of current and electromotive force will be almost in phase, and in the motor the counter electromotive force will be almost 180° different in phase. Single=Phase Synchronous Motor. — A single-phase generator and motor connected are shown in the diagram, Fig. 270. They are of identical construction. The current generated by the generator is indicated by the heavy arrows. This current causes the rotor of the motor to turn in exact synchronism with the generator. The rotation of the rotor of the motor generates 372 ELECTRICIANS' HANDY BOOK. counter electromotive force. The polarity of this is indicated by the lighter arrows. For synchronous rotation the conditions of phase in the two armatures must be exactly opposite, if one is to be a generator and one a motor. Therefore, torque is not to be looked for until sj^nchronism is attained. For this reason a synchronous single- phase motor has no starting torque, and has to be started in some way until it moves as fast as the generator. After that is done it will go on exercising torque, and absorbing electrical and developing mechanical energy. To start it the current is divided, and one branch by a capacity Fig. 271.— Three-Phase Generator and Synchronous Motor. or inductance is thrown as nearly 90° out of phase with the other as possible. The two leads are then connected to the ma- chine and establish a rotary field, and the synchronous motor is thus converted into an induction motor. It is speeded up, as it now has starting torque. When going fast enough the inductance is cut out, and it continues in motion as a synchronous motor. Synchronous Polyphase Motor.— As far as revolving is con- cerned, a polyphase generator and motor may be identical. In a rotating field magnet permanent poles are maintained by a direct current. The polyphase alternating current is passed through the windings of a stationary armature. This creates in it a rotary field. The condition is illustrated in diagram in Fig. 271, a permanent magnet representing the electro-magnet of the description. The generator on the left produces a rotary ALTERNATING CURRENT MOTORS. 373 field in the motor, which causes its rotor, which is its field mag- net, to revolve in exact synchronism. The stator and rotor m.ay be reversed in the construction. Self-Starting S>nchronous Motor. — To make a polyphase synchronous motor self-starting, the following arrangement is sometimes adopted. Copper bars connected at their ends are bedded in the faces of the pole pieces of the field magnet. The rotary field acts upon these, and induces current in them exactly as in the induction motor. To start the motor, the direct-current field circuit is opened, and the alternating circuit is closed. The motor is now an Fig. '^72.- Synchronous Motor Rotor with Starting Armature. induction motor, and the rotor begins to turn. It is given no load, so that the rotor soon turns almost at the speed of the rotary field. The direct current is now turned on, and the motor becomes a synchronous motor. The elements of the induction motor are still present, but no torque is exercised by them be- cause there is no slip. The same principle is carried out by m.ounting on the same shaft with the synchronous armature a smaller induction-motor armature. When in place each armature lies in its own field, and the induction motor is used to start and to bring up to syn- chronism the larger armature of the synchronous motor. When 374 ELECTRICIANS' HAXDY BOOK. this is effected, the synchronous motor takes up the load, and the induction motor ceases to act. Fig. 272 shows the armature of a 'synchronous motor with the squirrel-cage armature of the start- ing induction motor on the right-hand end of the shaft. In the above lines the use of direct current and of alternating current for the motor has been spoken of. This refers to the field and armature currents respectively. An alternating-cur- rent generator has its field excited by a direct current, and gen- erates an alternating current from its armature. An alternating current synchronous motor goes a step further, as it has to be connected to two distinct circuits for its operation, each one sup- plying power. One circuit possesses direct current, which excites the field; the latter may be rotor or stator. In the diagrams it is shown as the rotor, and to avoid complication a permanent magnet is used as its representative. The other circuit, entirely distinct from the first one mentioned, passes alternating current to the armature windings. This is the true power circuit, cur- rent from which actuates the machine. In the diagrams the armature is shown as the stator, but the relation of stator and rotor can be changed. CHAPTER XXII. TR.:lNSFORMERS. Basis of Transformer Construction.— If a current passes through a conductor, it establishes around it a field of force. Energy is expended in producing the field, but none in maintain- ing it. If a second wire or conductor lies parallel to the first during the time that the field of force is being built up, electro- motive force will be impressed upon it by the growth in number of the lines of force. This electromotive force will be of such polarity that it will tend to produce a current in the opposite di- rection to the original current. If the other cur- rent weakens, energy will be drawn away from the field, and the electromotive force impressed upon the neighboring wire will be of the reverse polarity. Current is only produced during the period of change of intensity of field. The transformer contains two coils of wire insulated from each other. One, the primary, receives varying electromotive force; the other, which is the secondary, has elec- tromotive force impressed upon it by variations in the current passing through the primary. In Fig. 273 C represents a bundle of iron wire wound with two coils of insulated wire. The circuit from one coil, which is of small relative length and large current capacity, contains a bat- tery a and key h. The other coil of long fine wire has on its outer circuit c' c' two electrodes d d. On depressing and releasing the key, a spark will jump across the air space between the two 375 Fig. 273.— Action of a Transformet?, 376 ELECTRICIANS' HANDY BOOK. electrodes, if all proportions are right. The first-named coil is the primary; the other is the secondary. The Object of a Transformer is to receive a given alternating voltage from an alternator delivered at one pair of terminals, and to deliver at another pair of terminals a different alternat- ing voltage. The transformers seen on house fronts and power line poles may have a comparatively fine wire deliver a small current at 1,000 to 6,000 volts potential to their primary terminals, while from their secondary terminals a current twenty to one hundred or more times greater is taken off with a potential difference at the secondary terminals of 50 or 60 volts only. A small copper wire might thus deliver a li/4 -ampere current to a transformer with 6,000 volts between the primary terminals. f IG. 274.— Alternator and Converter Delations. Fig. 374a. -i(i>G Tr^\nsf()h:.i. This would be about 10 horse-power. But if this 10 horse-power had to be delivered with only 50 volts potential difference be- tween the primary terminals, the wire would have had to be of twelve times the cross-sectional area, and consequently of twelve times the weight, and approximately twelve times the cost. Choking.— Another function is performed by transformers. The current passed through them between their primary ter- minals is almost nothing if none is taken from the secondary. Hence they act to "choke" or hold back the current when de- sired without any considerable ensuing loss of energy. The construction is simplicity itself. The apparatus is so sim- ple and efficient that it appeals to the electrician as one of the most perfect of all electrical appliances. There are no moving TRANSFORMERS. 377 parts to wear out, except in a special type of transformer, and its action is absolutely automatic and perfect. Sylvanus P. Thompson very aptly says that a transformer may be regarded as a dynamo with stationary field and armature, in which the alternating magnetism of the iron coil induces the desired current in the secondary coil, representing the armature. The Limitation of a Transformer is that it has to have a varying cur- rent; practically, it is used only on alternating current circuits. It pro- duces a secondary alternating current, which can be made to give a direct one by special mechanism. The Principle of a Transformer is shown in diagram in Fig. 274. An alternating current from an alternator on the left goes through the primary coil, which is wound around a sort of iron ring. Another coil entirely dis- connected from the primary is also wound around the ring; this is called the secondary coil. A ring trans- former with adequate coils is shown in Fig. 274a. As current goes back and forth in the primary, it produces lines of fnrce, in the iron core principally. As the current starting from zero in- creases to a maximum, the lines of force increase in number, and are of polarity corresponding to the direction of the current. As the current re- cedes to zero the lines of force die away, and as the current goes to a maximum in the reverse direction, lines of force of opposite polarity to the first are produced. These changes in intensity and polarity of the field impress upon the secondary an electro- motive force varying from zero to maximum, and of constantly changing polarity. If there are ten times as many turns of wire in the primary as tkere are in the secondary, the electromotive Figs. 275, 2T6, and 277.— La- minated Shell Type Transformer Cores. 378 ELECTRICIANS' HANDY BOOK. force impressed on the secondary will be one-tenth that in the primary coil. The direct proportion of voltage impressed to relative number of turns will hold for all ordinary conditions. Shell or Jacket Type Transformers are those in which the coils are surrounded by masses of laminated iron. The material of the cores has to be of metal of good quality and quite thin. Insulation of some kind is used between the plates out of which the core is built up. The cuts, Figs. 275, 276, and 277, show the construction of mod- sHim IIHIIHIBBBBai Fig. 278 —Section r*? a ?HFLii Type Transformer. ern transformer cores, which are built up from the plates after the coils have been wound upon a form. The plates in the ex- ample shown are cut in such a shape that they can be pushed into the openings of the coils. The primary and secondary coils can be wound on top of each other or side by side. The cuts, Figs. 278 and 280, show the coils on top of each other. Step=Up and Step=Down Transformers. — If the transformer raises the voltage of a system, it is called a step-up transformer; if it lowers it, its more usual service, it is called a step-down transformer or a transformer without any qualification. TRANSFORMERS. 379 Ratio of Transformation.— The ratio of voltage impressed on the primary to that impressed on the secondary in the working of the transformer, as determined by the relative numbers of turns of wire in each, is called the ratio of transformation. It secondary turns is expressed by a fraction and is often desig- primary turns nated by k. Shell Type Transformers. — The construction of the working Fig. 279.— Shell. Type Transformer. Fig. 280 —Section of Shell Type Transformer. parts of a typical shell type transformer is shown in Figs. 279 and 280 in section and elevation. The apparatus is a transformer of the shell or jacket type, so called because a hollow laminated mass of iron, electrically speaking its core, surrounds the coils of the transformer. In the section the coils are seen imbedded in the hollow core. In this particular coil, the coils primary and secondary are wound separately and placed one above the ether. 380 ELECTRICIANS' HANDY BOOK, Another shell type transformer with coils partly exposed is shown in Fig. 281. Core Transformers.— Transformers whose core is surrounded by the insulated primary and secondary coils are thus named. In Figs. 282 and others such coils are shown. In Fig. 283 the con- struction of such a transformer is shown, and in the next cuts, Figs. 284 and 285, another form is illus- trated. The ring transformers shown in Figs. 274 and 274a are really core type transformers, al- though the term is usually applied ^ «., ^ r^ rr, FiG. 3S2.-CoRES OF CoRE Type Fig. 281. -Shell Type Transformer. oilCooled Transformer. to those with straight cores, as the term ring transformers covers the other case. Disk-Wound Transformers.— Sometimes the coils are in a number of sections wound separately into disks and piled one on top of another in alternation, as shown in Figs. 283 and others. By connecting the sections in series or in different parallel con- nections, the primary can be made to serve for a voltage of varr^ TRANSFORMERS. 381 ing amount, and the secondary can be connected to give -different potentials. A series connection of the sections is used for the high voltages, and parallel connection for the low voltages. Pancake Coils. — Coils such as those sl:own in process of con- struction and completed in Figs. 286 and £37 are called pancake coils. They are insulated, taped, and shellacked so as to be quite sirong. Such coils are often wound of copper ribbon as wide as a coil is high. The coils illus- trated are used in shell-type trans- | g: formers cooled by air blast. Fig. 283.— Construction of a Core Transformer. Fig. 284.— Core Type Oil-Cooled Transformer. The Auto=Transformer consists of an iron core wound with a single coil which virtually constitutes the primary and secondary. The secondary circuit is taken from it at two points; one connec- tion is made at one end of the coil, the other at an intermediate point. The portion of the coil comprised between these points may be of wire of extra thickness. It represents the secondary coil. The voltages will be to each other as the total turns in 382 ELECTRICIANS' HANDY BOOK. the coil to those in the secondary portion of the winding. It may be a step-down or step-up transformer. In the latter case the short section of coil is connected as the primary. A similar connection is sometimes made to the secondary coil in the three-wire system as applied to the working circuits, which are the secondary circuits, of alternating current systems of dis- FiG. 285— Coils and Core OF Co-re Type Oil-Cooled Transformer. Fig. 3S6.— Manufacture op Pancake Coils. tribution. Three wires are connected to the secondary, one in the center and one at each end. If a voltage of 220 is given by the secondary, then 110 volts will be formed between each end lead and middle lead. The centrally-connected wire is the neutral wire. This arrangement supplies current on the two-wire system until points are reached where it is to be used. At these points the transformers perform a double function, changing Toltage and instituting a three-wire distribution. TRANSFORMERS, 383 Action of the Transformer.— When the secondary circuit of a transformer is open, the inductance acts to keep back the cur- rent in the primary, and the transformer becomes virtually what Is called a choke coil. Some electric energy is wasted upon it, as it is not absolutely without current and the full voltage must be expended. When the secondary circuit is closed, a change of current intensity in the primary sends a current through the secondary, but in the opposite sense. Inductance is due to the energy required to increase the intensity of a field of force. The Fig 287. -Pancake Coils. primary sends a current of changing intensity in one direction, which produces lines of force through the core of the transformer when left to itself, and as it expends energy on doing so is choked back. But if the secondary circuit is closed, a current in the reverse direction goes through it, and demagnetizing to great- er or less extent the core of the transformer, facilitates to that extent the passage of a current through the primary. A closed secondary circuit causes current to go through the primary; with an open secondary only a very small current can pass through the primary. Transformers must have as good permeance as possible, and 384 ELECTRICIAN8' HANDY BOOK, hysteresis being a source of loss of energy must be avoided by the selection of an iron with low hysteretic coefficient, and by the use of laminated cores. Heat in Transformers — Transformers become heated when in use, partly from the eddy currents in the masses of metal of which they are constructed, partly from the current in their coils. Cooling of some sort has to be adopted. For small coils this is effected by the circulation of air about them and by the natural radiation of heat. Outdoor transformers are generally of the smaller sizes, and frequently the above agencies are depended on to cool them. The radiating surface of any solid varies with the square of its lineal dimensions, while its cubic contents varies with the cube of the same. The cubic contents increases with linear size more rapidly than does the surface. Therefore, the smaller a body is, the more favorable is the ratio of its radiating surface to its cubic contents. The heat present in it from any cause varies with the cubic contents. Its output is proportional to the same, and the heat imparted to it varies with the output. A small transformer will cool more quickly than a large one, all other elements being equal. The tendency of electric engineering practice is to use special means for cooling transformers. Oil Cooling, — Small converters are frequently oil-cooled. The converter is placed in a liquid-tight case, which is filled with oil. As the coil rises in temperature the oil becomes heated, and by circulating conveys the heat to the outside case. The air cools this, and thereby cooling the oil keeps down the temperature of the coil. The cut. Fig. 284, page 381, shows an oil-cooled trans^ former with its coils lifted out of its case. This is a core-type transformer. In use the coils and cores are lowered into the case, and oil is poured in until it is full. Fig. 285, page 382, gives a separate view of the primary and secondary of the same type of transformer, with Its coils surrounding the cores. Oil in a converter case performs another function, as it im- proves the insulation. A thermometer as shown in Fig. 288 is sometimes set into an oil-cooled transformer, in order to show how hot it is getting. TRANSFORMERS. 385 As the size increases, the heat imparted rises with the cube of the linear dimensions, and the superficial area rises only as the square. The cooling power is pretty closely proportional to the superficial area. Notwithstanding the wasteful heating of trans- formers, large-sized ones are exceedingly economical, often giv- ing over 98 per cent of return, a waste of less than 2 per cent. There would be no difficulty in making the transformer so large in proportion to its out- put that special cooling would not be required. But this would be so expensive that it would cost more than would the use of smaller artificially- cooled transformers. A characteristic feature of many transformers is the cor- rugated case. The shape is given to increase the area with which the air comes in contact. Water Cooling. — Water can- not be directly applied for cooling transformers, on ac- count of its effect on the in- sulation. It is applied indi- rectly by using a coil through which water circulates to cool the oil. The cut, Fig. 289, shows the interior parts of a shell-type transformer lifted out of the case. The core and coils are surmounted by a coil of pipe. In use the whole apparatus, core, coil, and water pipe, is immersed in oil in the transformer case. In the operation of the transformer, as the oil gets hot, the hotter oil rises to the surface. Here the hot oil would naturally accumulate. The coil of pipe is immersed in this portion of the oil, and occupies the most effective place for cooling the oil. Water is kept circulating through it. Fig. 288.— Thermometer in Trans- former. 386 ELECTRICIAXIS' HANDY BOOK. Air-Blast Cooling. — The cooling power of an air blast is often used for transformers. A current of air in rapid motion possesses far greater cooling power than when it is left to its natural cir- culation. The cut, Fig. 289a, shows an air-blast cooled trans- former. The air enters from below through a pipe communicat- Fia. 289.— Water-Cooled Oil-Filled Transformer Coils and Core. Fig. 2S9a.— Air-Blast Trans- former. ing with a fan or other source of air blast. The primary and secondary are wound in flat coils separated from each other by diaphragms. The core is so built up as to leave air ducts regular- ly spaced throughout. On the top there is a central damper to regulate the draft of air between the coils, and the damper on the side near the top regulates the draft through the core. At the TRANSFORMERS. 387 bottom semicircular doors give access to the secondary terminal. The primary terminals enter on the top. The power required to operate the fan-blower is exceedingly small, about one-tenth of one per cent of the output. The fan is driven by an electric motor. Disk Winding. — Constructors of transformers often wind the low-tension coils disk fashion or concentrically, with one set of turns per layer, while its high-tension coils are wound out of wire of rectangular cross section. In the large transformers a number of wires are connected in parallel. This subdivision prevents in a great degree eddy currents in the conductors, just as lamination prevents it in the cores. The system used in the high-tension windings brings about another result. As one set of turns only is used for the width of each flat or disk coil, the electromotive force between neighbor- ing turns is never more than 25 volts, and sometimes is only 10 volts. The principle is the familiar one used in the higii-tension winding of induction coils. It is called disk winding when applied to this class of apparatus. Ducts are arranged all through these coils, so that the oil with which they are charged starts into vigorous circulation at once when the heating due to service begins and no part of the iron is more than an inch distant from oil in motion. In larger sizes of transformers the cast-iron covers may simply le put in place without bolting down. A case could hardly arise in which a large transformer would be placed on its side. With small transformers, their covers are bolted on, so that they can be subjected to considerable jolting and inclined posi- tions without disturbance. Constant^Current Transformers.— A constant-current trans- former is one in which there is not a constant ratio of electro- motive forces between the terminals of the primary and second- ary coils, but in which a constant current is maintained by the secondary as long as a constant electromotive force is maintained at the terminals of the primary coil. This represents the require- ments of series lighting. An ordinary transformer gives on the secondary an almost constant virtual voltage and varying intensity of current. If [^88 ELECTRICIAXS' HANDY BOOK. the coils of the transformer be so constructed that the induct- ance of the primary and of the secondary portions are high com- pared with the mutual induction between them, the coil will give a constant virtual intensity of current to its secondary. One way of effecting this result is to have a choke coil in series with the primary coil. Special constructions of coils may be con- structed to answer the same end. A long core with the coils on the ends is one desigr. Electrical constructors have also devised transformers in which the result described above is produced by changing the distance between the coils. The diagram. Fig. 290, illustrates the principle. C represents c s 8 P mm S s p .-.;».•.-. ■-;.•. :^: :.:-:•.;. Fig, 29).— Action of Constant Current Transformer. the iron core, P, P, the primary, and S, S, the secondary coil. The secondary coil is movable and suspended at the end of a lever with counterpoise, so that a little force will move the sec- ondary coil up and down. By Lenz's law (page 213), the induction of a current in the secondary coil will cause repulsion between the coils to be ex- erted. This varies in degree with the current induced. There- fore, in the apparatus any tendency to an increase of current in the secondary repels it from the primary, thereby diminishing the induced current. If the current grows less, the repulsion diminishes and the coils come nearer together, and the induc- tion is increased. The next cut. Fig. 291. shows the construction. In this the TRANSFORMEFS. 389 fixed coil is seen at the bottom. The movable coil is suspended as shown above the fixed coil. It is held in equipoise by a lever with counterweights. When a small current is taken from he secondary, the movable coil drops, and may even rest upon he fixed one But as more current is taken from the fixed coil, the repulsion drives them apart, so as to dimm- ish the induced current. In this way a con- stant current is maintained with changing resistance on the outer circuit. The cut shows the sectional view of the transformer in the upper portion of it, with the plan below. If the resistance of the outer circuit supplied by the secondary coil is re- duced by the operation of arc lamps or by cutting one of them out of the line, the cur- rent increases momentarily, the repulsion drives the coils apart, the induced electromo- tive force falls in value, and the current through the new and less resistance under less electromotive force is unchanged. In larger transformers of this type there are two primaries, one at the top and the other at the bottom, both fixed in place, and two secondaries poised between them. With- out any output one rests against the upper, the other against the lower primary. One characteristic feature of this appar- atus is the counterpoising of one movable yiq. 291.-Constant coil by the other one. Currekt Trans- An auxiliary lever is provided for adjust- former. ing the effects of attraction or repulsion between the coils. By adding or removing counterpoising weights, the adjustment is made. The apparatus shown has its coils immersed in an oil tank; the oil not only acts as a cooling agent, but damps the movements of the coils. Oil for Transformers.— The oil for filling transformers should be of low viscosity, so as to rapidly penetrate any interstice. High flashing point and high insulating value are also requisite. 390 ELECTRICIANS' HANDY BOOK. Sometimes sparking will make a little tube of carbonaceous mat ter through oil which will constitute a permanent source ol trouble. Insulation in Transformers • — The most elaborate care has to be taken in insulating the windings of transformers. Tape, shellac, and mica are used. The laminations of the core or core plates are insulated from each other also in order to prevent Foucault currents. Direct Current from Alternating Current. — By special con- nections to collecting rings on the shaft, an alternating current can be taken from an armature wound for direct current. The illustration. Fig. 292, shows a diagram of a Gramme ring wound for direct current. If rotated in a bipolar field with the connections shown in the cut, an alternating electromo- tive force will be impressed upon the circuit, if closed through the brushes and col- lecting rings. For the Gramme ring the general rule is that for single-phase current the connections must be taken from its windings at angular dis- tances equal to the pole spaces. For four poles there should be four connections, for six poles six, all evenly spaced, and con- nected alternately to one or the other collecting ring. In these windings, whether alternating currents are taken from them by means of connections to collecting rings, or wheih- er direct currents are taken from them by a commutator, the coils are subjected to precisely the same inductive influences, an I identical electromotive forces are impressed upon the windings in bot^"> cas^s. Rotary Converter. — If an armature of a dynamo is provided with t"'o S'^^s of connections, one to a commutator for direct cur- rent and another tg two, three, or four collecting rings for alter- nating current, a machine results which can receive one kind of Fig. 292.— Gramme Ring Giving Alternating Current. TRANSFORMERS. 391 current and act as a motor and deliver the other kind of current acting as a dynamo. Such a machine is called a rotary con- verter. The term continuous alternating transformer is applied to it in England. The machines can be driven by an alternating current as a synchronous motor, either for driving machinery or for generat- ing direct current, or for both. The latter current can be taken from the brushes bearing on the commutator. Use of the Rotary Transformer. — It is settled that for long- distance transmission of power the alternating current is to be preferred. It is in connection with such transmissions that the rotary converter is principally used. For many purposes direct Fig. 293.— Theory op Rotary Converter. current is preferable. Especially in high-voltage transmission is the rotary converter useful. Thus, a power station may generate electric energy, and trans- mit it any desired distance at a high voltage, so as to need only a small transmission line. It will in such cases practically always be of the alternating-current type. When it reaches a center of distribution, the current may go through step-down transformers, thereby giving an increase of current and diminu- tion of voltage. The current from the secondaries of the step- down transformers may be used to drive rotary converters, so as to produce direct current. Such an arrangement may be cited as particularly available for electric railroads on which direct- current motors are employed. 392 ELECTRICIANS' HANDY BOOK. Principles of Construction.— The diagram, Pig. 293, shows the principle of construction. An armature is indicated by its wind- ings, and is supposed to rotate in a magnetic field. The ends of the windings are connected to collecting rings and commutator segments. In the diagram each end of the winding connects to one of the collecting rings and then to one of the two commutator segments. Four brushes are provided; one pair for the direct current bear against the commutator, the other pair for the alter- nating current bear against the collecting rings. Alternating current received by the pair of brushes bearing against the collecting rings will cause the armature to turn when brought up to speed. It becomes a synchronous motor. Direct current can then be taken from the other pair of brushes, which bear against the commu- tator surface. In this case it op- erates as a converter of alter- nating into direct current. It may have its commutator brushes connected to a source of direct current. It then, turns as a direct-current motor, and alternating current can be taken from the collecting-ring brushes. This latter use is com- paratively rare in engineering. The arrows in the diagram indi- cate the current relations. The next cut. Fig. 294, shows a drum armature with a regular commutator at one end of its shaft, and three collecting rings at the other. From each collecting ring a wire connects with the winding of the armature. The connections are 120° apart. One result is a three-phase current if the armature is rotated in a field by a direct current. The other result is a direct current if the machine is driven as a polyphase synchronous motor by a three-phase current. Relations of Voltage and Current. — The single-phase rotary converter operating to convert direct into alternating current im- presses a maximum voltage on the alternating-current circuit equal to that of the direct-current circuit. By the law of sines Fig. 294.— Drum Armature of IlOTARY Converter. TRANSFORMERS. 393 the effective voltage on the alternating circuit is 0.707 of the direct-circuit voltage. If the rotary converter is operating to convert alternating into direct current, the direct-circuit voltage will be 1.41 times the effective alternating-circuit voltage. The effective current and the direct are in great degree the inverse of the proportion indicated above. All losses are neglected in the above general statement. Analogous ratios hold for polyphase rotary converters. The current in the armature of a rotary converter is made up of tv/o currents. One is that which passes through it by the col- lecting ring brushes, the other is that which is induced by the poles, and which is delivered to the outer circuit by the armature brushes. The algebraic combination of these two constitutes the total, and as these two are generally opposite in sign, the actual current is small. This gives a small armature reaction and a small heating effect in the coils. Whether or not it is fair to call the distortion of the field of force by armature reaction the cause of the torque, there can be no electro-magnetic torque without such reaction and cgnsequent distortion. The nature of the distortion determines the direction of torque, concentrating the lines of force under the leading horns of the pole pieces. The armature windings of the rotary converter, when it is per- forming its function of conversion of alternating into direct current, are traversed by a smaller current than when it is oper- ated as a direct-current dynamo. The output in power of which it is capable in its different roles, which is its working capacity, may be basea upon the current it can carry with equal heating of the armature windings. The following power ratings are for a rotary converter used in the functions described: Six- Phase Converter. 1.96 Rotary Converter in the Three=Wire System.— The Edison three-wire system can be supplied by a rotary converter on the following system, applicable for a three-phase original current. The three secondaries of the step-down transformer on the high- Continuous- Single- Three- Current Phase rhase Generator. Converter. Converter 1.00 0.85 1.34 394 ELECTRICIAXS' HANDY BOOK, tension circuit are Y-connected (page 359). The free ends of the coils are connected to the three collecting rings of the rotary converter. The electromotive force between the junction of the coils, which is the natural point of the Y connection, and either of the armature brushes on the direct-current commutator is constant and equal to one-half of the electromotive force between the brushes. The neutral wire of the three-wire system may be connected to the neutral point of the Y, the other two wires to the direct-current brushes. Starting a Rotary Converter.— If receiving power on the alternating side, the rotary converter has to be brought into syn- chronism. This can be very simply done by a small direct-current dynamo, which connected to the direct-current brushes will effect the result, when the alternating current can be substituted by way of the collector ring brushes. Functions of a Rotary Converter.— This machine can convert alternating current into direct current or the reverse. It can be used as a motor on either direct current or alternating current. It can be driven by power, and deliver either direct or alternating power or bolh at once. It may receive one kind of current and act as a motor, and generate the other kind of current simultan- eously. The Rectifier. — The alternating current rectifier is an appli- ance for converting an alternating current into a pulsating cur- rent of uniform direction, giving a series of half waves of identi- cal direction. Its use is principally for field excitation of alterna- tors. One or more of the armature coils is disconnected from the rest, and its ends are connected to the rectifier. The latter by its brushes delivers direct pulsating current to the field wind- ings, providing field excitation. The rectifier is a modification of commutator and collecting rings. It consists of a drum whose construction resembles that of a commutator. One bar or division is provided for each mag- net pole, giving an even number of bars. The bars are electrically connected in two sets, so that if they were numbered consecu- tively, the odd-numbered bars would be connected together, and the even ones also. Bach set is insulated from the other set, and both from the shaft. The rectifier is mounted on the commuta- TRANSFORMERS, 395 tor shaft of the alternator. Each set of bars is connected to a terminal of the coil. There are two brushes, which are so ad- justed that one will be in contact with an even-numbered bar when the other is in contact with an odd-numbered bar; and if the two brushes are connected, then the alternating current from the armature follows this path. It goes to one brush, by a bar of one commutator set, passes through the wire of the outer circuit, including the magnet coil of the machine generally connecting the brushes, thence through the other brush and other set of commutator bars to the original coil. The entire current from an alternator may be passed through a rectifier. The alternating current from vhe armature of the machine is rectified, passes from one brush through its cir- >mit, including, it may be, lamps, field vnagnet of the alternator, and other things, and returns to the other brush of the rectifier. From the other end of the rectifier the original alternating cur- rent circulates through the armature. A rectified current may be used for di- rect-current operations, such as charging storage batteries, supplying direct cur- rent lamps, etc. It is not perfectly sat- isfactory for some uses, on account of its pulsatory character. A simple rectifying commutator is shown in the cut, Fig. 295. Two cylinders cut like crown gear wheels are nested together as shown, and are insulated from each other and rotate with the main shaft of the alternator. The heavy black lines indicate insula- tion. One is connected to one end of a wire from the armature coils; the other to the other end of the same wire. This wire may be an independent parallel winding, for the purpose of giv- ing current to excite the field. The brushes bear one on one tooth, the other on the next tooth of the commutator. Wires from the brushes go to the field, if it is to be excited, and connect in circuit with it. As the current in the wire from the armature changes in direction, the rotation of the commutator brings the Fig. 295.— Rectifier Commutator. 896 ELECTRICIAXS' HANDY BOOK, brushes to the other te-eth. The effect is to send the rectified current through the outer circuit. An ordinary commutator can be used with its bars electrically connected into two sets of alternate bars, each set insulated from the other, provided it has one bar for each field-magnet pole. Operation of Transformers. — It has been impossible within the limits of the space at our disposal to go into full details of the theory of transformers. Owing to hysteresis and other factors the actual operation of a transformer is not so simple as the dis- cussion of it given here might make it appear. But the full treatment of the subject involves the application of the higher mathematics and is very intricate. The theory of the -action is only given in outline and the statements are subject to qualifica- tion if the field of full investigation is entered on. CHAPTER XXIII. MANAGEMENT OF MOTORS AND DYNAMOS. Starting Motors. — The current must be given to a motor with some degree of slowness, or the armature may become overheated. After the motor is in rapid motion, the counter electromotive force protects the armature to an extent more or less consider- able. A stationary armature will be burnt out under conditions of voltage and current of the outer circuit, when it would be per- fectly safe if in its full rotation. The Starting Boxes.— Protection is given in their starting by the use of resistance. The resistance used is generally contained in a case with^ switch handle on its top and contact points. By swinging the handle from point to point, resistances are cut out one by one until none are in circuit, and the motor receives as full voltage as is possible. The motor is started with all the resistances in circuit, and in series with the armature, and they are cut out as described until the motor is in full motion. Re- sistances cannot be economically used for running the motor. A simple construction is shown in Fig. 296. The switch is an arc of a circle shown in the middle of the cut. When turned clear to the right, the arc is out of contact with any of the four tongues. On turning it from the open circuit position, it first makes contact with contact No. 4, which is connected to the line. This contact is without effect. It next makes contact with No. 3. This sends the full current through the field. The next contact is No. 2. This sends current through the starting coil and armature, and the latter begins to rotate under the influence of the reduced current. Another contact remains. No. 1, which when made short-circuits the starting coil, and the armature receives the full working current. The long arc keeps all the con- tacts closed when in the last-described running position. 398 ELECTRICIAXS' HANDY BOOK. Magnetic Release Starting Box.— A series of resistance coils are connected to a set of contact studs. An arm is arranged to swing on a pivot. In its motion its outer end moves over the row of studs, making contact with them one by one. Each stud repre- sents a resistance held in a frame, which is back of the face of the apparatus. When the handle is swung to the left, as shown in Fig. 297, all the resist- ances are in series with each other. As the switch is moved to the right, it cuts out the resistances one by one until none is left in cir- cuit. On the switch handle there is an armature of soft Ff£LO COILS Fig. 296.— Simple Starting Box, Fig. 297.— Motor Starting Box WITH MaGNLTIC HeLEASE. iron, which when the resistance is all cut out is brought by the motion of the switch arm directly in front of and against the poles of an electro-magnet. This magnet is secured to the face of the box, and is connected so as to receive part or all of the current received by the motor. A spring is arranged to pull the switch arm away from the magnet and across the face of the box to the position where the MANAGEMENT OF MOTORS AND DYNAMOS. 399 current is entirely cut out, where it strikes a stud and has its motion arrested. The attraction of the magnet for the armature is great enough to hold it against the pull of the spring. If the voltage of the circuit should increase, and thus produce an over- load, automatic cut-outs or fuses would presumably open the cir- cuit. The ctirrent would cease to excite the magnet, and the armature would no longer be attracted by it; the handle would fly off to the other end of its arc and come to rest with the motor circuit open. When the circuit breaker was replaced, new fuses put in, or in general terms when current was again turned on, the motor would be cut off and would only start by the regular process of moving the starting-box arm across the resistance contacts to the no-resistance running point. If the starting box Fig. 298.— Diagram of Starting-Box Connections. is not provided with the feature described, the current when turned on again would be apt to burn out the motor armature, unless some one had had the thoughtfulness to turn off the switch arm. Various constructions and arrangements are possible to carry out this principle. Starting-Box Connection.— The starting box is placed in series with the armature. The field if shunt-wound receives the full current which it is capable of passing; if compound-wound, the shunt winding receives its full current, the series winding re- ceives the current diminished by the starting-box resistances. These act upon the entire armature current, but only on part of the field current in compound-wound machines. In Fig. 298 B is tlie starting box, S is the shunt coil, and A is the armature of the 400 ELECTRICIANS' HANDY BOOK, motor. The switch handle is horizontal when the motor is idle. It is turned clockwise. It connects the field first; then keeping this connected, current is passed through all the coils in series and the armature. Then the coils are cut out one by one as the handle is turned until the full current passes. Changing Voltage,— It is often desirable to transmit electric power at one voltage and transform it before use to another volt- age. For direct current this is done by a machine called a motor transformer, motor dynamo, or dynamotor. A Motor Transformer is a combined motor and generator. It has a single field magnet or set of field magnets, and a single armature is mounted in their field. The armature has two inde- pendent wiudings and a commutator for each winding; each commutator has its pair of brushes. Generally, the two com- mutators are placed at opposite ends of the armature. Action of the Motor Transformer. — The current from the original station passes through one of the armature windings and through the field coils. The terminals of the line are connected to one of the pairs of brushes. The machine, as far as this current and connection are concerned, is a motor, and its arma- ture rotates. As the armature rotates, it carries the other inde- pendent winding around, and electromotive force is impressed upon it. A circuit connected to its brushes has electromotive force impressed upon it, and if closed has a current induced. One of the independent windings is of a greater number of turns than those of the other winding. To decrease the electro- motive force, the winding of the greatest number of turns is used as the motor winding, and its brushes are connected to the actu- ating circuit. To increase the electromotive force, the winding of fewest turns is the motor winding. The relation of the original electromotive force to that im- pressed upon the second circuit by the generating coil is deter- mined by the relation of the turns in the one winding to those in the second winding. The cross-sectional area of the wires of the two windings is inversely proportional to the voltage ex- pended on the first coil and impressed on the first one. Step-Down and Step-Up Transformation.— A long fine wire of many turns in the first coil and a short thick wire of few turns MANAGEMENT OF MOTORS AND DYNAMOS. 401 in the second coil give a diminution of voltage and an increase of possible amperage. This is a step-down transformer. A short thick wire of few turns in the first coil and a long thin one of many turns in the second coil has the reverse effect, and the combination is that of a step-up transformer. For the first coil the machine is a motor, for the second it is a dynamo. The first coil is the primary and the other the second- ary. In operation the primary coil passes a current actuated by the voltage from the station. Electromotive force is impressed upon the other coil, and any current up to the current-carrying capacity of the secondary wire, multiplied by the number of leads in parallel in it, may be taken from the brushes of what may be called the secondary commutator. riotor Transformer Practice. —Motor transformers may be dis- tributed all through a district. The current may be generated at a distant source, by water power for instance, and sent by several thousand volts potential through a small and conse- quently cheap wire circuit to any desired points in the district to be supplied. Or it may be sent to a single centrally-located trans- forming station in the heart of the district. Here it may actuate any number of motor transformers, and independent circuits can be taken off from each. These circuits radiating through the district will supply electric power most advantageously at low voltage. The Economy of flotor Transformers running at full load ex- ceeds 90 per cent. They are cheaply run as regards maintenance. The commutators need attention and ultimate replacement. New brushes have to be put on when the old ones are too far gone to yield to trimming and adjustment. The great expense is the personal attendance required. A moving machine should not be left without someone to look after it. It needs attention some- times, even if it runs for hours without being touched. When attention is needed, it is apt to be rather urgent. A little neglect may lead to extensive injury. The expense of the labor item represented by the cost of the attendant workmen or engineers has operated to restrict the introduction of these machines. In Europe the system has been quite extensively employed. 402 ELECTRICIANS' HANDY BOOK, The other items of expense connected with it are estimated as considerably less than the interest and depreciation and energy loss charges in the direct-current low-potential distribu- tion from a distance. Where it is possible to concentrate the transforming under one roof in the heart of a district, the condi- tions are most favorable for its employment. Parallel Coupling of Dynamos. — Dynamos have to be coupled in parallel when the current to be sent out from a station ex- ceeds the capacity of one dynamo. When the current approaches the capacity of a single generator, if it seems probable that more cur- rent is to be required, a second dynamo must be connected in parallel with the other. It is. necessary also when the dynamo is to be replaced by another with- out interrupting the current. Parallel Coupling of Shunt Dynamos is shown in diagram in Fig. 299, in which the dynamos connected from bus-bar to bus-bar have their armatures indicated by Ai Ao, the shunt coils by Si S., and the switches in the leads to the bus-bars by Bi B.. To throw a dynamo in, it is brought up by use of the field rheostat to a volt- age two or three volts over that of the system, its main switch B being open. When the voltage is attained, the main switch is closed. A voltmeter not shown in the cut is connected to the armature brushes or to the conductors near thereto. Parallel Coupling of Compound Dynamos.— The general con- nections are shown in the diagram, Pig. 300. A^ A2 are the arma- tures. Si S, the shunt coils, Fi F. the series coils, B, C, and D are the switches. P Q is the equalizer. They are supposed to be provided also with voltmeters, ammeters, and rheostats for their shunt coils. Fig. 299.— Shunt Dynamos in Parallels. MANAGEMENT OF MOTORS AND DYNAMOS. 403 The operation of starting a dynamo in parallel is thus con- ducted: The switch D is closed. The machine to be thrown into parallel is started and regulated by speed and field excitation until its potential is one or two volts lower than that of the m.achine already working. Then the switch B appertaining to the new machine, and which has hitherto been open, is closed. To throw out of action a machine running in parallel with another, the field excitation is reduced by the rheostat on the shunt coil until the load is only a few amperes. If this does not bring down the current enough, its speed of rotation may be re- Fio. 300.— Compound Dynamo in Parallel. duced. Then the main switch B is opened and next the switch D on the equalizing wire. Trouble may follow from a machine accidentally stopping, as by a belt breaking. The machine thus freed of its load may take current from the other one, and begin to work as a motor. In each machine's circuit an underload circuit breaker should be included, which will break the circuit and prevent the motor action. Shunt^Wound flachines in Series, — These dynamos are some- times connected in series. The only object of this connection is 404 ELECTRICIANS' HANDY BOOK. to increase potential. The current capacity of the two will bo limited to that of the smaller one. Thus the two may have less current output than one. The potential is equal to the sum of the potentials of the two machines. Reversal of Direction of Armature Rotation. — The diagrams, Figs. 301 and 302, show how the connections of a dynamo must be reversed to change the direction of rotation of the armature of a series-wound machine. The diagrams represent a bipolar, shunt-wound dynamo. A and B are the field coils, and R is the regulating rheostat. The brushes are changed in position so as to give the reverse lead, and their connections are changed so as to connect them in the reverse sense with the two field coils. Figs. 301 and 302,— Reversal of Directiox of Afmatube Rotation. This throws the rheostat out, so its connections have also to be reversed. The two cuts are self-explanatory. If the dynamo is separately excited, simply reversing the con- nections from the exciter will effect the requisite alteration of direction of armature rotation. This may be of special use in installations where polarity or direction of current is the critical point in operation. Such are storage-battery charging plants, electro-plating works, and direct-current arc lamp systems. In the latter the upper carbon must be the positive one. Otherwise, the greater portion of the light is radiated upward. Polarity Tests. — Blue litmus paper moistened and held against the positive wire gives a red color. Paper dipped in potassium- iodide solution gives a black color at the same pole. Paper dipped in a solution of starch containing a little potassium iodide MANAGEMENT OF MOTORS AND DYNAMOS. 405 dissolved in it gives a blue color. Other test apparatus and appliances are on tlie market. Alternators in Step, — lu running alternators in parallel, not only has the potential to be kept the same for all the ma- chines, but the frequency of alternations or number of periods per second must be the same, and the machines must be in phase with each other. In throwing an extra machine into action in parallel with one or more running machines, all these three fac- tors have to be kept in view. The potential is brought up to the right point by changing the excitation of the field magnets; the frequency of alternations is brought to the proper point by changing the speed. •Synchronizing.— Two transformers, T^, T„ Fig. 303, have their Fig. 303.— Synchronizing Alternators in Parallel.. secondaries connected in series, one lead may include a voltmeter, the other an incandescent lamp, L. The primary of one o^ the transformers is connected to the terminals of one machine or else directly across the bus-bars; the primary of the other is con- nected to the terminals of the machine which is to be thrown into action. If the new machine, B, is operating in synchronism with the system or with machine A, the two transformers will co-operate in lighting the lamp. As the new machine is started, the lamps are lighted by the combined effect of the two ma- chines. The new machine is speeded up by turning on power, and as the frequency of the machines approaches equality, the 406 ELECTRICIANS' HANDY BOOK. light of the lamp begins to vary in brightness. At first the variations are very quick in following each other. As the hith- erto idle dynamo is speeded up, the frequency of its phases increases and approaches closer to that of the other machine. The lamps now vary more slowly, rising and falling regularly. The rising and falling grows slower and slower until a point is reached where it ceases and the lamps burn steadily. Meanwhile the voltage must have been kept right by adjusting the excitation of dynamo B. The voltmeter, not shown in the diagram, is used to direct this. At the instant when the lamp burns steadily the switches So are closed, throwing the machine into the work- ing circuit. Its voltage must be as nearly as possible that of the circuit when the switch is closed. The parallel working of alternators is made possible by the following fact: When running in phase with each other, alterna- tors tend to preserve their phase relation, or to run in synchron- ism. If one has a tendency to change its synchronism, reaction with the other pulls it up. Regulators or Boosters.— The potential given by a primary or secondary battery is increased by placing extra cells in series. In the secondary battery these are called end cells. If a dynamo gives insufficient voltage, an extra dynamo may be placed in series with it to add to the voltage. The second dynamo is called a regulator, compensator, and less elegantly but far more frequently, a booster. The ways of arranging booster circuits either with or without storage battery are numerous, and are subjects of a number of patents. Booster Connections.— A very usual method of connection is to place smaller dynamos as boosters upon the various feeders as required. The principal current is supplied by one or more dynamos running at constant voltage, which is the minimum required. From this dynamo the lines run directly to the bus- bars. From one bus-bar the feeders are led directly to the dis- trict. From the other bus-bar leads run to one set of terminals of smaller dynamos, and the other terminals of these dynamos are connected to the other leads of the feeders. The smaller dynamos are the boosters. MANAGEMENT OF MOTORS AND DYNAMOS. 407 The cut, Pig. 304, shows a typical arrangement. The principal dynamo is shown at D, and B and B are the feeder dynamos or boosters. The armatures of the dynamos B and B may be in series each with its own feeder. In such case the fields are sepa- rately excited. Often current is taken from the main dynamo for this purpose. On varying this current by a rheostat, the intensity of field and consequent electromotive force given by the boosters are made to vary. The main generator has to pro- duce the full current and more than one hundred (two-wire sys- tem) or two hundred (three-wire system) volts electromotive force; the boosters have to pass only a fraction each of the full Fig. 304.— Boosters. current, and impress a few volts electromotive force on the cir- cuit. Their armature resistance may be quite low. Hand Regulation of Booster. — The boosters have to be regu- lated so as to add more or less potential to that of the system in accordance with the R I drop. The field of the boosters may be excited by independent dynamos, and the field current in the boosters can be increased or diminished by rheostats or other appliances. Thus a rheostat may be placed in series with the field of the booster, and may be used to let more or less current flow through its coils, or the exciting dynamo may be regulated by its own field rheostat so as to give more or less current to the booster's fields. The operative has to shift the rheostat handle from time to time or otherwise modify the field excitation 408 ELECTRICIANS' HAXDY BOOK. of the booster to suit the requirements of the supply for the district. To carry out the hand regulating system, the armature of the booster is connected in series with the feeder line which it regu- lates. The feeder is connected directly to the brushes, and the armature in its separately-excited field is driven by the engine. Automatic Regulation of Boosters. — The regulation of the po- tential added to the circuit by boosters can be made automatic as well as very accurate by winding their fields and armatures in series with each other and connecting them in the feeder circuit. The feeder current goes through both field and armature. As the current increases, the series-wound booster responds, because of the increased current in its field. Its field excitation grows with the current, and a higher potential is developed. As cur- rent is diminished less goes through the field-magnet windings of the boosters, and they give a lower voltage. By modifying the field windings of the boosters, all sorts of effects can be secured, some analogous to those due to over- compounding. The feeders in the district connect with mains, and these with leads. As current increases, the drop on the feeders is not all that is decreased. The mains and leads also feel the loss in potential. The boosters can be so proportioned as to give some volts more than those of the drop on their respective feeders, so as to take care of the mains and leads also. If the drops on the feeder at maximum load were three volts, there might be one or two volts additional drop beyond the point of attachment of the feeder among the leads of the system. It is often advisable to give more potential increase within the feeder than its own drop, to compensate for the drop beyond it. It is a sort of over-compensating. This may apply to any feeder system. BDOster Construction. — The booster to act as described needs a high range of adaptability and power of varying its field strength. It may be said to require flexibility of action. The main point is to give it large field cores, so that the iron of the cores will never approach saturation, or else to have fewer turns than usual in the field coil. Hotor Dynamos as Boosters.— The current from a dynamo MANAGEMENT OF MOTORS AND DYNAMOS, 409 may be used to actuate a motor dynamo, and the current from the generating coils of the latter may be used as a booster cur- rent. The motor dynamo is practically a dynamo driven by an electric motor. The current from the station dynamo would pass through the motor to the feeders and mains of the system. The subsidiary dynamo driven by the motor would be connected to the feeders. As the line drew upon the station dynamo for more current, the motor would turn faster, because more cur- rent would go through its coils. This would cause the subsi- diary dynamo to rotate faster and to impress more electromotive force upon the system. This system would contain the automatic regulating feature. Nowhere in the field of electric engineering does the inter- I I I G - ^ - Fig. 305.— Equalizing Dynamos in Ihree-Wire System. changeability of dynamo and motor appear more clearly than in the uses of dynamos now being described. Their application to regulating lighting circuits is comparable to that of storage bat- teries, such application being based on the double role which such machines can play, at one time taking power from the system and acting as motors, at another time giving power to the system and acting as dynamos. Compensators.— This word has been used as a synonym for boosters. When a purely compensating action, and not a dis- tinctively intensifying action, is performed, it is specially appro- priate. The diagram, Fig. 305, shows the use of compensators on a three-wire system. The compensators B C are shunt-wound dynamos, coupled together mechanically, so that they rotate at the same speed. They are connected across the system, each dynamo being between the neutral and an outside wire. 410 ELECTRICIAXS' HANDY BOOK. In parallel with them, and between them and the main dynamo A, a resistance P G is placed across the outside leads. The neutral wire does not extend back of the compensators; it runs from between them out to the system of distribution; there are only two leads from the main generating system. The resistance is so arranged that but a slight current flows through the dynamos when the two leads are equally loaded. If by extinguishment of lamps or other appliances the wires re- ceive unequal current, the compensator connected to the wire carrying the lighter current acts as a motor. Turning under the influence of the current, it drives the other dynamo and gener- ates current for the other more heavily loaded line. Floating Battery. — Boosters are often operated in conjunc- ?L B Fig. 306.- Floating Battery. • tion with storage batteries. A storage battery connected across the two or three leads of a system, as shown in Fig. 306, is termed a "floating battery." It works automatically. When the voltage of the system tends to rise because of small consumption of current, the battery receives current and is charged from the main dynamos. When the district needs a heavy current, the battery discharges into the leads and assists the dynamos. This arrangement is the simplest, and is supposed to work automatically. An auxiliary dynamo or booster is generally used to assist the regulation. It acts to raise and lower the voltage of the system. Booster and Storage Battery Connections are shown in Figs. 307, 308, and 309. In Fig. 307 G is the station dynamo, B is a series-wound booster, S indicating its series coil; E is the storage battery and MM indicate motors in the district. At normal load the generator supplies just the right current, the voltage of the battery is equal to and opposed to that of the line, MANAGEMENT OF MOTORS AND DYNAMOS. 411 and no current goes through the field S of the booster, and the booster voltage is zero. When the load increases and more cur- rent is taken from the station dynamo G, the voltage of the sys- tem falls a little, the battery begins to discharge through the field S of the booster, and the latter adds electromotive force to the system. If all is in proper proportion, the electromotive force added will be just enough to compensate for the drop due to the loss of potential of the main generator. The battery dis- charges through field and armature of the booster, and the latter Fig. 307.— Booster and Storage Battery Connection. having its field excited with current with the polarity due to the battery's discharge, adds its voltage to that of the battery. If the voltage in the outer circuit due to the generator rises, this holds back the battery current and the booster field becomes inactive, and the booster ceases to generate current. As the voltage in the outer circuit rises still further, it exceeds that of the storage battery, and a charging current flows. This "ener- gizes," as it is called, the field of the booster, but with opposite polarity to the original, so that now it acts to help charge the battery. The whole arrangement works like a floating battery. The booster reinforces the action of the battery. It may be termed a 412 ELECTRICIANS' HANDY BOOK. floating booster. The system can only be used where voltage falls with load increase. In Fig. 308 a compound-wound booster B is supposed to be Fig. 308.— Booster and Storage Battery Connection, employed and is connected as shown. At normal load the excita- tion of the series field S is equal to that of the shunt field +. These two field coils are oppositely wound, so that they counter- FiG. 309^ -Storage Battery* and Boosteb Connection. act each other under this condition, and the booster generates no current. If the external load is increased by more power being taken in the district, the series field coil S of the booster receives more current than the shunt coil, and the preponder- MANAGEMENT OF MOTORS AND DYNAMOS. 413 ance excites the booster, so as to cause it to generate current in direction the same as that of the battery current. The booster and battery now add to the voltage of the line. If the external load decreases, the series coil gets less current than the oppositely-wound shunt coil. The polarity of the field of the booster is thus the reverse of what it was. The booster sends current into the battery and charges it. In the two last arrangements the booster and storage battery are in series with each other. The next cut, Fig. 309, shows a booster B with shunt field coil f and series field coil S, opposed to each other in winding, but with the storage battery in parallel with generator and motors, while the booster is on one of the leads between battery and generator. The booster voltage is added directly to the generator voltage. At normal load the magnetization of the shunt coil f exceeds that of the series coil S, and the electromotive force of the booster is of the same polar- ity as that of the generator, so that it reinforces the current due to the generator. The battery is of such number of couples that its normal voltage is equal to that of the sum of the genera- tor and booster voltages. If an excess load comes on the system, more current flows through the line, and consequently through the series coil S. This coil works against the shunt coil f. Therefore the voltage of the booster is diminished, and the bat- tery discharges on the line and takes up its share of the work. On decrease of load the field due to f preponderates, and the booster increases the voltage on the line until at low enough load this voltage exceeds that of the battery, and the battery receives a charge. Crushers. — This term is sometimes applied to a motor used to reduce the potential on a feeder line. Assume that there are several feeders running out from a station, and that some re- quire higher potential than others. The main dynamo can be run so as to give a higher electromotive force than that required by some feeders. On such feeder lines a motor would be placed which would absorb the extra voltage. The main generator's voltage would be lower than that reouired by other feeders, and on these feeders boosters would be placed, which in whole or in part would be driven by the motor. The latter would be a 414 ELECTRICIANS' HANDY BOOK, "crusher." The term is inelegant, and something better should be found for it. The same is to be said for booster. Abbott applies the term compensator so as to include all such appli- 80 \'OLT ARMATURE 120 VOLT ARMATURE 40 VOLT ARMATURE The Crocker=Wheeler System of Speed Control is especially designed for use in machine shops. It utilizes three dynamos, A, B, and C, Fig. 310, connected in series and with the three arma- tures on one shaft. The three armatures are practically con- nected in series across the circuit. Suppose the circuit to have a potential difference of 240 voles. Then the three armatures are wound for 40, 120, and 80 volts respectively. Four leads are taken from the ma- chines. One is at one end, another at the other end. and two intermediate ones are taken from between tke machines. These leads are carried through the shop T^Viiere power 16 to be uti- lized. The speed of the motors driven is regulated by changing the voltage ab- sorbed by them. A two- wire power lead of definite voltage is by the rotary transformers converted into a four-wire system. It will be noticed that the machines vary in voltage, and that the machine of highest voltage is placed between the others. The object of this will be seen. If a machine is to be run slowly, its terminals are connected across the 40-volt leads. The next degree is the 80-volt, and then the 120-volt lead. Each of these voltages can be taken off a single machine. Next the 40-volt and 120-volt machines can be put in series, giving 160 volts, then the 120 and 80 volts, giving 200 volts, and finally all three machines in series, giving 240 volts. This gives six voltages. The tool to be driven is provided with F]G. 310— t^cci^Mi-WHEEtiisR Multiple Voltage bPEED Control. MANAGEMENT OF MOTORS AND DYNAMOS. 415 its own motor, controller, and resistance coil. The six voltages give six speeds. Each voltage can be modified in its action by the resistance coil. Thus twelve speeds are obtained by a single resistance coil added to the four-wire system. Any motor can be caused to vary in sp-eed within certain limits by the use of a rheostat, which changes the current received by it. A motor with the rheostat control superadded to the multiple voltage control can be made to vary in speed by so many degrees of change as to work almost by insensible gradations. The rheo- stat takes the place of the resistance coil spoken of above. Accidents to Motors. — There are two principal causes of acci- dents. One is the burning out of the armature. This is guarded against by giving current slowly, by the use of a rheoctat or starting box. The other is destruction of the armature windings by too high speed. A run-away motor may have the binding wires on the armature break by centrifugal force acting on the windings. The latter are then driven against the pole faces, wrecking the machine. Too high speed should he guarded against- CHAPTER XXIV. CARE OF DYNAMOS AND MOTORS. Reversing the Direction of Current in a direct-current dy- namo or of motion in the same type of motor is effected by revers- ing the armature connections. This reverses the polarity of the core, and causes it to be subject to torque in the reverse direc- tion. If metal brushes or considerably inclined carbon brushes are used, their direction of inclination or *'rake" should be re- versed, to prevent the ends from catching on the commutator. Radial or even steeply-inclined carbon brushes need not be reversed. In multipolar machines the conneciions are shifted an angular distance equal to that, intervening between the poles. The easiest way is often to simply rotate the brush yoke through the arc of this number of degrees, carrying all the brushes with it. Sometimes such reversal cannot be allowed, as it in- terferes with regulating apparatus. Changing the main connec- tions reverses the polarity of both field and armature, and leaves the direction of revolution unchanged. Stopping a flachine.— When a machine is being stopped, the brushes should be kept on the commutator until it is running rather slowly. Then they are lifted off the surface. The object is to remove any chance of injury from a possible reverse move- ment of the armature. Strictly radial carbon brushes are almost free from danger in this regard. Too High Speed. — This is a cause of trouble. It may involve a strengthening of the field, so as to doubly raise the electro- motive force and cause sparking, which is to be cured by weak- ening the field. But too weak a field is in itself a cause of sparking. A field regulator may be used to adjust the strength of field. If so, it is a good precaution to use one without any zero point, or ''infinite resistance," especially in the case of 41t) CARE OF DYNAMOS AND MOTORS. ill motors. A motor without load and with current passing through the armature and the field cut out, will infallibly wreck Itself by racing. Loss of Magnetic Polarity. — A field magnet may lose its polarity. This may be due to long standing, so that the residual magnetism is lost, and the machine refuses to build up. It may also be due to wrong polarization of the field by means of a cur- \rent of wrong direction. Such a current may be produced when !so intense a current passes through an armature with advanced brushes that the armature reaction changes the polarity of the field. This may be due, in a shunt-wound machine, to a short circuit in the field. This wrong direction of current in the magnei coils is especially to be feared in compound-wound machines. Its results are especially bad in storage-battery work. Reversal of the magnetization of a field when the machine is charging a battery converts it into a motor, and the current from the battery drives it. Thus the battery loses any charge which may have been given it. The battery as it becomes more highly charged in regular working may be the agent in reversing the polarity of the field. Wrong Polarity iOf Field.— Sometimes it happens that the winding of a machine is such as to give the wrong polarity to the pole pieces of the field. This happens especially with mended machines. One thing to do is to recall the law of polarity, page 210, and to try to follow it out in the connections. Another is to try reversing the magnet connections. There should be no diffi- culty in arranging the connections so as to alternate north and south poles all around the field. The thing to remember is that in settling whether the current runs with or against the clock, the observer must conceive himself as facing the hollow in the pole piece which embraces the armature. The. brushes are raised from the commutator, and a current of proper direction is sent through the shunt winding for a few seconds. The machine can then be started again. Refusal of Motor to Start.— Connect an incandescent lamp or voltmeter between one of the leads and one of the binding posts of th€ motor. The lamp is the best, as it operates to some extent as a current tester as well as potential tester. If the lamp 418 ELECTRICIANS' HANDY BOOK. on one side shows no light, try a connection across from binding post to binding post, and then if the lamp lights, current passes, and the trouble is in the motor. If the line shows no current, a safety fuse is probably blown out or loosened. See if the brushes touch the commutator. If the line and motor seem to be all right, shift the brushes back and forth in search of the working point. If the motor will not go, it probably has too great a load. If a shunt motor is too heavily loaded, the armature refusing to start, develops no counter electromotive force, and practically short-circuits the field so as to impair the magnetization of the field. When a motor 'vill not start, ^nd the connections seem to be ail in order, the current should be cut off, and the clutch opened, or belt thrown off, so as to take the load off the machine. The arma- ture must then be set in motion by hand, and the current turned on while the armature is turning. Do not turn on the current while the hand is touching armature pulley or belt. When the machine is rotating regularly, throw on the load gradually. Re- member that a motor which refuses to start is in great danger of burning out its armature windings if the full current is left on for any appreciable time. The field coils al^o may suffer from overheating. This is another reason for starting slov/ly. Give current very slowly, and never anything like full current if the motor does not start. Slow Speed Without Load indicates in a motor an insufficient field ma'^retizing current or that the connections are inverted. Idle Motors.— When a motor is doing no work the current should be rut off. A motor running without load consumes cur- rent, and this if a meter is used, has to be paid for. Speed Regulation of Motor Without Load.— On suddenly throwing the load off a series-wound electric motor, as by shifting a belt or loosening a clutch, its speed will be suddenly and perhaps dangerously increased. The rheostat or starting box should be manipulated so ps to prevent this sudden increase of speed. A shunt-wound motor does not act thus, and does not need the above precaution. Starting: pvd S^oopins: Hotors. — These onpr'^^'ons should be performed gradually. A sudden throwing on or off of a load on CARE OF DYNAMOS AND MOTORS. 419 a motor affects the circuit sometimes to quite remote points. Large motors should for this reason alone be started and stopped slowly. In sudden, jerky starting there is also involved a great waste of power. Such wasteful manipulation is often very notice- able on trolley cars. The duty of the superintendent of power plants is to prevent all sudden starting and stopping of motors as far as he possibly can. Bad Contacts Between Winding and Commutator Ears. — The wires of the armature windings in some machines, especially those of earlier date, are connected to the commutator bars by means of screws. If a screw gets loose, resistance is introduced Avith danger of sparking, which will occur between the brushes and the badly-connected commutator bar. Thus, on stopping the machine the defective place can be located by the appearance of the bar. Properly-soldered connections in modern machines rarely fail, unless too many wires are bunched into one soldered joint. Temperature of Commutator. — The commutator should not rise to a temperature exceeding 185° F. (85° C.) A usual cause of heating of the commutator is too great pres- sure of the brushes against its surface. Relieving this pressure by weakening the action of the springs will contribute materially to the duration of both commutator and brushes. Collector Rings on alternators and alternating-current motors must be kept bright and clean. A little vaseline can be applied from time to time. If the surface is rough, the machine must l)e stopped, the brushes lifted off, the armature or rotor started turning again, and the rings may be sandpapered. Use a hollowed block of wood to hold the sandpaper. Materials of Commutator. — To withstand the action of carbon brushes, the commutator bars are made of hard copper (unan- nealed). But however hard the copper may be, it is apt to be more subject to wear than is the mica insulation which lies between the bars. Too hard or unwearable mica tends to project beyond the copper after a machine has run some time, and thus impairs the commutator surface. The projecting mica tends to cause the biushes to jump up as it passes, and occasions the worst kind of sparks, with lots of "extra current" behind them. Com- 420 ELECTRICIANS' HANDY BOOK, mon sandpaper is often not able to cut down the ridges. If not afraid of injury, emery or carborundum cloth or paper may be tried. The worst of this trouble is that it is slow^ to reveal itself. Loose Commutator Bars. — ^Sometlmes these are a source of trouble. By holding a somev/hat wedge-shaped piece of wood on each bar and striking it with a hammer, looseness can be de- tected. The internal insulation of the armature or the rings holding the commutator together may be in fault. The cure is to be intrusted to a competent person only. It may differ for different cases to an indefinite extent. Oval Commutator. — Especially if made of cast metals, commu- tators sometimes wear irregularly and become oval in cross sec- tion. The only cure is to turn them down in a lathe. A Gummy or Sticky Commutator Surface will cause the brushes to chatter or execute a series of little jumps. Cleaning is the remedy, with a very little oil. Do not attempt to stop it by lubrication, as this will make resistance at the contact of brush with commutator. Lubricating the Commutator Surface. — Sparking often fol- lo"^s as a result of this practice. If the surface is in good order and the brushes are properly shaped and trimmed, hardly any lubrication should be required. Electric contact of the best quality is required betw^een the brushes and the commutator surface. Anything in the nature of grease acts as an insulator. A drop or two of oil carefully rubbed over the w^hole surface should be sufficient lubrication. Brushes and Brush Holders.— These should not be so heavy that they will not readily yield to the Inevitable inequalities of the commutator surface. The width should range from % inch to 11/4 inch. The thickness prescribed by the manufacturer of the machine should be adhered to. For holders drawn copper is one of the best materials. Cast-metal holders are not generally recommended. Brush Pressure. • — A carbon brush may press with a weight of 2 to 214 pounds on the commutator; a copper brush should not press much over a pound. Good contact between carbon brushes and brush holders must be secured. For this object carbon brushes are copper-plated. CARE OF DyNA?/I03 AND MOTORS. 421 Replacing Brushes* — In putting new brushes in place, the surface resting on the commutator should be' made to fit accurate- ly. A sim-ple way of shaping them is to hold a sheet of sand- paper, rough side out, on the surface of the commutator, and to rub the bottom of the brush back and forth thereon, the brush being held firmly in the brush holder. If the machine shows any inclination to spark with a new brush, it is well to run it without load for a while until the brush shapes itself. Position of Brusties« — Opposite brushes should be placed so as to bear upon different portions of the commutator surface. If in the same plane at right angles to the shaft of the machine, they will wear a groove in the commutator. As far as possible, the entire surface of the commutator should be rubbed by the brushes. Copper Brushes must be cut square at their lower end ; especial- ly is this to be done for wire gauze brushes. They should be pressed just enough and not too much against the commutator. They should not vibrate when the armature is turning. Once a week they should be washed with benzine to remove grease and oil, and should be put in service again only when perfectly dry. Carbon Brushes are of Icvs-er conductivity than copper brushes. More carbon brushes are required for a given machine than copper brushes, which exacts a longer commutator. There must be no lost motion in brush holders or yoke. The screws and other movable parts of these portions of the machine must be watched. * If anything is loosened, it must be repaired or tightened. Setting Brushes« — In putting in the brushes, they must come in contact with the properly-spaced commutator bars. The gen- eral rule is to divide the number of commutator bars by the num- ber of poles, and set the brushes that number of bars apart. An- other way to get at their position is to lay a strip of paper around the commutator and cut it to exactly the circumference thereof. This can be divided with dividers accurately into as many equal parts as there are poles in the machine, and the divisions marked with a pencil. A simpler way is to do it by folding the paper. By placing it again around the commutator the pencil marks or the folds will show how to space the brushes. This applies espe- cially to setting tangential metal brushes. Direct-bearing carbon briishes tend to find their own place. 422 ELECTRICIANS' HANDY BOOK, If metal brushes are used, the greatest care should be exercised to avoid the commutator turning in the wrong direction, as this bends up their ends, and may injure or short-circuit the commu- tator bars. The difficulty in setting brushes arises from the fact that differ- ent machines require different setting. Once set wrong, enough sparking may occur to so deteriorate the commutator that spark- less adjustment will be impossible. Another source of damage may be the simple heating of the commutator on account of wrongly-set brushes. There is no rule to be given for setting brushes. For each t>T3e of machine it must be learned. The information can be ob)tained from the manufacturer if it is a new machine, or from tJie engineer who ran it if it is an old one. In the old two-pole machines it was a general rule that the "brushes should be 180° apart. In more recent two-pole machines the angle between the brushes is often little more than the angle Subtended by one of the pole pieces. On such machines a mark is generally made for the brushes to be set by. The older ma- chines had two marks, one for load, the other for no load. The general rule is to shift the brushes in the direction of the rota- tion, as a dynamo receives its load, and vice versa for a motor. Hard Carbon Brushes have sometimes to be rejected. One may be what some engineers call "glass hard," often harder than glass, or may be of high resistance. Such must be rejected and replaced by good ones. They cannot be made to work satis- factorily with brushes of the proper degree of hardness. Lifting Brushes. — When a machine is in operation generating current, a brush should never be lifted from the commutator. If there are several brushes on the same side, a single one may be lifted, but the best practice is not to do so. If there is only one brush and it is lifted, it may make an arc and burn the com- mutator. A good way to test the heating of the armature is to hold the hand in the draft of air comin;; from it. If the armature is hot, the air will be heated. Break in the Armature Windings. — This accident causes a TT.otor to spark very badly and may increase its speed. On stop- CARE OF DYNAMOS AND MOTORS. 423 ping it, the insulation between the commutator bars between which the broken coil is connected will show the effects of the sparking. If a dynamo refuses to build up, and this trouble is suspected, the machine can be run as a motor so as to identify and locate the place by burning the insulation as just described. If there are a great many commutator bars, the two involved can be temporarily connected by solder, so as to short-circuit the defective coil. The real remedy is to connect the ends of the broken wire with silver solder. A break will sometimes only show itself when the machine is running, when it will produce flashing between the commutator and brushes. When motionless, the severed ends may spring together. By determining the exact resistance of the armature the trouble may be found, as the break will probably increase the resistance if it does not absolutely break the circuit. A short circuit may exist under like condi- tions. If a commutator gets too hot, it will heat carbon brushes and get a coating from them, which will increase its superficial re- sistance and aggravate the trouble. The commutator blackens, and the carbon holders get hot and may become discolored. Such heating sho'ild not occur. End Muaon in an Armature Shaft is generally desirable. With the usral cylindrical commutator this motion causes the brushes to coii.e in contact with the entire surface of the commu- tator if the range of motion is sufRcient, and such contact favors even wearing, and the cylindrical contour of the armature is thus favored. If the armature shaft has end play, the belts are pretty sure to have irregularity enough to keep it in constant motion back and forth. In' some machines, end motion is given by mechanism for the purpose. Short Ci ;. iLs in Armature. — The windings may get their insulation ri:L.ed off and connect with each other or with the iron core of the armature. Copper or carbon dust may be the cause of short circuits between commutator bars. A commutator brush may te in electric contact with the frame of the machine. If a machine were perfectly insulated from the earth, such a single contact with core or frame would be without any effect. 424 ELECTRICIANS' HAXDY BOOK. If the frame of a machine is grounded and a ground exists in the commutator or armature, then such a contact of winding and arm- ature core causes a short circuit, which may burn out the arma- ture windings. When any short circuit of this character exists it is a menace, although it may do no harm for a long time. The short circuit can be sought for with a galvanometer and a source of current, such as a dry battery. The armature windings are disconnected from the field windings, and one end of the wire from the galvanometer and battery is kept in contact with the iron core of the armature or with the frame of the machine. With the other end of the wire the armature bars, brush connections, etc., are touched. A movement of the galvanometer needle indi- cates the contact and locates it. It is obvious that the magnet windings may be tested also by touching the exploring wire to their ends, as contact may exist between magnet core and magnet windings. Repairs have generally to be made at the factory. Frequently a battery with wires is sufficient to detect these troubles. A spark will show when contact is broken, or the tongue may be placed between the- wires, and the taste will reveal a leakage. A single contact between the armature winding and the iron core of the armature does no harm as long as no other contact or grounds exist. Of course, it should not be tolerated, as it is a constant menace. A short circuit due to the contact of two wires of the same coil of the armature winding may have serious con- sequences. A dynamo with this trouble will not build tip or excite itself. If the attem-pt is made to start it with an outside source of current, it will not absorb its full voltage, and the arma- ture windings will begin to get hot. This will be indicated by the smell of heated insulating materials. On stopping th^ ma- chine, the defective coils can be found by feeling the surface of the windings. The hottest part will be where the short circuit is. A motor will show such a short circuit by loss of power and speed. Sometimes it will not move at all. Entire or partial rewinding of the armature is the cure. If the short circuit is between two wires of different coils, the trouble is intensified. The whole armature may be burned out if the machine is not stopped in time. CARE OF DYNAMOS AND MOTORS. 425 What is said of this class of short circuits applies to the arma- ture bars. A contact between contiguous ones represents short- circuiting within the limits of a coil. If remote armature bars are connected, it represents the more serious case of short-circuit- ing of different coils. Already a temporary cure for a broken coil has been described. This was the soldering together of its two commutator bars. Such soldering must never be done unless there is absolute cer- tainty of the break. It would be better to cut the wire and bend the ends apart, to make sure of disconnection, and then to solder as described. Sparking; of the Commutator is a very serious evil. As the brush leaves a commutator bar, if all is not rightly adjusted, sparks will pass from the commutator bar to the brush. Every- thing in a direct-current dynamo or motor depends upon the accurate co-operation of the commutator and brushes. If spark- ing is allowed to go on, it deteriorates the metal parts of the com- mutator, and the edges of the bars cease to be straight and they lose their definition. The effect of this is to increase the spark- ing and with it the damage to the commutator until no remedy short of turning off the surface of the commutator in a lathe will restore the machine. The brushes may suffer in the same process of sparking. Their trimming and shaping is comparatively easy. The commutator is the critical thing. The sparking between a commutator and the brushes is injuri- ous to the commutator. It is trouble enough at the best to keep a commutator in good order. To turn it down on a lathe, to trim the brushes and set them is a piece of work requiring time and trouble. The dynamo is also out of commission during this time. A commutator loses metal every time it is turned down, and if this is often necessary, it will sooner or later succumb. The main cause of sparking is very simply presented. Fol- lowed out to its full scope, the subject may become rather in- tricate. The cut, Fig. 311, represents conventionally a part of a com- mutator whose bars are lett-red c cl e f g. n' n gives the position of ihe end of the neutral line. In the position shown, the arma- 426 ELECTRICIAXS' HA:SDY BOOK. ture windings indicated at W and X are sending current in the direction of the arrows. They lie in the left-hand half of the field. The coils U and Y lie in the right-hand half, and send current in the opposite direction. The currents join at the neutral point, and flow off through the brush. It will be observed that one coil V is short-circuited and is "dead," because the brush bridges over the space between the commutator bars. As the armature turns, this coil is suddenly thrown into series with T and U and the whole of the right-hand division of the armature. Owing to self-induction, the coil V resists the passage of the current, and a spark goes across from E to the brush which has t now left it. Such sparks ruin the commutator sur- face. But now suppose the brush advanced a little. When a coil is short-cir- cuited, it is no longer at the neutral point, but is in say the right-hand half of the armature. It is cutting lines of force, and electromotive force is generated in it of the same polarily as that in the other coils of that half. As it leaves the brush, it is ready under the influence of that electromotive force once more to start into action and carry its current. The sparking now does not take place. If on the other hand the. brushes are swung back, then the idle coil is still in the left-hand half. It cannot be called idle any more, as it is generating an opposite electromotive force, and it intensifies the sparking action by its own supplementary spark- ing. These facts must be firmly fixed in the mind. A neutral line exists in any active armature, which may or mav not correspond with the position of the brushes. If the brushes are advanced Fig. 311.— ?HORT-riRCUITING OF AX Armature Coil by a Brush. CARE OF DYNAMOS AND MOTORS. 427 from this position, they carry the neutral line in their direction, but not as far as they themselves go. If the brushes are advanced, they therefore throw some coils into the wrong division of the armature. Such coils generate counter electromotive force and reduce the output. If the brushes are retarded, there is bad sparking; if they are advanced too far, it is less than if retarded. The ideal would be to have the brushes in the neutral line if there were no sparking there. If a dynamo has a weak field, the distortion of the field will be excessive. We have seen how the current induced in the coil V of Fig. 311 operates to prevent sparking. The distortion of the lines of force throws this coil into a weak portion of the field, and its action is greatly weakened. The sparking may be very hard to avoid in such case. A relatively strong field acts to prevent sparking. The fewer the commutator divisions on a given armature, the greater will be the electromotive force - represented between any two of them, and the greater will be the inductance of the wind- ings included between them. One of the causes of sparking is disposed of by giving plenty of divisions to the commutator. The brushes keep the coil V of Fig! 311 short-circuited while it is passing the neutral line and electromotive force is being im- pressed upon it. This electromotive farce as already described acts to prevent sparking. It is therefore an. object to keep this coil short-circuited long enough to get it working before it is thrown into series with the others. One cause of sparking is too thin a brush or a badly-trimmed one. The idle coil must be in a strong field to be efiicient in prevents ing sparking. The pole pieces should be so shaped as to give a strong field at the positions of the brushes just forward of the neutral line. Short-circuiting of one of the sections of the winding may cause sparking. The mere mechanical disturbance of shaking or jump- ing of the brushes is a cause of sparking. Starting a flachine. — Before starting a new machine or one which has been for some time out of service, the armature should be turned by hand to see if it is free to turn and has not got gummed bearings or too tight bearings. As it thus turns, the 428 ELECTRICIANS' HANOV BOOK. windings should be observed to see if any v/irGs rub against tto pole pieces and if the axle of the avinat^^re is ceii'tere<'i at both ends with respect to the cavity between tlu poles. The oiling apparatus must be examined to see i? it is in pe. feet condition, as lack of oil in a ma^lii i3 ot* ^lIcL LioL^peed type may be disastrous. An iron oil can is not e gccd o-ij tc use, on account of the attraction the field may exert upon it. If oiling rings are used, they must be watclied to see if tliey operate freely. In oiling, no drops of oil must be alio •,.::! to fall oiitside of tht proper places. Especially none must luucL the brush holder, windings, or commutator. In starting, do it slowly, especially vvitli a^^r/ machines and new belting. It is well to run a ma.hj,^^ for a whil^ without load to ascertain if the media nifnl parts are perfect and in ad- justment. This will show whether u.e beamings are slack or the armature out of balance. If the machine ha.s to be moved along its base to tighten the belts, its axis must be kept at right angles to the line of belting. Starting a Dynamo. — ^The brushes shouid be lifted from the commutator, and only .brought down on the commutator after the full speed is attained. The brushes must not be pressed too strongly against the commutator. If metal brushes are used, cop- per dust is apt to be formed, which may short-circuit the commu- tator bars. After a run, when the machine goes out of action, always lift the brushes. Armature Running. — If a machi^ie's 'd'-matiire is running at full speed, it should, if the power is thrown off, continue in mo- tion for a minute or more by its own inertia. Balancing of Armature. — An armature may be perfectly bal- anced for all positions when stationary, and yet not be in balance when in motion. Want of such balance causes vibration, which may shake a whole building. If a machine runs quietl5% there is no need of further investigation of its rotary balance. To test it, however, the machine may be suspended in the air by its field eye-bolt and run as a motor, or less advisedly with a vertical belt. Any lack of balance will throw it into vibration. Vibration often produces sparking on rhe commutator. Cente ing of the Armature. — The magnetic pull exercised by CAKB OF DYNAMOS AND MOTORS. 429 the field of a machine or. its armature may pull the latter out of center by springing the shaft. The pull has been known to spring the field magnet frame so that It gripped the armature and held It mechanically, an-esting its motion. This point may be investi- gated in a new machine by passing current through the field and observing the ?rtio]) or. the armature and field. Armature Out of Center. — In bi-polar machines this trouble is of less account than in four-pole machines. In the latter, gen- erally tour windings In parallel with each other are on the arma- ture. If the armature is out of center, the electromotive force impressed on the VA'indings will be unequal. Local currents will occur to restore balance; the wires wi'l carry varying currents; sparking will ensue, and even on open circuit the armature may become heatecL. It fcllovrs that too great attention cannot be given to this point. Foucault or Eddy Currents may exist in the armature core, due to insufficient lamination, too thick disks, or to bad insulation between the disks. Nothing can be done for it except to rebuild the armature. They may exist in the armature windings, due to too massive conductors. Subdivided or laminated conductors tend to minimize ihe trouble; sinking the conductors in slots in the core, and rounding off the edges of the pole pieces, thus alter- ing the distribution of lines of force, also prevent it. The latter cure may introduce other troubles. Foucault currents are de- tected by the heat they produce. The hottest place is where they exist. Heating of Field Coils. — In shunt-wound machines this is liable to occur from too high a current being sent through the field. Ever^- machine i?^ built for a definite maximum voltage at a definite maximum speed of rotation. If a machine operates with proper \oltage, yet lequlres too high a speed for operation, or with proper speed h^s too low a voltage, the weakened field thus indicated v:iU be apt to cause sparking at the commutator. But the reverse trouble of too low a speed for the given voltage or too high a voltage for the given speed indicates too strong a field. In charging storage batteries, the latter trouble may arise. Reducing the current is useless, as the field magnet current de- pends on the potential difference at the terminals. The voltage of 430 ELECTRICIANS' HANDY BOOK. the outer circuit must be reduced. The trouble makes itself evi- dent by heating of the field coils and pole pieces. If one field coil is hotter than the other, look for a short circuit in the colder coil. If there is such, it will cause an excess of current to pass through the field coils, thereby heating the per- fect one. Break in the Field Winding, — This simply brings a dynamo out of action. A shunt motor may be ruined by such an accident, as its speed will, unless restrained, increase until it wrecks the armature. The trouble is easily found by a galvanometer and dry battery or by a strong current and electric lamp. If no cur- AAAA/^v\AAV\/V\AAr 'JWV\NWvVA^Wv- FiGS. 312 AND 313.— Short Circuits in Dynamo. rent will go through the magnet windings, there is evidently a break. Its repair may involve rewinding of the field. Siiort Circuits in Field Winding. — These are best detected by measuring the resistance of the field. They operate to weaken the field, to lower its resistance, so that it takes more current at equal potential and may give the remaining wires more load. The weakening of the field lowers the potential of the machine, and thus may often save the excessive load being given the wind- ings. The most complicated cases of short-circuiting in the field occur in compound-wound machines. Diagrams of some such accidents are given in which dotted liiies indicate short circuits. Fig. 312 shows a short circuit between the middle of the shunt winding and the beginning of the series winding. In this case one half of the shunt coil is thrown out of action, and the other CARE OF DYNAMOS AND MOTORS. 431 half may get heated from excess of current. Such a short circuit affects the compounding of multipolar machines more than it does that of bipolar machines. It results in over-compounding. The next diagram. Fig. 313, shows the whole shunt winding thrown into parallel with the series winding. This is due to a short circuit between the beginning of both windings. The short circuit also operates to cut out the armature, and thus throw the machine out of action. In Fig. 314 is shown the outer end of the shunt winding con- nected by a short circuit with the inner end of the series winding. If the short circuit is of low resistance, it short-circuits the series winding, and the machine has to operate as if shunt-wound. -M^AAA/WV\AMAA/W ({D^mwm -V\MMWV\A, Figs 314 and 315.-Short Circuits in Dynamo. Sometimes a shunt-wound machine has the outer end of the shunt winding connected to the inner end of the series winding. In such a machine a short circuit between the outer ends of the two coils will short-circuit the series coil, and the machine acts as if shunt-wound. This condition is shown in Fig. 315. The exact condition shown in Fig. 314 is brought about. Earthing Dynamo Frames.— The windings of a dynamo or mo- tor must be carefully insulated from the earth. The frame, on the other hand, is to be connected thereto. It is pretty sure to have such connection in any event. Small motors may be in- sulated altogether. To test grounds in the windings, a wire, with an incandescent lamp in circuit, may be connected to one of the bolts of the frame of the machine, and its other end brought into contact with any exposed part of the winding. The hands 432 ELECTRICIANS' HAXDY BOOK. should be carefully protected, either by using thickly-insulated wire or by India-rubber gloves. If the lamp becomes hot, a ground exists. Two lamps in series can be connected across the two main leads with the wire betw^een them connected to the earth. If the windings on both legs of the field are in contact v/ith the core, the lamps will light feebly. Short Circuits in Outer Circuit. — Sometimes a short circuit on the outer circuit may happen when no current is being de- livered, and the generating machinery is at rest. This takes away current from the shunt field circuit, and the generators will not build up a current. Nothing can be done except to find the trouble and rectify it. It is a very disagreeable occurrence, as the short circuit may never be even suspected until the time arrives for starting the generators affected by it. A short circuit on the outer circuit may be occasioned by neg- lect of customers to turn off their lamps or motors. A number of such left connected in parallel on an inactive circuit will inter- fere with the starting of the generator, just as if an accidental short circuit occurred between two leads. If lamps are kept burning and motors going until the current is shut off at the generating plant, they should be switched out by those in charge. Automatic cut-outs can be used for motors, w^hich will cut then out when the current ceases. A temperature of 72° F. (40° C.) above the atmosphere is given by some authorities as allow^able for dynamos and motors in action. Wrong Connections in Compound Dynamos. — When a com- pound-wound machine refuses to work, if a dynamo will not build up, and if a motor does not turn unless a short circuit such as has been just described can be detected, there is reason to suspect that the series coil of the field is wrongly connected. If inverted in connection, it will work in opposition to the shtmt coil. This destroys the field excitation. Such wrong connection simply re- duces a compound dynamo to inaction. Btit in the case of a compound motor there is danger of an accident. It is necessary to use great care in starting compotmd-wound motors lest s-uch an accident should occur. Turning Down a Commu'a'^or requires special care, r.s copper CAKE OF DYNAMOS AXD MOTORS. 433 is a very tough metal, and the breaks between the bars may make the tool jump a little, causing irregularity in the work. A dia- mond-pointed tool taking a very fine cut is recommended. The fine cut is desirable, not only to secure finish, but to avoid using up the commutator. Were there enough of such work to be done, a milling machine would be useful. After the turning, the spaces between the bars should be brushed over to remove copper dust from the mica. Sandpapering and Smoothing a Commutator should be done when it is cold. If the mica tends to project, it will project more when the commutator is cold than when it is hot. This enables the sandpaper to better remove it. Another good point to be noted is that the increased projection of the mica causes its re- moval by sandpaper to be done with less expenditure of the copper of the commutator bars, which it is highly desirable to save. To Sandpaper a Commutator, a block of wood cut to the curve of the commutator, so as to take in at least one-third of its length, may be employed with which to apply the sandpaper. Tallow must be applied to the paper or armature surface, to enable the sandpaper to cut the mica. Rarely use a file, as this is apt to wear the surface of the commutator unevenly. If a commutator has to be turned off, care must be taken lest short circuits form by copper being crowded across the intervals between the bars. If one or more bars of the commutator show excess of wear, it indicates bad contact between the connection from the bars in question to their respective coils, or some other bad contact to be sought for in or about the coil in question. Filing a Commutator, — This is sometimes done. A bastard- cut file is used, and the armature is slowly turned during the filing. Filings must be carefully removed from between the commutator bars; a sharp hook-shaped tool will do this. A bel- lows will blow away all loose filings. Sliort Circuits. — A short circuit between the primary and secondary coils of alternators, alternating current motors or trans- formers, is dangerous to life. It may be brought about by light- ning stroke. It will destroy lamps, motors, and the like by the high potential thrown upon the line. The cases in which 434 ELECTRICIANS' HANDY BOOK. this may occur are such as the following: The primary or high-tension line of a system may have a single ground connec- tion, and the low-tension or secondary may also have one. As long as the two circuits are insulated from each other, no harm need result, although a ground on a high-tension circuit is a perpetual menace to life. If by any cause such as a stroke of lightning an arc is started across from primary to secondary, this arc will connect electrically the two circuits, will burn out lamps and motor windings, and blow out fuses on such portions of the low-tension secondary circuit as lie between the arc and ground connection. Alternating-current dynamos are free from one source of trou- ble, the commutator. If short circuits are produced in their coils, the absence of a commutator involves the absence of sparking. The latter, while injurious to the commutator, has the attendant merit of being an indicator of trouble. There is no such indi- cator in alternators. Heating of the coil affected by the short circuit is the sign of trouble. Modern alternating-current generators generally have a re- volving field and stationary armature, sometimes called the stator. When the stationary armature is used, breaks in the windings very seldom occur, and are easily found by exploring with a source of current, and any current indicator. A dry cell and a simple galvanometer may suffice, if the machine is absolutely out of action and disconnected from every possible source of cur- rent. Such disconnection should be regarded as imperative in al- most all cases where explorations have to be conducted. Short circuits in the armature windings bring about local and dangerous heating. Such reduce also the current output of the generator, as the defective coil is no longer effective, and by its high current intensity operates to demagnetize the field. If the place where the windings are short-circuited has been found, and is accessible, a temporary repair may be made by pushing in mica between the wires. Sometimes the coil can be safely cut out and the neighboring coils can be connected across the in- terval. This is safe when there are enough coils on the arma- ture. Short circuits in modern alternators can hardly occur be- tween separate coils. CARE OF DYKAJIOS AiYD MOTORS. 435 Short Circuits Between Ariralure Windings an J Frame are dangerous not only to the generator, but to the lives of the oper- atives it a ground circuit is on the line. A 200-volt tension has killed in several recorded cases. A common practice is to ground the framework of alternators. Then if a man touches the frame of a machine in which such dangerous short circuit exists, he is merely in parallel with a portion of the frame and receives no injury. Were the frame not grounded, he might be killed. If no ground circuit exists on the line, such a contact of windings and frame may remain undiscovered indefinitely unless watched for. It would be well to explore inactive and disconnected ma- chines to detect such. Alternator Brushes. — Oh an alternator never lift a brush whilo the "machine is working, except where there are several in the set. Then one may be raised at a time, some being kept al- ways in contact. It is bad policy to work much around an al- ternator when it is in action. Trouble in Rotors of Alternators may be caused by bad condi- tion of the collecting rings, which may be dirty, unevenly worn, or oval. A periodic breaking of the circuit at the brushes due to such causes will occasion sparking. The cure is to attend to the rings and brushes and remedy their defects. The large num- ber of poles of the usual type of alternators makes short-circuit- ing unlikely in them, as the copper is distributed in smaller masses, and the insulation is not so subject to wear. Another effect of this subdivision of windings is that if one winding gives out, it can often be cut out and short-circuited, and the machine kept in action until the chance occurs to replace or re- pair the defective coil. Cutting out a pole merely exacts a little higher magnetizing current, and no sparking results. Self-Starting One-Phase flotor.— To make a one-phase alter- nating-current motor self-starting, a capacity is introduced into the main exciting circuit or an inductance into the shunt circuit. This splits the current and delivers a two-phase current to the motor, establishing a rotatory field. One of the most frequent sources of trouble is to be found in this capacity or inductance. If either of these is out of order, the current will not be properly split, and the motor will not start. A condenser with iron plates 436 ELECTRICIAI^I^' HANDY BOOK. immersed in canst ic-soda solution is sometimes used as a starting capacity. By evaporation it iid liable to change. The inductive eoii is to be preferred to such a capacity apparatus. Local Heating of the Windings of the Stator in an induction motor indicates a double siiort circuit eitiier in a single coil or in two neighboring coils. If wound for Y distribution, an interrup- tion of one phase will interfere with the running of the machine. If the load is light, it may go on ay a synchronous motor. Some- times the beginning and end of a coil are interchanged in their connection, so as to reduce the phase difference to 60°. This inter- feres with the running of a three-phaf^^e Y-connected motor. Induction flotor Rotors. — The short-circuited disconnected ro- tor of induction motors seldom gives trouble. In the older type the rotor would sometimes l)ecome so hot as to melt the solder on the connections of the windings. This by opening the circuit would bring the machine to rest. For this reason it is the Lest practice to use no solder on the joints, but by hard metal coup- lings or brazing to secure heat-proof joints. Synchronous Motors, v/hether single or polyphase, should be speeded up before loading, and the load should be gradually ap- plied only after the full speeel has been attained. They mtist not be overloaded, or they will be brought to rest. These motors have the feature of maintaining a constant speed as long as they run. Polyphase Induction Motors, which can endure an overload within certain limits, lose in si^eed as the load is increased, but are self-starting even with a load. Field riagnets of Alternators often constitute the rotating part, which exposes their windings to a certain strain and wear, from which they are exempt in direct-current machines, Btit their windings per pole are generally lighter than in direct cur- rent machines, which operates to save them from the action of centrifugal force and shocks and strains at starting and stopping. If sparking appears at tbe collecting rings and brushes, it may be dtie to periodical breaking of the field circuit, bad material of rings or brushes, dirt or oval rings. If the trouble is in one magnet pole, it can be sliort-circtiited as a temporary expedient. Two-Phase Operation.— If a two-phase station is operated by ■"wo single-phase transformers in parallel with each other, if one CARE OF DYKAiUOS A:: D MOTOR^i. 437 breaks down the station must be run on one phase until repairs are effected or the transformer is replaced. If a three-phase sta- tion is operated by three single-phase transformers in mesh or delta connection, and one breaks down, the station can still be run three-phase with two transformers, although on full load they will be greatly overloaded. Breakdowns in Transformers seldom occur. The principal on3 to be apprehended is a short circuit in the high-tension wind- ing. This in a dry transformer will slowly carbonize the insula- tion, and short-circuit a more or less considerable part of the winding. With disk-wound transformers, such as those using pancake coils, this trouble is minimized. If the short circuit operates to throw many turns out of action, the potential of the low-tension coils will increase perceptibly if such are the sec- ondary coils. This is an indication of trouble which will be seen on the voltmeter or possibly in lamps. The transformer will be caused to rise in temperature, and will pass more current when unloaded than it should under normal conditions. If it is a step- up transformer, a short circuit in the high-tension secondary will lower the potential. Care of Transformers. — These must be kept dry. If immersed in oil, the oil will take care of them. If air-cooled transformers are to be stored in a place where moisture is to be feared, chlor- ide of calcium may be placed on trays or saucers inside the cases. This has to be renewed as it gets moist. Quicklime is also avail- able for the same purpose. These precautions do not apply to dry storage apartments. Dust is injurious especially to oil-cooled transformers, as it tends to thicken the oil. Dust may make short circuits. Transformers long out of use must be started with the precau- tions exacted for new ones. As long as transformers are working, they should not be touched. If anything is to be done to them, they should be dis- connected. This applies especij\lly to operations on the high-ten- sion side, such as putting in safety fuses. Oil for Filling Transformers is a petroleum product specially prepared for the purpose. It should have a very high flashing point; 500 "* F. (260^ C.) is given as a proper temperature of 438 ELECTRICIANS' HANDY BOOK. flash. Before pouring it into the tanks of the transformers, it should be heated to 160° F. (71° C.) It must be poured or pumped in carefully, and must be kept off the cable ends. It must rise over the tops of the coils. The insulation absorbs more or less of the oil, on which account it is advisable to in- spect a newly-filled transformer from time to time, to see if it needs more oil. After three or four weeks all should be absorbed that will be taken up. Oil filling must be done under cover if there is any fear of water getting in. Every two or three years, renewal of the oil is advisable. New transformers should be inspected, to see if any foreign bodies have lodged within them. Such should be removed. Moisture in Transformers must be watched for carefully in their first run of ten or twelve hours, as any present is expelled during this period and there is danger of its collecting into drops. If such appear, the transformer must be cut out of the circuit and dried as well as possible with a dry cloth or blotting paper. After such partial drying the heat of the core will complete the oper- ation rapidly. Inspection of Transformers. — Every four weeks transformers should be inspected. The high- and low-tension safety fuses should be examined. After dusting it off the cover should be removed, the contents cleaned, and the capacity of the safety fuses should be noted. Short Circuits in * Transformers. — If a transformer strikes across from hi^h tension to low tension or to the case, supposed in this instance to be grounded, the safety fuses should melt and cut out the transformer. Short circuits in the same coil, which are generally to be apprehended in the high-tension coils, are not so easily detected. The noise of the transformer is apt to increase in such case, the oil gets hot, and the iron case is warmer than usual to the hand. The heat impairs the quality of the oil. The transformer should be disconnected and repaired. There is even a possibility of an explosion, as the heated oil may give off gas enough to be inflamed on the melting of a safety fuse. Transformers need to be well protected from lightning. CHAPTER XXV. STATION NOTES. Temperature of Dynamo or Motor.— It is not easy to accurately determine the temperature of the coils of a machine. The best that can be done with ordinary appliances is to put a thermom- eter as well in contact as possible with the part to be tested, and to cover the place of contact with cloths to keep in the heat. The cloth must be so disposed as to form a little chamber for th■^ thermometer bulb. The temperature of 122° F. (50° C.) is given as the maximum allowable. Cleaning New flachlne. — A new machine should be cleaned before being set at work. Chamois is a good material to wipe it with, as it leaves no threads or lint hanging to the bolt heads, nuts, and screws of the machine. Dust can be blown out of inac- cessible parts, as among the end wires back of the commutator, with a bellows. Anything in the shape of filings may go to short- circuit adjacent commutator bars. Interchangeability of Parts is expected almost as a matter of course in buying a standard American machine. It is of the greatest convenience, and often of the greatest importance, to be able to get parts on short notice. But the moment this system acts to discourage improvements, it exercises a sterilizing effect. A first-class manufacturing company should keep in stock a full line of parts of all their principal machines, but the inevitable accumulation of parts should not discourage the course of im- provement. Cotton Waste, — Never use cotton waste in cleaning a machine, as threads from it will catch in and stick to the commutator and other surfaces. Access of Air. — Air is so constant a cooling agent, that free access of it to motors and dynamos is highly to be recommended. 439 440 ELECTRICIANS' HAXDY BOOK. Boxing up or inclosing in closets of dynamos and motors, except for very intermittent service, is to be condemned. Oiling, — Before starting a machine, turn the armature by hand. This will disclose any friction. This precaution should be taken for generators or motors, unless they are in such constant use that the operative is certain that their oil feed is in perfect order. Want of oil will wear the journal boxes, and throw the armature out of center. The greatest care of the oil-feeding apparatus is requisite, because this class of machinery runs at high velocity. The best and purest oil should be used, and once a satisfactory oil is found, its use should be adhered to. If the oil is fed by wicks, the drip of oil from the bearings will show that they are in order. Oil once used should be filtered or water-purified before being used a second time. Oiling rings must be watched to see if they move properly around the axle. When the old oil gets thick, new must be added after drawing off the old. Before adding the new oil it is a good plan to wash out the oil trough with kerosene oil. A small syringe is useful for this purpose. Oil should carefully be kept off the brush holders, commutator surface and wire windings of field and armature. Ring Oiling is much used. Rings several times the diameter of the axle hang on it. Their lower portions dip into a tank of oil. As the shaft revolves, they travel around it and feed oil to its upper surface from the tank in which they dip. Bearings,— If new machines are started without due attention being given to their bearings, heating, burning fast, or even melt- ing (if babbitted) of the journals is liable to occur. A good plan is to pour kerosene through the bearings until it comes out clean. This leaves them ready for lubrication, by washing out dust or dirt which may have collected. Too stiff or tight belts cause heating of the journals. These are especially to be antici- pated in new installations. If belts have to be tightened, it should be done slowly and a little at a time, with periods of run- ning between, until they are just right. Belts too tightly stretched may even occasion melting of the bearings. Bearings are generally so constructed that with proper manage- ment they will not heat. Insufficient oil, too thick oil which does STATION XOTES. 441 not penetrate, dirt and dust getting into the bearings, are causes of heating. Cleaning with kerosene followed by the application of good lubricating oil is the cure. Safety Fuses should be inspected to see if they are tightly screwed down or clamped, and if their contact faces are clean. Insulation of Windings. — Watch for bare spots or weak spots on the insulation of the wires of the winding. If any such appear, they must be taped or insulated in some way. Broken Wires, even if thickly insulated, can be detected by the feeling, when the wire is slightly bent or moved. If the insulation shows signs of burning or of overheating, a fracture may be suspected there. Soldering. — Never use acid in soldering wires together. Anti- corrosive soldering fluxes are sold, which operate on iron and copper as well as acid, and whose use is not followed by any bad after effects of corrosion. Nails, Tacks, and Iron Filings may do harm by being at- tracted to a machine by its field magnets. Bronze spanners are recommended in place of iron ones. An iron object suddenlj' drawn to the field may, by rubbing against the armature or strik- ing the screw heads or windings, near the commutator, do harm. Screws in Binding Posts and connections should be looked after; imperatively so if their color indicates overheating. Covering Machines, dynamos and motors when not running is recommended. Emergency and Danger Signals. — A sudden rise of voltage or of current, sudden sparking at the commutator or elsewhere, heating of the windings, or smell of heated insulation should be a danger signal, and the current should be cut off from the ma- chine showing any such manifestation. A working test of dan- gerous heating is the ability of the hand to stand it. If the hand can be held on the windings, they are reasonably safe from overheating. For each dynamo the allowable heating should be learned, so that by holding the hand on it, any unusual heating can be detected. In throw^ing open the main switch in such a case, do it quickly to avoid arcing, and have the engine watched to prevent its racing as the load is taken off. 442 ELECTRICIANS' HANDY BOOK, If a safety fuse blows out, do not put in a new one until the cause of the blowing out is known and overcome. Always use the regular fuse wire. Never substitute anything except in emer- gency. Keep all switches perfectly clean as regards their contact sur- faces especially. Whenever a machine goes out of action, it should be examined and cleaned. If any oil has fallen on the coils, it should be wiped off, and oily places generally should be cleaned. Dampness is bad for machines, and dripping water may make dangerous short circuits. Rheostats and resistances are often strongly heated in use, so no combustible substance, oily waste especially, should be allowed near them. Forgetfulness and Negligence are justly said to be the cause of many troubles. Thus, in stopping a motor, or in case cur- rent is cut off without notification, it is absolutely necessary to bring the handle of the starting box back to the starting position, so as to throw in the full starting resistance in series with the armature, to save it from burning out when the current is again turned on. Keep One Hand in Your Pocket is an old and a good rule to follow when working around motors and dynamos. If good new India-rubber shoes are worn, the safety is increased. When work must be done around a high-tension active machine, such precautions as wearing India-rubber shoes are eminently proper and not at all extreme. Treatment of Electric Sliock.— The first thing to be done when a man is injured by contact with an electric circuit is to get him out of contact therewith. If possible, open the circuit or stop the generator. Drag the man away from the conductor or conductors. Do not touch his hand or any part of his skin in doing this, handling his clothes only. If they are wet, throw a dry towel or other cloth over them before touching them. Other- wise, the one helping may be shocked. Especially dangerous is touching the surface of the body directly. A physician should be called as quickly as possible. It is a grave r3sponsibility to neglect this. Do your utmost to rescue STATION NOTES. 443 the man, but get professional help instantly or as soon as prac- ticable. If the man is in contact with one conductor and cannot be re- leased, break the ground connection of his body. Push coats, a blanket, wooden boards, and the like under him. In doing this,, stand upon a dry board or coat; if you have them, put on India- rubber gloves before handling even his clothes. When thoroughly insulated from the ground; the rescuer can try to open the man's hand if that is cramped upon the conductor. This is very dan- gerous unless India-rubber gloves are worn or a dry towel is used to protect the one doing it. The two leads can be short-circuited by throwing a chain or bare 'Wire across them. This undoes the effect of a ground. It is a rather desperate thing to do. When the senseless man is free from the contact, remove cloth- ing from around his neck and his waist, so that he will be free to exercise his breathing organs and muscles of the trunk and diaphragm. He is then to be treated as if he were to be resusci- tated from drowning. Artificial breathing is to be started. Place him on his back, with a pillow or other support under his neck and shoulders, not under his head. His head will drop back, and the top will almost touch the floor. Open his mouth and hold it open, seize the tongue between finger and thumb with a handkerchief between, and draw it slowly forward. The root of the tongue must move outward as this is done. It is useless to merely elongate the tongue itself. When satisfied that this action has taken place, let the tongue slowly go back. Repeat this double movement about fifteen times a minute. While this is being done, a second person can assist the breath- ing by moving the arms up and down. He should kneel back of the man's head with face toward him, seize his arms at the forearms, press them strongly against the breast-bone, and then lift them slowly in the arc of a circle over his head, and after a short pause return them and press them against the chest again. Let the man manipulating the tongue call "one" as he draws the tongue forward. The arms are now pressed against the chest. This represents expiration of the breath. 'Two" is called, and 444 ELECTRICIANS' HANDY BOOK. the tongue is slowly allowed to go back and the arms are raised as described. This represents Inspiration. Again "one" is called, and the tongue is drawn out and tiio arms returned. Thus every four seconds the double movement i*j repeated, arm motion and tongue motion keeping time with ea(*ii other. It is easy to experiment with one'a owu tongue, and thus study the effect of manipulating it. It will be found that a drawing down over the chin as well as outward opens the windpipe. The first sign of resuscitation is natural inspiration. See that the tongue is drawn forward, so as not to hinder the access of air to the lungs. By no means pur.sue the movement of explication until the incipient natural inspiration is completed. Keep the arms raised and tongue drawn forward until it ceases. Then re- peat the expiratory movement. Do not get excited, but do all slowly and with clocklike regularity. \Vhen he breathes regu- larly, he may be brought into a more upright position, and re- moved to a bed or other better resting place. Before this a phy- sician should be at hand to treat him further. CHAPTER XXYL SWITCHBOARDS. Switchboards. — These pre vertical partitions, generally made Oi marble, on one side of which are installed rheostats, bus-bars, voltmeter switches, and connections of station apparatus, and on whose other side are installed handles for operating the rheostats and voltmeter switches, automatic cut-outs, safety fuses, switches, voltmeters, and other appliances. The general system is to make the switchboard in panels. Each panel is about two feet wide and six to eight feet high. It stands vertically, being supported by braces at its top running back to the wall of the building, thus liaving a space behind it, so that its back is accessible. Any desired number of panels are joined side by side. Panels* — Of panels there are various kinds. Some are for mo- tor control, others for dynamo running, others for operating the outer circuit, others for storage-battery charging and the like. As many as are required by the station are set up side by side, so as to present a long front. The variety of panel chosen de- pends on the work it is to do. The description of a switchboard is little more than a descrip- tion of the apparatus which it carries. Every engineer must study the switchboards in his own station, as the varieties are numerous. The General Electric Company manufactures a line of standard panels, which are so varied in design as to cover almost any de- sired case. The front and rear views of a generator switchboard panel are given in Figs. 316 and 317. Toward the bottom are triple con- tact switches, which close both leads of the circuit, and also the equalizing conductor if compound dynamos in parallel are used. 445 446 ELECTRICIANS' HANDY BOOK. On the real the horizontal flat conductors are bus-bars. On the sides of the rear view are seen rheostats operated by handles in front. Voltmeters are placed near the top. The central connection J I I 1 Figs. 316 and 317.— Front and Rear Views of a Direct Current Switchboard Panel of the switches, intended for the equalizer, is left unconnected if the generators controlled by them are shunt-wound. Reference may here be made to page 403, showing the use of the equalizer. Switchboard panels are named, from their uses, generator pan- SWITCHBOARDS. 447 els or feeder panels, and also from the current they are arranged for, whether direct or alternating. Air Switches. — Of these there are a large variety. The prin- cipal working contact made by closing them is a knife-edge contact, made by a thin copper bar on the switch going edgewise between two leaves of copper that spring against it. This makes one of the best kinds of connection, but in breaking it an arc is apt to form. To prevent this, switches are often provided with auxiliary contact blocks of carbon. These are so arranged as tO' be the first to make and last to cpen. An arc between carbon sur- faces will not draw out as will one between metal surfaces, and if it does form, it does no harm. Metal electrodes are burned and. rapidly injured by arcs. The subject comes up again under the automatic circuit breakers. Pig. 318 shows a two-pole knife- blade switch for use on switchboards. It has metallic contacts only. Oil Switches. — To avoid the for- mation of arcs, and to insure definite opening of a circuit when the switch is opened, oil switches are employed. These have the part of their mech- anism which opens and closes the circuit immersed in oil. This fea- ture insures definite action, and is rarticularly applied to high-voltage alternating-current switches. The principle of construction is shown in Fig. 319. On the right and left hand are seen two metallic rods, which descend through insulating blocks and carry springs at their lower end projecting therefrom. Through an- insulating block another me- tallic rod descends between these two, and carries at its lower end a cross piece with beveled carbon contacts C C, facing upward. This rod moves up and down. It is connected to one lead of the circuit; both side rods are connected to the other. When the central rod is raised, its carbon blocks enter between the springs and make the contact, closing the circuit. When lowered, it opens the circuit. A tank of oil, indicated in section in the diagram,, contains oil in which the mechanism shown is immersed. Fia. 318.- Two-roLE Switch.. 448 ELECTRICIAXS' HANDY BOOK. 9^^^ + The tank is placed behind the switchboard. A handle on the front of the switchboard raises and lowers the central rod. Overload and Underload Cut=Outs.— These are of two types, safety fuses and mechanical cut-outs. The subject is more ac- curately treated than formerly, and cut-outs are now expected to operate with a high degree of accuracy. In ordinary parallel circuits overload cut-outs are placed so as to open their portion of the circuit if the current becomes too strong. In series cir- cuits the cut-outs are arranged to operate by short-circuiting any lamp which may be ex- tinguished. The object is to preserve the continuity of the circuit; it is exactly the oppo- site of the function of a paral- lel-circuit cut-out. The under- load cut-out operates to open a circuit when the current w^eakens, ceases, or is reversed. Such an appliance is used in charging storage batteries. If the current falls to zero, it in- dicates that the counter elec- tromotive force of the battery is equal to the electromotive force of the charging appli- ance or circuit, and there is danger that it may increase, when current would flow back from the battery and discharge it. The underload circuit breaker opens the circuit as the current falls to zero. Cut-out and circuit breaker are practically synonymous. Safety Fuses are strips of fusible metal, whose resistance will develop heat enough to melt when a current goes through it too strong for the rest of the circuit. An ordinary type of fuse is a wire or strip of a specified cross section and length with ears at the ends by which it is screwed down to the circuit terminals. A very usual system is to mount it on a block of porcelain. The lugs or ears at the ends should be clean before it is inserted. Fig. 319.-OIL Switch. SWITCHBOARDS, 449 They may be scraped with a knife or may be filed or sandpapered before being put in place. The terminals to which they are screw- ed should also be bright. In screwing the screws in or out, care must be exercised to avoid making a short circuit with the screw driver. Sometimes the fuse is mounted in a screw cap, and is screwed on a plug in somewhat the way an incandescent lamp is screwed to its socket. Screwing the cap into place makes the con- tact of the ends of the switch with the circuit terminals. The Figs. 330 and 321 —Safety Fupe and Holder. plug cut-out, as it is called, is a very safe form. If the short circuit still exists, the socket being screwed in the plug will simply blow out again without the operator's incurring any dan- ger. The inclosed fuse is a fusible wire or metal strip embedded in porous non-combustible material within a tube. It is sprung into place between clips in some constructions, which is a very convenient and safe arrangement. The fuse being protected from the air is supposed to be more constant in its action than is the exposed fuse. It is also claimed that it does not blow out so 450 ELECTRICIANS' HANDY BOOK. quickly, requiring a sensible time to fuse. This is an advantage generally, as an excess of current lasting only a second does no harm. The fuse does its work better if it is a little slow about blowing out than if it yields instantly. The inclosed fuse does Fig. 323— I.T.E. Overload Circuit Breaker. not throw melted metal about, which is another advantage. In- closed fuses in and out of their clips are shown in Figs. 320 and 321. A small wire in parallel with the main fuse is exposed in a little circle seen on the surface of the fuse case. If the fuse melts this also melts, so that the operative knows what has hap- pened. SWITCHBOARDS, 451 Overload Circuit Breakers. — These are switches operated by electro-magnets directly or indirectly, so as to open if the cur- rent becomes too strong. The cut, Fig. 322, shows a section of the I. T. E. circuit breaker, whose initials stand for inverse time element. The instrument is shown w^ith the switch closed, and connection made by knife- blade contact. The switch arm is pivoted at the bottom and works in a vertical plane. When it is pushed up into the vertical position as shown, it is held there by a catch, seen just below the handle. The upper end is forced outward by a horizontal plunger actuated by a spiral spring. This is contained in a tube at the top of the apparatus. In the illustration part of the tube is seen broken or cut away, so as to show the plunger. The full current goes through the magnetizing coil back of the pivoted switchbar. The plunger armature of the coil is shown partly below it, and with its upper end within it. Above the armature at some distance from it is a rod which if lifted trips the catch which holds the switchbar in place. Its upper end bears con- stantly against or almost touches the back of the catch. If the currenc becomes too strong, the armature is drawn upward with increasing velocity, and strikes the loose plunger a sharp blow, driving it upward. This releases the switchbar by tripping the catch. The horizontal plunger, forced out by its spring, pushes the svv^itchbar backward, breaks the knife-blade contacts, and the bar falls back into a position about 45° above the horizontal, resting on the bracket or stop seen behind it. The position of the armature is regulated by a screw below it with jam nut, all of which is shown in the cut. This can regulate the circuit breaker so that it will open at different current strengths. It is evident that the greater the excess of current, the more rapidly will the opening occur. The armature is more strongly attracted as it rises. Therefore a very small excess of current will operate it, because if a current is strong enough to lift the plunger from its seat, it will act upon it more energetically from the moment it leaves its seat and rises toward the coil. The circuit breaker shown in Fig. 322a operates on a slightly different principle. If the handle projecting to the right is pushed upward, the circuit is opened. Two contacts, one of 4a2 ELECTRICIANS' HANDY BOOK. metal and one of carbon, the carbon placed dire tly above the metal contact, are opened if the handle is pushe . vj. The con- tacts are shown closed in the full lines of the c The dotted lines show the position of the movable parts • the contacts when they are opened by pushing up the hanJ.. The circuit can be opened and closed by hand. To give : the overload i r ' 1 Li 1 L -^oV- (O) O U'"^ > ^^ ^ c O ) Fig. 322o.— G.E.Co. Overload Circuit Breaker. Fig. 3?3.— Magnetic Release Ukd::r- LOAD Circuit Preaker. automatic action an electro-magnet is attached to the base of the apparatus, which attracts, when excited, a pivoted armature. YvTien attracted it flies up, strikes the handle, driving it up, and opens both contacts, the carbon one last. The pivoted armature is seen in the cut just below the switch arm and electro-magnet. An adjusting screw is provided to adjust it to act at any desired current wit'hin the range of its action. Underload Circuit Breakers are designed to open a circuit if the current weakens. This is often requisite, especially in charging storage batteries. A weakening of the current indi- SWITCHBOARDS. 453 cates increased counter electromotive force in the batter5^ If this increases beyond a certain amount, the battery will discharge itself through the dynamo and drive the latter as a motor. Magnetic Release Underload Circuit Breaker. — This is a form used often on motor starting boxes, as explained elsewhere. The illustration, Fig. 323, shows a switch-arm held in place by an electro-magnet against the attraction of a spring which pulls it back. A series of contact studs are shown. In the position shown in the cut, one is under its end, and current goes through the magnet. If the current weakens, the spring will prevail and will jerk the handle back and open the circuit. The spring is not drawn of the full length. Often a spiral spring like a heavy clock spring is used at the pivot end of the switchbar instead of such a one as that in the figure. Mechanical Release Underload Circuit B r eakers.— These are constructed on the lines of the overload circuit breaker just described. The principle is shown in Fig. 324. A pivoted bar carrying an armature R is held in the position shown in the dotted lines against a spring, omitted in the cut, by a magnet M actuated by the working current. When the current ceases, the magnet releases its armature, which, drawn back by the spring, trips the catch a and releases the switchbar. Often both overload and underload coils, each with its own tripping mechanism, are embodied in the same switch. To release at no load, the magnet is in series with the main current. At the handle end of the switchbar is a pivoted lever. This is pulled back by hand when the switch is to be set. Its projecting end pushes up the bar. Fig. 324.— Underload Mechanical Release Circuit Breaker. 454 ELECTRICIAXS' HANDY BOOK. so that its armature R is pushed up against the magnet M. The catch a then locks and the pivoted lever at H drops out of action. A is a metal contact and B B are carbon contacts. The contact A opens first, and then B B open. This breaking the contact on car- bon is done to avoid the formation of a metallic arc on the break, as spoken of in the case of switches. Reverse Current Circuit Breaker. — In this type the contact is kept closed as long as a difference of potential exists on the line, although it may be on open circuit. A shunt coil surrounds the magnet core, to give the reverse current release. If any potential difference is maintained on the leads of the circuit, a current goes through the shunt circuit and keeps the magnet excited, so that it cannot release its armature. A reversal of current de- magnetizes it for an instant; during the change of polarity the armature drops and strikes the switchbar catch. The bar drops and opens the circuit. Combined Circuit Breakers. — It is very usual to combine two circuit breakers in one, an overload and underload one, both actu- ated by the same outer circuit. Such circuit breakers as shown in Figs. 322 and 324 are often combined in one. Other contacts than the knife blade are used. Sometimes a series of leaves of laminated copper slightly bent, something like a carriage spring, are arranged to make contact by pressing their ends against a flat surface of copper. Circuit Breakers as Switches. — Frequently circuit breakers are used as switches, regular switches being dispensed with. Whole switchboards are fitted up in this way. Alternating = Current Potential Regulator.— This consists of an induction coil whose secondary is tapped at a number of points. For each tap a contact is provided on the dial face of the appar- atus shown in Fig. 325. The contacts are arranged in .a circle, and an arm turns on the same center, so as to make connection with them in so doing. The arrangement forms a multipoint switch. By cutting in or out parts of the secondary by means of this switch, the potential of the secondary circuit is changed. The potential of feeders is controlled by the regulator, which adds its voltage to that already impressed by the alternator upon the feeder circuit. SWITCHBOARDS. 455 Reversing switches are provided on the faces, so that the poten- tial of the regulator may work in counter and lower the voltage of the feeder circuit if this effect is required. One lead from the generator bus-bar goes straight to the dis- trict. The other lead goes to the reversing switch, and passes through some of the turns of the secondary of the regulator coil. The number it passes through depends upon the position of the Fig. 325.— A1.TERNAT1NG Current Potential. Regulator. L>-^L / ^/^ •V. ^ Fig. 326.— Direct Current Ground Indicator. multipoint switch bar. If the switch is set in one direction, the regulator adds potential to the circuit. If set in the other direc- tion, it subtracts potential. The primary coil is connected across the two main leads before the regulator is reached. In some stations the generator is run at potential sufficient only for the line having the smallest drop, and regulators are used to add to it. The action is likQ that of boosters in direct current work. Sometimes the original potential is enough for the highest drop of the system, and the regulator with reversed switch lowers it, acting like a crusher in direct-current work. 456 ELECTRICIANS' HANDY BOOK. Direct^ Current Ground Indicator. — If two wires of a lighting or power circuit, shown in Fig. 326, are connected with each other through two lamps, L L, each one of the voltage adapted to the circuit, they will show a dull red. This is because being in series they will receive far too little current. Their com- bined voltage is twice that of the circuit. From the conductor at C between the lamps a connection / is made to the ground. A ground plate or water pipe may be used for this purpose. If there is no ground upon the circuit, the lamps will take one- L©J m^ »^((sC Figs. 327 and i Alternating Current Ground Indicator. half their normal current, and will show a dull red as described. If a ground should occur on either line, one of the lamps will be short-circuited by the accidental ground and the ground between the lamps and will decrease in brightness, while the other will increase perhaps nearly to its normal. Generally speaking, the lamp which is reduced in illuminating power is the one connected to the grounded line. Alternating - Current Ground Indicator. — Alternating-current lamps in permanent connection are not favorite ground indicators, as they necessitate the grounding of the circuit. By a switch and transformers lamps can be arranged to show a ground whenever SWITCHBOARDS. 457 /^ c c C m 11 the switch is closed. The arrangements are shown in the dia- grams, Figs. 327 and 328. One embodies the use of two coils and two lamps, L L; the other that of a single lamp, L. The grounding is indicated by T. The single-lamp arrangement does everything which the double-lamp arrangement does. Ground Alarm.— Neither of these is an alarm properly speaking. They disclose nothing until the switch is closed. This is an undesirable feature. The next arrangement. Fig. 329, is a constant-alarm appar- atus. At C C are two plates of metal forming condensers. They are connected through m n as shown with a telephone t on the line going to the ground T. As long as there is no ground on the line, no cur- rent goes through the telephone. If a ground occurs, a current goes through it, causing it to pro- duce a humming sound loud enough to be heard through a good- sized room. ^7. Fig. 329.— Telephonic Ground Indic tor tor Al- ternating CuRRE^T Circuit. CHAPTER XXVII. VOLTMETERS AND AMMETERS. The Voltmeter is a galvanometer whose scale is graduated to read directly the potential difference at its terminals in volts. Certain conditions wnich it has to fulfill are determined by the use it is to be put to. It must be so constructed as to give the voltage between different points of a circuit over which a cur- rent is passing. This it must do without affecting appreciably the current. As it has to be connected in parallel with the por- tion of the circuit to be tested, it follows that a certain propor- tion of the original current will pass through the voltmeter, and the main current will be diminished by that amount. Therefore it must operate with an exceedingly small current — one so small that it will count for nothing. Voltmeters are used principally in engineering practice and on reasonably large current circuits. The current which goes through the voltmeter in such cases is treated as infinitely small. Other cases arise in laboratory practice where the cur- rent passing through the coil of the instrument has to be taken into consideration. One type of voltmeter is actuated by coils of wire thrcvigh which a very small current passes. The wire of the coil is ex- ceedingly thin, and the apparatus is so delicately made, balanced, and journaled that it operates under the effect of an almost infin- itesimal current. The elements of the Deprez-D'Arsonval galvanometer embodied in a portable instrument constitute the essential parts of the voltmeter generally adopted in American practice. The field is established by permanent magnets, with a circular opening be- tween the poles for the coil to rotate in. Within the coil, and concentric with it and with the cylindrical opening, is a cylinder 458 VOLTMETERS AXD AMMET:t]RS. 459 of iron, which operates to reduce the air gap, thereby intensifying the magnet field. The iron cylinder is fixed in position. This leaves an annular or ring-shaped opening between the core and the pole pieces, unobstructed except where the support of the core comes. This only takes up a few degrees of the circle. The ceil turns in this space. In Fig. 330, SS and NN indicate the field poles, PPthe pole pieces, and the cylindrical core is shown be- tween them. The coil is also indicated. It moves freely in the space between core and pole pieces, touching neither. There is no difficulty with the core support, because the coil never turns through a full half-circle, and therefore never touches the sup- port. To understand clearly the relations of core coil and pole pieces, Fig. 468 on page 611 may be referred to. This shows the Deprez-D'Arsonval galvanometer, in which the same system of field, stationary core, and re- volving coil is used that appears in the type of voltmeters de- scribed here. In old practice the voltmeter was a high-resistance galvanom- eter with a compass needle actu- ated by the earth's field and by the coils of the instrument. Such a galvanometer had to be placed horizontally with its needle in the magnetic meridian when no current was passing. The mod- ern instrument is independent of the earth's field, so that it can be set up without regard to the points of the compass, and ver- tically or at an angle. Weston's Voltmeter. — This instrument is very . extensively used. Its moving part is a small rectangular coil of wire car- ried on a shaft whose ends are supported by jeweled bearings. To the ends of the shaft are attached the inner ends of spiral springs, exactly like the balance or hair spring of a watch. The ends are insulated from the axle to which they are attached, and one end of the coil wire is connected to the inner end of one spring and the other end to the inner end of the other spring. Fig. 330.— Field or the Weston YOLTMETEll. 460 ELECTRICIANS' HANDY BOOK. These connections are electrical, and the springs serve to conduct the current to the little coil without preventing it from rotating as the current passes through it. They are leading-in springs, in the sense that the platinum wires in an incandescent lamp are leading-in wires. If a balance wheel of a watch were replaced by the coil and two hair springs were attached to the axis, one at its top and one at its bottom, it would give the mechanical combination of the apparatus. It is shown in Fig. 331. The field is that of a horseshoe magnet strengthened by a cyl- inder of iron held within the coil and concentric with it as regards its axis of rotation. The cylinder is carried by an arm extending from the base or frame of the instru- ment. It is so placed as not to interfere with the swings or partial rotations of the coil. The instrument has a long index whose end moves over a graduated scale, the arc of a circle, and divided into di- visions representing and calibrated for volts. The soft-iron core and the shape of the pole pieces secure a uniform field and tend to give a uni- form motion for increase of current in the working coil. Damping Coil, — To render the instrument **dead-beat," which means that it shall at once give its reading without having its needle swing back and forth a number of times before coming to rest over the proper mark on the scale, a special damping coil is used. This consists of a coil of insulated wire short-circuited on itself. Over this is wound the active winding, which consists of a number of turns of fine copper wire, whose ends connect with the springs. This double coil is mounted in the field, and its sides move through the annular space between the soft-iron fixed core Fig. 331.-^ore, Coil, Field Poles and Leading-in Springs of Weston's Galvanometer. VOLTMETERS AND AMMETERS, 461 and the magnet poles. As it moves under the influence of a cur- rent passing through its active coil, eddy currents are induced in the closed circuit of the damping coil, and these oppose its motion and "thus prevent swinging back and forth. The instrument goes at once to its proper reading, and shows the voltage at once. Air=Vane Damping.— An aluminium vane is sometimes attached to the index. As the latter moves, it sweeps the vane through the air. The resistance of the air operates to mechanically damp the movements of the index, and to make it still more aperiodic. Fig. 3;j2.— Plan op the Empire Voltmeter. The critical point about calibrated instruments of this type is to secure a permanent and unchanging magnetic field. This depends on the permanent magnets retaining their magnetism unchanged, year after year. To secure this feature, they must be made of a proper quality of steel. Much secrecy is observed as to this point. They also are not magnetized to saturation; about three-quarters saturation is good practice. Empire Vol meters. — The cuts, Figs. 332 and 333, show the con- struction of the Empire voltmeter. Its general construction re- calls the D'Arsonval instrument more than does that of the Wes- ton voltmeter. The characteristic feature is that the needle is 452 ELECTRICIANS' HANDY BOOK. carried by straight phosphor bronze wires kept strained hy spiral springs. These wires by their torsion act to keep the coil in a neutral position, and to bring it back to zero if it is turned away from it. They are also the leading-in wires for the cur- rent. Fig. 333.— The Empire Voltmeter. From the inner surface of the field-magnet poles four flat plates : of iron project. These form a strong field, made still stronger by a disk-shaped core supported between them. The coil includes the disk within its open center, but touches no part of the field. The suspension wires, which are also the leading-in wires, as stated above, are kept strained by springs contained in little tubes at the opposite ends of the support. Standards attached to VOLTMETERS AXD AMMETERS. 4G3 the four pole pieces carry cross pieces, to which the spring sup- ports of the suspension wires are attached. The connections of the instrument are omitted in the diagram, which is designed to show the charar^teristic features only. Graduation of Voltmeter Scales. — The instrument is put in parallel with a standardized voltmeter, and the value of its full reading is noted. This may read widely different from the truth. Suppose the scale is to be graduated to 150 volts, and that 30 volts bring the needle to the end of the scale as yet un- marked. At this point a mark is made. The potential is now lowered to 28 volts, and another mark is made, then to 26 volts, and so on. This gives fifteen divisions on the scale. Each di- vision is evenly divided into ten divisions, thus giving 150 di- visions. The 150 divisions correspond to 30 volts. Resistance in series with the coil is now placed in the interior of the case, in the shape of spools of fine insulated wire. It is tested and added to or reduced until a potential of 150 volts carries the index ex- actly to the end of the scale. The instrument is then standardized and ready for use. General Notes on Voltmeters, — The index is counterpoised so as to be perfectly balanced. Its one end forms the pointer. Its other end, prolonged beyond the suspension axis, sometimes has a thread cut upon its end, on which counterpoise nuts are screwed back and forth to secure perfect balance. Sometimes this end of the index is bent at right angles, and has counterpoise nuts on the bent portion as well as on the straight portion, to give greater power of securing a perfect balance. To secure an approximately even motion of the pointer, so that a given change of voltage shall cause the pointer to move over the same number of degrees at all parts of the scale, the com- bined effect of the even magnetic field and of the springs is relied on. Cardew Voltmeter. — This instrument, which would seem pecu- liarly well adapted for alternating currents, is not as much used as is Siemens's dynamometer. It is really an ammeter of very high resistance. Its action depends upon the expansion of a wire through which a current passes. This wire expands with heat and contracts with reduction of temperature, and the temperature 464 ELECTRICIAN^ S' HANDY BOOK. changes depend on the current passing through it. Thes^ changes depend on the changes in voltage at its terminals, and it is a voltmeter in practice. In its essentials a wire is attached to a rotating shaft which carries an index. The other end of the wire is attached to a point a foot or more distant from the shaft, and is stretched. As it changes in length, it turns the shaft. The latter is provided with an index, which indicates the changes in length of the wire. The simplest construction has a straight wire stretched through the center of a brass tube. It has to be calibrated by trial for various electromotive forces. After enough readings for definite values have been found, others may be intercalated between them. If calibrated for continuous electromotive force, its readings for alternating electromotive force will give the effective value. Its sensitiveness is greatly increased by using a longer wire. To keep the size of the instrument within practical limits, the long wire is carried back and forth over pulleys made of bone. In a recent example, the wire of platinum-silver alloy was 0.0025 inch in diameter and 13 feet long. It passed up and down eight times. This brought each stretch of it to a length of about 18 inches. The terminals of the instrument connect with the ends of the wire. When connected to the circuit whose voltage is to be measured, the thin wire very quickly acquires the full temper- ature due to the current produced by the voltage. The thinness of the wire enabling it to grow hot or cold with great rapidity makes it very quick-reading, or almost dead-beat. Such an in- strument can measure voltages from 30 to 120 volts. For higher voltages a resistance is added. This is sometimes made of ex- actly the same wire and stretched through metal tubes, as in the instrument itself. There is considerable vagueness in the readings near the zero point, and it is considered inaccurate in the upper part of the scale. In the construction shown in Fig. 334 a long wire C, carried up and down a frame four times, is used. The current passes through this, and its changes in length draw the little pulley at its upper central bend or bight up and down. By wheel mechan- VOLTMETERS AND AMMETERS. 465 ism these movements cause an index like a clock-hand to re- volve on a dial, which in the cut is facing away from the reader. A pulley P, around which the wire passes and to which it is se- FiG. sai.— Cardew Voltmeter. cured, turns clockwise as the wire lengthens and vice versa. This pulley actuates the index. The spring S drags the pulley around clockwise; the contracting wire drags it the other way. A larger view of the pulley P is given, to show how the wires are attached to it. The index and scale are omitted from the cut. 466 ELECTRICIANS' HAXDY BOOK. Hot=Wire Instruments. — The Cardew voltmeter is the parent of the hot-wire instruments. It has tended to go out of use of late years. It is affected by alternating as well as by direct cur- rents, and this operated to keep it in use. Hot-wire instruments have had qtiite extensive application and are still in use. The Stanley Hot=Wire Voltmeter. — A wire is secured across Fig. 335.— The Stanley Hot- Wire Voltmeter. the upper part of the instrument, and is held horizontally in general position. Two leading-in springs or connections descend from above and carry current to it, so that the current passes through the few inches of wire between the ends of the leading- in connections. From the center of this actuating wire another wire descends to the bottom of the instrument, and is secured there. This wire is approximately at right angles to the actu- ating wire. To the left of its center is the index with horizontal axle, carrying a pulley or drum ^xed to it. A filament from the VOLTMETERS AND AMMETERS. 467 center of the vertical wire passes around this drum, and has its other end secured to a spring. Thus this spring keeps the sys- tem of filament, vertical wire, and actuating wire in tension. A current passed through the actuating wire heats it and causes it to expand. A very slight expansion causes its center to descend a measurable distance, on the elbow-joint principle. This magni- fication of motion is repeated by the vertical wire, so that an infinitesimal change in length of the actuating wire by means of the two magnifications causes the index to move a visible dis- tance. Tig. 336.— The Stanley Hot- Wire Voltmeter. The cuts, Fig. 335 and 336, give the general view of the working parts of the instrument. The hot-wire instrument is unaffected by any electro-magnetic fields, and hence is peculiarly well adapted for places where such fields exist. Ammeters.— The word ammeter is an abbreviation for ampere- meter. It is an apparatus for measuring current rate. Any cali- brated galvanometer with its scale marked so as to read amperes is an ammeter. Total-Current Solenoid Ammeter.— The first instruments were constructed so that the entire current passed through the actu- ating coils. The cut, Fig. 337, shows a modern total-current in- strument. A coil of heavy wire is secured to the base of the in- 468 ELECTRICIAXS' HA^^DY BOOK. strument. Its axis is vertical. Through its center a core of iron is free to play up and down, being suspended from the end Fig. 337.— General Electric Company's Total CaRRENT OR Solenoid Ammeter. of a bent lever. The latter has one end prolonged to form the index. As more current passes, the core is drawn downward, and the needle moves over the scale in one direction. If the cur- FlG. 33Ta.— ^MMET- H C >nkec- 1JOS Wt I H MIUNT. Fig. 338.~Ammeter Shunt. rent diminishes, the core rises and the needle moves the other way. Although such an instrument may work with total cur- rent, it may be connected in shunt with a conductor, so that VOLTMETERS AXD AMMETERS. 469 only iDart of the current will pass through it. In this way it Fig. 339.— Transformer for Stanley Hot Wire Ammeter. can measure a much larger current than its coil could carry. The coil attracting a plunger is called not quite correctly a sole- noid. Shunted Ammeter. — This instrument is a shunted galvanometer, which is calibrated to read amperes. The am- peres are those which go through the shunt and actuating coils. The indicating portion of the apparatus is identi- cal with a voltmeter. A heavy shunt sufficient in carrying capacity for the full current is con- nected in parallel with it, and the calibration is made to fit these con- ditions. The resist- ance of the shunt is very Fig. 340.— Alternating Current Volt- meter Compensator. low compared to that of the instru- 470 ELECTRICIANS' HANDY BOOK, ment. The diagram of the connection is given in Fig. 337a. Various forms of shunt are employed, one of which is shown in Fig. 338. Transformer Ammeter. — Sometimes on alternating current work a transformer is used to take off current for a voltmeter. By properly proportioning the coils and instrument the voltmeter becomes an ammeter. Fig. 339 shows a transformer mounted on a bus-bar which forms its primary. The terminals from the sec- ondary go to a Stanley hot-wire ammeter. The whole is so cali- brated that the readings of the instrument give the amperes pass- ing through the bus-bar. Wattmeter. — A modification of the construction of the volt- meter gives a wattmeter. In this instrument there are two ac- tive coils. One is fixed and the other is movable. The fixed coil increases the field in which the other one moves, and the index readings are a product of the voltage and amperage of the circuit. This multiplying action is in line with the action of magnet poles on each other, the intensity of which is the product of the two intensities, and not the sum. Pressure Lines or Pilot Wires. — Sometimes small conductors are run to various points in the district, are tapped into the system, and their ends in the station are connected to the ter- minals of voltmeters. These voltmeters give the potential differ- ence at the distant points to which their wires lead, and the read- ings of the voltmeters are the factors for operating the machinery in the station. Compensated Voltmeter. — A voltmeter wound so that its read- ings practically solve Ohm's law is sometimes employed, con- nected directly to the mains in the station. It is a voltmeter containing an auxiliary coil wound in opposi- tion to the main coil. This auxiliary coil is proportioned to the main coil as the feeder drop is to the total potential difference. Such an instrument gives pretty closely the potential difference at the end of the feeder. Compensators. — A compensator is an instrument for use on al- ternating-current circuits which indicates voltage between dis- tant points of a circuit. The compensator is installed in the station, it may be a mile or more from the place to which its in- VOLTMETERS AXD AMMETERS. 471 dications apply. In constant-potential lighting, for which it is specially applicable, pilot wires are sometimes used to give con- nections for voltage determinations at distant points. Such wires are connected to any desired point on the circuit, are led into the station and there connected to a voltmeter. The readings of the instrument give the potential difference or voltage at the more or less distant point on the circuit from which the wires come and to which they are connected. The compensator gives the same voltage reading without the use of pilot wires. The Ohmic Compensator includes a transformer v/hose pri- mary is connected in series with the supply line. The active turns of the primary can be varied by a switch, with a number of contact studs, each one corresponding to and throwing into action a greater or less number of turns in the primary. The secondary, also adjustable, connects with a voltmeter. This con- nection may be a simple series connection, but sometimes it is connected to an auxiliary coil, which is wound around the volt- meter-actuating coil. This coil is so wound or connected that it opposes the action of the voltmeter coil. The action is like that of the series coil on a compound-wound dynamo. The action of the auxiliary coil increases with the current which passes through the main conductor. This increase of current indicates the need of higher voltage, and to make the voltmeter read the same, the voltage of the circuit has to be increased to make its own proper coil pull harder against the auxiliary coil. Thus, if the voltmeter is kept at a constant figure, more voltage must be given to the line as more current is given it. By adjustment with the switch and contact plugs the readings of the voltmeter can be made to correspond with any desired drop on the line per given intensity of current. The Inductance Compensator has a second switch with a num- ber of contact plugs, by which the adjustment for inductance on the line is made, so that the total impedance is taken into account. The instrument is shown in Fig. 340. A compensator is without action of any appreciable degree upon the circuit. Its action on the voltmeter is such that in order to maintain a constant reading of the voltmeter, the pres- sure on the circuit must be increased as the current increases. CHAPTER XXVIII. DISTRIBUTION. Two Distribution Systems. — The systems of distribution of electric power may be divided into two main divisions — the con- stant-current and the constant-potential systems. In the constant- current system the central generators force an unvarying current through the circuit. The potential of the dynamos -must rise and fall as the resistance of the circuit varies under different conditions, but the same number of amperes must pass over the line. In the constant-potential system the generators are oper- att-d to keep a constant average voltage between the two leads of the circuit; the amperage may vary from almost nothing up to very high values; the station voltage may rise a little in its readings as more current is taken, and may fall a little as less cur- rent is taken. These variations compensate for the distance from station to district. Arc and Incandescent Lamp Circuits.— Lamps and motors are the principal appliances for utilizing electric power. All lamps require a constant current. Arc lamps without individual re- sistances or reactances can only be operated on constant-current systems. By the use of these individual attachments, which are described later, arc lamps are used on constant-potential sys- tems in very large number. Incandescent lamps can be used on either constant-current or constant-potential system. Motors can be connectert so as to work on either system. Constant" Current Systems. — A constant-current circuit con- sists of two leads carried through the district to be supplied. The leads are without branches or deviations properly so called; they unite at the most distant part, and form a simple closed me- 472 DISTRIBUTION. 473 tallic circuit. Lamps to be lighted are placed in their circuit, so that the entire current from the station goes through every lamp, and every lamp gets the same current. The potential on a constant-current system may vary consid- erably. A lamp may be removed from the line, and the ends of the line may be directly connected without resistance being in- serted in place of the lamp. In such case the potential of the lamp will be taken out of the system, and the generators will have to be run at a potential lower than the original potential by an amount equal to that of the lamp taken out. The amperage will be the same as before the removal of the lamp. The constant-current system is also called the series system; it supplies power by se- ries distribution. It is ^ O Q O shown in diagram in Fig. 341. Constant - Potential Systems. — The two-wire constant-potential system begins with two leads, which may divide into any number of branches, each branch consisting of two parallel leads, one from each original lead, and the two leads are not united at their ends, but are on open circuit, ex- cept as closed by the lamps or other appliances. The three and other multiple wire systemjs do the same, except that instead of two parallel lines three or more as the case may be are carried through the district, branching whenever it is necessary and always on open circuit except for the lamps or other appliances. The constant-potential system operates by parallel distribution. The lamps or other appliances used are sometimes said to bear to the main leads the relation of the rungs of a ladder to its sides. It is shown in Fig. 342. Series Distribution. — In series distribution of electric energy the lamps or other appliances to be supplied with current are placed in series with each other. The illustration, Fig. 341, Fig. 841.— Constant Current Distribution. 474 ELECTRICIAXS' HAXDY BOOK. shows a diagram of series distribution to a number of lamps. The simplicity of the system is obvious. A wire circuit, whose capacity for current is equal to that of a single lamp, can sup- ply any number. There is no question of increasing the size of the wire as more lamps are put into service. In these respects its advantage over the constant-potential circuit is very great. With one exception its limitations are not very great. Limitations. — One limitation is that each lamp must be con- structed for the same current. The potential drop for each one would normally be the same, but this is quite unnecessary. An- other limitation is that the total potential difference existing be- FiG. 34S. Constant Potential Distribution. tween the ends of the line shall not be too great. This is a prac- tical consideration affecting safety to life and possibility of in- sulating adequately. The next limitation to be noticed is one which has relegated series lighting to a very limited field. It is not practicable to put out one light without substituting for it some equivalent resist- ance or inductance. The latter can only be used for alternating- current systems, and high-voltage alternating-current systems are not supposed to have single lamps extinguished by hand, on ac- count of danger to life. It follows that domestic illumination cannot be organized on the lines of series system of distribution, because single lamps cannot be extinguished. It is not practicable to supply every DISTRIBUTION. 475 lamp in an incandescent lighting circuit with a resistance equal to its own, to be substituted for it when extinguished. If prac- ticable, it would be uneconomical. Features of Series or Constant= Current System for Arc Lamps.— It will be evident that as every lamp receives the same current, the wire should be of one size throughout. It is also evident that were there one or a hundred lamps on the circuit, the same current would pass and the same sized wire would be required. It follows that the economy in wire is increased by placing as many lamps as possible on the one circuit. The simplicity of the system is seen in the cut. A single line runs out from the generator and returns to it with as many lamps put on it as the voltage of the machine can take care of. There are no real branches or other complications. The management is also of the simplest. The dynamo is to be made to supply a constant current and to give the potential re- quired to keep up the current strength. Fifty to one hundred arc lamps may be placed on one circuit, which may be several miles in length. An ordinary arc lamp would require ten amperes of current and would develop a poten- tial drop of fifty volts. Calculations. — The calculation for a plain series distribution is simplicity itself. Take as an example fifty arc lamps, each of R 10 amperes and 50 volts. By Ohm's law, R =_, the resistance of one such lamp would be 5 ohms. There is a total drop in the lamps of 50 (volts) X 50 (lamps) or 2,500 volts to be provided for. Besides this, a current of 10 amperes has to be forced through the line. Take one mile as the length of the line, and assume that a loss of 5 per cent of the lamp energy on the line is admissible. As the current is a fixed quantity, the watts of energy are pro- portional to the resistance, because IE (watts) = R P. The re- sistance of the lamps would be 5 (ohms) X 50 (lamps) = 250 ohms. Five per cent of 250 ohms is 12.5 ohms. Consulting a wiring table, we find that No. 14 wire would give a resistance per mile at ordinary temperatures of 13.31 ohms, and No. 13 wire would give a resistance of 10 (> ohms. If we wished to have an exact resistance of 12.5 ohms, we could 476 ELECTRICIAXS' HAyDY BOOK. use both sizes of wire in the line. Calling x the relative length of No. 14 wire required, 1 — x will be the relative length of No. 13 wire required. Multiplying the relative length of each wire by- its resistance per unit of length, we have the equation U.^x + 10.6 {1 — X) = 12.5, which being solved gives: X =z 0.7 for No. 14 wire and 1 — o:^ = 0.3 for No. 13 wire. Multiplying each factor by its resistance, we have 0.7 X 13.3 = 9.31 No. 14. 0.3 X 10.6 = 3.18 No. 13. 12.49 The decimals 0.7 and 0.3 refer to a unit of 1 mile, and multiply- ing 5,280 feet (the feet in one mile) by them, we have: 3,696 feet of No. 14 wire. 1,584 feet of No. 13 wire. 5,280 feet or 1 mile. But it would be unnecessary to work so close as this. Mech- anical considerations apply also under the head of good practice. No. 8 wire is the smallest that is approved on arc-light circuits. If the mile of wire were of this size, its resistance would be about 3.3 ohms. This would bring the energy absorbed by the line to q q -rJi. == .0132, or 1.3 per cent of the lamp energy. X 5U Where the percentage is so small, this would be almost exactly the percentage if the total energy of line and lamps together were taken as 100. Keeping in mind the law that resistance is to be concentrated in the appliances in which heat or light energy is to be devel- oped, the object of keeping the line resistance low is to avoid waste of power on the line. Advantage of High Potential.— In the early days of electric lighting, a number of deaths occurred from contact with arc-light circuits. The higher potentials involve the greater danger. High po- tential of a circuit makes the adequate insulation more difficult than it is for the lower voltages. The idea is that it is good practice to keep voltage as low as is consistent with economical installation. In general terms, the high voltages are more economical in the DISTRIBUTION. 477 wire required for the distribution of electric energy. The unit of rate of energy Is the volt-ampere or watt. With high voltage the amperes for a given number of volt-amperes will be less than with a low voltage. Thus 100 volts multiplied by 100 amperes gives 10,000 volt-amperes or watts of power, which is also given by 1,000 volts multiplied by 10 amperes. But the larger number of amperes need a larger conductor than do the small number. Increasing the voltage and decreasing the amperage saves capital invested in lines. Standard Series Lighting Current. — For arc lighting on the series system a sort of standard has been established in the 10- ampere current. A station supplying circuits of this type simply has to send out 10-ampere currents, and as long as they pass to the line, the engineer can be almost certain that all is well in the district. The voltmeter will indicate the extinction of a lamp on the circuit. Series Incandescent Lighting. — What has been said about arc lighting applies to series incandescent lighting. The lamps are made of dimensions adapted to the current. If the arc-light cur- rent of 10 amperes is used, the incandescent lamps must have very thick filaments of length adapted to establish a relatively low potential difference. Thus, were it proposed to put one hun- dred ordinary incandescent lamps on one circuit in series, the potential difference due to them alone would be over 11,000 volts, and only one-half ampere of current would be required. One hundred thick and short filament lamps, on the other hand, would replac© about double the number of arc lamps and would work with the same current and potential difference. An incandescent lamp for this work passing 10 amperes of cur- rent by the expenditure of 10 volts would give at 3 to 4 watts to the candle power about 32 candle illumination. About thirty such lamps would give the light of a single arc lamp of 10 am- peres and 40 to 50 volts. The economy is poor, but is offset by other considerations, one of which is the evenness of distribution. A large number of small lamps give a more even light than that afforded by a few more powerful lamps. For outdoor lighting a 10-ampere, 100 carndle-power lamp is a standard. 478 ELECTRICIANS' HANDY BOOK, Film Cut=Out. — In all the series systems the entire electromo- tive force of the circuit would appear if the circuit were broken. This applies to the three-loop Brush system as much as to any other. This gives a simple method for constructing a cut-out which will short-circuit a broken lamp through w^hich no cur- rent can pass. It is called the film cut-out. The ends of the line connected to the lamp are bifurcated, as shown in Fig. 343. Between the free ends a piece of paper or other film is interposed, the ends pressing against it, thereby sending the current through the lamp. But if the lamp breaks or the filament parts, the voltage due to the entire electromotive force of the system is developed -on the two ends of the con- ductor separated by the film. This is at once pierced, the ends of the conductor spring together, and the current passes. The resistance of the lamp is gone from the circuit, so the current has to be re- duced from the central station to save the lamps from overheating, with consequent breaking down. Relief Lamps. — On each circuit in the station one or more idle or relief lamps are provided. The attendant watching the ammeters recognizes the breakage and cutting out of a lamp by the increase of current on the line con- taining it. He then throws one of the relief lamps into the cir- cuit. This reduces the current to the normal, and the broken lamp has to be dispensed with until replaced. The relief lamp is one of many cases in electric engineering where a lamp is used as a resistance. As a lamp sooner or later burns out, it is an expensive resistance. It has one good side, however. The bright lamp shows that something is wrong. A common resistance would disclose nothing except by the position of its switch. riultiple-Series System. — Incandescent lamps for street light- ing are sometimes made for a lower current, 3 to 3.5 amperes. To enable a larger current to be used, several series of such Fig. 313.— Film Cdt-Out. DISTRIBUTIOX. 479 lamps may be placed in parallel with each other, as shown im Fig. 344. Each series must be of the same resistance, or it will not receive the proper current. This is termed the multiple-series system. As the same number of lamps are on each circuit, it fol- lows that if the lines connecting them are of identical resistance, the circuit can be operated on constant potential. The property possessed by some makes of incandescent lamps of increasing in resistance as they rise in temperature operates in multiple-series distribution to even the currents received in the parallel lines of lamps. This self-regulating quality works in one way disadvan- tageously. A very slight rise in voltage makes the lamp work at an exceedingly great economy in consumption of electric en- ergy, but a slight fall dims the light very badly. Fig. 344.— Ml L.TIPLE Series, This feature may to some extent provide for the contingency of a lamp breaking down and being automatically short-circuited. The potential drop in that series is distributed among less than the proper number of lamps, and they burn too brightly, but not to such an extent as if their resistance was unaffected by heat. But incandescent lamps as a rule are made without this self- regulating quality. ♦'Municipal" Series Incandescent Lighting is often carried out on these lines, lamps using 3 to ^l^ amperes being employed. Thus with a 10-ampere machine three series could be operated in parallel. Series -Multiple System,— Another system of distribution for incandescent light is termed series-multiple. In it the lamps are put in parallel in groups, and any number of these groups ac- cording to the potential available are put in series. The cut. Fig. 480 ELECTRICIANS' HANDY BOOK, 345, shows the arrangement. By selecting lamps of suitable voltage and grouping them in parallel, each group can be made to represent any resistance equivalent to or calling for any current desired. All the lamps in one set must agree in voltage rating — all the station is called upon to give is a constant current of known amount, and the voltage must be enough to produce this current. It is possible to introduce lamps of different candle-power in this system, provided (a) that they are of the same voltage as the others and (h) that the current required for each group of lamps is the same. Thus a group could be composed of five 16- 3andle-power 50-volt lamps or of three 16-candle-power 50-volt Fig. 345.— Series-Multiple Connection. lamps and of one 32-candle-power 50-volt lamp. Many variations can be made in a group, provided the requirements as outlined are fulfilled. If one of the lamps breaks down, it will cause the entire group of which it is a part to receive too much current, and will tend to burn out the lamps. This can only be met by having an auto- matic device of some kind which will switch in a new lamp in place of the other, or which will cut out the whole group. In the latter case the voltage of the system will be suddenly re- duced. The station must take care of this, and maintain a con- stant current or all the lamps will receive too much current and be in danger of burning out. This system is very little used. The difficulties to be overcome in providing for the contingency of lamps breaking down militate ac:ainst it. DISTRIBUTION. 481 Objections to Series Distribution. — Series distribution for in- candescent lighting involves several features that militate against its use in houses. It requires too high a potential. A high-potential system is a cause of danger to life and property. It exacts that a number of lamps be operated as a unit. A single lamp cannot be turned on and off without disturbing the whole of its group, unless an equivalent resistance or inductance for alter- nating circuits be substituted. A resistance for every lamp would involve expense in installation, and would absorb just as much energy as a lamp and make no return. An inductance ab- sorbs but little energy and is an important adjunct in outdoor circuit work in alternating-current lighting. But the great dan- ger of a considerable voltage on an alternating-current system absolutely proscribes indoor alternating-curuent series lighting. Parallel Distribution is constant-potential distribution. In parallel lighting pairs of mains or wires from the electric station are kept by the station machinery at a constant difference of po- tential. The lamps are arranged in parallel across them, as has been said, like the rungs of a ladder, as far as their representa- tion in a diagram is considered. Incandescent lamps are con- structed for a specific current by being made of adequate thick- ness of filament and of length sufficient to operate at the de- sired current with a specified drop of potential. One hundred and ten volts is a sort of standard. For half an ampere of current, the filament has to be of two hundred and twenty ohms resist- ance. Lamps are made, however, of the most various voltages, and are generally rated by the voltage required to operate them. If for 110-volt lamps the mains are kept at a constant difference of potential of 110 volts, perfect independence of action of all the lamps is established. They may be lighted or turned off one by one without affecting each other to any noticeable extent. The maximum difference of potential in two-wire circuits is that of a single lamp, which cannot hurt anyone and is treated as a safe potential as far as fire risks are concerned. The electric energy of the system is draw^n upon in almost exact proportion to the number of lamps lighted. The conditions of safety, simplicity, and economy of energy are adequately fulfilled by the parallel circuit. 482 ELECTRICIANS' HANDY BOOK, Disadvantages of Parallel Distribution. — On series connec- tion one hundred lamps could be supplied with current through a wire of one-hundredth the cross section of that required for cur- rent for the same lamps in parallel. This is a most important ad- vantage. Heavy currents in electric engineering involve expense of installation at every part, and the interest on the capital in- vested is to be treated as a part of the fixed charges of the sys- tem. The independence of each lamp of the circuit has made the parallel lighting system universal for indoor illumination. Where the street mains already exist, it is used for arc lighting with the attendant sacrifice of economy involved in the use of an individual resistance for each lamp. Elementary Case of Parallel System. — The simplest case is shown in the cut. Fig. 342, where two leads of even thickness are c^Tried out through the district, and have as many lamps con- nected across them as they can carry current for. This is waste- ful of copper, because the wire which comes between the first lamp and the dynamo determines the size of the outer end of the wire. Current for all the lamps has to go through the wire next the dynamo, while at the outer end current for only one lamp is 10 be carried, and it is wasteful of copper to use too large a con- ductor. The size of a conductor is determined by several considerations. It must carry the current without undue heating. It must be of low enough resistance to pass the current at a low enough drop to secure economical working. The latter consideration is the" controlling one, as under its requirements the wire is sure to be large enough to carry the current with safety from overheating. Potential Drop in Parallel System. — Incandescent lamps for a variation of one per cent down or up in potential drop lose or gain a little over one-sixteenth of their illuminating power. A drop of one volt in a 110-volt 16-candle-power lamp will reduce its candle-power to 15 candles. The consumer's payment is based on light and only indirectly on electric energy. A drop in voltage deprives the customer of the light he is paying for, and the re- duction in fuel consumption due thereto is too trifling to be con- sidered. It amounts to failure in carrying out a contract and to injury of the customer without benefiting anyone. The utmost DISTRIBUTION, 483 care should be taken in planning a system to obtain good dis- tribution of potential. Inevitable variations in potential drop can be allowed for by using lamps of different voltage. But after all calculations are made, the results in practice will vary, be- cause various numbers of the lamps may be lighted at once. The calculations have to refer always to all the lamps or to some fraction of their total. Their results will not stand for any other number. Feeders, flain and Leads. — Districts are not supplied by a sin- gle pair of conductors. Feeders run out from the station to points in the district, and are not supposed to be tapped for lamps. These are and should be of uniform thickness. Their ends connect with other wires, called mains. Between the ends of the first lines ^ ^ ^ Fig. 346.— Loop System. of feeders and the lighting mains secondary feeders may inter- vene. By tertiary and other feeders the system may be made quite complicated. From the mains run other wires called leads, and the lamps are supplied by them. Classification. — The calculations for supplying a district are based on Ohm's law, and whatever arrangement of mains and feeders is adopted, the calculations are simple. Classification of the systems of supply may be elaborate, but they all are subject to Ohm's law, worked best perhaps by the drop system of calcula- tion. Loop System.— The loop system of distribution arranges the circuits so that the current for each lamp goes through the same length of wire. There are two loop systems shown in the cuts, Figs. 346 and 347, the straight loop and the spiral loop. If the reader will examine the length of conductor through which the current for each lamp passes, he will find that the lengths are 484 ELECTRICIANS' HANDY BOOK. Fig. 347.— Spiral. Loop System. identical. With a eonstant current the line drop for each lamp would be the same; for a dimin- ished current, due to the extin- guishment of some of the lamps, the line drop varies. The amount of copper required for loop system conductors is greater than in other systems, but the potential is much better main- tained than in systems more eco- nomical of copper. Tree System.— In the early installations a pair of mains was carried from the station through the district. From this pair a quantity of minor conduct- ors were carried to supply the lamps. The plan laid out in simplified form re- sembles a tree, with the sta- tion as the root or pot out of which it grows. The two mains are the trunk and the branches, with minor branches to carry the lamps. The cut, Fig. 348, elucidates the origin of the name, the **Tree System," given to it. Closet System. — Another system is the "Closet Sys- tem." In it the lamps are collected into groups. Each group has its own circuit running back to the dyna- mos. The method is used in interior wiring. An interest- ing example is shown in the -pzo. 348. cut, Fig. 349, where two feeders are connected to op- posite sides of a double circle of mains, across which the lamps O Lamps Switches caSafstj' Cu^Oa^« -Treej System of Parallel Distribution. DISTRIBUTION. 4S5 are connected by their individual leads. In this arrangement the length of main for each lamp is identical. This length is half the circumference of the circle, assuming that the lamps are so close to the mains that the circles of wire virtually coincide. In prac- tice this scheme would be carried out by two more or less irregularly-shaped cir- cuits of mains. The feeders would be tapped in at opposite points. In Fig. 350 the closet sys- tem is shown as carried out for a number of lamps ar- ranged in four closet connec- tions, with voltmeters and fuses for each group. Cylindrical and Conical Conductors. — Wire is normally of one diameter throughout, and is almost always of circular cross sec- tion. Where such wire is used throughout a circuit or division of Fig. 349.— CroRFT System of Par- allel Distribution. Fig. 350.— Closet System. a circuit, the term cylindrical system is applicable. If the wire is reduced in diameter as the distance from the station inrrpases, 486 ELECTRICIANS' HANDY BOOK. it represents a cone, and the term "conical" becomes applicable. The reduction in diameter may be, and practically always will be, by reduction of diameter at various places, so as to constitute a step-by-step reduction. The cylindrical system secures the most even effects as regards potential difference, while the conical system saves copper, and if properly carried out secures good enough results in evenness of potential difference. It is important to keep in mind the statement of the last para- graph; conical distribution, Fig. 351, does not secure even poten- tial difference between the lines. What it may secure if properly calculated, and if the number of lamps or other appliances as- sumed in the calculation are operating, is an even potential drop per unit length of line. A drop of this description is simply to be OTT Fig. 351.— Conical Mains. accepted as an indicator of good practice and as giving a basis for calculating the sizes of conductors. Calculation for Conical Conductor.— Assume that lamps to be supplied by a main can be divided into three groups for the purposes of the calculation. Let the initial difference of poten- tial be 115 volts. Suppose the first group of lamps are of an average voltage of 114 volts, the next group 112 volts, and the last group 110 volts. The wire is to be reduced in two steps. What should be its resistance at the three divisions? Suppose 50 lamps are in the first group, 60 in the next, and 30 in the last, and that each lamp takes i/. ampere of current. . 50 + 60 + 30 The total current is = 70 amperes. By Ohm's law 2 ■p 1 R = j*^ and substituting we have R =. — ohm. This is the resist- I 70 ance of the first portion of the mains, or 1/140 ohm for each ]ead, DISTRIBUTION, 487 to give a drop of one volt for the 114-volt lamps. The next section has ^^ "^ ^^ = 45 amperes to supply at a drop of 2 volts; its re- O Of) sistanee is __ ohm, or 1/45 ohm for each lead. The third has — 45 . 2 2 = 15 amperes at 2 volts, giving a resistance of ohm for both 16 leads. In diagram the above conditions would be indicated as in the cut, Fig. 352. Suppose that each section of conductor is of the same length, and that it was a cylindrical conductor, one of the same diam- eter throughout, and that the diameter was that of its first or largest section. The drop for the first group of lamps would be 1 1 140 OHM. -45 OHM. A- n ^ ■115V0LTS 70.AMP.. ^114 loLTS 45 AMP. f112, c, or d, by w^hich it can be thrown out of action when de- sired. M M are the mains. Sornetimes an effort is made to con- nect feeders symmetrically. This means that each one shall feed the same number of lamps. This plan is of little value, because O O ^ M F' Fig. 358.— Feeders With Rheostats. the same number of lamps can never be assumed to be burning at all times. A general estimate is all that can be made. The next cut, Fig. 358, shows an attempted refinement on the last described connection. Here each feeder has its own rheostat. DISTRIBUTION. 495 This makes it possible to vary the resistance so as to maintain an even potential drop in the feeder. This method is opposed to the general law to the effect that resistance should be con- centrated in the lamps, or wherever heat energy is to be used. The putting resistance voluntarily into a feeder or any other transmission line is on its face at least bad engineering. The voltage should be increased or reduced at the dynamo. Resist- ance such as indicated in the diagram is called **dead" resistance. Auxiliary Feeder Connections at higher voltage than that of the station dynamo are sometimes used. The next diagram, Fig. 359, shows this method. To the left is the station dynamo yiG. 359.— Auxiliary Bus-Bar Connection. delivering current to the main feeder F'. For auxiliary feeders a special bus-bar is provided. This is connected to a special dynamo D, which maintains any desired potential in the feeder circuit F. A is connected to D by the switch a. Transfer Bus-Bar.— Sometimes a feeder supplied from one bus- bar of a given potential has to be shifted to another at a higher or lower potential. A transfer bus-bar is used for this purpose. In the diagram, Fig. 360, A is a high-potential, C is a low- potential, and B the transfer bus-bar. Suppose that the low- potential feeder M, as the circuit is drawn upon for current in the lighting hours of the evening, has to be shifted from C to A. The switch c is closed upon B. At cZ is a rheostat. As shown, the circuit at d is open. The arm of the rheostat is swung to the left, thus closing the contact through the resistance of the 496 ELECTRICIANS' HANDY BOOK. rheostat. The switch arm d is moved on slowly until an ammeter shows that B is taking all its current from A. This can be brought about by reducing the resistance by moving the switch d. Fig. 360.— Transfer Bus-Bar. The switch b is next opened, and d is swung to the end of its course, so as to cut out all resistance. Example. — As an example of parallel and feeder distribution LINE OF BOULEVARDS Bl mi,i,6i?^V^F' ^^l?if^^U^ =£C Tl T^ OX, Fig. 361.- Example of Parallel Distribution. embodying conical leads, Fig. 361 is given, showing a district of Paris, which illustrates much which has been described. Feeder Economy. — When capital has been invested in tons of copi)er in order to keep resistance down all through a lighting dis- trict, it seems crude to regulate the action of feeders by volun- DISTRIBUTION. 497 tarily increa.sing their resistance by a rheostat. It seems still worse to make them absolutely useless by opening a switch, and utilizing an expensive feeder line perhaps only during an hour or two of peak. With rheostats some good is got out of the lines. With an open switch the line does nothing. It is fair to assume that most stations are operated largely for light. It therefore follows that for some tw^enty hours out of the twenty-four their mains will be comparatively idle. Hence if a main is switched on for only one hour, it is hardly fair to say that it is idle for ^ of the day. Relatively speaking, it would 24 be fairer to refer its action to the lighting period, and treat it as idle for three-fourths of the time only. Three- Wire System.— The three-wire system, like the rest of the parallel systems, is a concession in the direction of economy A c I J Fig. 363.— Three- Wire system With One Generator. of copper. The direct source of this economy lies in the doubling of the initial voltage of the system. For lamps of 110 volts a potential difference of 220 volts between the mains is employed. The station dynamo may run at 230 volts. A further economy in the expenses of the leads or conductors is based on the probability that lamps can be so distributed into two groups that all the lamps in one group will never be lighted at a time when all the lamps in the other groups are extinguished. rn the three-wire system three leads are carried through the district, Fig. 362. A potential difference of 220 to 230 volts is maintained between two of the wires; the third wire lies half way between the others in potential. The third one is called the neutral wire. One dynamo, as in this cut, or two, as in Fig. 363, may maintain the power. Saving in Copper,— The saving is due to the fact that the cir- cuit has its two outer leads maintained at double the potential 498 ELECTRICIANS' HANDY BOOK. difference of that which would be required in the two-wire sys- tem. Hence for the same number of watts, and consequently lor the same number of lamps, one-half the current would be re- quired. The two outer conductors could be made one-half the size of those in the two-wire system. This would be one-half the copper. But the neutral wire has to be provided. This may be Fig. 363.— Three- Wii?e Systtm With Two Generators. smaller than either of the others, but it is always of some con- siderable proportion of the size of the main wires. If the lamps were always lighted in even number on each side of the neutral wire, it could be dispensed with. If all the lamps on one side were lighted and all on the other were extinguished, the neutral wire would have to be as large as the main wire. Its -5 -^3 i I 2*-^ Fig. 3^.— Action of the N^-utral Wire. relative size is a matter sometimes of calculation and sometimes of judgment. The diagram. Fig. 364, shows a case typical of the three-wire system. The neutral wire here has two currents going through different parts of it, in opposite directions. It is like two tides coming around an island and impinging against each other. A portion of the neutral wire in this case receives no current what- ever, yet other parts of it are passing current and keeping the system balanced. DISTRIBUTION, 499 Two-Dynamo Three=Wire System.— In first-class station work the lliree-wire system is operated by two dynamos, each of the requisite potential to supply a single set of lamps. The cut, Fig. 365, shows the system. It is clear that each dynamo could sup- ply the lamps between its main and the neutral main. The neutral wire or main connects with a line connecting the two dynamos, one positive and one negative brush, as shown. /wzm O T2 22?^ ■0- o. o ss: 2Z2: ^31 315: o- ■oo oo ^^^^^^ TT^ Fig. 365.— -Chrbe-Wire System With Two Dynamos. Single-Dynamo Three-Wire System.-Various modifications of the three-wire system are em-ployed in special cases. One is shown in Fig. 362, in which the neutral wire does not connect with the single dynamo used. This dynamo must have twice the voltage required for a single lamp in addition to that required for the drop. Three-Brush Dynamo. — Another modification consists in the introduction of a third brush on the dynamo, placed midway be- tween the regular ones. The neutral wire is connected to this brush as shown in the cut. Fig. 366. The system is apt to give a great deal of sparking on the commutator if the two circuits take different currents. The normally idle neutral wire at least supplies a security against the obligatory shutting off of two 500 ELECTRICIANS' HANDY BOOK, lamps at once. Where there is little chance of great inequality between the two groups, such an arrangement will work very well. It is not to be regarded as a standard method, on account of the liability to sparking on the commutator. Storage Batteries in the Tliree=Wire System can be used to advantage. The cuts, Figs. 367 and 368, show three-wire systems with storage batteries. When lamps are extinguished, the sur- plus current from the dynamo goes through the battery and charges it. When the current from the dynamo is drawn upon beyond its fullest extent, the stor- age battery comes into action, and -O- -o ^> -o- ITT ITT Fig. 366.— Three-Brush Dynamo. Figs. 367 and 368.— Storage Batteries in Three- Wire System. supplies the deficiency. Its action is regulated by the use of end cells, counter electromotive force cells, or rheostat, as elsewhere spoken of. distributio:n. 501 Storage Battery Equalizer in Three-Wire System,— In Fig. 369 the storage battery S is connected to the neutral wire N and to the outer wire M. If lamps are extinguished on one side of the system, the current thus thrown upon N is taken care of Fig. -Three-Wire System With Storage Battery Equalizer. by the battery. It will charge or discharge according to which group A or B is using most watts. A rheostat is provided to regu- late the dynamo field. The battery could be connected to P in- FiG. 370.— Balancing Dynamo in Three-Wire System. stead of to M, but not to both without abandoning this particular arrangement. Balancing Dynamo. — The illustration, Fig. 370. shows two 502 ELECTRICIANS' HANDY BOOK, dynamos in a three-wire system, one, indicated by A, being of double the voltage of B. Both are driven from the same counter- shaft E. At even load on both branches, P and M, the dynamo B runs idle. If the branch P has most load, current going through the neutral wire goes through B and actuates it as a motor. If M has most load, B operates as a dynamo to supply the M side of the system. Motor and Booster,— In the cut, Fig. 371, A represents a dyna- mo running at a high enough potential to make the loss between G and R comparatively small. A is in the central station, R snd C are in the district. R is a motor, and its functions are to drive the booster C. Fig. 371.— Motor and Booster in Th.^ee-Wire System. Five and Seven=Wire System. — The three-wire system is the first step in multiple wiring, as a two-wire system does not fall into the category of multiple wiring, where it etymologically should belong. The next step is to add couples of wire. Thus the five-wire and the seven-wire system are developed. In the five- wire system the potential is four times that of a single lamp; in the seven-wire system it is six times that quantity. If stand- ard incandescent lamps are used, the voltage of the systems will be 120 X 4 = 480 volts, and 120 X 6 = 720 volts, allowing for the drop of the lines. The central wire is the neutral wire, but the current may be variously divided among the wires by the consumption varying in different groups. The high voltages are not very safe, and it can be readily seen DISTRIBUTION, 503 that such a multiplication of wires complicates the station ma- chinery and the distribution of lamps on the circuits. The at- tendant high voltage exacts better insulation and more careful laying of mains and leads. In America the three-wire system has obtained by far the greatest extension. In Europe the five- wire system is used in a number of places. Examples of five-wire systems are shown in Figs. 372, 373, and 374. The last two illus- trate the use of storage r— i batteries at the station J y y O^ end of the system. They f~^\ b i are susceptible of many I ) i ' ^ variations. V_V^ ^^— t — L o: o; High - Voltage ParaN [ ^ ^ Q lei Systems, — The manu- facture of 220-volt lamps . ^ — , has been considered a an- y) -r y y O ficult problem to solve /^ ^ —^ A A ^ under commercial limits. i ) _£2 1 i p^ With such, a three-wire ^^^^ "T 1 . T ^ system could be operated [ -p y 9 ^ at 480 volts minimum, re- ducing the copper used amount for 110-volt ^v/' 1^^ 6~6 Q lamps. Some authorities ^~ ~i ^ i ^^ consider that the three- ( ] "P 1 ■ T ^«- wire system with 220-volt ^ ^ -p i y Q lamps is destined to pre- Tigs. 37-, 3:3 and 3T4.-Five-Wire Systems. vent the extensive use of the five-wire system. Multiple-wire systems possess a fea- ture which may be of value. There is nothing in the system to interfere with the possibility of connecting apparatus such as motors across from main wire to main wire, thus utiliz- ing the double voltage of the system with the exclusion of the neutral wire. A 220-volt motor can thus be used on a three-w^ire 110-volt circuit. On a five-wire or seven-wire sys- tem the entire potential difference will approximate respective- 504 ELECTRICIANS' HANDY BOOK, ly 480 and 720 volts. This gives the conditions for a high- power motor with small conductors. The voltage in. such cases is about that of a trolley car system, and the system repre- sents a combination of high and low voltage parallel distribu- tions. TRANSFORMEHS ARRANGED IN SERIES, WITH LAMPS IN PARALLEL. TRANSFORMERS ARRANGED IN SERIES, WITH LAMPS IN aERiES. TRANSFORMERS ARRANGED IN PARALLEL» WlTii LAMPS IN J^ARALLEL.. Figs. 375, 376 and 377.— Examples of Transformer Distribution. AIteriiating=Current Distribution.— The use of the transform- er to change voltage is the characteristic feature of this class of distribution. Fig. 375 shows in diagram transformers in series, each absorbing a portion of the voltage of a dynamo and trans- forming it into voltage adapted for lamps, which are supplied in parallel from the secondaries. Fig. 376 shows a series of trans- formers as before, but each one supplying a set of lamps in series. A full parallel system is shown in Fig. 377, where the transformers DISTRIBUTION, 505 are in parallel, their primaries connecting to two leads from a dynamo, and lamps in parallel being supplied from each trans- former. The lamps as in both the preceding cases take current, from the secondaries. The latter arrangement is shown more in detail in Fig. 378, where arc lamps absorbing 104 volts each are supplied by means of a converter from a 1040 or 2080 volt circuit. PBiMARY ciRcurr 1040 0« 2090 VOLTS PRIMARY SWITCH 4 CUT^eUT TRANSFORMER SECONDARY CIRCUIT 104 VOLTS D.P.CUT-OUT D.P.SWITCH M Fig, 378.— Transformer Connection for Arc Lamps. Individual Transformers. — Small transformers are used for single motors and lamps. In Fig. 379 is shown a motor supplied from a high-tension circuit by means of a transformer. This and the preceding cut have the names of the different parts noted on the illustration. Although only one motor is shown, the ex- tension of the secondary circuit to right and left indicates that more motors may be supplied by the same transformer. Choke Coils. — In Fig. 380 is shown a single incandescent lamp carried on a bracket with a receptacle at its base in which there 506 ELECTRICIANS' HANDY BOOK. is a choke coil. This is virtually a transformer without any sec- ondary. It is connected in parallel with the lamp. An alternating current as often thus connected lights the lamp because the inductance of the coil sends current through the lamp. If the lamp filament breaks, the cur- rent goes through the coil. Thus the breaking of the lamp does not break the circuit. The arrangement is adapted for lamps in series, as shown in Fig. 381. Y Connect/on for Alter- nating Current. -Three-phase alternating current is often distributed by the Y connec- tion, so called because the three leads are connected as if by a letter Y. The diagram, Fig. 379.- -MoTOR AND Individual. Transformer. Fig. 380. —Choke Coil for Incandescent Lamp. Fig. 381a, shows the system. At the generator end the arma- ture windings A, B and C are connected at a central point n. DISTRIBUTION. 507 This is described elsewhere under the subject of alternating cur- rent generators. From the ends of the three windings three leads are carried through the district and lamps or motors are connected as indicated. A motor is indicated on the right hand witn Its three armature coils. A, B and C, also connected at a sin- 10 100 VOLT L>*MP5 IN SERIES. Fig. 3Si.— Incandescent Lamps in Series With Choke Coils. gle point n. The lamps are connected between any two leads as shown. If there are more lamps on one pair than on another the system will be out of balance, and a fourth neutral wire con- necting n and n will be required. This is sometimes called star connection. Delta Connection. — This is also spoken of under alternating 508 ELECTRICIANS' HANDY BOOK. current generators and is illustrated in Fig. 381&. A, B and C represent the three armature coils of a three-phase generator and L^MPQ GENERATOR MOTOk Figs. 381a & 381b.— Y and Delta Connections for Alternating Currents. -aMM^s^lMM^ Figs. 382 and 383— Iron Wire Joint and Tib. motor respectively connected as shown. No neutral wire is used in this system. Joints in Line Wire.— It is beyond the scope of this work to give the details of line construction, which is becoming more DISTRIBUTION, 509 complicated as aerial and underground distribution systems ac- quire more extension. In the illustrations, Figs. 382 to 390, some examples of joints and ties in wire conductors are given. Figs. 382 to 386 show how iron wires are joined to each other and how they are tied to glass insulators. The joint shown in Fig. 382 i& sometimes called the Western Union joint. The tie wire in Fig. 383, it will be observed, is carried around th© insu- lator, and it§ ends are then twisted around the line wire. Other mm WIlF= Figs. 384 and 385.— Iron Wire Ties. Fig. 386— Puttinq ox Ties. ways of tying are shown in Figs. 384 and 385. In one the tie wire does not go entirely around the insulator, in the other it completely encircles it and is twisted once around itself before the ends are twisted around the line wire. Fig. 386 shows the operation of making such joints. For copper wire, sleeve joints have met extensive use. The Helvin joint was made with a brass double sleeve receiving th'^ ends of the wire. One way of using a sleeve is to twist the ends 510 ELECTRICIANS' HANDY BOOK. of the wires projecting beyond the sleeve around the line wire outside of the sleeve. The ends of the sleeve are closed with solder. Fig. 387 shows such a double sleeve used in the Mclntyre joint. Here the wire is passed well into the sleeve, and then wire and sleeve are twisted together as shown. Sometimes solder is ap- FiGS 387 AND 388.— Sleeve Joints. plied, holes being made in the sides of the sleeve to admit the solder. A simple strip of copper bent so that its cross section is S-shaped is used as in the Mclntyre tubular sleeve. It is shown in Fig. 388. A simple joint made with a small wire seizing is shown in Fig. 389. Soldering may be applied to this joint. Ends of wires in cables are joined by twisting, as shown in. Fig. 389.— Seized Joint. Fig. 390, care being taken to prevent the wire at the joint in one wire from touching that in another. When ends of cables are to be connected, a lead sleeve is placed over the end of one cable, is pushed back, and the wires are connected and the joints are insu- lated by paper wrapping or other material. The sleeve is then DISTRIBUTION, 511 pulled over the joint and soldered to the ends of both cables in- closing the joint, so as to make it perfectly water-tight. Such a sleeve soldered in place is shown in Fig. 391. In Fig. 392 is shown the transposition of wires on a pole top. This is done in order to avoid induction; the induction inevitable when an active telegraph or telephone wire is near another one, Fig. r90.— Joining AVires in a Cable. Fig. 391.— Sleeve on Cable. Fig. 392.— Transposition in Aerial Line Work. being of opposite polarity as the leads are changed. Thus the in- ductive effect from one length of wire counteracts that from the other. Insulators. — These are now made in a great variety of forms. As typical of modern practice two insulators are given in the cuts. Fig. 392a is an insulator with a groove in its top to carry the wire, and constructed to withstand a potential difference of 80,000 volts. By doubling the projecting flanges or "petticoats," 512 ELECTRICIAyS' IIANDY BOOK, the insulator shown in Fig. 39 2Z; is made, which is good for a po tential difference of 120,000 volts. These are extreme cases. Figs. 392a and 392&.— High-Tenston Insui.ators. In former practice there were comparatively few forms of in- sulators, but the recent development in the use of high-tension circuits has brought a great many forms into the field. The problem of adequately insulating a line with a potential difference of thousands of volts backed up by a heavy current is widely dif- ferent from insulating a telegraph line. CHAPTER XXIX. ELECTRIC METERS. Electric Meters may measure current irrespective of voltage T\hen they are current meters. They may measure the current and voltage v^hen they are wattmeters. Wattmeters operate correctly where electric power is sup- plied, but not for incandescent light unless a constant voltage is maintained. They only correct for about one-fifth of the de- ficiency in light suffered by the customer or excess obtained by him on changes in voltage. An over-compounded wattmeter would seem to be the best for light-supply metering, one which for a change of one volt would change the reading about six per cent. Edison's Meter, — This meter was conceived on the somewhat heroic principle of the collection and weighing of metal deposited in meters by electrolytic action. The meters gave no direct read- ing. To get at their results, small quantities of zinc had to be weighed for each meter periodically, and the current supplied was taken as being proportional to the weight of this zinc. For years the meters in cities supplied by the Edison system were thus taken by the operative in charge. Baskets filled with electrodes were transported to the station, and the electrodes were individually weighed, and the current supplied was calculated on this electro- chemical basis. The cut, Fig. 393, shows its construction. It contained two cells, each containing a pair of amalj?-amated zinc electrodes. They were made of as pure zinc as possible, and before amalgama- tion were coated with zinc by electro-deoosition. The cells con- tained a solution of zinc sulphate of 1.11 sn-ecific gravity. The meter had in series with the plates a coil of copper wire. The resistance of copper Tvire increases as the temperature rises, just 513 514 ELECTRICIANS' HANDY BOOK. as does that of other metals. This was to compensate for the fall of resistance with rise of temperature which occurs in the solution. The meter was placed in shunt with a known resistance on the line, and its own resistance being known, it received a fraction of the total current equal to the quotient of its own resist- ance of the portion of the line in parallel with it divided by the resistance. The weight of zinc deposited gave the coulombs of electric- ity used. An incandes- cent lamp was automatic- ally lighted by an expan- sion bar when the tem- perature fell, and extin- guished as it rose. The same principle was applied to a register- ing meter. The plates were hung at opposite ends of a scale beam, and were alternately subjected to one or the other action, so as to move the beam from time to time. Each movement was due to a definite deposition on one plate and dissolving of the other. As the beam swung it reversed the current, and after a cer- tain amount of coulombs had passed, it swung back. The swings were registered by clockwork or geared mechanism of the regular type. A counter electromotive force of 0.001 to 0.003 volt caused the readins-s at low current to be erroneous. Forbes rieter.— This meter was actuated by the heat produced by a current. In the lower part of a glass shade there was a flat Fig. 393.— ^dtson's Chemtcal Meter- Sectional Diagram. ELECTRIC METERS. 515 coil of wire which occupied a horizontal position. Above it was a vane with four inclined wings like a little screw propeller. This vane worked in very delicate bearings. The current to be meas- ured or a known fraction of it passed through the coil and heated it. The heat caused an air current to rise from the wire, and this turned the vane windmill fashion. The turns of the vane were registered by machinery. Fro. 394.— Thomson's Induction Meter. Thomson's Meter.— This meter, due to Elihu Thomson, is a wattmeter and is shown in part section in Fig. 394. It consists of two field coils without iron core, through which the entire current which is to be measured passes. Within, the coils an arm- ature coil without iron core is mounted. It has a commutator. It receives current from the wires of the circuit, being connected across them with high resistance interposed. It receives current proportional to the voltage existing between its places of attach- ment. The field coils of low resistance receive all the current prac- tically that passes. The armature rotates and drives an indicating 516 ELECTRICIANS' HANDY BOOK. train of wheels like that on a gas meter. A horizontal copper disk rotates on the vertical axis which carries the armature, and steel magnets with poles brought near together embrace the outer portion of the disk between their poles, and constitute a brake on the rotation of the armature. The speed of rotation is due to the field acting on the armature. The strength of the field is due to the amperes of the current; the strength of the armature is due to the voltage of the circuit; the reading of the meter is due to the combined effect or to the volt-amperes or watts. The meter is primarily a shunt-wound motor. An auxiliary field coil in series with the armature gives it the character to a limited extent of a compound-wound meter. This field with the armature develops alone almost enough torque to turn the arma- ture. It therefore takes care of the friction of the meter in great part, so that the magnetic brake opposes all the resistance to its motion, a resistance increasing v/ith the speed. It will be seen in the cut, Fig. 394, that the permanent magnets are held in position by screws going through a horizontal bar, a portion of the frame of the meter. These can be loosened if desired, and the magnets can thus be moved in and out. This operates to regulate the meter and make it move faster or slower. It can thus be tested Vvith lamps, and adjusted over a range of about 16 per cent. An alternating or direct current can be meas- ured by this meter. For three-wire systems, one of the field coils takes the current of one active wire; the other coil that of the other active wire. The coil in circuit with the armature is connected across from the neutral wire to one of the outer wires, thus getting the voltage of one lamp, or customer's voltage. Sometimes the shunt field coil and armature are connected across the outer wires, thus laking twice the voltage. A transformer can be used in alternat- ing current supply where the voltage is too high for the resist- ance of the meter. In meters for heavy currents a single copper bar passing between two armature coils constitutes the field. For two and three-phase alternating-current circuits a combina- tion of two or three meters in one is made. One dial gives the reading. Otherwise, two meters can be connected to give the readings of three-phase systems. The sum of their readings is ELECTRIC METERS. 517 taken. If the lag exceeds 60°, giving a power factor of less than one-half (cos 60° = VL>) one of the wattmet-ers will have a negative reading, in which case it must be subtracted from the reading of the other one. For series systems the field is in series with one of the main conductors, so that the full current, which is not a very high one, goes through it. The meter gives watt hours. Shallenberger's fleter. — The entire current passes through a fixed coil of few turns. Within this coil is a second one with self-contained re-entrant circuit, constituting an induction motor armature, as it has no outside connection. Its axis is at an angle with that of the outer coil. When an alternating current passes through the outer coil, it induces a current in the closed circuit of the inner coil. A reaction is established with a resultant field between the two fields, one of the outer and the other of the inner coil, which fields are not coincident in position, but lie at an angle to each other, equal to the angle between the axes of the coils. There is also a difference of phase between the two coils, which causes the resultant of the fields to rotate, thus constituting a rotary field. A vertical arbor or spindle carries a horizontal metallic disk which lies in the field, and is acted on by the rotary field when current passes, and caused to rotate. To retard its motion, air vanes are carried by the spindle. The principle of the meter is that the torque increases with the square of the current, being due to the energy expended. The resistance offered by the vanes varies with the square of the speed. Thus, the speed of rotation of the disk is directly proportional to the cur- rent strength. This meter is a current measurer, taking no di- rect cognizance of the volts of the circuit. CHAPTER XXX, LIGHTNING ARRESTERS. Lightning Protectors. — Atmospheric electricity produces dis- turbances in electric apparatus unless means are taken to give it a way of escaping to the ground. Whatever the nature of the disturbance, so great a voltage is established that the current due to the atmospheric electric- ity can jump across an air gap quite impassable for working electrical currents. Comb or SawTooth Arrest* er.— This was one of the early protectors. Attached to the line to be protected was a plate with a series of saw teeth on one edge. The plate might be an inch long. A similar plate faced it tooth to tooth, both being screwed flat on a board. The s€cond plate was connected by a conductor to the earth. Ordinarily the working elec- trical apparatus would contain electro-magnets or similar ap- pliance of high inductance. If a disturbance occurred, produc- ing a discharge on the line, the regular apparatus by its induc- tance would choke back the discharge, which would jump across the gap from one set of teeth to the other, and so escape to the earth. riagnetic Blow=Out Arrester.— This is shown in Pig. 395. The 518 Fig. 395.— Magnettc Blow-Out Lightning Arrester, LIGHTNING ARRESTERS. 519 two flaring plates of metal approach each other closely at the lower end. One is connected to the earth, the other to the line. An electro-magnet is in the line circuit. Lightning on the line is choked back by the magnet, owing to its inductance springs across the gap, and goes to the earth. Any arc which it may form is blown out by the magnet. It is driven toward the diverg- ing ends of the plates, and breaks. The three connections to line, earth and machine are indicated in the cut. Non-Arcing Metal Arrester. — This arrester is made up of a number of cylinders of metal of the cadmium group or near it, which does not readily main- tain an arc if in the po- sition of electrodes. Fig. 396 shows the construc- tion. The seven cylinders have about one-thirty- second inch of air be- tween each two. The ex- terior or end cylinders are connected with the line, and the central cyl- inder is grounded. The other four serve to form the additional gaps. With alternating currents this arrester forms no arc after a discharge; with direct current it may form a harmless one. Discriminating: Arresters. — This name is due to Mr. A. J. Wurtz, the inventor of the last described as well as of this ar- rester. Two brass terminals an inch wide are laid in grooves and flush with the surface of a block of marble. Their ends come within half an inch of each other. A piece of lignum vitae fills the gap between their ends and across it are made a series of charred grooves about one-tenth of an inch wide and one-thirty- second of an inch deep. A cover of marble is secured over it. Fig.396.-Non-At?cino Metal Lightning Arrester. 520 ELECTRICIANS' HANDY BOOK, One plate is grounded; the other is connected to the line. No ordinary current can pass over the charred surface, which acts to Pigs. 397 and 398.— Wtjrtz's Carbon Lightning Arrester. LINE G =^ ^ .i .//fr ^/ Shunt esistance Fig. 399.— Westing house Lightning Arrester. Fig. 400.— AI.TERKATING Current LIG^T^'I^G AllREfeTER. LIGHTNING ARRESTERS. 521 conduct the atmospheric discharge to the earth. No arc forms in this apparatus. The resistance of the apparatus may be as high as 50,000 ohms. Sometimes no marble is used, the electrodes being screwed directly to the wooden block of lignum vitse. It is shown in Pigs. 397 and 398. Westinghouse Lightning Arrester.— A disk-shaped choke coil is carried on an insulator, as shown in Fig. 399. This coil has sufficient inductance to oppose the passage of a lightning dis- charge, yet not enough to seriously affect the current. To the r~®~^ Dynamo Ground tt Ground ^ Line Line Line Linfe Figs. 400a and 400b. -Double-Pole Light- ning Arresters. right are non-arcing spark gaps. The line is connected above and below the coil; the lateral connection gives the path for the lightning discharge, which goes to the earth through the ar- resters, which are of one of the types already described. Low -Equivalent Alternating=Current Lightning Arrester.— In Fig. 400 is given a diagram of an alternating-current lightning arrester for high-voltage currents. Its action is as follows: The discharge springs across the gaps and goes to the earth. Any arc formed in the shunted gaps is destroyed by the path for the cur- rent offered by the shunt resistance. The series resistance is made as non-inductive as possible, and acts to reduce any current which follows the discharge. A certain amount of the discharge goes through the shunt resistance. 522 ELECTRICIANS' HANDY BOOK. Double-Pole Lightning Arrester. — The diagrams, Figs. 400 a and &, illustrate double-pole connection of lightning arresters, where they are connected like lamps across the two leads of a circuit. Tank Lightning Arrester. — This arrester is found particularly serviceable on electric railways. Choke coils carried on a slate or marble base are put in the circuit, as shown in the upper part 111'' ''I'-..' ' . ^L4.,v^V^ / '" Bus SorJ line. Fig. 4G1 —Tank Lightning At^resteh. of Fig. 401. Conductors from the coils run down to a tank of water shown in the lower part of the cut. Water is run through the tank when a storm threatens. A slight current leakage constantly takes place, but is trifling. If a lightning discharge occurs, it goes to the earth by way of the tank. The choke coils force it to the tank. CHAPTER XXXI. THE INCANDESCENT LAMP. Incandescent Lighting. — The incandescent lamp is the expres- sion of a fundamental law of electric supply, which is to the effect that resistance in an electric circuit should be concentrated at the point where energy is to be developed. If a circuit is de- voted to running machinery, the resistance should be in the ma- chines, and as little resistance as possible should be in the lines. If lamps are to be lighted, as much of the total resistance as pos- sible should be concentrated in them. In the case of incandescent lighting, the useful resistance is that which is produced by the filaments of the lamps. All re- sistance not manifesting itself through heating the thin fila- ments represents lost power and w^aste of energy. It is a ctirious thing that the useful energy of every horse-power in an incan- descent electric-light system is represented by the ignition of only five or six feet of carbon filament. The Incandescent Lamp comprises a filament of carbon of various shapes, approximating to a letter U. The filament is inclosed in a glass bulb within which a vacuum is produced. Wires passing through the glass are connected to the source of current, which heats the filament bright red or white hot, so that it emits light. Tamidine Filaments. — Weston made for the basis of filaments a substance which was named tamidine. It was prepared from solid massive nitro-cellulose, the substance left by the evapora- tion of collodion, so familiar to the old-time photographer, and now used for surgical treatment of minor cuts and the like. The nitro-cellulose was reduced by a chemical reducing agent such as sulphureted hydrogen, converting the mass completely or nearly into cellulose. This material resembled transparent 523 524 ELECTRICIANS' HANDY BOOK. horn. Filaments were cut out of it, were carbonized, and used in lamps. Squirted Filaments. — Filaments are now made also by forcing the proper material through a die. A thick solution of nitro- cellulose, which is a syrupy collodion, can be forced through a fine aperture and evaporated, giving a thread. This after reduc- tion could be used as a basis for filaments. Cotton can be dis- solved in a solution of zinc chloride, giving a syrupy transparent solution. This can be forced through an aperture into a vessel of alcohol. This hardens the thread so that it can be handled. The zinc chloride is washed out of it as far as practicable, and it is eventually wound on drums as a long thread, resembling the fisherman's silkworm "gut," which is attached to the fishhook. The thread made as described is cut into the proper lengths ready for carbonization. Various practical details have to be followed. Bubbles are one of the troubles. The thick solution retains these with some persistence, and heating the solution in a vacuum is sometimes used to expel them from the solution. Per- fect evenness of the solution is secured by thorough stirring, and an exact formula for the solution is followed. The purified cot- ton prepared for physician's use under the name of absorbent cotton is the best material for the process. Filaments made by this method are called "squirted filaments." Carbonization is effected by heating the thread to redness in an oven. It is protected from the air by being imbedded in powdered charcoal, or by some method by which no oxygen can reach it while heated. It would instantly burn if air had access to it while at a red heat. Calibration. — As the process is usually carried out, the thread from the circular die is still somewhat soft when wound off upon the drum, and the winding fiattens it a little. It is necessary to have filaments of exact dimensions, so the filament of oval section is calibrated in two directions to determine its cross-sectional area after carbonization. Filaments are thus sorted out for various resistances. The length is not so conveniently changed, as the bulb is supposed to be suited for a certain sized loop of fila- ment. Flashing.— The filaments from the carbonizing oven are next THE INCANDESCENT LAMP. 525 flashed. This process was a very early conception. The electric- light filament is increased in density, elasticity, and hardness by it, its pores being filled and its surface being coated with graphitic carbon. A number of the filaments are fastened by holders of metal to the stopper of a jar. This jar is filled with vapor of naphtha or other hydrocarbon, and the stopper is inserted with the filaments on its inner side protruding into the jar. A current is passed through them, igniting them to bright redness. The thin parts get hotter than the thick ones. The hydrocarbon is decomposed when it comes in contact with the hot filament, and more of it is deposited where the filament is hottest, which is where it is thinnest. Thus fiashing not only solidifies the fila- ment, but builds up its thin places. Occlusion of Gases by Filament. — A porous solid has some- times a peculiar action on gases which is termed occlusion. Gases will thus be retained much as water is retained by a sponge. The thread of cellulose or cotton b-efore carbonization is as absolutely without pores as anything can well be, but in the carbonization process it becom-es full of pores, and these may occlude oxygen. When such a filament is placed in an exhausted bulb, all of the gases may not be given up until ignition is applied by the cur- rent. If gas is thus introduced into the bulb, it will have a bad effect upon the filaments. The fiashing process fills the pores, and gets rid of occluded oxygen by combustion as well as ignition. Lowering of Resistance by Plashing.— The flashing process lowers resistance 10 to 15 per cent, so due cognizance must be taken of this action in selecting the size of filament for any given lamp. It is easy to bring about any desired resistance by flash- ing, and the ^resistance can be determined if desired from tim.e to time during the process. The logical Way of determining resist- ance is to do it while hot, as the resistance of a lamp when cold is only an indirect factor as far as its use is concerned. Making Joints by Flashing.— The filament has to be fastened to a wire at each of its ends, and an Interesting application of flashing is the making of a joint between these wires and the filament. It is made by flashing the filament in a hydrocarbon vapor or even in liquid naphtha through the wires held against the ends of the filament. A solid coating of hard graphite is 526 ELECTRICIAX8' HANDY BOOK. thus formed around wire and filament end, just as if a soldered joint were made. Pasted Joints. — An easier and cheaper way to make the joint is to put a little putty-like mixture of finely-powdered carbon and molasses around the junction of filament and wires. On ignition this hardens and forms a secure joint. Electroplated and Other Joints.— These are made by electro- lytic soldering. A coating of copper is deposited over the junc- tions of wires and filaments by electroplating, forming a conduct- ing coating over wire and filament ends. The joint has often been made by a very small bolt, which passes through holes in the enlarged ends of the filament and wire, and has a nut screwed on its end. Another system is to have sockets in the ends of the Avires, into which the filament ends are thrust. The fiash joint and carbon paste joint are the principal ones used in recent prac- tice. Leading^in Wires.— The solution of the problem of passing a wire through glass and then melting the glass around it so as to form an air-tight joint hinges on the coefficients of expansion by heat of the metal and glass. These must be practically the same, or else the wire will work loose from the glass, forming cracks, perhaps very minute yet sufficient to admit air. All sorts of com- binations of different kinds of glass and metals have been tried. The practice has now settled down into the use of platinum lead- ing-in wires, which are passed through holes in the glass. The glass is then melted around the wires. The metal platinum ex- pands and contracts under changes of temperature almost exactly as much as glass. It possesses another property of considerable importance, which is that it is inalterable under any ordinary range of temperature. -It will not oxidize at any temperature, and melts only at very high heats, far higher than any to which it is exposed in the construction or operation of the Incandescent lamp. This use of platinum has drawn very lar^^ely upon the supply, and its tendency is to rise in price. The lamp maker uses as little as possible, electrically v/eldin*^ copper wire to the platinum, so as only to use enough of the rarer metal to pass through the glass. flaking the Lamps.— The methods differing in details, the fol- THE INCANDESCENT LAMP. 527 lowing cut, Fig. 402, gives a typical process. No. 1 shows a glass tube closed at the upper end, with the leading-in wires passing through it, melted in, and with the carbon filament attached. No. Z shows the globe with long exhaust tube with the filament thrust into it. No. 3 shows the melting together of the two pieces of glass with a blow-pipe flame. No. 4 shows the lamp with filament tube melted in ready for exhaustion, and No. 5 shows the lamp after exhaustion with its exhaust tube melted off, the lamp being ready for use. Vacuum.— The bulb of an incandescent lamp after the carbon Fig. 402.— Making Incandescent Laxps. is in place is exhausted until a very high vacuum is produced in it. The vacuum was originally designed to prevent the carbon from burning, but it accomplishes other results also. It keeps the filament hotter. If the bulb is filled with an inert gas, the gas under the effect of the hot filament enters into active circula- tion, cools itself against the sides of the bulb, gets heated by the hot filament, and then is cooled again. The filament has to heat the gas over and over again, and the temperature is materially lowered by the process. The efficiency is thus diminished. An exhausted bulb is much cooler when the filament is giving light than if it were filled with inert gas. As a mere matter of 528 ELECTRICIANS' HANDY BOOK, convenience this is desirable. It is a good feature about the in- candescent lamp that its bulb cannot burn the hand, or set fire to anything und^r normal conditions, although it is not altogether safe to leave burning lamps wrapped up in a combustible wrap- ping for a considerable period. Production of Vacuum. — The Torricellian vacuum, Fig. 403, such as exists above the mercury in a barometer tube, is one of the best vacuums produced without special care or for special ends. The Sprengel and the Geissler pumps are based upon the production of this vacuum. In these air pumps the piston is represented by a column of mercury, and the force driving the piston is represented by the pressure of a column of mercury over 30 inches high. A quantity of the lamps are sometimes exhausted to a pretty high vacuum by a mechanical air pump, and the ex- haustion is finished by the use of a mercurial pump. This re- moves the last air, whose removal is facilitated by passing a current of electricity through the filament, heating it as the close of the operation is reached. This expels any occluded and other gas held by the wires, glass or filament. Sometimes a little phosphorus is put into the exhausting tube and is h-eated from the outside by applying a fiame or other source of heat to the glass on which the phosphorus is lying. It combines with any trace of oxygen present. To prevent danger to the health of the operatives, and to avoid liability of ignition of the phosphorus, the modification called red phosphorus is best for this purpose. E?xternal heat can be applied to the lamp during the last of the exhaustion to assist the operation. The exhaustion is done through a tube extending from the top of the bulb. This tube is melted off in the blowpipe fiame when the exhaustion is com- plete. The point seen on the end of the bulb shows where the sealing was effected. The riercury Air Pump.— The Sprengel pump utilizing the Torricellian vacuum is shown in Fig. 404. At the top of the pump is a horizontal pipe, through which mercury is passed. At D D are cocks admitting it to the pumps, one of which is shovm on the left. The mercury descends through B, goes through the inclined tube down T and out through D' D', to be repumped into the upper pipe. R is a glass vessel containing a THE INCANDESCENT LAMP. 529 drying agent, such as phosphoric oxide or sulphuric acid. At O is an opening into which the exhausting tube a on the upper end of the lamp L can be sealed. It has another opening at S communicating with T. The mercury as it leaves the inclined tube, if there is a trace of air in R, breaks up into little columns Fig. 403.— Torricellian Vacuum. Fig. 404.— Sprengfl's Atr Pump. and draws the air down and out. The filament is heated during the process by a current adjusted in intensity by the resistance coils F F. In modern works various kinds of special pumps are employed to work on the large scale, exhausting a number of lamps simul- taneously. 530 ELECTRICIANS'- HA'NDY BOOK. The Geissler air pump is operated by the agency of a column of mercury, but involves the raising and lowering of a reservoir of mercury. The Sprengel pump is described as a typical mercurial air pump. Luminescence is the quality of giving light when heated. All substances possess more or less of this quality, some in higher degree than others. Luminescence of a very high degree is shown by the Welsbach incandescent gas light. A filament of its ma- terial would represent an almost ideal substance for an incan- descent lamp filament if it was heated so as to become a con- ductor. Metallic Filaments have been tried for incandescent electric lamps with very little success. Their fusibility is the principal objection to their use. At present the metals osmium and one or two others are being tried. Oxide Filament. — There are substances which are free from most of the objections which attach to carbon and the metals, except that they normally do not conduct electricity. These are the oxides of the metals of the earths, lime, magnesia, and others. They are in the full sense non-conductors when cold, having enormously high specific resistance, but on heating they become conductors. The Nernst Lamp is an incandescent lamp whose filament is made of earth oxides. These are absolutely incombustible, so that they can be ignited in the air, providing the condition for an open-air incandescent burner. The Glower. — The Nernst lamp filament is a straight bar of -earth oxides and is termed the glower. To its ends are attached wires. The current once made to pass through the glower raises it to a white heat, produces light, and keeps it in the conducting state. The composition of the glower is not disclosed. It is said to be composed of the rarer -earths, resembling the Welsbach gas mantle in composition. The standard glower for 220 volts is almost exactly an inch in length and 0.025 inch in diameter. It is formed from a putty- or dough-like mixture of the earths, by squeezing them through an aperture in a die. This produces a thread, which is dried and baked. The cuts of the Nernst lamp all show the glowers. THE INCANDESCENT LAMP, 531 Glower Terminals. — The connection of the wires with tho glower was originally effected by winding platinum wire around the ends, and puttying over the ends with cement. This did not work very well, as the wires were apt to become partly detached, and thus had their contact with the glower made imperfect. The result of this Vv'as that the glower soon broke near the terminal, where the bad junction caused a concentration of heat. Another method is based on the reverse principle. A globule of platinun at the end of each wire , lamp terminals is embedded in each end eTj ^ of the glower. Any shrinkage in the ma- terial of the glower causes it to grip the bead still tighter. It cannot shrink away from it, as it tends to shrink from the wires wound around its ex- t e r i o r. Conducting wires fused to the plati- num globules project an inch or two from the ends of the glower. The ends of the conducting wires are fastened to the body of the lamp by little aluminium plugs. The ends of the wires are thrust into holes in the two contact blocks of the lamp, and the plugs are forced into holes, wedging them fast. The glower becomes a conductor when heated to about 1300'' F. (700° C). When cold it is a non-conductor. Heaters. — The glower can be heated by a match or alcohol flame, in order to make it conduct current. In the lamp as now made electric heaters are us«d, also shown in the cuts. These are of various shapes, consisting of platinum wire wound upon a porcelain form and imbedded in refractory paste. W)*en the Fig. 405.— Diagram of Nernst Lamp COKSIRUCTION. 532 ELECTRICIAXS' HANDY BOOK, lamp current is turned on, none can go through the cold glo^yer, and a slight current only passes through the heater. It is enough to make it quite hot; and as it is in close proximity to the glower, it heats the latter, which in a few seconds begins to pass a cur- rent strong enough to excite a magnet, which attracts pivoted armatures cutting out the heater, and thereafter all the current goes through the glower or glowers. The heater has to heat the glower up to a temperature of about 1742° F. (950° C.) Ballast.— The glower is in series with a steadying resist- ance, which is called the ballast. The resistance of the glower di- minishes with increase of tem- perature. The resistance of iron wire increases with increase of temperature, and the two bal- ance each other approximately, which prevents the glower burn- ing out. The case is analogous to the use of the individual re- sistance in a constant-potential arc lamp. The Nernst lamp has to be employed on fixed-potential circuits. Iron wire is selected for the ballast because it pos- sesses in a high degree the prop- erty of increasing in resistance with increase of temperature. It is inclosed in glass tubes hermetically sealed and filled with nitro- gen gas, and is shown in Fig. 405. The Cut-Out, also showm in Fig. 405, is an electro-magnetic switch which opens a circuit when its magnet is excited. This circuit is normally closed, and only opens by the action of the electro-magnet as described above. The magnet winding is in series with the glower; the circuit which it opens contains the heater. Direct-Current Lamps.— If used on direct current, a blacken- ing of the glower near the negative end takes place, which causes Fig. 406.— Nernst Lamp Beady FOR Insertion in its Socket. THE IXCANDE8CEXT LAMP. 533 the efficiency and candle power of the glower to fall off. Its dura- bility is also impaired. On alternating current this action does not take place, and its life is much longer. Vacuum Lamps — If the glower is inclosed in a vacuum, its efficiency as far as the glower is concerned is increased. But this increase is accompanied by a very rapid rate of diminution of resistance with increase of temperature. This has to be met by a larger ballast, which reduces the efficiency. It is considered preferable to inclose it in a globe with access of air. This gives Fig. 407.— Spiral. Heater and Single Horizontal Glower OF Nernst Lamp. Fig. i^'S.— Spiral Heater and Single Vertical Glowf b op K ERNST Lamp. enough cooling to lighten the work of the ballast, and yet to give higher efficiency than in the open air. Before a glower breaks, the voltage rises rapidly until the rupture occurs and the lamp goes out. Sometimes as many as six glowers are put into one lamp, in which they are simultaneously ignited. The efficiency of such is higher than that of a single-glower lamp. The Efficiency of the Nernst Lamp is about double that of the ordinary incandescent lamp. The cuts, in the light of what has been said, are self-explana- tory. Fig. 405 shows the parts of a lamp in diagram. The magnet coil being inactive, the pivoted armatures are not yet attracted. When attracted they open the circuit at their lower 534 ELECTRICIANS' HANDY BOOK. ends, one of which is marked '^silver contact" in the diagram. Fig. 406 shows the heaters and glowers of a lamp ready for in- sertion into the socket, the parts being marked. The next cuts, Figs. 407 and 408, show spiral heaters surrounding the glowers. Distribution of Light. — The diagram, Fig. 409, shows the Fig. 409.— Distribution of Light From the Nernst Lamp. distribution of light of a Nernst lamp and of other lamps in the vertical plane as by the following table: 1. 110-volt, A. C. constant potential arc, 6.3 amp. 2. 110-volt, D. C. constant potential arc, 4.9 amp. 3. G.3 amp. D. C. series arc, 71.6 volts. 4. G.6 amp. A. C. series arc, 65.4 volts. 5. 6 glower Nernst lamp, 220 volts. Arc l^mps — Opalescent inner -^nd clear outer globe. Nernst lamp — 8-inch sand-blasted globe. CHAPTER XXXII. THE ARC LAMP. The Voltaic Arc. — If two rods of carbon are connected to a source of current and are brought into contact with each other, and are then separated a fraction of an inch, the current will continue to pass across the interval. An intense heat is produced, and the space between is filled with carbon vapor and minute particles. The heat makes the carbons very hot. As carbon is not a very good conductor of heat, almost all the heat concentrates on the ends. The arc may be produced by direct current or alter- nating current, which gives two divisions of the subject, direct- current and alternating current arc. Positive and Negative Carbon. — When a direct-curent arc is produced in the open air between two carbon pencils, both wear away, but do so differently. One keeps a pointed end, like a sharp- ened lead pencil, and is the negative carbon. The other has a little crater or cup formed on its end, and is the positive carbon. The latter gives far more light than the other. Naturally, the interior of the crater radiates the most light. In direct-current arcs the crater of the positive carbon is made to face as nearly as possible in the direction in which the light is to be utilized. Thus, for overhead lamps the positive carbon is placed uppermost, so that its crater radiates light to the ground. Striking the Arc— The arc will not strike across a space filled with air unless a very short one. The carbons may be arranged to stay in contact when idle and to be pulled apart the instant the current starts. As they separate, the arc forms across the gap or space between the ends of the carbon rods. This is the universal way of operating arc lamps, although it can be done otherwise. If a spark can be made to strike across the gap. the arc will start over the path thus made for it. The air between 5»3o 536 ELECTRICIAN^' HANDY BOOK. the poles is intensely heated, and is a tolerably good conductor, so that once the arc is established, it can be drawn out to a con- siderable length — greater than the striking distance of the po- tential utilized. Heat of the Arc— The resistance of the arc is not great enough to account for its intense heat. The positive pole is hotter, 7200^ F. (4000^ C.) than the negative, 5400° F. (3000° C.) to 6300° F. (3500° C). Counter electromotive force is set up, due to thermo- electric effect, or to condensation of carbon vapor, and is equival- ent to resistance, and the heating effect results. The higher temperature of the positive pole causes it to wear away the faster. With alternating currents the poles wear evenly, and with almost flat ends, if the arcs are inclosed in a glass globe so as to be partly protected from the air. Voltage Drop.— In a direct-current arc the voltage drop be- tween the positive carbon and the arc has been determined to be about 40 volts. In the arc itself a drop of 2^^ volts was observed, and a 2''2-volt drop between the arc and the negative carbon. These determinations are not to- be considered accurate. They indicate the distribution of voltage, of r-esistance, and of light- giving areas or volumes with a good degree of approximation. Counter Electromotive Force is believed to exist in the direct- current electric arc, and to account for part of its apparent re- sistance. The cause is not certain. The different temperatures of the carbons producing a thermo-electric effect has been assigned as its cause. The alternating current arc, both of whose carbons are of identical temperature, exhibits apparent resistance enough to have counter electrom.otive force attributed to it. Solidifica- tion of carbon vapor may be the cause of its production in both direct and alternating current arcs. It would be possible to imagine the rapid volatilization and condensation of carbon vapor in the successive cycles of an alternating current as produc- ing an alternating counter electromotive force. The counter electromotive force for a 10-ampere 45-volt arc with pure carbons has been put at 35 to 39i^> volts. This is ap- proximate only. All determinations affecting the internal physics of the arc must from the nature of things be difficult to execute, and the results will generally be approximate. THE ARC LAMP. 537 The Resistance of the Arc Proper has been placed at about 5 ohms per inch of length. The 10-ampere arc, which is a standard, varies from 1/10 to i/^ ohm in resistance, the arc length varying from 1/16 to % inch. Questions in which length of arc is in- volved are only to be valued approximately, as there is nothing accurate about the determination of its length. The resistance of the arc varies inversely in some ratio with the current. A heavy current diminishes the arc's resistance. This is the reason an arc lamp without a resistance or inductance for alternating currents cannot be used on parallel or constant potential systems. This diminishing of resistance is partly due to reduction of the resistance of air by heat, for the more intense current heats the air to a higher degree and heats more of it than does the smaller current. Another cause is the presence of carbon in the arc, probably as vapor, possibly as particles, which is in- creased in relative amount by greater heating. The old modifica- tion, which has recently been experimented with, of introducing alkaline earth salts or the like into the arc diminishes its resist- ance by supplying it with vapor of these salts or of their con- stituents. Increase of pressure increases the resistance. This applies to pure carbon arcs, and is by some thought to produce this effect by preventing the production of the full amount of carbon vppor. Efficiency of the Arc Light. — Of the efficiency of the arc as a light producer nothing can well be said beyond the comparison with other sources of light. The two reasons are that the arc is very seldom photometered, and that the absolute unit of light is as yet undetermined. If light is defined as that which affects the retina of the eye, its mechanical equivalent may be exceed- ingly small. What we know about odor tends to ratify this be- lief. An almost inconceivably small quantity of matter is re- quired to affect the olfactory nerves. A very minute amount of energy is represented in the action of light upon the optic nerves. The arc is one of the most efficient sources of artificial light. The magnesium light is put next to and very close to it, and by modifications might be made to equal or exceed it. It is 8.66 times as efficient as candle light, 13 times as efficient as gas light. 538 ELECTRICIANS' HAXDY BOOK. 5.2 times as efficient as the Welsbach light. These all are so variable that the relative figures given are only approximations. The reason of its efficiency is that its heat is so intense. There is a possibility that there is a considerable loss by some of the heat producing ether waves of so short a period that they do not affect the optic nerve or are not visual. Quality of Carbons. — The nature of the carbons affects the efficiency. The great agent of economy is the concentration of heat at the ends of the carbon. Too hard a carbon is apt to be a relatively good conductor of heat and therefore uneconomical. A small diamet-er of the carbon pencil favors concentration of the heat at the point, and small carbons give higher results. The efficiency diminishes approximately in inverse ratio with the diameter of the carbon. A soft core in the carbons reduces the efficiency. In order to give better surface contact between the carbon clamps and the carbons, the carbon pencils are often copper-plated, and nickel plating has been applied. This dimin- ishes the light a little by improving the conductivity of the carbons for heat. Power Consumed in Arc. — A consum^ption of 480 watts is usual in a nominal 2000-candle-power lamp. The ratio of volts to amperes in the production of the watts expended in an arc lamp affects its efficiency and consequently its light. Carhart found that 45 volts and 10 amperes gave a maximum light of 450 eandles or 1 candle to the watt. With 8.4 amperes and 54 volts the maximum candle-power was doubled. There is nothing definite about these figures, as the size and quality of the carbons would affect them materially. Effect of Air Blast.— A blast of air will blow out an arc as it will a candle flame. This principle is utilized in the Thomson- Houston alternating-current dynamo. A blast of air is there produced by a rotary blower, which is directed on the ends of the brushes to blow out any arc which may form in the operation of the machine. Effect of riagnet. — A powerful magnet deflects the arc to one side, and if near enough thereto and strong enough, will blow it out as a blast of air will. Voltage Drop and Arc Length, — Where the arc produced be- THE ARC LAMP. 539 tween tv;o carbons attains a certain length, it has to be increased in length to keep a fixed voltage. This is in line with the prop- erties of the arc, which makes it impossible to operate arc lamps on constant-potential circuits v/ithout auxiliary resistance coils. The resistance falls with increase of current, and the lengthening of the arc is necessary to bring its voltage back to its original figure. Wearing of Carbon <^. — With a direct current the positive car- bon wears away about twice as fast as the negative. The latter has a little accretion of carbon particles form upon it, which may increase its length. This amounts to nothing from the practical standpoint. In open arc practice, when the arc is produced in the open air, the accretion burns away. With an alternating current the wearing of the carbons, other things being equal, is the same for both. But when they are placed vertically, as they always are now, the upper carbon has been found to wear away about eight per cent faster than the lower one. This is due to the uprising currents of air and to gravity acting on the transfer back and forth of carbon particles. This uneven wearing away of the carbons affects the operation of arc lamps for some special purposes. Such occur in its use in searchlights and lighthouses, where the center of light must be at the level of the focus of the lens or reflector. Different feed- ing rates for the two carbons may be used to keep the light-giving gap in its proper place. Arc Light Carbons. — Carbons are made from a mixture of finely ground and ignited carbon with some carbonaceous cement- ing material such as pitch. They are molded into shape and baked for a long period at a red heat with exclusion of air. Two general systems of molding them are followed. In one grooved plates are the molds. The plates contain straight grooves of semi-circular section spaced equally on both plates, so that when the plates are laid face to face the grooves form a series of cylindrical molds. The composition is molded in these. Carbons which have been made by this or analogous methods sometimes show on their peripheries the mold print. The filled molds are heated in an oven until the mixture softens. Avhcn they are subjected to a hydraulic pressure of several hun- 540 ELECTRICIANS' HAXDY BOOK. dred tons. They are then removed, and any fin left where the joints between the molds come is scraped off, and they are ready for baking. In another system the carbon composition is forced through a die by a hj'draulic or other form of powerful press. The die which is at the foot of the apparatus has a circular aperture of the size of the carbon. The cylinder is filled with composi- tion which is forced out through the aperture or die. As the cylinder emerges it is cut into the correct lengths and the green carbons are baked. In modern practice the mixture is made into cylinders fitting the press cylinder. The size may be about six inches long and two to six inches in diameter. The cylinders are horizontal. To produce cored carbons a circular mandrel extends through the aperture of the die, and the carbon is forced out in the shape of a hollow cylinder. The central opening of the carbons is then filled with a composition, which on baking gives a softer carbon. The object of the cored carbon is to hold the arc in inclosed lamps in a central position. The baking of carbons has to be sufficient in temperature and duration to completely decompose the cementing pitch or syrup, and to give them good conductivity. Too much baking may make them too hard. Too rapid application of heat may warp them, and it is essential to good operation that they should be perfectly straight. To keep them straight during the baking and to ex-' elude air, one method adopted is the imbedding the green car- bons in sand, one layer of carbons above the other in the furnace. From seven to fourteen days may be consumed in charging a furnace, baking the carbons and cooling. The crooked carbons are sorted out from the lot by rolling on a plane surface. If not too crooked, the ones thrown out by the rolling test are sold as seconds. From crooked carbons, short ones useful as bottom carbons can sometimes be cut. The forced carbon, as the one made by the die process is called, is used in inclosed arc lamps, especially in carbon feed lamps. The Direct=Current Open Arc is the arc produced by direct current between two carbons in the open air. It varies in current from 6 to 10 amperes, and in electromotive force expended or THE ARC LAMP. 541 drop from 42 to 52 volts. This refers to ordinary or standard size lamps, such as are in general use. Larger lamps with car- bons of greater diameter use more current. Some very large lamps have used carbons of an inch or more in diameter. A very large number of open-arc lamps are still in use. The new installations are almost universally fitted with inclosed-arc lamps. One of the great expenses of conducting an open-arc light system is the frequent trimming of the lamps. This requires time, which involves a labor charge. The carbons require fre- quent replacing as they burn out, which is another item of ex- pense. Distribution of Light in Direct- Current Open Arc, — The gas engineer has always tested the light given by a gas flame in the horizontal direction. It has never been the practice to try it at various angles from the horizontal. With gas this would be far from easy, because the gas flame must burn vertically, and the construction of a photometer to test its value as a light giver at different angles would be somewhat difficult. The electric light, arc as well as Incandescent, is far from being as sensitive to change of position as is the gas flame, and by inclining the lamp in different positions, candle-power at various angles is deter- mined. This is spoken of more at length elsewhere. With an arc lamp with carbons end on to each other, now the invariable position, the following variations of candle-power to angle exist with direct current. The horizontal direction gives a low candle-power. The crater is screened by its edges from contributing its due share to the light. As the angle is depressed, the light given increases, until in the neighborhood of 40 deg. depression the greatest light is given. After this it decreases rather rapidly to zero directly underneath the lamp. Typical distributions of illuminating power are shown in the cuts, Figs. 410 to 414. The radius vectors of the curve indicate the relative illuminating power of the arc at the different angles indicated by the figures from 0° at the upper vertical to 180° at the lower vertical. 130° from the vertical is 40° from the horizontal, and this angle marks the line of greatest light. 542 ELECTRICIANS' HANDY BOOK, The lower carbon, cutting off light by its shadow, is respon- sible for the diminution that increases so rapidly once the IBS'" ingle from the vertical is passed. It will be evident that it is impossible to express the value of an arc lamp in candle-power unless the same course is followed which is outlined above. It is taken in different directions, vary- Hissing Arc Over Feed 300 Max. C.P. bed eOO Max. CJ>. Normal Arc 48 Volts Extremely Long Arc Sluggish Feed Figs. 410 to 414.— Distribution of Light from Open-Arc Lamp. ing from horizontal to vertical, thus giving eventually what is known as the spherical candle-power. Commercial Rating of Arj Lamps. — A practice has arisen of calling the illuminating power of an arc lamp of standard street size 2,000 candles. This is a 480-watt lamp. Another standard size is the 300-watt lamp, rated at 1,200 candle-power. These values are far in excess of the spherical candle-power. But as it is the earth and not the sky which is to be illuminated, the THE ARC LAMP. 543 above figures are nearer the truth than they are usually supposed to be, if the value of the lower hemispherical candle power is taken. Hissing Arc. — On being driven too hard, or with too much, current, the arc makes a noise. Some change occurs at this point, because the voltage drops suddenly 10 to 2t) volts, and with varying current gives a straight-line characteristic for the voltage, the voltage remaining unchanged for wide variations of current. No explanation that is satisfactory has been offered for this phenomenon. Light Given by the Arc Proper. — It has already been noted that the positive carbon gives the most light. It gives 85 per cent of the light, the negative 10 per cent, and the arc proper only 5 per cent. Resistance of Short Arcs. — When the current passing between carbons within less than 1/25 inch of each other is increased, the resistance does not decrease in the same proportion, and the product, R I, which by Ohm's law is equal to E, increases. There- fore the voltage drop with such short arcs increases with increase of current. If this condition held for commercial arc lamps^ they could be used on parallel circuits at constant potential with- out wasteful resistance coils. The length of 1/25 inch seems to mark a point where the voltage remains constant for a wide range of current. This is because in the arc of this particular length the resistance diminishes exactly in proportion as the cur- rent increases, giving a constant value to the product R I or to E. The Resistance of Longer Arcs on increase of current dimin- ishes in more rapid proportion than with short ones, so that as current increases, the product R I grows less. This is why a resistance coil for each lamp has to be employed for constant- potential lighting, as explained on page 553. From what has been said, it follows that on constant-current supply the energy expended in maintaining an arc will increase as the length increases, and with constant leneth will do the same as the mr^ent intensity increases. On trial the increase is found to be a nronortional one in both cases. Sta^'onar^' Sat^.— This is the state of normal burning of an arc lamp. When it starts, its constants varv until it reaches the 544 ELECTRICIANS' HANDY BOOK. degree of heat clue to the current and distance between the car- bons. When this heat is reached, voltage and resistance remain constant as long as the length of the arc and the strength of the current are unchanged. In practical operation arc lamps are best operated in series. In this system the current is kept con- stant by the station management, and the regulators or lamp machinery maintain the distance between the carbons unchanged. Alter nating= Current Arc. — This type of arc consumes about the same watts in effective reckoning as the direct-current arc does. The 480-watt standard divides into 15 amperes and 30 or 35 volts. The volts are given in effective value, so the maximum value of the electromotive force is greater than the voltage of the same direct-current arc. The current value is greater in the al- ternating-current arc than in the direct-current arc. This com- pensates for the alternations, which would tend to produce flick- ering. Power Factor in AIternating=Current Arc— In the alternat- ing-current arc the current lags about 30° behind the electromo- tive force. This introduces a power factor of 85 per cent of the apparent watts or product of effective current and potential drop. Influence of Wave Form. — The efficiency of the alternating-cur- rent lamp is greater as its current curve avoids peaks and as its frequency is increased. It will be seen that the period of change of direction is a time when the carbon gets so little energy that it has to give light from its own acquired heat. The shorter this period, the greater is the efficiency, and therefore a high fre- quency is advisable for efficiency as well as for steadiness. A flat-tipped' wave with quick or steep changes from one extreme to the other favors efficiency. Distribution of Liglit of Alternating=Current Arc Lamps.— The light from a lamp does its work generally in the lower hemi- sphere of its distribution; tlie light cast out horizontally and at all downward angles is the useful light. This distribution is 2:iven by the direct-current arc lamp. The cratered upper carbon, in which the heat is concentrated, gives most light, and its light is principally thrown downward, and out horizontally. The light cf the alternating-current arc is distributed alike up and down, and for this reason this arc is less advantageous than the THE ARC LAMP. 545 other. A reflector is often used to reflect the upward rays down- ward, but its effect is small. Reactance Coii or Economy Coil.— Alternating-current arc lamps can be used on constant-potential circuit by the introduc- tion in the circuit of each lamp of an inductance, a coil of wire with laminated iron core. A single coil with several interme- diate connections may be used. These operate in a manner analo- gous to that of the individual resistance coil of a constant-poten- tial direct-current arc lamp. They work by inductance, which is exceedingly economical as a reducer of current strength. It com- pensates in the case of constant-potential lamps for the otherwise low economy of the alternating-current lamp, and for its uneco- nomical distribution of light, as spoken of in the preceding para- graph. Efficiency of Alternating=Current Arc Lamps. — The mean spherical candle-power for equal watts is put at one-half that ot the direct-current arc lamp. Noise. — The alternations in the current and the effects of the corresponding induction on the laminations of some of the parts caused considerable noise. In modern construction the latter noise is prevented by clamping fast all vibrating laminations af iron, and by the use of springs and India-rubber supports for such parts, so as to prevent anything like sounding-board action. The application of the inclosed-arc principle operates to greatly diminish the hum of the arc. Duration of Carbons. — The alternating-current inclosed-arc lamp with a 6-inch lower and 9iA-inch upper carbon burns about 80 hours before the carbons need renewal. The direct-current in- closed-arc lamp may run 100 to 150 hours before the carbons need changing. This is to be compared with 8 to 10 hours' dura- tion f jr open-arc lamps. Length of Arc. — The alternating-current arc is in practice about % inch with a 6-ampere current and 70 to 75 volts. This is quite different from the direct-current factors of working. Inclosed- Arc Lamps. — The original arc lamp of the early days of electricity, with charcoal electrodes made conducting by im- pregnation with mercury, was of very short duration as regarded its carbons. It was only experimental, was actuated by a primary 546 UJLECTRICIANS' HANDY BOOK, battery which soon expended itself, and awaited the development of some cheap source of electricity to become practical. When the modern dynamo gave large quantities of electric en- ergy, many forms of arc lamp were devised, depending for the durability of their carbons on the composition of the same. These were made hard and relatively incombustible, but in the intense heat of the arc they burned away quite rapidly and had to be frequently replaced. On every replacement a stump of more or less considerable length was lost and thrown away. About 1882 attempts were made to follow in the wake of the Incandescent lamp, and to inclose the arc in an air-tight globe. In 1894 successful inclosed-arc lamps were produced, and now the movement is for their universal use. It is evident that an hermetically-sealed globe is almost an impossibility for an arc lamp. The carbons are certain to be re- duced as the lamp burns, irrespective of combustion. The arc wears away the carbons mechanically by its transfer of carbon particles from one carbon to the other. The problem of inclosing and protecting the carbons is solved by using an approximately tight globe. The carbon is fed through a hole in the top, which it almost fills. The globe is otherwise closed. A very little air gets in by diffusion, but the duration of the carbons is increased very greatly. On standing idle, the globe slowly fills with air. On starting the arc, combustion of the carbon begins, and in a few minutes the oxygen in the globe is exhausted, being replaced by carbonic-oxide and carbonic-acid gases, and the wasting of the carbons is now mechanical for the most part. The Action of the Inclosed Arc is to transport carbon particles from the positive carbon to the negative. These particles im- pinging on the hot negative carbon stick there, and tend to form a little lump upon it. The positive carbon wears away, with a slight tendency to become concave or a very little hollowed at the end. The negative has the opposite tendency, becoming slightly rounded. It the carbons are started with the usual pointed ends, ihey soon become almost flat-ended. The object being to preserve the shape of the carbon ends, the more or less irregular deposit of carbon particles on the negative THE ARC LAMP, 547 electrode is a disadvantage. The carbon particles do not all de- posit on the negative, but also tend to form a blackish coating on the glass. If the globe were hermetically sealed, the glass would inevitably blacken. Recourse is had to the air as a cleaning agent. Enough finds its way into the globe to burn up the car- bon particles and vapor and prevent it from forming the deposits on the glass and on the negative carbon. The present successful form of inclosed-arc lamp is the product of years of experimenta- tion and gradual development. The General Electric Company inclosed-arc lamps have a com- bined globe and lower carbon holder. The lower carbon is held stationary, all the feed being done by the upper carbon. This feature enables the trimmer to remove the lower carbon and globe together and replace them by a clean globe and new lower carbon. The globe removed can be returned to the station for cleaning. The inclosing globe is comparatively small, 6i^ inches high by 3 inches diameter. A small passage several inches long in the cap connects the space inside the globe with the air. The idea is to have the passage act as a gas chamber to prevent the direct access of air. Globe and Carbon Holder.— This is shown in Fig. 415. B is the globe holder which is seen rising from it, and in its center are shown the lower carbon and the lower end of the upper carbon. The head of the screw for fastening the lower carbon faces the reader. A is the holder for the outer globe, which is held from shaking by the spring h &, which goes inside it. The frame carry- ing it is drawn down. "When the lamp is in use, the frame is pushed up, the clamp C enters the slot at D, and by turning the clamp through 90° all is secured. Inclosed=Arc Lamp Carbons .-—In inclosed-arc lamps of stand- ard size, i/^-inch cartons are used. The upper one is for direct- current lamps 12 inches and the lower one 5 inches long. The upper one wearing away twice as fast as the lower one, becomes too short after long burning. Before It loses more than half its length, the lamp has to be trimmed. This is not only necessary for the replacement of carbons, uut also for cleaning the small globe. A certain amount of carbon dust and ashes collects in the inclosing globe, which has to be cleaned. 548 ELECTRICIANS' HANDY BOOK. The short lower and negative carbon becomes reduced to a mere stump, and can without much waste be thrown away. The upper carbon, about half its former length, is preserved and is cut down to the standard length of 5 inches, and is used as the lower carbon. A lamp of this type burns for 125 hours or for 10 or 12 nights without trimming. One characteristic of inclosed arcs affects the shape of the carbons. They burn with approximately flat ends. This undoubt- edly hurts their efficiency by screening off the incipient crater or hottest point on the positive. The flat opposing negative and the projecting area of the positive operate to produce this screening. On 110-volt constant-potential circuits a lamp will take 80 volts; on 120-volt circuits it will take 85 volts. The remainder is taken up and lost in the resistance coils which are used on constant-po- tential systems. The Clutch. — The development of arc lamps was marked by the most important application of the clutch to the carbon feed. The cut, Fig. 416, shows one of the original forms, the old Brush clutch. It is designed as shown to feed two carbons. It consists of a flat plate W with a hole slightly larger than the rod or carbon R which passes through it. When its end is lifted by the mech- anism operating K, the plate binds or grips the rod and raises it. When the end is lowered, rod and clutch descend together, the clutch not losing its grip until the outer free end is arrested in its descent. The clutch is then said to trip. As it approaches the horizontal position on being tripped, the grip ceases, and the rod descends through it. In Fig. 417 is shown a modern clutch. When the lever F is raised or as long as the weight of the carbon is sustained by it, the upper end C of the shoe is pressed against the carbon or carw bon rod A, and grips it between Itself and the upper end of D. As tbe carbon burns and is fed down a little, the tripping piece E touches the tripping platform G, and the lever F descending, the grip opens and the carbon can drop down. The lifting posi- tion is shown in the left-hand figure, and the releasing position in the right-hand figure. A still simpler clutch is shown in Fig. 418. It is used in the General Electric Company's arc lamps. THE ARC LAMP. 541) Tripping Platform. -This name is given to the little platform or plate which the clutch comes in contact with on its descent, and contact with which trips it, and causes it to relax or lose it& grip upon the rod. In Fig. 417 a tripping platform is seen at G. Fig. 415.— Globe IIolder or Inclosed^Irc Lamp. Fig. 417 —A \c Lamp It is also shown in Fig. 422 directly above the lamp globes and below the clutches. Carbon=Feed Lamps.— The clutch is used for almost all com- mercial lamps. It may grip a brass rod or tube to whose lower end the carbon pencil is secured. The advantage of this arrange- ment is that the clutch has always the same cylinder to act upon. In many lamps the clutch grips the upper carbon directly. Such 50 ELECTRICIANS' HANDY BOOK. Fig. 4ia— Arc Lamp Clutch lamps are said to have a carbon feed. The carbons for such must be of uniform shape, and but a very slight variation in diameter is admissible. Concentric Magnets. — In some lamps a single magnet coil is used, placed directly in the axis of the lamp. A plunger works up and down within the coil and operates the feeding mechanism. In other types of lamps the magnet coils are placed to one side of the axis. In Fig. 419 is shown the type used in some of the General Electric Company's lamps, a U-shaped magnet with double-plunger arma- ture. Dash Pots. — A dash pot, Fig. 420, IS a cylinder with a piston. Owing to the slow escape and ingress of air, the piston cannot move rap- idly up or down, although it fs devoid of friction as far as possible. Thus, it is quite free to take any position, but is not free to do so rap- idly. The piston is attached lamps, so as to rise and fall without effect upon the posi- it presents all sudden Fig. 419.— Arc Lamp Magnet With Double-Plunger Armature. to one of the armature parts in arc with the armature. While entirely lion ultimately taken by the armature, movements and secures steady carbon feed. Carbon Holders. — The lower carbon does not move, and is set into a socket, as shown in Fig. 415. This may have a setscrew to retain the carbon firmly in position. An upper carbon holder is often a short tube. Fig. 421, whose lower end is slotted and springs over the upper end of the lower carbon; the wire from the line connects to the top of this holder. The holder slides up and down in a long vertical tube in the axis of the lamp. Its mo- THE ARC LAMP, 551 FiQ. 420.-DASH Tot. tions are governed by the magnets and clutch mechanism. In the diagram, Fig. 422, the tube is seen in a vertical position directly- over the carbons, and within it is seen the carbon holder with the leading-in wire coiled in the tube above it. Constant=Current or Series Arc Lamps.— In operating arc lamps on series, a constant current is forced through the line. The length of the arc cannot therefore be regulated by the current in series, as the latter is invariable. The carbons in a single lamp may be drawn so far apart as to greatly increase the re- sistance and disturb the working of the entire series of lamps. This possible trouble has to be pro- vided for also in constant-current lamps. Two magnets are used to operate the clutch. One magnet is in series with the lamp, and when the current is turned on, the clutch is raised, lift- ing the upper carbon, which when the lamp is idle rests upon the lower one, and causes the arc to strike. This ends the functions of the series mag- net until the next period of lighting comes. A second magnet is placed in par- • allel with the arc. As the arc in- creases in length, the resistance of the arc also increases, and more current is shunted through the shunt magnet. This mag- net is so connected to the clutch that as it attracts its arma- ture and the latter rises, the clutch descends, thus shortening the arc. If for any cause the arc should become too long, so as to re- quire two or three more volts for its maintenance than proper, the mechanism closes a cut-out, which operates by closing a cir- cuit in parallel with the lamp. If the carbons descend, and the ends come in contact because the clutch trips and refuses to act. the cut-out also closes. Thus the cut-out becomes operative at either extreme. The diagram. Fig. 422, illustrates the action. It shows the lamp Fig. 421.— roppER Carbon Holder. 552 ELECTRICIANS' HAXDY BOOK, CUT-OUT STARTING RESISTANCE inactive, the carbons in contact, and the cut-out closed. If cur- rent is turned on, it goes through the cut-out. In series with the cut-out is a coil which provides the starting resistance. Its resistance shunts sufficient current through the series magnet to cause it to attract its armature and raise the clutch. This sep- arates the carbons^ the arc strikes, and current is shunt- ed through the shunt mag- net. This at once begins to regulate the length of the arc. The armatures of the shunt and series magnets operate a rocker arm which is pivoted between the mag- nets, so that the series and shunt magnet have reverse effects on the movable upper carbon. As the shunt-magnet armature is drawn up, the clutch de- scends, owing to the action of the rocker arms, and the reverse action takes place when the shunt-magnet ar- mature descends. In this way the increase of arc length shunting more cur- rent through the shunt mag- net causes the clutch to de- scend and the arc shortens. The dash-pot is shown to the left of the central tube above the rocker arm. Immediately below the clutch is the tripping plat- form, seen extending over the top of the globe. Adjusting Weight.— This slides back and forth upon the rocker arm attached to the two armature rods. This is fastened in any desired position by a setscrew. For variations in current exceed- ing 0.2 ampere above or below the rated current of the lamp, Fig. 423.— Diagram of Constant- Current Series Arc Lamp Mechanism. THE ARC LAMP. the weight must be shifted. Moving the weight toward the clutch rod reduces arc voltage, and moving it away increases arc voltage. Fig. 423 shows the lamp with the cover removed from the mech- anism. The parts can be identified by the diagram, Fig. 422. Action of an Arc Lamp on a Constant^ Potential Circuit. — The resistance of the arc decreases as the current increases, and vice versa. Therefore on a constant- potential circuit, where the current is practically unlimited, an arc lamp can- not be used williout auxiliary apparatus. The resistance coil in the case of the direct-current arc lamp, and the indue-' tance coil in the case of th-e alternating- current lamp, are the auxiliary apparatus preventing this action. Action of the Resistance Coil in a Constant 'Potential Arc Lamp. — A mo- mentary increase in the current through a lamp without a coil would lower its re- sistance so that too much current w^ould pass, and the current would increase until some damage would ensue or until a fuse would blow out. But a fixed resistance in series with the lamp prevents this trouble. By Ohm's law, E = RI, with fixed resist- ance the drop required to force a current through the resistance will increase or de- crease in proportion to the current. The drop of potential expended on the lamp alone is a fixed amount, that of the po- tential of the system minus the drop ex- pended on the resistance coil. The moment the current increases in the lamp, the drop in the resistance coil is increased and that in the lamp is diminished. This reduction of drop cuts down the cur- rent again to its proper amount. A momentary decrease in the current, on the other hand, in- Ftg. 423.— Constant- Current Series Arc Lamp. 554 ELECTRICIANS' HANDY BOOK. creases the resistance of the lamp, which cuts down, the current still further, and the lamp may be quite extinguished if without a series resistance. But with a resistance in series the action converse to that just described takes place. A reduction of cur- rent through the lamp and resistance coil can only be due to an increased resistance in the lamp. This decreases the drop in voltage due to the resfstance coil; and as the dynamo maintains a constant voltage, the drop at the lamp is increased. This oper- ates to give more current to the lamp, compensating for its in- crease in resistance. The reactance coil in series with the alternating-current lamp acts fn the same way, except that reactance of induction plays the principal role in steadying the lamp, resistance being quite sec- ondary. An additional regulating action of the series resistance may be sought for in its variations in temperature. As more cur- rent passes, it gets hotter and increases in resistance. This is exactly what is wanted; but whether this action is ever sufficient in extent to play any part in the actual regulation is problematical in most cases. A standard potential for constant-potential systems is 110 volts. This, of course, varies considerably in different parts of a district, but it gives a basis for parallel circuit arc lighting. Forty to fifty volts are the drop for a commercial open-arc lamp. Two in series with a steadying resistance will meet the voltage of the incandescent system. This has become a general method of dis- posing of them. They take some ten amperes of current, so that each group of two in series represents in current consumption twenty incandescent lamps in parallel. The Parallel=Circuit System of Electric Supply is very ex- travagant in first cost of installation. A district could have its illumination supplied by incandescent lamps in series of twenty or more through a network of comparatively small wires. Roughly speaking, one extreme would be the case where the copper mains which would supply the lamps would be of but one-twentieth the size of those required on parallel circuit for the same lamps. First cost of installation is capitalization, and interest has to be paid upon it, so that heavy copper mains and large current THE ARC LAMP, 555 machines are a source of annual expense just as much as coal consumption. An arc lamp gives far more light per unit of power than an incandescent lamp. Placed in parallel circuit it exacts large mains, and the resistance in series consumes energy. The series connection is the ideal system for arc lamps. They are used for continuous or periodic illumination, are not supposed to be lighted and extinguished by con- sumers, and the use of them on parallel circuit is a concession to an existing state of things only. No engineer would pri- marily establish a parallel system of arc lighting. Constant^Potential Arc Lamps. — This •class of arc lamp operates on constant- potential circuits, and its regulating mag- net is operated by variations in the cur- rent. An increase of current causes the magnet to lift its armature, and thereby to lift the upper carbon. This increases the length of the arc and its resistance and rediT'Ps the current. A diminution of current permits the magnet armature to descend, the upper carbon descends with it, and the arc is shortened. This reduces the resistance of the arc and in- creases the current. The increase of cur- rent arrests the downward movement of the armature, and may cause it to rise a little. These converse actions keep the length of the arc approximately the same. The diagram, Fig. 424, shows the principle of construction of a constant-potential arc lamp of the General Electric Company. The one illustrated is an alternating-current lamp. For the pur- poses of this description the principal difference betw^een it and a direct-current lamp is the use of a reactive coil instead of a resisiance coil. The current enters by a binding post, passes through the reactance coii, the lower carbon, arc, and upper car- FiG. 424.— Diagram op Constant Potential ar.terxatino -cur rent Arc Lamp Mechanism. 556 ELECTRICIANS' HANDY BOOK, bon in the order named. It then passes through the magnet coils and out on the line. The armature is of double-plunger type, with the lower end of the plungers connected by a cross bar which carries a downwardly projecting rod at its center, which oper- ates the clutch as shown. Immediately below the clutch is a trip- ping platform. When the clutch strikes this it trips, and the carbon drops a little. Another distinction betwe-en the constant current lamp and the constant potential lamp is that the latter has only one regulating magnet, which is in series with the arc. The reactance coil is shown in horizontal diagram above the lamp. The lettered places indicate points of connection for the Fig. 425.— BEACTA^"CE Coil for Constant Potential Alter- nating-Current Arc Lamp. Fig. 426.— Resistance Coil FOR Constant Potential Direct - Current Arc Lamp. wires. The wire T, the right-hand one in the cut, is always con- nected to point A or H. The wire S, over the top of the lamp, is connected to any of the other points according to the voltage on the circuit and the frequency of the circuit. The arc voltage is taken at 70 to 73 volts. For 60 cycles and 104 volts on the line, S should be connected probably to J; for 125 cycles and 104 volts, to P. To increase arc voltage fewer coil divisions must be brought into series. Thus, changing the wire S from M to L or to K increases the voltage of the arc by cutting out part of the reactance of the coil. The direct-current lamp is of the same construction, except that it contains no reactance coil, but a resistance coil wound upon a grooved porcelain block occupies the same place. A sliding con- tact arranged in a groove shown on the side of the porcelain I THE ARC LAMP, 557 block enables the resistance to be regulated by rheostat action. Fig. 425 shows the reactance coil, and Fig. 426 the porcelain block for resistance coil. The groove on the side receives the Fig. 427.— Constant Poten- tial. Alternating-Cur- rent Arc Lamp. Fig. 428.- Constant Poten- tial! DIRECT-CC RRENT Arc Lamp. sliding contact piece used to cut out resistance as desired. The alternating-current lamp is shown in Fig. 427, and the direct- current lamp in Fig. 428. In both types of lamp the arc is liable to travel from side to side of the space between the carbons. The effect on the dis- tribution of light is quite different in the inclosed and open arc 558 ELECTRICIANS' HANDY BOOK, lamps, and is illustrated in Figs. 429 to 434. The distribution of light from the open arc lamp with central and side arc is shown in L-oi Figs. 429to431.— Distribution of Light ix Direct- Cuukekt Open-Arc Lamps. Figs. 43 i to 434.— Distribution OF LlGDT IN DiRECT-CURRJlNT Inclosed-Abc Lamps. Figs. 429 to 431. The crater in the upper carbon is so displaced by the migrations of the arc as to make a great difference in the amount of light given on the side where the arc is, compared with that given by the other side. Figs. 432 to 434 show the ef- I THE ARC LAMP. 559 feet of the migration of the arc in direct-current inclosed-arc lamps. As only a slight crater forms in the carbons of this type of lamp, the unevenness of distribution of light due to shifting of the arc is very slight. Management of Inclosed=Arc Carbons. — To get the longest life out of the carbons, the following rules should be observed. The lamp should not be run on a circuit of frequency or voltage different from that for which the lamp was adjusted. A lamp when burning should have at least 100 volts drop at the arc. The carbons in inclosed-arc lamps are separated by twice the interval which obtains in the open-arc lamps. The inclosing globe must fit perfectly. Its upper edge must * make a virtually air-tight joint with the cap. The mechanism must work freely, so as to insure correct feed. The old upper carbons can be cut to proper length if too long, and used as lower carbons. Carbons should not be used of length greater than that specified for the lamp. Adjusting Lamps— Lamps are usually sent out adjusted for the voltage or current which the purchaser has specified in his order. A variation of a quarter of an ampere above or below the rated current calls for adjustment. In the General Electric Company's lamps a weight is sometimes mounted on the working lever. This weight can be shifted so as to adjust the lamp for different currents. Moving this w^eight toward the clutch rod reduces the voltage or drop at the arc; moving it in the other direction increases it, as it acts to pull the carbons apart. The Inclosing Globes.— The directions for installing lamps is- sued by the manufacturing company sometimes specify that the lamps should not be started without the small inclosing globe being in place. This instruction should be rigorously followed. The inclosing globe is as much a part of the lamp as the carbons are. Not only the rate of consumption of carbons is reduced by the presence of the globe, but the carbon ends take a different shape. The reduction in consumption of carbons is important as an economy in supplies, and because it diminishes the labor bill for trimming. The inclosing globe is subjected to strong heat. It must not be clamped so tight as to break for lack of room to expand. The hole for the upper carbon must be a good but perfectly loose fit. The little air which works in through it 560 ELECTRICIAXS' HANDY BOOK, is rather a benefit than otherwise, as it tends to keep the lamp cleaner. Negative and Positive Connections in Inclosed= Arc Lamps.— If a direct-current arc lamp is to be installed, the upper carbon must be connected to the positive terminal of the line. If there is any doubt about the connections, the current may be turned on for a few minutes and then turned off. If properly connected, the upper carbon will be the hotter, and consequently will remain red hot longer than the lower one will. If improperly connected, the lower carbon will be the hotter. In such case reverse the con- nections. Putting a Lamp Into Service. — After unpacking a new lamp re- move the upper casing. This is sometimes secured by a bayonet joint, sometimes by screws. Sometimes wedges and packing are used for safety in shipping. Such will be seen inserted in the machinery, if present. Remove them carefully, brush out the machinery if necessary, examine it for loose parts, and see that the movable parts work freely. When all is in order, replace the casing. Sometimes lamps are shipped with the lower carbon holder and its rod removed from the lamp. If so, it must be put into place. Care must be taken in doing this to center ac- curately the lower carbon. This is effected by putting the lower carbon rod in its right position. Then perfectly straight carbons should be- used. If they do not come in line, the lower carbon holder may in some lamps be used to rectify their position by twisting. Oil. — Do not oil the dash pot or other mechanism of an arc lamp. Its parts are so exposed that lubrication is inadvisable. Clutch Stop Adjustment. — The clutch stop should be so ad- justed that with the carbon of smallest allowable diameter the upward movement of the clutch is arrested when the armature is within one-eighth to one-quarter inch of the magnet pole faces. Cut=Out. — The cut-out is adjusted to close when the stem of the clutch is about one-sixteenth inch below the tripping point. The lamp should with this adjustment cut out when the voltaro is two or three volts above the feeding voltage. Carbons for Inclosed-Arc Lamp. — For satisfactory operation of an inclosed-arc lamp, one cored and one solid carbon should THE ARC LAMP. 561 be used. In ordering, half of the order should be for solid and half for cored carbons. They may all be of the same length, say twelve inches. When the lamps are started, the lower carbons can be got by cutting 12-inch carbons into pieces. Afterward the partly-burned upper carbons will act as lower carbons. The car- bons must be smooth and of even diameter. The upper one is supposed to act almost as a stopper for the upper hole in the inclosing globe's metallic cap. Any friction at this point will interfere with the feed of the upper carbon and may put the lamp out. To Carbon a Lamp,— The following directions are given by the General Electric Company for their series inclosed-arc lamps. Be sure that the current is switched off. Hold the inclosing globe firmly and swing the bail to one side after pulling down on it. The globe will come off. Loosen the setscrew and remove the lower carbon. Remove the upper carbon, and put in a new one, inserting it in the spring carbon holder of the upper carbon tube. Put a lower carbon of proper length in the lower holder, and secure it with the thumbscrew. Replace the inclosing globe, being careful to set the upper edge squarely against the finished surface of the cap, so as to exclude the air from the arc. Secure the globe by placing the supporting ring of the bail around the projection on the bottom of the globe. To insure proper electrical connection to the upper carbon, it must be well inserted in the spring carbon holder on the inside of the carbon tube. The insertion of the carbons into the holders is facilitated by their having beveled ends. The inclosing globe should be cleaned at the station periodically, or the dirt w^hich collects on its inner sur- face will reduce the light. The above directions have to be modi- fied for lamps with inclosing globes of other type. The modifica- tions are obvious on inspection of the lamp. Lamps Without flechanism. The Jablochkoff Candle. —At one time the efforts of inventors were directed to the end of producing an arc lamp without mechanism, but all such have practically gone out of use. The Jablochkoff candle, illustrated in Fig. 435, had very extensive use at one time. It consisted of two parallel rods of carbon separated by an insulating material, such as gypsum. They were used necessarily with an alternating ctir- 562 ELECTRICIANS' HANDY BOOK. rent. A small bit of carbon was laid across the top to comic :^ the carbons. This enabled the current to start, and in a few seconds the carbon slip burned away and the arc was formed. In the cut d d are the line connections, & is a spring keeping pressure upon the socket holding the base a of the candle. Once The arc was formed, it was supposed to continue until the candle burned out. If the arc went out, it would not form again. The Wallace Lamp, an American invention, is deserving of notice, although it w^as never much used. The carbons were in the form of two rectangular plates. By regulating mechanism they were kept edge to edge within a fraction of an inch of each other. The edges were sensibly parallel to each other, but inevitably one place would mark a slightly closer approximation of the carbons. Here the arc sprang across, and as it burned, increasing the dis- tance, it shifted a little, and eventually traveled the whole length, several inches in extent, of the edges of the carbon plates. As the distance be- tween the edges increased, the uppei* plate was fed down so as to diminish it. The Sun Lamp had two inclined rods of carbon occupying a position like that of the two arms of the letter V. They descended throu::h holes in a biock of refractory material by their own weight. Open 'Air Incandescence. — One modification of the true arc lamp has disappeared from the field. Open-air incandescence was the name given to the principle on which this class of lamps oper- ated. This principle utilized the loose contact between a carbon point resting on a carbon surface as the seat of incandescence. This secured a simple gravity feed, and to a considerable extent got rid of mechanism. Gradually all these lamps died out, and at the present time arc- lamp lighting is fast settling down into the use of the inclosed- arc lamp with positive downward feed of the upper carbon. d FiG.435.— The Jablochkoff Candle. CHAPTER XXXIII. PHOTOMETRY. Standards of Illuminating Power.— The light given by a source of illumination, such as a gas flame, oil lamp, or electric lamp, has in the existing state of science to be referred to and meas- ured by some standard. The usual standard in this country is the candle. This is a sperm candle burning 120 grains per hour. Many other units of illuminating power have been proposed, and in other countries they have been adopted to a greater or less extent. A number of the more prominent are summarized below, with their relative values. The light given by a lamp is called indifferently its illumin- ating power or its candle power. The latter term does not apply to French practice, where the Carcel lamp (Bee Carcel) is the standard. Principle of the Photometer. — The principle on which the test- ing of lamps for candle-power is based is the following. The source of light is assumed to be a point. As the distance of the observer from it is increased, he receives less light. The degree of light received is dependent on the area over which its effect is spread, and like all radiations its intensity varies inversely with the square of the distance. The cut, Fig. 436, shows this clearly. The larger area has distributed over its surface the exact amount of light which lights the smaller area. One is twice as far removed from the source of light as is the other, and its area is four times as great. Therefore, a portion of the area of the distant surface equal in area to the nearer one receives one-auarter the amount of light, because it is at rionble the distance. Suppose two lisrhts are placed at a distance of 90 inches from each other, and a screen is placed at a point on the line connect^ 563 564 ELECTRICIANS' HANDY BOOK, ing them where it will receive an equal amount of light from each. Suppose that this point is 60 inches from one light and consequently 30 inches from the other. The ratio of 60 to 30 is as 2 is to 1. As the light given varies inversely with the square of the distance, it follows that the nearer light is of one-quarter the power of the distant one. Bar Photometer. — The above is the principle of the bar photo- meter, the instrument universally used for testing the candle- power of lamps, as well as of the shadow and other less used ap- paratus. Fig. 436.— Law of the Inverse Squares. Photometric Screens. — A screen is used to determine the point on the bar at which an equal amount of light is received from both sources. Several devices have been employed or suggested for this purpose. The Bunsen Disk. — The Bunsen disk is founded on the follow- ing principle. If a spot upon a sheet of paper be treated with grease, it will become more translucent and less reflective than it was before. Therefore, if seen by transmitted light, if held between the observer and a candle, for instance, it will appear lighter in color than the rest of the paper. If light is caused to shine upon it, then the spot will appear darker than the rest of the paper, because it does not reflect light so well. If such a piece of paper is held between two sources of light and receives the same amount of light from each, the spot will PHOTOMETRY, 565 tend to disappear. It may not disappear completely, but the posi- tion of greatest faintness is easily found with considerable ac- curacy. The disk i8 made of rather heavy white paper, and the spot in the center i? made by melting paraffin wax into the paper. Any kind of greasy matter will do as an expedient for temporary pur- poses. A hot bit of wire will answer for melting it into the paper. The translucent spot should be about an inch in diameter. Sometimes the spot in the center is the untouched paper, and the paraffin is melted in a ring surrounding it. The Leeson Disk.— This screen is of the simplest description also. A star is cut out of a piece of heavy note paper. It is laid between two pieces of thin note paper. In use |||Slil^^^~" \^ , the screen is moved to such a posi- ^\ ^ 1 tion that the star appears equally j ./- i bright on both sides of the disk. | ^- n, riounting the Disks.— The disk or rJ|| f screen should be three or four inches ^^^^J^ i^ in diameter. It is mounted on a block ^^fc^ -^ - ~ ^/ of wood which slides upon the bar. ^^^^^^^^^^^^^^^S^^^^B^^ The observer then looks first at one ^^''^JJJJJiilllin^ side and then at the other, shift- ^«; <37.-Th„ Bunsen Dior Mounted Between Two ing it back and forth until the spot Mirrors. nearly disappears. It is then receiving the same amount of light from both sources, and the reading on the bar gives the relative intensity of the lights. Sometimes the disk as shown in Pig. 437 is mounted between two mirrors. A B is the frame carrying the disk, M N and M' N' are the two mir- rors. This enables the observer to see both sides of the disk without moving his head. A disk mounted in this way between mirrors is often carried in a little car which runs along the bar. The center m of the disk should be on the level of the two lights which are being compared, and directly between them. The Lummer=Brodlum Screen.— In this screen the observations are made with a single eye, eliminating it is claimed any chance of error due to unequal sensibility of the two eyes. The diagram, Fig. 438, represents the horizontal plan of the apparatus. It is 566 ELECTRICIANS' HANDY BOOK. supposed to be mounted on the photometer bar. C indicates the standard lamp, X the light which is being tested. S is an opaque white screen of plaster of Paris; both sides are illuminated, one by each light, C or X. At M and N are mirrors which reflect the light to the prisms, the beams falling normally or perpendicularly upon the face of the prism receiving it. Each prism has a spherical Fig. 438.--Lummer-Brodlum Photombter Screen. face, and a circle is ground upon the center of each face, one circle being larger than the other. When placed in contact, flat side against flat side, there is a circle of contact, surrounded by the outer parts of the larger circle, which is not in contact with the other prism. Light reflected from the mirror N passes through the circle of contact to the observer's eye at the end O of the sighting tube. The circle of contact shows the degree of illu- mination of the side of the screen S facing C and lighted by it. The light from the side of the screen S, due to X, reflected from PHOTOMETRY. 567 the mirror N, goes to the double prism also. The portion of the beam which impinges on the outer flat circle is reflected to the observer's eye at O. The observer therefore sees a circle through which light from C passes, surrounded by a circle from which light from the screen S due to X is reflected. If the screen is moved back and forth upon the bar, a point will be reached when each side of the screen will receive the same intensity of light. At this point the central circle and outer circle will appear of equal brightness. The back of the outer circle is blackened, and total reflection of the light falling on it ensues. It is estimated that the mean error of setting this screen does not exceed flve per cent, and that it is four or five times as ac- curate as the ordinary Bunsen or Leeson disks. Ftg. 439.— The Bar or Bunsen Photometer. The Standard English Candle.— This, which is the American standard also, is a sperm candle burning 120 grains of sperm per hour. It is the commercial article made of a mixture of wax and sperm, and with a plaited wick. When in good condition the wick should bend over and have a red end. If it burns more than five per cent too much or too little, the readings are to be distrusted. The standard candle in hot weather is apt to burn too much sperm, and give too high a value to the lamp which is being tested. This is sometimes overcome by putting the can- dles on ice for an hour before they are used. At best, it is so very poor a standard that the w^onder is that it has so long been used. The Apparatus, — The general disposition of a photometer is shown in the cut. Fig. 439. In it are shown the divided bar. 568 ELECTRICIANS' HANDY BOOK, with an electric lamp to be tested at one end of it and the candles at the other. The box holding the disk and mirrors runs upon wheels along the bar. The apparatus is contained in a room with blackened walls. A curtain may be used to further inclose it. Calculating the Scale of the Bar. — It would be a simple matter to use a bar divided into inches and fractions of inches. Then by placing the screen at a distance where it would be equally illuminated on both sides, the distances of the two lights from it could be squared, and their inverse ratio would give the relative illuminating power as above. It would be more conve- nient to have the bar so divided as to give by its direct reading the relative value of the two lamps. This system of dividing is frequently followed. It may be done by the following process: Let 1 = value of standard light at one end of bar. Let V = value of lamp to be tested at other end of bar. Let 100" = length of bar. Let X = distance from 1 to screen. Then 100 — a? = distance from v to screen. The light-giving value varies inversely with the square of the distance; the more powerful light gives an equal illumination at a greater distance than does the weaker one. This gives the proportion and resulting equation: (100 — a:')^ 1 : V :: (100 — x)- : x- ov v : (100 — a?)' Let i; = 2; then _ x' 1 To obtain the place on the bar where this ratio holds, the square root of both members of the expression — must be ex- tracted. This gives 1^ . as the ratio in which 100 inches must 1 be divided. It may be done by proportion, thus: 2.414 : 1.414 : : 100 : ^ = 58.58 inches. Therefore, at points 58.58 inches from each end a 2 is to be marked on the bar. (100 — :r)= 3 Next let t; = 3, and we have x^ 1 PHOTOMETRY, 569 The square root of 3 is 1.73; the ratio of parts of the bar is ilZ5_ which by the proportion 2.73 : 1.73 : : 100 : a; = 63.37. gives 63.37 inches as points measured from right and left ends of the bar on which the figure 3 must be marked. This is the simplest method as regards the arithmetic of the process by which the division can be effectually effected. As exe- cuted above, the decimals are not carried out as far as they should be. It is a case in which the work should be done by logarithms, not only for the sake of expedition, but to avoid errors in the operation. The Observation. — The candles — for in modern practice two are generally used simultaneously to give an average — are lighted and allowed to burn some five or ten minutes. They are placed on a balance and weights adjusted so as to make their end of the balance beam a few grains the heavier. As they burn they get lighter, and soon overbalance. The lamp to be tested is lighted and a voltmeter and ammeter are arranged for reading. The instant the candles overbalance, the time is taken and written down. The candles are carefully placed in position at their end of the bar, and the readings are taken every half minute until ten readings have been taken. At exactly five minutes from the time noted, the candles are carefully blown out. If the ends stay red, they must be bent down with a pin until they absorb melted sperm, when they will at once expire. If the candles are not carefully blown out, grease will fly about, and the candles will lose weight. The candles are now weighed, and their percentage error is deducted or added to the average of the photometer read- ings. Suppose the candles burned 19.2 grains. This is an error of four per cent, for two candles in five minutes should burn 20 grains of sperm. The candles gave too little light as they should have burned 10 grains in five minutes. Therefore four per cent has to be subtracted from the average reading. The candle balance is often mounted at the end of the bar, so that the candles are weighed there, and never need to be moved from their position. 570 ELECTRICIANS' HANDY BOOK. Other Standards. — The French standard is the Carcel lamp, accurately defined as to its dimensions, and burning 42 grammes of colza oil per hour. Many precautions to be observed with the Carcel. lamp have been formulated by MM. Dumas and RegnaulL The German standard is a paraffin candle burning with a flame of 50 millimeters (1.98 inches) height. The Munich standard is a stearin candle consuming 10.4 grammes of stearin per hour. The Violle standard, adopted by an international conference of electricians, is the light emitted by a square centimeter (0.39- inch) of platinum at its temperature of solidification. It is not adapted for ordinary use, and it is questionable if it should ever have been adopted. The Heffner-Alteneck lamp is a simple round solid-wick lamp burning amyl acetate with a flame exactly 1.57 inches high and regulated for each reading to that height. The French have another standard, the star candle, burning 154 grains per hour. Table of Photometric Standards. — The following table gives the relative values of the more important standards of light: Violle Violle 1.000 Carcel 0.481 Star candles 0.062 German candles . . . 0.061 English candles . . . 0.054 Heffner-Alteneck . . 0.053 Shadow Photometer.— If a rod or bar is placed upright, and two lights are placed a few feet apart and a few feet back from it, they will cast two shadows upon an adjacent wall or while paper screen. The lights or one of them are moved' back and forth until the shadows are of equal intensity; then the distance of each shadow from the lamp diagonally placed with reference to it must be measured. The illuminating power of the lamps wilJ be in inverse proportion to their distance. Suppose a lamp which was being tested was 48 inches from the shadow appertaining to it, and the standard candle was 12 Star G crman English Hcffner- Carcel candles candles candies Altcneck 2.08 16.1 16.4 18.5 18.9 1.00 7.75 7.89 8.91 9.08 0.130 1.00 1.02 1.15 1.17 0.127 0.984 1.00 1.13 1.15 0.112 0.870 0.886 1.00 1.02 0.114 0.853 0.869 0.98 1.00 PHOTOMETRY. 571 inches from the other shadow. Then the illuminating powers of candle to lamp are as~48^ : 12^ or as 16 : 1. In Fig. 440, E is the lamp under trial with voltmeter V and am- meter A, It is held on an arm carried by the spring clip H. C is Fig. 440.— Shadow Photometer. the standard candle on a scale G G. R is the rod whose shadows from lamp and candle are seen side by side on the paper screen S. This principle can be applied roughly in the street or elsewhere Fig. 441.— Bouguer's Photometer. by comparing shadows thrown by two lamps, and pacing off or measuring the distances. A gas lamp can thus be compared with an arc lamp with some approach to accuracy. Bouguer's Photometer.— The cut, Fig. 441, shows another sim- 572 ELECTRICIANS' HANDY BOOK, pie apparatus. The two lights under comparison ar2 placed on opposite sides of an opaque screen, and illuminate a translucent one of paper or ground glass placed at right angles to the sep- arating one. When both halves appear equally illuminated, the distances from lights to screen are measured, and the values are calculated by the law of inverse squares. The observer is sta- tioned on the further side of the translucent screen. Foucault's Photometer. — This is a modification of the one just described. The opaque screen is moved back until the dark line or band at the junction of it with the translucent screen dis- appears. The cut, Fig. 442, shows the principle. In this way the comparison of the two divi- sions of the translucent screen is X) much facilitated. ,/ Direct Photometering of an ^.-'''' Arc Lamp is not very satisfactory, I on account of its richness in violet ^^^^ rays. The standard against which \ it is" tried gives a light of a far \^ different character. A very simple ^. and practically efficacious instru- FiG. 4^2.— FoucAULT Photo- ment for testing the relative quali- METKR. ties of arc lights is the luminom- eter. In this instrument the human eye in its every-day action of reading is made the measurer of the light. This is very logical, because the object of artificial light is to enable the eye to see, and the light may be measured by the ability of the eye to see things illuminated by the light examined. The Luminometer. — It is a box, Figs. 443 and 444, containing a card of printed matter. Two tubes open into it. One receives the light from the lamp. The observer looks into the other, and sees the card illuminated by the light under trial. The light falls on it at such an angle that light is not reflected directly into the observer's eyes. The distance at which the card can be read is called the luminometer distance. The illuminating power is determined by this distance. The test gives the practical power of the light tested. PHOTOMETRY. 573 Two features characterize this instrument. One is its porta- bility. It can be taken anywhere and used in the open street. By using cards printed in various sizes of type, it can be ac- commodated to different distances. The other feature is its direc: appeal to the eye. A light is produced to enable the human eye to see. This instrument tests the power of the light for this purpose. It is the invention of Mr. W. D'A. Ryan, of the General Electric Company. Pupillary Photometer.— The pupil of the human eye expands Figs. 443 and 444.— Luminometer. and contracts virtually under the effect of varying intensity of light. The iris, in other words, acts like a diaphragm of a photo- graphic lens, and affords a larger or smaller opening according to the light acting on the retina of the eye. The pupillary photo- meter is based on this principle. It measures the diameter of the pupil of the eye when affected by different lights. This gives a coefficient of intensity of the light. Around the edge of a disk a number of pairs of holes are made near the outer ends of the radii. The holes of the different pairs vary in distance from each other. One pair of holes are sep- b74 ELECTRICIANS' HANDY BOOK. arated by a space of 0.07 inch. These are the closest spaced. The widest spaced are 0.38 inch apart. A second disk is pivoted over the first. It has a radial opening, which exposes one pair oi' holes at a time. The light to be tested is looked at through a pair of holes. One pair after another is tried, until a pair is found whose edges seem to touch. There is a scale marked on the screen with a value for each pair of holes. It gives the diameter of the pupil which brings the two holes apparently in contact. The reading gives the relative brightness of the light, on the basis of the relative size of the pupil of the eye. The standard light is first looked at, and the holes which seem to touch are found for it. Then the light to be tested is examined, and the corresponding factor found for it also. Fig. 445.— Diffraction Photometer. Diffractive Photometer.— In testing powerful lights a concave lens is sometimes used to increase the diffraction of the rays and make it possible to use a shorter bar. The cut, Fig. 445, illus- trates the principle. The light given by the lamp E is diminished by the lens L in the inverse ratio of the squares of A and A'. Spherical Candle Power. — The electrician often takes a num- ber of photometric observations at different angles. To do this a standard or rated incandescent lamp is used as a standard. The value of this is known in candles when its voltage or am- perage are at a known value. The lamp to be tested is mounted so that it can be rotated horizontally or vertically. A number of observations are taken at angles numerous and diverse enough to represent the surface of a sphere, and the average of the observations gives the spherical candle-power. The lamp is PHOTOMETRY. 575 mounted on a support which can be rotated in all directions, and a number of observations at many angles are taken and averaged. The horizontal candle-power is averaged by rotating the lamp rapidly and photometering it while in motion. There are various methods of averaging the observations at different angles. A system employed at the Paris Expo- sition of 1881 consists in di- viding the surface of the im- aginary sphere into horizontal zones. The candle-power is determined for angles cor- responding to the center oi; each zone. These candle- powers are multiplied by the relative areas of the zones to which they respectively be- long. The sum of these products is divided by i 7t to get the mean spherical candle-power. The factor 4 3t is taken as the area of a standard sphere. The cut, Fig. 446, shows an apparatus for taking spherical candle-power of an incandescent lamp. The lamp is mounted so as to be rotated rapidly by an electric motor. This gives an aver- age illumination all around, and the candle power is determined while it rotates. It is mounted so that it can be inclined at various angles from the vertical while still rotating. Observations of candle-power are taken while it is in various positions, as indicated on the scale D. An average of the observations is taken as giving the candle-power. If the candle-power is determined at different angles in the Fig. 446.— Apparatus for Spherical CanjjLE-Power. 576 ELECTRICIANS' HANDY BOOK. horizontal plane, it is generally enough to determine one set of vertical-angle candle-powerg — the candle-powers at various angles on one meridian. The corresponding candle-powers on the re- maining meridians may be calculated from the relations of the different candle-powers on the horizontal plane. If the lamp is rotated as described, the average is given directly as far as the different horizontal angles are concerned. Candle=Powers of Incandescent Lamps.— The horizontal candle- power of an incandescent lamp is its maximum, but the ratio of horizontal to spherical varies greatly according to the shape of the filament. The table gives the mean spherical and mean horizontal intensity of several incandescent lamps. Mean Spherical Mean Horizontal Candle-Power. Candle-Power. Edison 15.49 18.83 Stanley 13.56 16.54 Woodhouse and Rawson. . . 15.09 19.11 White 12.44 15.08 Weston 16.27 17.87 The candle-power at different vertical angles varies very greartly. The tip on the top of the bulb diffracts light, and reduces the vertical candle-power at that end, while the base of the lamp reduces it to zero at the other end. It will be sufficient to give a set of candle-powers for an Edison lamp taken at vertical angles of 0°, 30°, 60°, and 90° all around the lamp. 0° gives the hori- zontal plane. 0°, 16.70; 30°, 15.02; 60°, 9.54; 90°, 3.57; 120°, 8.25; 150°, 14.96; 180°, 16.82; 210°, 14.84; 240°, 9.07; 270°, 0.00; 300°, 9.84; 330°, 15.06. The Photometry of the Arc Lamp is far from satisfactory. The carbons are never perfectly homogeneous, are almost certain to be a little out of center, and this causes the horizontal candle- powers to vary greatly. After burning a little while, carbons are apt to bend a little, which throws the ends out of line with each other. The candle-power in one direction on 'the horizontal plane may be twice or three times as great as in the other. The maxi- mum candle-power is found many degrees removed from the PHOTOMETRY, 577 horizontal. This varies far less at different meridians than does the horizontal candle-power. The variations at vertical angles are very great. A direct current arc gave the following candle-powers at different vertical angles: Above the horizontal, 60°, 48; 30°, 110. At the horizontal, 0°, 208. Below the horizontal, 10°, 401; 20°, 612; 30°, 871; 40°, 1,000; 50°, 807; 60°, 457; 70^. 188. As arc lamps are used, the mean spherical candle power is of little importance, and it is not often determined. It is a laborious operation, as the great irregularity of the distribution of light requires a large number of observations at small angular distances from each other. A short road to the result is that proposed at the Paris Electrical Exposition of 1881. The average horizontal candle-powder is divided by 2 and added to the maximum candle- power divided by 4. The sum is taken as the spherical candle- power. Thus a Brush arc lamp gave a mean horizontal candle-power of 909 candle-power; a maximum candle-power of 4651 candles; and a spherical candle-power as calculated, 1776 candles; and spherical candle-power by observation, 1675 candles. The formula reads thus: ^ H M in which S is the spherical candle-power, H is the average hori- zontal candle-power, and M is the maximum candle-power. From a number of observations it is found that the formula gives an error of 1 to 14 per cent. Mechanical Equivalent of Light.— Light is the action of cer- tain ether waves upon the retina of the eye. If light is decom- posed by means of the prism, the visible spectrum will be em- braced within relatively narrow limits. The violet end of the spectrum has its color produced by the shortest waves that affect the eye. A musician would say that violet was a very high note, or at the top of the scale. Beyond the violet there are waves which are so short that the eye does not take cognizance of them. These rays act with great energy on chemical agents such as salts 578 ELECTRICIANS' HANDY BOOK, of silver. A photograph can be taken by means of them. If separated from the other rays, they would enable a photograph to be taken in a dark room. Going to the other end of the spectrum, the red appears due to relatively long waves or high heating power. Below the red is a long stretch of spectrum quite invisible, but producing heat. By a sensitive thermometric apparatus the spectrum can be fol- lowed out a long distance below the scale of visibility. A micron is about one twenty-five-millionth of an inch or one one-millionth of a millimeter. The shortest wave length of vis- ible light is 0.360 micron for normal eyes. Dark red light has a wave length of 0.810 micron, and 1.000 micron is the utmost range of visible light. This is a range of 0.640 micron, within which all visible rays must lie. Above this range is the ray of invisible actinic radiation, ''invisible light" it is sometimes paradoxically called, due to the spectrum of radiations less than 0.185 micron long. Below the spectrum we have heat radiations, due to waves less than 30 microns long. Thus without Including Hertz waves it appears that in a range of nearly 30 microns only 0.640 micron is visible, or in decimals 0.021 of the entire scale of naturally-pro- duced ether waves. Light being a physiological effect of a natural cause can hardly be said to possess a mechanical equivalent. Yet if we determined the mechanical equivalent of the entire radiations of a given spectrum, and subtracted therefrom the proportion which was ob- scure, we would obtain a figure that might be taken as the me- chanical equivalent of light. This has been done. The total energy of the rays from a source of light was determined by an air thermometer. The air expanded under the influence of the total heat received. The luminous rays were screened out by a dark solution, such as one of iodine, and the heat imparted by the invisible rays was determined. A thermo-electric pile was employed for this. The experiment by Tumlirz is described in Wiedemann's Annalen. He found that the light given by the Heffner-Alteneck lamp, which is 0.98 standard candle, was 0.00361 gramme degree C. calorie per second, or 151.500 ergs per second. This corresponds to the energy rate of a current of 0.1226 ampere through a re- PHOTOMETRY. 579 sistance of 1 ohm. By Ohm's law E = RI. This gives a voltage of 0.1226 volt. The electric energy is 0.1226 X 0.1226 =: 0.0150 volt-ampere, or watt. The pupil of the eye covers a very small portion of the spherical area of illumination. If the eye were 1 meter (39.37 inches) from the light, and if its pupil were 3 millimeters (0.118 inch) diam- eter, the light it would receive on the above basis would require a year and 89 days to raise 1 gramme (15.403 grains) of water 1° C. or 1.8° F. If the physiological aspect of the subject is dropped, the above may be taken as of value. It gives with reasonable closeness the mechanical equivalent of rays which affect the human eye. The mechanical energy expended by a source of light may be divided by the units of light which it gives. The quotient is a practical figure expressing the relative economy of the source of light, and this figure is sometimes incorrectly called the mech- anical equivalent of light. Thus a 16-candle-power kerosene lamp was found to burn oil enough to represent 37 calories per hour per candle. This gives 428.6 meg-ergs per second, a rate of energy equal to 42. S watts. A gas burner required 68.8 watts per candle-power. An incandescent lamp is generally allowed 3.5 watts per candle-power. The arc lamp may go as low as 0.8 watt. The light of the spectrum is due to ether waves succeeding each othcx' approximately between 4 X 10^* and 7 X 10^* times per second. In a second they travel about 180,000 miles. If we divide this by the number of waves per second of any given light, we shall obtain as quotient the length of such wave. As, roughly speaking, light travels a little over 10^^ inches per second, the quo- tient of 10^^ -^ 10'* would be one 1/1000 of an inch. On the basis of 4 X 10'* waves per second, such wave would be about 1/4000 inch long. Watts per Candle=Power in Arc Light.— The watts per candle- power for direct-current arc lamps vary from 0.60 to 1.13 watts; for alternating-current arc lamps, from 1.13 to 1.80 watts. As an interesting example of the practice of some years ago, the Jabloch- koff candle may be cited. At 200 candles it used 2.80 watts per candle, and at 500 candles 1.81 watts per candle. 580 ELECTRICIANS' HANDY BOOK. Watts per Candle-Power in Incandescent Lamp.— In incan descent lamps at high efficiency 2.5 watts may be absorbed per candle-power. Lamps run at this efficiency soon break down. A low efficiency is 3.5 watts, when the light given is expensive with regard to the power absorbed. The mean figure of 3 watts to the candle-power represents good average practice. Quality of Arc Liglit.— The diagram, Fig. 447, taken from Abney, shows the proportions of the different rays of the spectrum in gas, arc, and sunlight. The curve of gaslight may be taken as practically that of the incandescent carbon-film electric lamp. To obtain a light pleasing to the eye, too much of the light of the BCD E F G H Red Violet Fig. 447.— Qualities of Ditfebent Lights. violet end of the spectrum should not be present. The sun may be taken as giving the mixture which it should be the object of the engineer to imitate in producing artificial light. The arc's light, it will be seen, approaches closely to the composition of the light of the sun. A convenient way to remember the succession of colors in the spectrum is by the combination vihgyor, indicating violet, indigo, blue, green, yellow, orange, red. Lithium chloride gives a bril- liant red light when a wire dipped into it is held in an alco- hol or Bunsen-burner flame. Copper gives a green, salt a yellow light. The rays of short wave length, such as violet, are not easily produced except when accompanied by other rays. The mixture of light of all colors gives white light. This is what is needed by mankind for illumination. PHOTOMETRY, 581 In photometering arc lamps, as we have seen, values widely differing are found at different vertical angles. These values for a given lamp, with specific carbons, current, and other factors, are reasonably constant. The horizontal angle should make no differ- ence if the lamp works perfectly. But invariably the departure from centering of the arc shifts the hottest point of the carbon to one side, so that in practice a difference may always be antici- pated. Arc lamps have received a sort of trade valuation — that of 2,000 candles. This has long been recognized as grossly inaccu- rate and in excess of the truth. The so-called 2,000-candle-power lamp is one of standard size using less than 500 watts. The present standard is 10 amperes and 48 volts, or 480 watts. From such a lamp by manipulation at the photometer 1,700 or 1,800 candle-power can be obtained as a maximum. The average maxi- mum candle-power for a direct-current lamp at a vertical angle of 45° is about 1,250 candles. The alternating-current lamp distributes its light symmetric- ally above and below the horizontal plane. The direct-current lamp distributes its light principally below the horizontal plane. There is a distinction between open-arc and inclosed-arc practice. The open-arc lamp works with its carbons much closer together than does the inclosed arc. As we have seen, about 85 per cent of the light comes from the crater in the positive carbons in direct- . current lamps. In alternating current, 95 per cent comes from the carbons. The adjustment of the carbons, if varied by the smallest amount, changes the distribution of the light. The arc is about one-eighth inch long. A small fraction of an inch makes a considerable difference in so short a distance. The inclosed arc is produced between carbons which are consid- erably farther apart. The slight changes in feed are referable to a longer distance, and hence affect the arc less in proportion than for the shorter-distanced carbons in the open-arc lamp. The carbons in the inclosed arc burn with flat ends. The arc travels about between the disk-shaped ends of the carbons. The arc in open-arc lamps also shifts about, but its movements affect the distribution of the light much more. Figs. 448 and 449 show results from photometry of open-arc and inclosed-arc lamps. 582 ELECTRICIANS' HANDY BOOK. The distances from center of carbon space to the curves give the relative values of the candle-power at different vertical angles, of the candle-power at different vertical angles. The long arc diminishes the screening effect of the lower carbon. If carbons are fed close to each other, the lower one will cut off part of the light which would otherwise reach the ground. Fig. 448.— Distribution of Light from an Arc Lamp on Pole. Distribution of Light from Arc Lamps in Service.— The illus- trations, Figs. 448 and 449, show the distribution of light in the vertical plane from arc lamps. The curve A in both diagrams gives the distribution of light from an open-arc lamp using 9.6 amperes of direct current. Of high illuminating power near the lamp, it rapidly drops off. The curve B is that corresponding to the light from an inclosed-arc lamp using 6.6 amperes of direct PHOTOMETRY. 583 current also. The distribution of light is far evener than in the case first cited. The curve C corresponds to the light from an inclosed-arc lamp using 7.5 amperes of alternating current. The diagrams are so fully marked as to be virtually self-explanatory. We are indebted for them to the General Electric Company. Distribution of Light from Injandescent Lamps.— The light given by incandescent lamps in dilterent directions varies great- ly. The single-loop filament gives the most irregular distri- bution, varying from an average for the horizontal plane of 16 candles down to 5.7 candles from the tip. The small quan- tity of light given from the tip is due largely to the glass tip or point refracting the light in all directions, which falls upon it. A lamp w^hose filament has two turns in it gives a much evener distribution from 16 candles down to 10 candles. It is not of great importance to have even distribution of light, because the lamp can be adjusted to give the most favorable as- pect to the reader or user of it, and because incandescent lamps are so often put in clusters, which tends to even matters. Fig. 449.— Distribution op Light FROM AN Arc Lamp. CHAPTER XXXIV. THE ELECTRIC RAILWAY. The Electric-Car flotor is constructed with a view to pro- tection from mud and water. It is accordingly inclosed in an iron case, and this case is used as part of the field magnet. From its interior the poles project inward, and field coils are placed on these poles. A drum armature revolves inside the case. On the end of the armature shaft a pinion is mounted. This gears into a large gear wheel on the driving axle of the car. Such is the general outline of the trolley-car motor as now constructed. Standard Voltage and Allowable Temperature. — The trolley systems have a standard voltage of 500 volts. The motor capacity is rated as horse-power, which refers to the power it can develop without getting overheated. The temperature of 167° F. (93° C) is considered a sort of standard allowable rise of temperature. Motors are often rated on the power which can be developed continuously for an hour with a rise of temperature of 167° F. ( 93° C). This rise is generally based on an atmospheric temper- ature of 45° F. (7 C.) as a starting point, thus giving the tem- perature of boiling water as the allowable temperature of a mo- tor. In practice the motor is cooled to a considerable extent by the motion through the air. It is thought that this is good for about 20° F. (11° C.) reduction from the above figures. Cause of Motor Heating. — The heating of a motor indicates core loss and copper loss. The first-named source is caused by eddy currents, and varies principally with the voltage. The Copper Loss is the heating of the wires by the current passing through them. The heating effect of a current varies with the energy rate or with the volt-amperes, or watts. 584 1 THE ELECTRIC RAILWAY. 585 We have as the formula for watts I E, and by Ohm's law E =1 RI and substituting for E this value we have Watts = I E = R I-. This states that with constant resistance the watts absorbed by a conductor vary with the square of the current, and therefore the heat developed varies with the same. The copper loss is determined from the current intensity and varies with the square of the current, and for a continuous current the practical determination is easily made by running the motor and ascertaining its heating under different loads. A thermom- eter gives the temperature. This is very simple; but when vary- ing currents are in question, the difficulty of reaching a conclu- sion as to the permissible average current is considerable. The heating effect varying with the square of the current, a mo- mentary increase of current produces far more than its direct proportion of heat. Suppose the current doubled for a few seconds. During that period it is developing four times the heat it did at the lower rate. The heat developed by an irregular current varies with the mean square of the current. The greatest allowable average cur- rent is equal to the square root of the mean square of the current. It is estimated that this quantity for ordinary street-car service will be about 35 per cent greater than the average current. Thus, if a motor could without overheating pass a steady cur- rent of 50 amperes, it could pass approximately an average cur- rent of 38 amperes under the conditions obtaining in street-car service as assumed under the above estimate. This would be a most valuable figure, were it not that it ap- plies only to an estimated condition of a particular service. Ac- curacy can only be reached by a determination of the average current for each specific case. To determine average current, the ammeter should be put in series with a single motor, where the series-parallel system is used. Where this system is in use, the current per motor is generally a good deal in excess of half the total current. The reason is that when the motors are put in series, each one takes the total current. Determining the Heating of Motors.— This may be done 586 ELECTRICIANS' HANDY BOOK. roughly by the use of thermometers on the outside of the coils. Another more satisfactory way is by determination of resistance before and after a run. The resistance of copper varies as the temperature varies, and from a table of resistance changes due to temperature changes, the heat to which the conductors are subjected can be calculated. Conditions Causing Heating. — An insulating material which is an especially poor conductor of heat and lack of ventilation of the armature cause high temperature in motor windings. Horse=Power of Car flotors, — A fair allowance of tractive power for average conditions is about 20 pounds per ton weight on a level, with an addition of 20 pounds for each per cent increase of grade. Within reasonable limits of speed these figures do not change greatly. A spring balance placed as a coupling between a motor car and a trailer would indicate on level ground a pull of about 200 pounds if the trailer weighed 10 tons. Horse-power varies with the product of force by space tra- versed per second. The horsepower with approximately con- stant tractive effort would vary approximately with the speed. If the speed of a moving car and its traction in pounds are known, the horse-power can be calculated. A horse-power is 550 feet per second multiplied by one pound, which is 1,980,000 feet per hour multiplied by one pound. (1,980,000 = 550 X 3,600; 3,600 = the seconds in one hour.) This can be put thus for a car in motion: Feet per hour X Traction in pounds Horse-power = 1,980,000 There are 5,280 feet in one mile, therefore Feet per hour = miles per hour X 5,280. Substituting this value in the last formula, we have: Miles per hour X 5,280 X Traction in Pounds Horse-power 1,980,000 Dividing both numbers by 5,280, we have: Miles per hour X Traction in pounds Horse-power = 375 Suppose a 20-ton car is going up a 2 per cent grade at 16 miles sin hour. The traction on the level grade at 20 pounds to the ton THE ELECTRIC RAILWAY, 587 Ln /' i / !/ / : 1/ i 4^ i 1 / / / / / 1 ^ O / i ! /l 1 / / z I 1 1 / 1 1 / 1 / / / 1 ' ^ 1 "-" / / 1 1 y^ \ 1 / %u. 1/! ' / i i/i 1 / / ^ . <^ ' mo O #Wr yf 1 / i / y Hr, fcTT^ Ai 1/1 1/ y ! ^ Tl / 1 / ■ /r i/d i/i y' 1 i>" ?K> / / / / Ufi i X\ y ^ 1 ^-^^ So 1 / / / / /i y >^ ^^..--^ / / / ' ^ ^f . ^ 1 CO ///^/^ iX l.^i^ L--^^ 1 o c- o o / / / / X^LX><^^n 1 1 i : ! //^/^-^! 1 1 J ^^^:^^ i i 1 1 T( ^AdrivE d FF( DRT LBS. ^ 1 ■- . - i - 1 1- |o U1 -P' o '-^ 1 o ^ O 1 o 1 ,o 1 o 1 o :o 1 8 8 8 :8 'B i 8 :<;^^---.^ ! ! 1 1 ^^^^^:N^-^1 ■~~+~--. ^>^ < /I'r ' i i i -P' \v \\]^^^^^>N ■~-<^ i^ L^f'? l^.\ 1 1 1\\\\\M ■v ^ ^ ^^ 5 i a. (\\\\\\X^ \ "^ ^ ^ M j 1 d o VK\HA^\ \ ^0 ^ K ^ ^^ . 1 i H \\\\i\ \\1 X^ .\ X "-vj ■\ "W^l 1 o X W^^ ^-^ \ \ V \ -S. 1 • 1 Tl \ \ V\\ \ V \ \ s V \ ■^s^ \ _ O ANcp ^ \ N \ \, \ \ \ 1 ^ -^ V V \ \ \ \ s \ \ 1 \ V o \ V \ \ \ \ \ \^ -^ ^ y\ \ s \ s \, \ \ \ sj 1^ o \ V \ \ \ V \ \ Ni cr. \ \ \ k \ \ i\ K 588 ELECTRICIANS' HAXDY BOOK. v.oiild be 400 pounds. Allowing 20 pounds more traction per ton per each per cent of grade, the traction on a 2 per cent grade would be 1,200 pounds. The formula is now applicable. 16 miles per hour X 1,200 pounds traction Horse-power =. 375 giving 51.2 horse-power. Traction Table. — The table. Fig. 450, gives traction data. The central column of figures gives the tractive effort. Car weights are given at the bottom on the left. The vertical line rising from any given car weight intersects the lines of grades. If from any such intersection the horizontal line is followed to the right, it will give the tractive effort to move a car of that weight up the grade in question. Thus, a 20-ton car on a 10 per cent grade will require a little over 4,300 pounds traction, or drawbar pull if it were a case of towing. On the right hand is given horse-power at given traction and speed. Thus, taking any given traction and following out the horizontal line, it intersects different speed lines. If from any intersection the vertical line is followed down to the base, it will give the horse-power. Taking 4,300 pounds traction of the last example at 10 miles an hour, the vertical line from its inter- section with the 10-mile-an-hour line leads to about 115 horse- power. Construction of Electric=Car flotor. — The general features of a standard railway motor may be thus summarized: There are four field poles projecting radially inward from the iron case, which constitutes in itself a portion of the field mag- nets corresponding to the yokes. The yoke or case is of steel casting; the projecting poles are of laminated iron or disks. These are fastened together, and are cast into the yoke. The yoke is made in halves, hinged at the side parallel to the car axle, so that the case can be opened like a box. The field is shown in Fig. 451, opened with the poles projecting as described. The field coils are wound upon molds in a lathe, and are in- sulated with mica and fuller board. Each coil is solidly made, und slips over a pole piece. Cast brass pieces bolted to the yoke /lold each coil in place. Two field coils are on poles in the upper )ialf of the case. Their terminals are soldered to insulated wire THE ELECTRIC RAILWAY, 58^ pieces several feet long, to keep them out of the way of the brush holders. The coils in the lower half of the case have me- tallic terminals. A slotted drum armature of disk or laminated structure is used. Holes are made through the assemblage of disks to secure ventila- tion, in order to keep down the temperature. A low temperature conduces not only to higher power capacity, but to efficiency and to security from injury. The winding of the coils is designed to secure ventilation. Three coils are wound together and are in- sulated in a casing, which is then placed in the slot in the arma- ture. No bending or hammering into place is needed or used. Fig. 451.— Car Motor Field Opened. Steel binding wires, themselves sunk into grooves running around the armature at right angles to the conductor grooves* hold the armature coils in place. The imbedding of these wires prevents them from cutting if the armature should become so badly disr placed as to strike the field poles. The commutator is of the regular mica-insulated type. The brush holders are fastened to the upper half of the case. The armature shaft carries a forged steel pinion. This works into a cast-steel gear wheel on the driving axle of the car. Stand- ard gear ratios are 58 to 24, 64 to 18, and 68 to 14. TlK^s^e are 590 ELECTRICIANS' HANDY BOOK. siieli that the teeth will constantly change in relation, the same teeth coming together but seldom as the gears rotate. A large air gap is allowed between field poles and armature. This, although disadvantageous from the point of view of the permeance of the magnetic circuit, minimizes the effects if the armature should get out of center. One of such effects is a strong side pull exerted by the nearest pole or poles. If the air gap is n_^ Fig. 453.— Car Motor Opened. large, a given displacement, a tenth of an inch for instance, is much less proportionately than it would be with a small air gap. If the air gap were one-tenth of an inch, such displacement might be termed 100 per cent; if the air gap were half an inch, it would be only 20 per cent on the same basis. A typical car motor with the field opened is shown in Fig. 452. Switch Boxes and Circuit Breaker.— The current from the trolley pole connection goes first to a switch placed over the plat- THE ELECTRIC RAILWAY. 591 form on the under side of the projecting roof or canopy. There is another of these switches at the other end of the car, and the two are in series with each other. The current enters by one switch, goes through it, and a conducting wire leads to the other switch, and from it the current is led to a fuse hox or mechanical circuit breaker. The two switches are called canopy switches, main motor switches, auxiliary or overhead switches. There is generally an electro-magnet in the switch box, which prevents any arc from forming when the switch is opened. The magnet repels the arc, and puts it out as a draft of air puts out a candle, although on widely different principles. It is called a blow-out magnet or magnet coil. Lightning Arresters, — After passing the circuit breaker, or else the fuse box if such is used, the lightning arrester is reached. The old lightning arrester consisted of two plates with saw teeth secured so that tooth faced tooth at a small distance. The circuit to be protected has one of its leads attached to one of the plates, and thence goes on its regular course. The other plate is grounded. If lightning enters the system, it easily breaks across the air gap and goes to earth before it reaches the controller, dy- namo or other appliances. Lightning has such high potential that ohmic resistance means little to it. But it is of oscillatory character, and a relatively slight inductance will resist its passage strongly. In the course of the circuit as it leaves the lightning arrester a choke coil is placed. This is of slight ohmic resistance, and has a negligible effect on the working current of the system. When lightning enters the circuit, this acts by its inductance to hold it back and to force it to the earth over the gap in the light- ning arrester. Another lightning arrester has two carbon terminals with their ends close together but not touching. The lightning gap is at this point. If lightning strikes the circuit, it springs across the gap and goes to the earth. A coil of wire surrounds the upper end of one of the carbons and extends some distance above it. Within the coil is an armature lying loosely in it. If the arma- ture is raised, the circuit is broken. If the main current follows the course of the lightning, it excites the coil, and the armature springs up. This breaks the circuit, and the arc is destroyed, and 592 ELECTRICIAXS' HAXDY BOOK. the armature dropping back to its place, the current goes on its regular course. Other lightning arresters are described elsewhere. Controllers. — The speed of rotation of a street-car motor and coincidently the speed of the car is regulated by giving it more or less power. The volts of potential difference which produce a current through the car connections and wiring are constant as near as may be. There are in modern practice always two or four motors in a car. For low power the voltage which acts upoa each motor or pair of motors is reduced to less than half that of the circuit. For high f)ower each motor or pair of motors is given the entire voltage of the circuit. There are several controllers in use, the Westinghouse and the General Electric Company's being very extensively employed. The general principle is the following: A vertical shaft is mounted in a case, generally placed against the dashboard of the car. The case is of sheet iron, approximately semi-cylindrical in shape, with a door wiiich opens its entire height. The shaft is square on top, and a crank handle fits on the square end. Upon the shaft are mounted a number of horizontal cams. In a typical controller there are eleven. They are insu- lated from the shaft, and are connected together in groups or pairs. The shaft is never turned through a full circle. The cams are of such shape that their w^orking or contact faces are arcs of circles, concentric with the center of the shaft. Some of the arcs are so long that their angular scope is equal to the extreme range of motion of the shaft. Thus, if the shaft moves through 200°, the largest cams would include 200° in their arc. Other cams are very short. They are distributed as regards their work- ing surfaces or contact arcs over the whole range of the angular movement of the shaft. They are secured to the shaft at even distances apart vertically. By the side of the cam shaft is a series of contact fingers. These are exactly similar one to the other, and arranged vertical- ly and spaced so that there is one finger for each cam. If the shaft is turned to the extreme right, no finger will touch a cam. If turned to the left, the fingers will make contacts. The order of the contacts and the duration of each one depends upon the arrangement of the cams and on their extent of contact surface THE ELECTRIC RAILWj±Y, 593 •r arc. If the cam surface is long enough, the finger will, once It is brought in contact with it, remain in contact for the full swing of the handle. If the cam surface is short, its finger may come in contact with it for a short period and then leave it. The construction of a controller is shown in Fig. 453. Controller Points.— On the plate which covers the top of the controller case are cast a series of short radial bars or ''points/' distributed on the arc of a circle concentric with the shaft and cams. Each point indicates a position of the handle. A hori- zontal wheel is fastened to the shaft immediately below the cover. This has rounded notches in its edge, one for each of the points. A sort of pawl drops into these notches as the wheel is rotated by the handle. The notch and pawl fix the shaft in place, and also disclose to the motorman that a point is reached. If he counts the notches, he will know where his handle is with- out looking at the points. A qualified motorman need never take his eyes off the road in front. If in doubt as to what point the handle is on, he can turn his handle clear back to the starting point and then return it, counting the notches one by one as he passes them until the desired one is reached. It is not a matter of indifference which points are used; there are pre- ferred driving points which should always be used. Driving Points. — Some points indicate a maximum of resist- ance in series with the motors. Other points indicate less re- sirjtance in series, and there are two or three points which indi- FiG. 453.— Trolley Car Coktroller. 594 ELECTRICIAXS: HANDY BOOK, cate no resistance in series. The general law for the con< centration of resistance in machines absorbing energy applies here. The points indicating no resistance are the ones on which the car should be driven. The energy is not wasted in external resistance as it is on the other points. The driving points are cast longer than the others, so as to be clearly indicated to the motorman. Series « Parallel Controller. — A large variety of controllers of this type are made, adapted for different-sized cars and mo- tors. Naturally, a high-powered car needs more regulating con- tacts than does a low-powered one. The term series-parallel indicates that the two motors on a car are operated sometimes in series and sometimes in parallel. This gives tw^o speeds. Intermediate speeds are produced by a set of changes, each one involving a definite step. There is no gradual transition, but a step-by-step progress from low to high speed. A nine-point controller controls by the following combina- tions: When the handle is turned to the first point, it brings into a series of three a resistance and the two motors, one behind the other. The current flows through the resistance, which cuts it down wastefully. Then it goes through one motor, and it is still further cut down, but here not w^astefully, and then goes through the other motor, and then to the ground. This connection gives the least energy to the motors that is possible as the connections are arranged. On moving the handle to the second point, a portion of the resistance is cut out; on moving it to the third point, fourth point, and fifth point, resistance is cut out each time, the motors remaining in series. As the system is run on constant potential, the movements described have increased the current given to the motors, and therefore have increased the power develop-ed by them, and the car under equal conditions increases its speed. At the fifth point all the resistance is cut out, and the motors are left in series. This is the first running point, as there is no wasteful resistance in series with the motors. . The rule thai resistance should be concentrated in the motor applies here. THE ELECTRIC RAILWAY. 595 The handle now swings through a transition stage m which (a) the motors again have resistance in series with them; {h) one is cut out, the other having the same resistance in series with it; c the same as &; and the sixth notch is reached. There are no notches for positions a, h, and c; the handle swings hy them to the sixth notch, at which most of the resistance is in series, and the two motors are in parallel. This gives more power. The seventh notch cuts out more resistance, the eighth still more, and at the ninth notch cuts out all the resistance and the 8 9 10 11 12 Fig. 453a.— Development on CoNTROiiLER Connections. motors are left in parallel, with the full potential and maximum current acting on them. The cut, Fig. 453a, shows the development of this controller. The cam faces are supposed to be straightened, and the successive points and the connections for each are indicated. The fingers only make contacts w^hen over the cams. Thus at point 3 and at all subsequent points finger No. 2 is cut off. It only makes con- tact at points 1 and 2. The cam faces are connected with each other, as indicated by the lines. If the description is followed with constant reference to the cut, the operation will be clear. There are other arrangements of controller. In some the con- troller throws a shunt in parallel with the motor fields, thus in- 596 ELECTRICIANS' HANDY BOOK. creasing the speed, the armature taking a still greater current. For high-power motors more points may be given, sometimes as many as thirteen, with the seventh and thirteenth as running points. Hot Resistance. — If a car is run upon the wrong point the resistance is heated, and a hot resistance indicates wasteful running. A car should be run on the driving points as much as possible, except when it is allowed to coast or drift with all power off. Blow"Out Magnet.— In the controller case is an electro-magnet whose function is to blow out arcs. As the fingers slip from cam to cam, there is constant danger that arcs will form. The electro- magnet has hinged to one pole a plate of metal, which shuts over the cam shaft and contact fingers like a door, and forms a prolongation or extension of one of its poles. On the inside face are secured a number of blocks of insulating material, cor- responding to the spaces between the successive cams. These go into the gaps, and separate each cam with its finger from its neighbor. In the cut. Pig. 453, already referred to, the hinged role piece is shown swung back, and the asbestos composition insulators are shown projecting from it. The magnetic field extinguishes arcs as fast as they form. Reverser.— To the right of the cam shaft is a reverser. This operates by reversing the relations of the field and armature con- nections. Board and Cut=Outs, — In the bottom of the case is a board, to which the wires from the motors and resistances are con- nected, as directed in the wiring plan, which the electric manu- facturing company supplies. Two knife cut-out switches are here. They have wooden handles, and are numbered 1 and 2. Each one cuts out or in its own motor, according to the number inscribed upon it. Rheostat Controller. — In this system the changes in current are brought about by changing the resistance in series with the motor or motors. Some resistance is always in series, ab- sorbing energy, except when the car is running at full spe^d in the rheostat system. This involves waste of energy. The sys- tem is out of date. THE ELECTRIC RAILWAY. 597 Motorman's Duties. — Various directions are given for running trolley cars. Several books are published devoted to the motor- man's work. Generally, more than one car is operated on a line, and it is fair to say that on all small roads where much business is done a broken-down car will be pushed to the car stable by the next car. General directions for making repairs can be given, but cars differ from one another in their electrical equipment. X.ii electrician in charge of the repairs of cars of a road will have to study their special machinery, and especially the connections used in the cars whose repairs come under his charge. The motorman will only be expected to make the simpler kind of re- pairs, and may be forbidden to do even that much. Outside of this function, the motorman has very specific duties to perform in running his car properly. It is stated by one author that by actual trial he found one competent motorman ran his car with one-half the power which an incompetent one required. Economical Running.— It is wasteful of energy to turn power on suddenly. A jerk involves waste of energy, and shakes the whole structure of the car. The power can often be shut off on slight down grades. When the track is obstructed, instead of running up to the obstacle under power and then putting on the brake, the power may be shut off a considerable distance be- fore the obstacle is reached, and a comparatively slight applica- tion of the brakes will suffice to stop or slacken the speed of the car as required. Excessive Use of the Brake is hard on the brake shoes and wheels. If the wheels are completely arrested, so that they slide on the track, it is apt to wear flat places on them. They then need grinding or turning to restore their circular contour. If a Nvagon is on the track, the car c^n be slowed by turning off the povv'er while it is still a good distance away, and the wagon may turn out while the car is still coasting. Waste of energy would result from running up to the wagon under power and suddenly turning off the power and putting on the brake at the last minute. Bad running exhausts the motorman also. The excessive use of the brake is hard work in the fullest sense of the term. Flat Wheels. — This term is applied to wheels which have had flat places worn upon them. They make a most disagreeable noise 598 ELECTRICIANS' HANDY BOOK. when the car is running, and expense is involved in grinding or turning them to shape. Sliding Wheels. — Wheels caused to slide by excessive braking do not stop a car as quickly as wheels which turn so as to con- stantly present a new surface to the rail. If held so that they cannot turn, the spot in contact wears smooth and slides along with less friction than in the other case. Skidding Wheels. — If wheels turn without moving the car, use a little sand. Turn the power off, and then slowly on again to the last notch. If wheels slide on slippery rails, when the brakes are put on, do not apply sand. First throw off the brake, start the sand, and then apply the brakes again. Reversing.— Never reverse the car until the controller handle is in the off position. The car should first be brought to a stop, the reversing lever turned, and then power should be slowly given. The trolley pole should always be shifted, except for very short distances. Leaving the Car, — If the motorman leaves the car, he should turn the controller completely off and take the handle with him, otherwise some unauthorized person may interfere, and turn on the power. Bad Ground. — If the rails are dusty, the car may refuse to start because it makes a very poor ground. Rocking the car by sway- ing and almost jumping on the platform may give a ground to a motionless car which has refused to start. The rail may be cleaned of dust a short distance from one of the wheels, and a ground can be made by touching a bar of metal or a heavy copper wire to the clean spot on the rail and to the tread of the wheel. The car will then start. Pouring water on the track may be enough to form a ground. If the ground is made with a wire or bar as described, the rail must first be touched, and th-en the wheel. The connection must be firmly held in place, or a shock will result. The motorman can shut off the power an instant to permit the bar or wire to be removed. The use of a thick glove, cap, or piece of heavy cloth for holding the connecting piece is ad- visable or imoerative. Refusinja: to Start. — If a car refuses to start with a good THE ELECTRIC RAILWAY, 599 ground, it may indicate that the rail bonds are gone. The rails can be connected electrically with a piece of wire attached in any way that seems best. Even a few nails may be driven between the ends of two rails to give some attempt at a connection. The lightning arrester may be a source of trouble. If dirt gets into it, it may establish a ground, and so short-circuit all the car connections between it and the motors. This may be of such low resistance as to melt the fuse. If cleaning the arrester is not possible, it may be disconnected, or its ground wire may be re- moved or disconnected. A lightning arrester making ground will blow out the fuse when the controller handle is in the off posi- tion. This is one way of recognizing it. Fuses. — If the fuse blows when the car starts, it may be due to so great a load that the armature turns abnormally slow, and generates insufficient counter electromotive force. If the brakes are set, or if a quantity of dirt has wedged in between brake shoe and wheel, the load on the motors may be increased thereby, so as to burn out the fuse. Another cause for a fuse blowing out is the grounding of the field coil of a motor. The cure is to cut out the motor. Short- circuiting of field or armature will do the same. Loose or bad contacts at the ends of the fuse may help to blow it out. The contact pieces at the ends of fuses should be bright and clean, as should the surfaces to which they are secured. Sandpapering or scraping may be resorted to if necessary. Screws holding fuses should be screwed down hard, and should be watched if they are liable to become loose. Never put in two fuses in place of one or a fuse heavier than the standard, as it might result in a burned-out armature or in- jury of wires or other connections from overheating. Examining Connections.— If electrical connections have to be examined, tightened up, or disconnected, either pull down the trol- ley pole and tie it down, or open the main circuit switch under the platform roof or canopy. Take no risks with a live circuit. The lamp circuit may be used to give a clew to electrical trou- bles. If the lamps light, then the current is on the line and the car has a ground connection. It may be only enou^gh for a •^mail current. While the lamps are burning, turn on the con- 600 ELECTRICIANS' HANDY BOOK. troller. If the lamps are dimmed or go out, it indicates a poor ground. There is a possibility of running the car ahead slowly and picking up a good ground again. Controller Troubles.— Sometimes the car will run with one controller and imperfectly or not at all with the other. A burn- out, a broken or loose connection, a bent contact finger, may be the trouble, or the motor cut-outs may have dropped out of their con- tacts. Broken=Down Controller. — If a controller breaks down and the cause is not obvious nor easily removed, run with the other com- mutator. The car must be run by signal from the front platform in this case, the conductor remaining in front to watch the track. Motor Troubles may be due to the causes which affect other motors, but greater in degree, because of the conditions under which railroad car motors have to operate. The carbon brushes may not play freely, the commutator may wear uneven or have a high bar, the carbon brushes may even be burnt into the holders from excessive current. There is every chance for dirt and oil to accumulate on the commutator surface. Sparking on the commutator when the motor is running may be due to one of these causes. Absolute flaming on the commutator indicates a broken, short-circuited, or wrongly-connected coil. Such troubles should be found out before the car goes into service. In the bottom of the controller case are two cut-outs, marked motor 1 and motor 2 or equivalently. If a motor is in trouble, €ut it out with its own cut-out and run carefully with one motor. Start the car very slowly and gradually under such conditions. On a steep grade it would be well not to stop at all. A short, steep grade should in such case be taken on the run. Avoid using sand. The point to be remembered with a single motor is to avoid running it slowly with the controller turned to high power. Emergency Stop. — If the brakes refuse to work, emergency methods must be resorted to. They should be avoided if possible. There are two. Throw off the power at the controller, reverse the reversing lever, and turn the controller to first or second notch. This method is quick, but is more of a strain on the machinery than the following. THE ELECTltIC RAILWAY, 601 Throw off the power at the controller, open main-circuit switch, reverse the reversing lever, and turn the controller handle to the last notch. Jerking Car. — If a car jerks or bucks, it may be due to water and mud which has reached the commutator, a bit of wire may have got into the motor case and have short-circuited the com- trolley wire TRACK RETURN CIRCUIT Fig, 454. — Feeder Coxxection for Electric Railway. mutator bars, or other short circuit may have occurred. The mo- tor in trouble may be detected by the smell due to overheated in- sulation. Cut it out at once and run with one motor. Car Heating.— Many uses have been found for electric heating, but the expense has restricted its use greatly, and its principal ap- plication is in trolley cars. For a car with twelve windows, from 2000 to 3000 watts are needed to supply the heaters, or about 4 horse-power. A car stove burns about 33 pounds of coal per day. TROLLEY IN SECTIONS -^ TRACK RETURN CIRCUIT Fig. 455.— Feeder Connection for Electric Railway. and the expense of a day's heating with allowance for repairs ta stoves, removal of ashes, and every incidental expense, was cal- culated at 19^/4 cents a day, with coal at $2 a ton. The expense of electric heating varies from 0.36 cent to 2.41 cents per hour. The showing is so favorable only because the electric heating system has comparatively few repairs. The heaters are not re- moved in summer, and there is little in the way of replacement needed. The fuel cost for a stove may be only l^o cents for a 602 ELECTRICIANS' HANDY BOOK. whole day; the principal expense is in the labor and repair items. Electric Radiators are simply resistance coils of iron wire sometimes protected by asbestos or equivalent coating. They are placed under the seats, and therefore take no room in the car; a stove sometimes takes the room of one passenger. In a crowded system the conductor may have to neglect a coal stove, and the Fig. 456.— Separate Feeder System for Electric Railway. passengers may interfere with his giving it proper attention. In such cases electric heaters are especially advantageous. Power Circuit and Feeders. — There are various ways of ar- ranging trolley-line circuits. The simplest consists of a single Fig. 457.— Interconnecting Feeder System on Electric Railway. wire with the rails as a return. Sometimes feeders are used to maintain the pressure. One of the oldest ways of using a feeder is to run it along paral- lel with the trolley wire, and connect from it at intervals to the latter. This is an imperfect svstem. Nothing is gained by it over the results which a sin2:le trolley wire of cross section equal to that of the two wires would give. A variation on this system is THE ELECTRIC RAILWAY, 603 to divide the trolley wire into sections, each corresponding in length to the distance between feed-wire connections. This makes it possible to cut out any section of the road, which might be use- ul in some cases of accident. These systems are shown in the cuts, Figs. 454 and 455. A true feeder's action would consist in keeping a definite poten- tial on a distant point of a line. The trouble is that if a feeder is drawn upon for current, its drop increases, and it fails to some extent in its function. An active feeder must act imperfectly. The following systems try to bring feeder action more into play. In Fig. 456 is chown a road supplied with current as usual through its wire, and with several feeders carried di- rectly from the station and connecting at distant parts of the line. The trolley wire may be di- vided into sections, and subsidiary feeders may be introduced. By intercon- necting feeders the system shown in Fig. 457 may be employed. Insulators. — These are of the most varied type in the great variety of electric railway work now carried out. Fig. 458 shows an insulator for carrying the trolley wire. It crosses the bottom of the insulator, and the weight is taken by a suspension wire from which the insulator is suspended by the double hook on its top. At the center of the top is a slot through which the sus- pension wire passes. Fig. 458.— Trolley Wire Insulator. CHAPTER XXXV. ELECTRICAL MEASURING INSTRUMENTS. The Galvanometer is an instrument for indicating the passage of a current. If used only as an indicator, it is more properly called a galvanoscope. Usually the first name is employed. Under Ampere's law we have seen the law of the deflection of a needle by a current illustrated. If a common pocket compass with good enough pivot and bearing is held above a conductor through which a current is passing, the needle will be deflected more or less according to the strength of the current. The fric- tional resistance may be great enough to hold the needle immov- able unless a strong current is flowing. Such a compass would constitute a galvanoscope. To increase its sensitiveness, two things can be done. The delicacy of pivoting can be increased. For extreme sensitiveness the magnetized needle can be hung at the end of a filament of silk instead of being poised on a pivot. To increase the action of the current, the conductor can be bent into circular or other closed curve, and go completely around the needle once or many times. A coil of wire of hundreds of turns may surround the needle. Proximity increases the action. The coil may be so close to the needle as to just leave it room to turn in. A galvanometer as usually constructed consists of a magnetic needle and a coil of wire surrounding it. Simple Galvanometer. — The simple form of galvanometer called originally a multiplier is shown in the cut. Fig. 459. A coil of wire wound upon a wooden or pasteboard spool or bobbin surrounds a magnetic needle. The instrument must be placed so that the coil will lie in the magnetic meridian or nearly north and south. The magnet will then lie in the coil in the position ishown. When a current is passed through the coil, the needle will 004 ELECTRICAL MEASURING INSTRUMENTS. 605 be more or less deflected. On such lines as this many galvanom- eters of widely varying sensitiveness are constructed. The needle as shown in this connection has an axis fastened to it. This may be prolonged upward through the coil, and have an index fastened upon it. Then the first movements can be seen. Otherwise they would escape notice because the needle is hidden by the coil. The needle may be above the coil, when it will move in direc- tions the reverse of those which it would have if within the coil, as shown in the cut. Astatic Galvanometer.— Sometimes two needles are used fast- ened to a central axis, with north and south poles opposed. This construction almost destroys the polarity of the two, and would Fig. 459.— Simple (JrAIiVA^fO]METER. completely were it possible to have them of equal strength. If one magnet is within the. coil, and the other with reversed pole is above or below the coil, the deflective action on both will be the same. This is because the poles are reversed. The cuts. Figs. 460 to 462, show an astatic galvanometer. On a frame F is wound a double coil of wire, whose turns lie in a verti- cal plane. The astatic needles are shown below the frame, their north and south poles being indicated by N N and S S. The whole arrangement as set up is shown on the right. A glass shade cuts off all air drafts, so as to prevent irregular movements. Fiber Suspension. — For sensitive galvanometers a thread or fiber is used to suspend the needle. It may be a fine thread of silk. It is sometimes a thread of silica. This is made by melting a piece of quartz, and diawin^: from it a fine thread on the prin- ciple of spinning g'ass. Sometimes the point of an arrow is G06 ELECTRICIANS' HANDY BOOK. touched to the melted quartz and shot from a bow, drawing out a thread of quartz which is so fine as to be almost or quite invisible. Reflecting Galvanometer. — The sensitiveness of a galvanom- eter could be increased by increasing the length of the index. The weight of the index might be a factor which would impair its sensitiveness, and it would be affected bj^ drafts of air unless in- closed. An index four or five feet long would be Impracticable. N J 3 S -S^ . 3 K. Figs. 460 to 46'^.— Astaiic G al\ aNOMETeiv. But a weightless index of any length can be provided by using a parallel beam of light. A concave mirror will reflect a beam of light, and will produce a focal image of the source of light at a distance from itself de- termined by the relation of the distance L o to the degree of con- cavity or radius of curvature of the mirror. The conditions are shown in Fig. 463. s s is the mirror, L the source of light, and Q the reflected image thrown upon a screen or scale m m. The ray Q o is the weightless index referred to above. In the reflecting galvanometer the mirror is attached to the magnetic needle indi- ELECTRICAL MEASURIXG IXSTRUMENTS. 607 cated by N S in the diagram. The center of the mirror lies in the line of the suspending fiber. The distance Q o may be as great m ^-"\ ^l / y Fig. 463.— Principle op the Reflecting Galvanometeb. as desired; the longer it is, the more sensitive will the instru- ment be. Arrangement of Reflecting Galvanometer. — The diagram, Fig. 4G4, shows a lamp in a case, with an aperture 7n m out of which a. t m i, m i 1 Fig. 464.— Arrangement op Lamp Mibror and Scale for REFLECTINa GaLVANOMETKR. a beam of its light emerges. At s is the mirror attached to a galvanometer needle. The latter with all other detail is omitted from the diagram to avoid complication. At ^ is a long scale^ 608 ELECTRICIAXS' HANDY BOOK. Fig. 465.— Arrangement of Tamp Screen and bcale for reflecting galvanometer. The light which falls upon the concave mirror at s is reflected upon the scale, and a focal image of the aperture tti m is produced upon the scale. The aperture m m may have a ver- tical wire across its center, which will ap- pear as a dark line across the spot of light upon the screen and will serve as the index. In modern practice an incandescent elec- tric lamp is often used instead of the oil lamp. In such a case the lamp is so placed with reference to the focal length of the mir- ror that its incandescent filament is projected upon the screen, and this image serves as the index. Referring again to the diagram, Fig. 463, it will be understood that if the mirror S turns left-handedly, the ob- ject reflected to and pro- duced upon the scale t will move toward the top of the page as shown by the dotted line Q o. The reverse will occur for a swinging of the mirror in the opposite direction. Fig. 466.— Telescope and Scale for Plane Mirror Reflecting Galvanometer. ELECTRICAL MEASURING INSTRUMENTS. 609 In the same cut a lens is shown in front of the lamp. By using this lens a plane mirror at S may be used instead of ihe concave one, and the dotted lines indicate the direction of the rays of light when such lens is used. The next cut. Fig. 465, shows how the oil lamp is placed below the scale and screened from the observer who is in front of it. The aperture through which the rays pass is s^en. A vertical wire is secured across it. Translucent Scale.— The diagrams have illustrated the use of an opaque scale which the observer looks at directly. Some- times a translucent scale is used, the observer being back of it and watching the index mark through it Plane Mirror Reflecting Galvanometer. — Another arrange- ment of the reflecting galvanometer depends upon simple reflec- tion of the scale in a plane mirror attached to the galvanometer needle. The cut. Fig. 466, shows a telescope mounted on a stand- ard with a scale below it. The observer looks directly at the galvanometer mirror and sees reflected in it a portion of the scale. This gives the reading of the instrument. A cross wire, cocoon flber, or equivalent may cross the mirror or be contained within the telescope, in order to give a line to fix exactly the scale division. The Thomson or Kelvin Galvanometer. — The characteristics of this instrument are the extreme lightness and small size of the moving parts, which are the needle or needlGo and a mirror, generally concave. The coil of wire is of rather large diameter as referred to the length of the needle. As used, the deflections of the needle are so small that the current is sensibly propor- tional to the deflections. It is constructed dead-beat, astatic, or according to any other requirement. In one type of instrument four magnetic needles 0.015 inch long are cemented to the mirror. The latter is 0.024 inch diam- eter, the total weight of mirror and needles being 1 grain. The object of having several needles is to get the m.aximum of mag- netization with the smallest weight. This reduces momentum and makes the combination more dead-beat. The mirror and needles are suspended by a cocoon fiber, unspun silk, and of extreme thinness. The mirror hangs in the center of a vertical ring of brass with closed back, and the front covered t)l(5 ELEUTRICIANtP HANDY BOOK. Tvjth a pane of glass. The coil surrounds the ring. A rod rises from the ring and carries a curved regulating magnet. Regulation of Sensibility. — The regulating magnet is turned with its north pole to the south. This counteracts to a certain extent the terrestrial magnetism. It is moved up and down until its action on the needles nearly deprives them of directive force. This is taken as the working position for sensitive work. An astatic galvanometer is made on the same lines. Two coils wound in opposite senses are employed, one above the other. One acts on one needle and the other on the second needle, whose poles are reversed. The coils are each divided into two parts, so that there are really four coils. The connections are arranged so that the coils can be connected in various ways. In some very accurate observations the scale is placed 20 feet from the mirror of the galvanometer. This is equal in sensitive- ness to an index 40 feet long. A small form of the Thomson or Kelvin galvanometer is shown In Fig. 467. The Ballistic Galvanometer is used to measure the quantity of electricity in an instantaneous discharge. In use the dis- charge is passed through its coils, and the extreme deflection of the galvanometer is noted. Various types can be employed. Ayrton and Perry have thus modified the Thomson galvanometer for ballistic work. Forty little magnetic needles of different lengths are, with the aid of segments of a hollow lead sphere, mounted as two spheres. The spheres are joined by a rigid rod astatically, or with the magnet poles pointing in opposite directions. The combination is sus- pended in place of the usual mirror and needles by a fiber. The galvanometer is very sensitive, and the air offers little resistance. It is corrected as follows: Call a' the first throw and a" the second throw on the same side of the zero mark. The arc a, which would have been attained by the first throw without the resistance of the air, would be expressed by the formula: . , a' — a'' a = a + . 4 The extreme limit of an oscillation is called its elongation or instantaneous deflection. ELECTRICAL MEASURING INSTRUMENTS. 611 The Depre2=D'Arsonval Galvanometer is much used for bal- listic work. The cut. Fig. 468, shows the construction of a simple form. A strong horseshoe permanent magnet is mounted on a base- board, its poles projecting directly upward. A rectangular coil of No. 40 silk-covered copper wire is the moving element. This is held symmetrically between the poles of the magnet. The magnet in the instrument we are describing is 7 inches high, and Fig. 467.— Sir William Thom- son's OR Kelvin TJeflect- iNG Galvanometer. Fig. 468.— Deprez-D'Arson- VAL Galvanometer. • Original Form. is formed of three magnets each 14 inch thick and bolted to- gether. The coil of wire is 21/2 inches long internally and l^A inches in internal width. Within the coil a hollow soft-iron :?ylinder is supported by an arm projecting from a standard at he back of the baseboard. The cylinder is a fraction of an inch smaller than the coil in all directions, so as to fit within it without :ouching it. Its sides are about 3-32 inch thick. The coil is sus- pended by a hard-drawn silver wire No. 32 or 0.008 inch diameter. k similar wire connects the center of the bottom member of the :oil to the base. The current goes through the coil, entering 612 ELECTRICIANS^ HAXDY BOOK. by one silver wire and passing out by the other. The resistance of the coil is about 150 ohms. A concave mirror is attached to the suspension hook directly above the coil, moving with it. Sometimes the wires are strained, and sometimes the lower wire Fig. 469.— Beprez-D'Arsonvaij Galvanometer. is left loose. In the latter case, as far as the action of the in- strument is concerned, it is only a conductor for the current. A modern form, as made by Leeds & Northrup of Philadelphia, is shown in Fig. 469. The coil and core are seen best in the left- hand figure, showing the suspension element removed from the magnet and with its glass front taken off. ELECTRICAL MEASURING INSTRUMENTS. C13 Ballistic neasurement. — The ballistic galvanometer is used to determine the quantity or coulombs K of electricity which pas:^ through its coils during a very short discharge. If the galvanom- eter needle moved as the discharge passed, it would receive a weaker and weaker current as the discharge approached its end. In such a case the motion of the needle or other movable element 01 the galvanometer would not tell anything, as there would be no practicable way of running up or integrating its motions. But the whole discharge may be completed before the needle begins to move. When it does move under such condition, the motion represents the sum of all the actions w^hich have been exerted upon it, whether great or small. The needle under their com- bined effect will be deflected suddenly, and the limit of its throw will depend upon the sum of these forces. If the charge passes before the needle begins to move, one ballistic condition will be present. The motion of the galvanometer indicator may be checked or (lamped by air resistance or by magnetic induction. Another con- dition for the ballistic galvanometer to fulfill is that this shall be very small. Generally, a reflecting galvanometer is employed for ballistic work. Its reflected light spot is received upon a scale four feet or more distant. Let K indicate the coulombs which produce an instantaneous throw of k° by ballistic action. Let A indicate the amperes of current which would produce the steady deflection a°. Let P be the time of vibration of the galvanometer in seconds. When lc° and a° are both small, the law of the deflection under ballistic conditions as given above is: P sin- K=: - X A X 2 ^ tan a^ The angle of deflection a° for a given current A amperes must be small. The angle Jc° must also be small to keep the light spot upon the board. Thus, if the scale board is 4 feet distant from the mirror of the galvanometer, a deflection of 2 feet corresponds to an angle of less than 30°. If the scale is divided into equal divisions, they may be read ^1314 ELECTRICIANS' HANDY BOOK. and used instead of angular deflections. Then if we part from degrees and call the two kinds of deflections k and a respectively, ^we may greatly simplify the formula and let it read thus: P A fc .-' ' This will give a result very nearly accurate. A galvanometer for ballistic work should be of slow periodicity, as P has to be determined, ana it is more accurately determined if it is a long period. A heavy moving element, whether needle or coil, lengthens the periodicity and also makes the needle slower in starting, which is a favorable condition. A usual method of working is to reflect a lighted incandescent lamp filament from the mirror, and to receive its image upon a strip of ground glass, through which the ignited filament shows. Before making a test the galvanometer must be absolutely at rest. This condition is disclosed by the image of the filament appearing motionless on the scale. It is impossible to measure with accuracy the time of a single swing of the needle or coil of the galvanometer. For this deter- mination the needle is set swinging, and the time of ten or more swings is taken. Dividing the time by the number of swings gives the periodicity P of the instrument. Ballistic Calcula^ ion.— The following example is taken from Ayrton's ''Practical Electricity": With a galvanometer, the needle of which executes eleven complete swings in 6% seconds, 1 Daniell's cell, having an E. M. F. of 1.07 volts and an internal resistance of 3 ohms, produces a constant deflection of 127 scale divisions when there is a resist- ance of 10,000 ohms in the circuit, excluding the galvanometer, v/hich has a resistance of 7,500 ohms, and which is shunted with the one one-thousandth shunt. AVhat number of coulombs is discharged through the galvanometer when an instantaneous deflection of 230 =: k scale divisions is produced? The solution is as follows: The periodicity P of the galvanom- 6 5 €ter is — or 0.59. The current in amperes A producing the de- ll E flection 127, which is a, is found by Ohm's law, I = — . ELECTRICAL MEASURIXG INSTRUMENTS. 615 The resistance of the battery is 3 ohms, that of the resistance coil is 10,000 ohms. The resistance of the galvanometer and the shunt in parallel with it is 2l!l^ or 7.560. But as the shunt in H;UU parallel with the galvanometer passes 0.999 of the current, one- one-thousandth of the current goes through the galvanometer 1 07 1.07 coils. The total curr-ent is therefore 3 + 10,000 + 7.56 ~ 10,000 approximately. As of this current acts upon the galvanom- 1000 1.07 eter, because of the shunt the current A is i a qqq qqq ampere. Returning to our formula and substituting these values, we have: 0^9 1.07 2S0_ ^"^ 7t ^ 2 X 10,000,000 ^ r^7 which gives as ansv/er 0.01822 micro-€oulomb. The answer, it will be observed, is given in micro-coulombs. This is done to avoid the six more decimal places which would be required were the answer given in coulombs. The above might have been put thus: Micro-coulombs = i^ ^ ^-^^ ^ 1^ It ^ 2 X 10 ^ 127 The Tangent Galvanometer is an instrument whose deflec* tions can be interpreted to give directly the intensity of the cur- rent which passes through them. Its construction is based on the following principle: If a magnetic needle is placed in a uni- form field of force due to a current, which field is at a right angle to the terrestrial field, it will be deflected at an angle greater or less as the strength of the field is greater or less. The law 01 the defiection will be that the tangent of the angle of deflection will be proportional to the strength of the current producing the field. The construction of the tangent galvanometer is shown in the cut. Fig. 470. A ring of large diameter stands vertically on a support. The current whose intensity is to be determined passes directly through an insulated conductor wound around the ring. For heavy currents a smgle turn of wire would be sufficient. A 616 ELECTRICIANS' HANDY BOOK. magnetized needle is supported at the center of the ring. The needle must he as short as possible. For a ring 12 inches in diameter, a needle 1 inch long may be used. As the needle will be too short to admit of a scale being used large enough to give good readings, a very light index is attached to it, which index is several inches long. A dial of corresponding diameter is under the index. The dial can be graduated in de- grees. Then a reference to a table of natural tangents will give the rela- tive value of the current intensity producing any given deflection. The dial can also be graduated so as to give tangent readings. Thus, the tang- ent of 5° is 0.08749, that of 10^ is 0.1804, that of 15° is 0.268, and so on. A direct-reading tangent scale might have the reading 8.75 correspond to its 5° point, with the interniediate ones filled in. The numbers put upon the scale would be integral ones, starting from 1 and extending on either side of the zero point. The angle of 45°, whose tangent is equal to unity, would on the above basis be marked 100. The point of maximum sensitiveness is at 45°. The tangent galvanometer must be placed in the plane of the earth's mag- netic meridian when it is to be used on the tangent pri nciple. This instrument is sometimes called the tangent compass. The 5ine Galvanometer is a galvanometer whose indications of strength of current passing vary with the sine of the angle read off its scale. It has a vertical coil with a magnetic needle in the center pivoted so as to rotate in a horizontal plane. Thus far it resembles the tangeni compass. In use the coil is turned into the plane of the magnetic meridian, as shown by the magnetic needle. The current is then turned on and the needle is deflected. FiQ. 470.— Tangent Galvan- OMtTER. ELECTRICAL MEASURING INSTRUMENTS. 617 The roil is turned in the direction of th^ deflection, following the motion of the needle, which still moves a little as the coil ap- proaches its plane of position. Eventually the coil is brought accurately into line with the needle, and the angular deflection from the original position or zero point is taken. The strength of the current is proportional to the sine of this angle. The proportions of length of needle to diameter of coil are without effect on the exactness of the sine law. The coil can be made of small relative diameter compared to that necessary in a tangent compass. This increases the sensitiveness. The sensitiveness also increases with the angle of deflection. This instrument is also called the sine compass. The Thomson or Kelvin Absolute Electrometer is based upon Fig. 471.— Sir William Thomson's Absolute Electrometfr. the attraction exercised between two electrified surfaces. An insulated metallic disk is hung from one end of a balance beam. It hangs horizontally in an opening in a larger annular metallic plate called the guard ring, which is also insulated. Sometimes it is suspended by a spring. When uncharged it hangs a little above the plane of the guard ring. Below the annular plate, a little distance from it and parallel with it, is another insulated metallic plate in electrical contact with the movable plate. The cut. Fig. 471, shows the disposition of parts. The principle is that two surfaces oppositely electrified attract each other with a force proportional to the square of the electro- motive force between them. When an instrument of this de- scription is calibrated for direct current, it can be used for alternating currents, and will indicate their effective values. To use it, the terminals whose potential difference is to be de- t)18. ELECTRICIAXS' HANDY BOOK. termined are connected, one to the lower plate and the other to the suspended plate. The force with which they attract each other is determined by weighing it, if a balance, or by deflection of the spring, if the spring construction is employed. The mov- able plate must be brought to accurat-ely lie in the plane of the guard ring. The distance between the upper and lower plates must be accurately known. The guard ring and circular plate must be of identical potential, and this is why they are in electri- cal connection through the suspension rods or spring carrying the movable plate. Let E= electromotive force between the upper and low^r plates. d = the distance between the same. F = the attraction between the plates in dynes. (1 dyne = 1 — ~ gramme) 9«1 a = the area of the movable plate in square centimeters. Then E= - 720,000^^^^ a Another method of using it is to keep the upper plate at a con- stant potential by some source of constant electromotive force, which may be a small influence machine, somewhat on the Wims- hurst type. The lower plate is alternately connected to the earth and to the terminal whose potential is to be measured. Each time the connection is made, the attraction of the disk for the lower plate is determined. The difference between the potentials of earth and terminal gives the potential of the body referred to the earth. Galvanometer Shunts. — A galvanometer may be too sensitive for some specific test. The voltage may be sufficient to pro- duce a current which would throw the light spot off the scale, or which would deflect it so nearly to 90° as to make its readings worthless. A right angle or 90° is the limit of motion of a gal- vanometer needle, and in the neighborhood of 90° its readings are very inexact. Thus a galvanometer too sensitive for the work it has to do would have its needle deflected so nearly over 90° that it would lose accuracy. If the current is split up, and only a portion is passed through the coils of the instrument, it can be reduced in sensitiveness so as to bring the readings within a good working portion of the scale. To split up the current, ELECTRICAL MEASURING INSTRUMENTS. 619 a known resistance is connected across the galvanometer term- inals, so as to be in parallel with the coils. These resistances are definite fractions of the resistance of th-e galvanometer coil. They are termed galvanometer shunts. A galvanometer shunt, such as supplied by instrument makers, is shown in Fig. 472, and the diagram. Fig. 473, shows the connection and its re- lation to the galvanometer. It will be seen that it is in parallel with the instru- ment, and passes a fraction of the total current, whose value depends on the rela- tive resistance of the galvanometer and of the shunt, as explained below. To reduce the sensibility of the gal- vanometer to 1/n of its normal value, the resistance must be equal to that of the galvanometer g divided by n — 1. Sup- pose that a shunt box is used which can re- duce the sensibility to 1/10, 1/100, and 1/1000 of the normal value. Then the 're- sistances of the three shunts are g/9, g/9d, ^/999, calling g the resistance of the .galvanometer. When a shunt is put in parallel with a galvanometer, the proportion of the total current which passes through the gal- vanometer is equal to the quotient obtain- ed by dividing the resistance of the shunt by the combined resistance of the two pieces, galvanometer and shunt. This quo- tient multiplied by the total current gives the galvanometer current. The resistance of the galvanometer and shunt is equal to the product of the resistances divided by their sum. Compensating Resistance — When a galvanometer is shunted, the decrease in resistance causes an increase of current. A re- sistance in series, called compensating resistance, is used to bring the current back to its former strength. The compensating re- FiG. 472.— Galvano- METEa Shunt. Fig. 473.— Shunted Galvanometer. 620 ELECTRICIAXS' HANDY BOOK. sistance is equal to the square of the galvanometer resistance di- vided by the sum of its resistance and the resistance of the shunt. Constant of a Galvanometer. — The French constant is the de- flection produced in a galvanometer by a Daniell's cell in a cir- cuit of total resistance of one meg-ohm or one million ohms. Tlie resistance of the battery and galvanometer are included. In England the constant of a galvanometer is the numb-er by which its indications must be multiplied to reduce them to a given unit of current. Another form of galvanometer constant applicable to tangent galvanometers is used in England. Let ni= number of turns of wire in the coil. ri= radius of coil. Then constant = -^ — 2 7t n It appears in the following formula: Let H := horizontal component of earth's magnetism in dynes. I = current intensity in C. G. S. units. S = angle of deflection of needle. n and r z=z values given above for them. Then: I = - X H tan S. 2 Tt n And this value multiplied by 10 will give the current in amperes. A still more general definition of the working constant as used in every-day practical work in this country is the following: The working constant of a galvanometer is the number of scale divisions of deflection that would be obtained by causing the cur- rant from the given battery to pass through the galvanometer and a resistance of one meg-ohm. Determination of the Constant. — Galvanometer constants as used in France, England, and America vary because it is an arbitrary working figure only. Its determination for practical use is now to be described. In the diagram, Fig. 474, G represents a galvanometer, B a bat- tery, S the galvanometer shunt, and R a known resistance. Sup- ELECTRICAL MEASURING INSTRUMENTS. 621 pose the shunt to be ^/999. This reduces the sensibility of the galvanometer to 1/1000 of its normal sensibility. Suppose the resistance R to be 100,000 ohms. When the circuit is closed, the galvanometer will be deflected. Without the shunt the deflection would be theoretically 1000 times as great, on the assumption that the deflections vary as the current. This assumption only holds true for very small deflections in reality, and is applicable in this case because in the actual test deflections within this limit are used. The shunt makes the deflection 1/1000 as great as if there were no shunt there; the resistance makes it 10 times as great as if a meg-ohm (1,000,000 ohms) were the resistance instead of 100,000 ohms. Therefore, for a meg- ohm resistance and without any shunt the resistance would be equal to that shown multi- plied by 1000 and divided by 10. The general rule for deter- FIG. 47i.-liETEaMiNATioN OF THE fining the working constant Galvanometer Constant. with the connections shown in Fig. 474 is as follows: The working constant is equal to the product of the deflection of the galvanometer multiplied by the multiplying power of the shunt, and by the meg-ohms resistance in series with it. In the case cited the multiplying power of the f7/999 shunt is 1000, the meg-ohm resistance in series is 1/10 meg-ohm. The de- flection is multiplied by 1000 X 1/10. Suppose a deflection of 250 scale divisions was given with the resistance and shunt as above. The constant would be: 250 X 1000 X 1/10 =: 25,000. Such would be the constant for a D'Arsonval galvanometer with 40 or 50 volts battery. As high a constant as 2,000,000 can be obtained for laboratory practice. A battery giving 50 volts is enough for ordinary work. By increasing this voltage the deflection is increased, and conse- quently the galvanometer constant is also increased. In deli- cate work a potential of 600 volts is sometimes used. In ordi- nary work 100 volts is not excessive. 622 ELECTRICIANS' HANDY BOOK. Figure of Merit. — This is the resistance of a coil placed in series with a galvanometer, so that a potential difference of one volt will produce a deflection of one division on the scale. Sometimes a Daniell cell (E = 1.07 volt) is taken as giving the potential for the figure of merit. Properly, the entire re- sistance of the circuit should give the figure of merit, not merely that of a resistance coil in series with the line. But in practice the resistance of the galvanometer coil so far exceeds that of i K^ S . E/ G\ Fig. 475.— Siemexs's Dynamometer. Fig. 476 —Connections of SiEMENs's Dynamometer. the rest of the circuit that it can sometimes be used directly with- out adding in the rest of the resistances. Galvanometer Resistance. — For thermo-electric work a gal- vanometer of about 1^ ohm resistance is used. Thomson's gal- vanometers have from 5,000 to 10,000 ohms resistance, with be- tween 2 and 3 miles of wire 0.004 to 0.008 inch diameter. Some galvanometers wound with German-silver wire have 50,000 ohms resistance, and a single Daniell's cell through 20 meg-ohms re- sistance will move the index through 200 divisions of the scale. Slemens's Dynamometer,— This instrument, whose construction ELECTRICAL MEASURIXG INSTRUMENTS. 623 is as simple as its theory, is the standard instrument for meas- uring alternating currents. It is shown in the cut, Fig. 475. A fixed coil of a number of turns of wire, 55 in one pattern, is mounted immovably as shown. The axis of its central opening is horizontal. A movable coil surrounds the central immovable one. The latter has comparatively few turns — often it has only one. The relation of the two coils is shown in Fig. 476, in which A, B, C, D represents the immovable coil, and E, F, G, H the mov- able one. The current enters at D or G, and passing through both coils in series, as shown, establishes fields, which tend to pull the coils into parallelism. The movable coil has its two lower terminals one above the other in line with the spring suspension S at K, and the three points are exactly in the axis of the coil, so that it is free to rotate under the smallest force. On this free- dom depends its sensitiveness. Near the base of the machine is a mercury cup, and imme- diately below it is a second one. The ends of the movable coil dip into these cups, which are vertically over each other. The current, entering by a binding post, goes through the immovable coil and then to one of the mercury cups. Passing through the movable coil, it enters the other mercury cup, which is connected to the other binding post, by which the current leaves the instru- ment. Thus the coils are connected in series with each other, and the entire current to be measured goes through each. The movable coil is kept in a vertical position by a spiral spring. The axis of this spring is vertical, it is fastened to a bracket directly over the mercury cups, and its lower end by a wire is connected to the center of the upper bend of the movable coil. Above the spring is a horizontal dial. A handle to which the spring is attached rises from the center of the dial, and an index is attached to it. By turning the handle the index can be moved over the face of the dial like a hand of a clock. The dial is graduated around its edge. A second index rises from the movable coil, passes by the edge of the dial, and is bent over across the graduated scale on the dial. Its position can thus be determined by the zero point on the scale. A plumb bob or level is used to set the instrument level, and sometimes connections are supplied, so that different numbers 624 ELECTRICIANS' HANDY BOOK. of turns of wire of the fixed coil can be thrown into the circuit. Normally, when the index on the handle is set at zero, the in- dex of the coil will also be there, the points of the indices facing each other and coinciding in angular position. This is when no* current is passing. The coils will then be at right angles to each other, and the spring will be without any torque or turning force (moment). If a current is passed, the movable coil will tend to turn and place itself parallel to the other one. By turning the handle this tendency is resisted, and the coil index is brought back to zero. This strains the spring, which now exercises torque equal* to that of the coil. The angle through which the index of the spring is turned is proportional accurately to the square of the current, whether it is an alternating or direct current. This is be- cause of the law that the action between two coils such as those of the dynamometer is equal to the product of the currents pass- ing through them. But as these coils are in series, the product of the currents is the square of the current. This instrument is as far as the coils are concerned a zero in- strument. Its indications depend, on the values of the squares of the deflections of the spring index. Hence if it is calibrated by passing a single current of known value through it, it is cali- brated for all currents within its range of action. The advantages of the instrument are several. The parts act- ing on each other occupy exactly the same relative positions when the reading is taken. Another is that it contains no per- manent magnet. The field established by such is liable to change, although as magnets are now made by makers of reputation, there is little danger of any such change. Its simplicity and approach in action to being an absolute instrument are also ad- vantages. Sometimes two stationary coils are used of different number of turns, and one or the other is used according to the current to be measured. The instrument should be set up so that the 0° diameter of the ^cale coincides with the magnetic meridian of the earth. This pre- vents it from being acted on by the earth's magnetism. Rheostats. — An early form of resistance for use in experi- mental work is the rheostat. It is still in extensive use in labor- ELECTRICAL MEASURING INSTRUMENTS, 625 atories. It consists in its most usual type of construction of a bare wire, often of iron or German silver, which is wound around a cylinder. If the cylinder is of metal — and it is often a piece of wrought iron pipe — it must be insulated from the wire. Asbestos paper is a good material for this purpose. The wire is wound around it, with the turns as close to each other as pos- sible without touching each other. One end of the wire is con- nected to the circuit. The other terminal of the circuit is con- nected to a sliding contact, which latter is mounted on a bar, so as to slide longitudinally up and down the cylinder making con- FiG. 477.— Laboratory Rheostat. tact with the wire. The farther it is placed from the end connect- ed to the circuit terminal, the more of the wire will be thrown into the circuit; and the greater this length of wire is, the greater the resistance is also. As described, the wire is brought into the circuit one turn at a time. By mounting the cylinder so as to rotate, the wire can be brought into circuit a fraction of an inch at a time. Many varieties of the rheostat have been constructed. The cut, Fig. 477, shows a very delicate rheostat for use with a potentiometer or similar apparatus. A helical line on the sur- face of the cylinder shows where a wire is secured. The small screw projecting from the center of the apparatus carries an arm, 626 ELECTRICIANS' HANDY BOOK. which has a contact point which can be brought in contact with any part of the wire. As shown in the cut, a scale is seen on the left. This reads one division for each turn of the handle. The screw rising from the center is of the same pitch as that fol- lowed by the wire on the drum. A circular scale on the upper edge gives the fractions of a turn. The position of the handle determines the point of contact with the wire. The position of this point brings an amount of the wire indicated by the reading of the two scales into the circuit. The resistance of all the wire being known, the resistance of the fraction is calculated. Resistance Coils.— A resist- ance coil is made of a length of insulated wire of known resistance. As the wire may- be of very great length, it is coiled compactly. To avoid inductance it is doubled be- fore coiling. The current goes through one half of the coil in one direction and through the other half in the other, and the two inductances counter- act each other almost perfect- ly. Insulated German-silver wire is a usual material for change of resistance by tem- FiG. 478. -Arrangement of Resist. ANCE Coils. of low. other alloys are used by the coils, as its coefficient perature variations is very different makers. Resistance Boxes. — A quantity of such coils are mounted in a single box called a resistance box. The resistance box should have the following qualities: Accuracy of adjustment, dependent on the individual coils being correct, and small sensibility to changes of temperature, dependent on the alloy of which the wires are made. The wire should be double silk-coated. The doubling of the wire and its connection to contact blocks on the top of the box is shown in Fig. 478. The wire is wound on spools or reels. Some makers use thin ELECTRICAL MEASURING INSTRUMENTS. 627 brass reels to facilitate cooling; others use ebonite or paraffined wood for the reels. The wire is liable to be heated by the passage of a current, and it is this heating which the brass reels are in- tended to dispose of. Wire is wound on each coil until the desired resistance is attained, the last corrections are applied, and it is then steeped in melted paraffin wax. The resistance of a wire is changed by bending. It is therefore necessary to test the resist- ance of the coil after it is wound. The object of the paraffin wax is to exclude moisture. Resistance Wire.— A typical composition of German silver is the following: Copper, 50 parts; zinc, 30 parts; nickel, 20 parts. All parts are parts by weight. The resistance of the wire is increased by increasing the per- centage of nickel. The wire should be well annealed to make it as soft as possible. The British Association Standard Ohm.— The alloy adopted for this standard was either German silver or an alloy of two- thirds silver and one-third platinum by weight. Arrangement of Coils. — On the top of a resistance box are seen a number of blocks of brass. To each block two terminals are connected by tapping into sleeves or into the undersurface of the block, and by soldering. One is a terminal of one coil, and the other that of its neighbor. Grooves are made in the vertical sides of the blocks facing each other. These are accurately reamed out to fit the slope of brass plugs with insulating handles. Referring to Fig. 478, it will be seen that if a plug is inserted, the coil beneath it will be short-circuited or cut out. Numbers mark- ed on the top of the box indicate the resistance of each coil below the number. The resistance of a box with the plugs out is equal to the sum of the resistances marked on its top. Frequently there is a pair of blocks, which if unplugged cut out the whole set of coils. This is often called the infinity hole or plug. The cut, Fig. 479, shows the top of a modern resistance box. The "Wheatstone bridge box is merely a special form of resistance box with the coils arranged for convenient bridge operations. The order of resistances of the coils varies according to the ideas of the makers. Siemens's Plan is one of the oldest arrangements of resistance 628 ELECTRICIANS' HANDY BOOK, coils in a resistance box. A series of coils are arranged between blocks. The successive values of the coils are 1, 2. 4, 8, 16, 32, etc., as far as desired. By plugging between all the coils, the re- sistances are all short-circuited. By removing any single plug the resistance named above it is thrown into circuit. Starting at the left, taking out the first plug throws in 1 ohm; taking out the second throws in 2 ohms, giving a total of 3; taking out the third also throws in a total of 7, and so on. By taking out any number of plugs, whether consecutively or not, the sum of the resistances marked will be thrown into the series. The Fig. 479.— Top of a Wheatstone Bridge Resistance Box. combination is very interesting, but it is obsolete, as it does not lend itself to easy decimal summation. Modern Arrangements. -The following are approved systems of resistances for 16-coil boxes. It will be seen that any number of ohms down to units can be obtained by different combinations: (a) 5000. (&) 4000. (c) 1, 1, 3, 5, 10, 10, 30, 50, 100, 100, 300, 500, 1000, 1000, 3000, 5000. The easiest way to plug in a resistance is to start with all the plugs in place. A glance at the figure indicating the ohms desired 1. 2, 2, 5, 10, 10, 20, 50, 100, 100, 200, 500, 1000, 1000, 2000, 1, 2, 3, 4, 10, 20, 30, 40, 100, 200, 300, 400, 1000, 2000 3000, ELECTRICAL MEASURING INSTRUMENTS. 629 will show what plugs to remove to make the sum of resistances thrown in equal thereto. A disadvantage is to he found in this type of arrangement. A plug is needed for every coil, and when a number of coils are cut out, a quantity of plugs equal to their number must be used. A single badly-placed plug will introduce unknown resistance. The trouble is emphasized by the fact that this last is most apt to occur when the greatest number of plugs are in place. This is when the resistance is lowest and when any additional resistance will be the largest per cent of the total. T/TA/S mMrgrgmrg /o UN/T3 mmaHBrargEiriiB w k Iwi mvi KvM mvH Fig. 480.— Decade Plan of Resistance Box. The Decade Plan is an improvement that has been recently introduced. The diagram. Fig. 480, shows one of the arrange- ments. The lower set of coils are of one ohm resistance each and con- nected in series. Each block has a plug groove in its side facing outward. A long bar of brass is mounted opposite the row of plugs, with grooves in it corresponding to those in the blocks. The next row of coils are of 10 ohms resistance each, and are arranged in series with blocks. A long brass strip is provided for them also. The connections for the circuit are marked + and — • in the diagram. In the above arrangement a single plug inserted in a hole be- tween block and bar will give any value in a decade. Thus the 630 ELECTRICIANS' HANDY BOOK, two plugs indicated by the black spots give 35 ohms resistance. No more plugs than these two are ever needed for this box. There is less danger of losing plugs, of loose contacts, and of straining the junction of the brass blocks and the hard-rubber box top. The latter trouble sometimes leads to warping the rubber, as the plugs are forced down between the blocks. The decade plan lends itself to the use of sliding contacts, either on a straight line or arranged dial fashion. This substitute for plugs is com- ing into use. The number of coils used on this system is rather large. The Leeds & Northrup Company have other combinations, of which the following, Fig. 481, is an ex- / (/) ample. A 1-ohm, 3-ohm, 3-ohm, ^AAAA/NAAAA/^-— -h and 2-ohm coil are arranged in (2jo - series as shown. The terminals sAArWVWVV\A ^""^ indicated by + and — . If 1 and 5 are connected, the resist- 3 ' ance will be zero. If 2 and 5 WWWWVV are connected, only 1 ohm will Wo ^ be left in circuit. If 4 and 1 are sAA A /v^A^A A A A^ .^ connected, 2 ohms will be left in /^ circuit. By following this out, Fig. 481.-DECADE Plan roii it will be found that every re- Resistance Box. sistance from 1 ohm to 9 ohms can be given by these four coils. By the block and plate arrangement a single plug does all the connecting for the nine values of the four coils. The above arrangement takes care of each decade with only four coils and one plug to the decade. Details in the Construction of Resistance Boxes. — The hard-rubber surface must be clean to avoid a diminution of re- sistance; therefore all parts of the rubber must be accessible for cleaning. A defect in some constructions of resistance boxes is that the surface of the rubber between the pairs of blocks can- not be conveniently got at for the removal of dust. The plugs should go down below the shoulder or top of their tapering ends to avoid the formation of ridges by wearing and friction against the edges of the contact blocks. ELECTRICAL MEASURIXG IXSTRUMEXTS. 631 Shellac for coil insulation is now preferred by the best makers to paraffin. It is put on in solution, dried and baked. Wire with a low-temperature coefficient must be used for the coils. The baking of the shellacked coils tends to equalize the winding strains and to artificially age the wire. Metal spools by their better cooling powers have the effect of reducing temperature errors and changes. Metal Spools are made by Leeds & Northrup in two parts, be- ing divided longitudinally. The halves are insulated from each other, and secured together by rings of insulating material at top and bottom. The spool is covered with silk, shellacked,' and wound with the wire. Each half of the tube is connected to its own contact block or plate, and the ends of the wire are soldered each to one-half of the spool. Thus there are no long ends of wire to be disposed of, and connecting the spool to its holding bolts or studs connects at the same operation the ends of the coil. Practical Notes. — The plugs must be perfectly clean. In con- structing tho box, the taper of the plugs must match that of the holes. Filing and rubbing with fine emery paper is some- times recommended, but such treatment should be sparingly ussd, as it will tend to spoil the shape of the plugs. Burnishing with the back of a knife is good. In inserting the plugs a slight twist should be given. Never touch the metal part of a plug with the fingers. In putting in or taking out a plug, be careful not to disturb the ones next to it. A plug from one box should not be used in another unless it has the same taper. A well-arranged bridge box will answer for the measurement of resistances from 1/100 ohm up to 1,000,000 ohms. A larger size with 10,000-ohm coils may extend oirer a range of 1/1000 ohm to 10 meg-ohms (10,000,000 ohms). Thicknesses of the wire for the different coils are given thus: Coils of 1 ohm No. 18 to 21 B.W.G. Coils of 10 ohms No. 20 to 29 B.W.G. Coils of 100 ohms No. 25 to 34 B.W.G. Coils of 1000 ohms No. 32 to 40 B.W.G. A high-class resistance box or Wheatstone bridge box can have its top lifted off and turned upside down for inspection of the 632 ELECTKICIAXiS' HANDY BOOK, coils, which are attached to the top and are lifted with it. A damaged coil can be removed and another put in its place, thus avoiding the necessity of sending the entire box to the makers. Wheatstone Bridge or Bridge Box.— This is a resistance box Avith its coils so arranged that the connections of the Wheat- stone bridge may be carried out with it. It has four binding posts for the end and galvanometer connections. It has already been shown in Fig. 479. Fig. 482.— Diagram of Wheatstone Bridge. The Wheatstone Bridge is an apparatus for determining the resistance of a conductor. If a conductor carrying a current is divided into two parallel conductors for a portion of its length, the following law will always exist: For every point on one of the parallel conductors there will always be a corresponding one on the other, between which, if they are electrically connected, no current will pass. Let the Wheatstone bridge be represented by a diamond, Fig. 482, with opposite points connected. Let the four arms of the ELECTRICAL MEASURING INSTRUMENTS. 632 bridge be designated by a, h, c, and d. If no current flows through the wire indicated by g, the proportions will hold: a : 1) : : c : d and a : c : : b : d. In a proportion if three of the quantities are known, the fourth one can always be found by the arithmetical *'rule of three." If therefore any three of the resistances are known, and if no cur- rent passes through g, the fourth or unknown resistance can be ^ calculated by the rule of three. Suppose. that an unknown resistance is to be determined. It Fig. 483. -Simple Whevtstone Bridge. is placed in the bridge connection, at d it may be; theoretically, the place is indifferent. The current goes through it, and it must constitute the entire resistance of the arm d. Known resistances are put in for a and b. Suppose they are a = 100 ohms and h = 5 ohms. Then one resistance after another is tried at c until no current passes through g. Suppose that this was 57 ohms. We then have the proportions: 100 : 5 :: 57 : a^ or 100 : 57 ,' : 5 : a; from either of which we find that X =: 2.85 ohms. This is the law of the Wheatstone bridge The apparatus is one of the most used in electrical work. To asr-evtah^ when no cur- 634 ELECTRICIANS' HANDY BOOK. rent passes through g, a sensitive gaivanoseope may be used. It need not be a galvanometer, that is to say, it need not be a meas- urer of current; it is enougli if it shows the presence of a current. It must be sensitive, as the slightest current must be shown by it if it exists. Fig. 483 gives a perspective view of a simple bridge to demon- strate the principle. Operation of the Wheatstone Bridge, — It may be operated on two principles. One of the resistances only may be changed until no current passes through the galvanometer connection, or two of the connections may be changed simultaneously, one being increased as the other is diminished. The latter method is used in a form called the meter bridge originally, but since its first use modified in various ways, so that it is no longer a meter bridge. The name meter indicated that two of the limbs, a and & for in- stance, are one meter in length when taken together, being rep- resented by a single straight wire. Another thing to be noted is that it is only necessary to know the value of one of the three resistances. If the proportional value of the other two to each other is known, it is sufficient. What is known as a Wheatstone bridge is usually a box filled with resistance coils and with connection points or binding posts representing the points of the diamond. If the cut, Fig. 479 representing the horizontal plan of a bridge, be examined, it will be seen that the points of connection of the wires from the bat- tery represent the ends of the diamond. The gaivanoseope is con- nected to points representing the top and bottom of the diamond. The loose wire running from the right-hand binding post, where the galvanometer is connected to the battery connection, is the wire whose resistance is to be measured. By putting in and taking out plugs, the relations of the resistances can be varied until the gaivanoseope reads zero. Null riethod.— One great advantage about the bridge method is that the gaivanoseope reads zero always when the resistance is determined. A calibrated instrument is not needed. It is what is called a null method. Tlie rieter Bridge has been used for the most delicate re- searches. The cut, Fig. 484, shows the connections. The character- ELECTRICAL MEASURING INSTRUMENTS. (535 istic part from which it takes its name is the wire in this in- strument stretched three times along its front. This wire repre- sents two of the arms of the bridge. A sliding piece K moves along it, and by depressing a key connects the conductor from the galvanometer to the wire. This point represents the top or bottom of the diamond. The position of the point read off on the scale gives the ratio of resistance of the two sides represented by the stretched wire. By using one or the other of the three leads of the stretched wire, or by using two or three of them Fig. 484.— Meter Bridge. simultaneously, all sorts of proportions between the parts to right and left of R can be brought about. The known and un- known resistance represent the other legs of the diamond, and the point where the other conductor from the galvanoscope is connected is the end of the diamond. The small figure shows the contact piece which is moved along the wire. Bridge Key. — In using the bridge, the current is only turned on momentarily for each trial adjustment, until the zero reading is reached. This would set the galvanoscope swinging, owing to the capacity of the elements of the bridge or of the conductor or to the capacity or Inductance of the unknown resistance. A 636 ELECTRICIANS' HANDY BOOK, key is used which makes two contacts in succession. The first connects the battery with the bridge circuits, and the second brings the galvanoscope into its circuit. Thus in the cut, Fig. 482, when A is depressed, it first makes contact with B, thus bringing the battery and the four arms of the bridge on closed circuit. This instantly cliarges all parts and expends any induct- ance. A further depression of the key brings A, B, and C in con- tact, which operates to throw the galvanoscope into its proper circuit across the diamond. Two separate keys may be connected, so as to effect the same result. Shunt to the Galvanoscope. — This is sometimes used to di- minish its sensibility for the first trials. As the work approaches its finish, the shunt key is opened, allowing the galvanoscope to operate with its full degree of sensitiveness. In the cut, Fig. 482, the shunt is indicated by s and the shunt key by k. Proportional Coils. — This term is applied to the arms of the bridge opposite to the unknown resistance and the arm in series with it. Thus in Fig. 482 if c or cl represents the unknown re- sistance, a and h are the proportional coils. Galvanoscope. — Although this term has been used, the gal- vanoscope actually employed is a galvanometer in most cases, and a highly sensitive reflecting instrument is adopted for delicate work. As a galvanoscope the telephone receiver is sometimes employed. Conditions of Sensitiveness.— The galvanometer must be sensi- tive. On inspecting the diagram, Fig. 482, on page 632, it will be seen that the battery and galvanometer can be interchanged. The one which has highest resistance should be placed so as to connect the junction of the two arms of highest resistance with the junction of the two arms of least resistance. Thus, if the resistances are a == 1 ohm. h = 100 ohms, c = 4 ohms, and d = 400 ohms, the higher resistance apparatus or appliance, whether it is battery or galvanometer, should connect the junction of a and c with that of h and cl. The galvanometer will almost al- ways have the higher resistance. With galvanometers equal in all other respects except in the thickness and length of wire winding, the resistance for greatest sensitiveness will be ex- pressed by the following expression, referring to Fig. 482: ELECTRICAL MEASURING INSTRUMENTS. 637 (a + b) (c + cl) a + b + c -\- d This is not a practical consideration, as the galvanometer cannot be changed for every new testing. Direction of Deflection.— The galvanometer will deflect one way for one change of relative resistances and the reverse way for the other change. This will hold only for the identical battery connections. It is recommended by some to mark upon the work table some indication for these deflections. Then by noting whether it is to left or right, the operator will know whether to increase or diminish the given resistance. Ordinarily, it will be Fig. 485.— Principle of the Potentiometer. one resistance that will be varied. The others will be* plugged in and left untouched, perhaps for a number of tests. Tlie Potentiometer is an apparatus for measurement of resist- ances, current strengths, and potential differences. It has ac- quired in late years most extensive application. Modern electric measurement practice tends or should tend in the direction of null methods. The potentiometer uses one of these. A reflecting galvanometer may be and generally is used with the potentiom- eter. Its function is simply as a galvanoscope, just as in the Wheatstone bridge method. When it shows no potential differ- ence, the reading of the resistance coils gives the result of the experiment. Principle of the Potentiometer.— In Fig. 485 W is a battery 638 ELECTRICIANS' HAXDy BOOK. giving a constant current, R is an adjustable resistance, A B is a resistance divided into 150,000 parts, and by movable contacts M M' different lengths of it may be thrown into parallel with the circuit containing the galvanometer G and at E a battery not shown in place, because various cells are used there, E indi- cating the binding posts for connecting them. For general re- quirements the drop between M and M' must be at least 1.5 volts under the action of the main battery W, which is not a standard one. jVWWWWIAMAAAA/VWiAW ^AWVW iw Fig. 486.— Potentiometer Connections* The standard cell is introduced at E, and the points E and E' are so set that the number of divisions of A B included between them represents the voltage of the standard cell. Suppose this voltage to be 1.434, then M and M' should include 1,434,000 di- visions of A B between them. The resistance at R is now varied until the galvanometer G shows a zero reading. For the stand- ard cell there is substituted a cell whose electromotive force is to be determined. The distance between M and M' is adjusted until the galvanometer again reads zero. The direct reading of the divisions gives the voltage of the cell. This merely gives the principle. In Fig. 486 is shown one of the developments. The galvanometer is seen at the bottom of ELECTRICAL MEASURING INSTRUMENTS. 63D the diagram. A double-throw switch throws it into circuit with the standard battery S or the battery to be tested, connected at E. The drop against which the standard cell S is balanced is the fixed resistance R s. By varying the resistance R, the zero read- ing of the galvanometer is secured with the connections shown in the diagram. The double-pole switch is then thrown to the right and M M' adjusted until a zero reading is obtained. The divisions of A B included between M and M' give directly the volt- age of the cell at E. R s is chosen of such resistance as to secure this relation. Fig. 487.— Potentiometer Connections. Another development is shown in Fig. 487, which approaches more closely than the last to the conditions of Fig. 485. The standard cell connects at fixed points on A B distant a number of divisions expressing as before its voltage. The cell to be tested is connected by a right-hand movement of the switch, and its voltage is determined as before for M M'. High= Voltage Determinations with the Potentiometer.— If the voltage to be determined exceeds that of the standard cell con- siderably, resistance is put in series with the cell to be tested. Connections to E are taken from known divisions of the resistance. Thus, suppose a 30-volt battery were to be measured, which is 640 ELECTRICIANS' HANDY BOOK. twenty times the capacity of the instrument. The battery would be connected through a resistance which might be 1,000 ohms. Taps from portions of the 1000-ohm resistance, including 50 ohms between them, would be connected to B. The reading be- tween M and M' multiplied by 20 would give the voltage of the battery. In Fig. 487a the source of electromotive force which is to be measured is connected at E. M. F. By means of the switch N different portions of the resistance Q Q' can be connected to the potentiometer at P. For a high electromotive force the fraction E Q' of the resistance could be connected, giving a fraction of O E.M.F. O ^ C D E O Fig. 487a.— High-Yoltage Connections for Potentiometeti. the electromotive force expressed by the quotient of the entire resistance divided by the resistance E Q'. For less electro- motive forces the switch N is swung so as to connect with D or with C. Current Measurement with the Potentiometer. — The current is passed through a standard low resistance. The drop between its ends is determined by connecting branches from its ends at E. Knowing the drop and the resistance in which such drop occurs, E the current is calculated by Ohm's law I = — . R To determine resistance by the potentiometer, the conductor under trial is put in series with a known resistance, and a bat- tery is connected in the circuit. The potential difference between the ends of the known resistance fs determined, which depends on ELECTRICAL MEASURING INSTRUMENTS. 641 the current strength. The potential difference between the ends of the conductor under trial is next determined. As the current strength is supposed to be the same as before, the resistance of the conductor is determined by the relative drop. Should there be any apprehension that the current strength has — £[ Fig. 488.~Resistance Determination by Potentiometer. changed between the two determinations, it is only necessary to make new determinations, and if there is only a slight dif- ference the average may be taken. In order to secure a virtually constant current put good resistance in series with the battery and two working resistances. The connection is shown in Fig. 488. CHAPTER XXXVI. ELECTRICAL ENGINEERING MEASUREMENTS. Voltmeter fleasurement of Resistance, — The following is a quick method of measurement with simple appliances. A knoy»^n resistance and voltmeter are all that are needed. Referring to the diagram, Fig. 489, D E is a source of current supposed to be constant during the time of the experiment, r is a known resistance, and R is an unknown resistance, which is to be determined. The voltmeter V is first placed across the ter- minals of one resistance, say of r, as shown, and its deflection giving the drop or voltage is noted. It is then connected across the other resistance — in this case it would be the unknown one R — and its deflection also noted. Suppose that for R the deflection is E, and for r is e. We then have E : e : : R : ^, or R = ^^ ^ e Voltmeter and Ammeter Determination of Resistance.— Sup- pose that we have a voltmeter and an ammeter, but no known resistance. Then we place the ammeter in circuit with the un- known resistance, and then connect the voltmeter across the ter- minals of the unknown resistance. The diagram, Fig. 490, shows the connections. E is the source of current, A is the ammeter. V is the voltmeter, and R is the unknown resistance. By Ohm's E law we have R = The ammeter reading gives I, the voltmeter reading gives E; the quotient of E divided by I gives the resist- ance of the conductor R in the diagram. Low -Resistance Measurements. — With a milli-voltmeter low resistances can be measured by the above method. The diagram, Fig. 491, shows it applied to measuring the resistance of the ar- mature of a dynamo or motor. The terminals from the circuit containing the source of cur- 642 ELECTRICAL ENGINEERING MEASUREMENTS. 643 rent E and ammeter A are connected to opposite bars of the com- mutator through the brushes, or directly by being pushed under the brushes between them and the commutator bars. With the milli-voltmeter m V connected as shown, the resistance of the armature can be measured, using the formula given in the last example. Hlgh=Resistance Measurements. — For high-resistance measure- ments the plain voltmeter may be used. Its resistance must be known, and figures as the known resistance r. The diagram. rA/vwwvs — mm-' V R Fig. 489.~'Vot.tmeter I)ETT:RMrisrA Fig. 490.— Voltmeter and Ammeter TiON OF Resistance. Determination of Resistance. Fig. 492, shows the arrangement of the apparatus. E is the source of current, R is the unknown resistance, V is the volt- meter of the resistance r, and K is a switch. The voltage E through r is given when the switch is closed; the voltage E' through R + ^ in series when the switch is open. E is greater than E'. The unknown resistance R is given by the formula E — E' R = r E' Line Insulation Tests. — The above method as applied to an active high-potential line to determine its insulation, is shown in 644 ELECTRICIANS' HANDY BOOK. Figs. 493 and 494. It is used to detect a ground. The voltmeter is connected with one terminal grounded first to one lead and Fig. 491.— Determination of Low Resistance with Voltmeter AND Ammeter. Fig. 492.— Voltmeter Measure- ment OF High Kesist- ANCE. then to the other line. If the line is dead, a battery can be put in circuit with the voltmeter. Let E be the difference in potential between the lines, and E' the difference in potential between one =^-=^ 3^ '^r^ 1 G ' / Figs. 493 and 494.— Line Insulation Tests. . r line and the ground. Then calling R the line insulation and the known resistance of the voltmeter, we have: E — E' ^ = r — _ — E' Rail -Joint Test.— The resistance of rail joints is an impor- ELECTRICAL ENGINEERI:SjG MEASUREMENTS. 645 tant factor in electric railroad practice. With a sensitive gal- yanoscope, such as a milli-voltmeter, reflecting galvanometer, or telephone receiver, it can be determined thus: Wires are secured to the rail just over the ends of the bond, or within a short distance of the joint and inside the limits of the bond. They are connected to the galvanoscope as shown in Fig. 495. The rail joint represents one arm of a Wheatstone bridge, the wire A B represents another arm, and the cross con- nection is made up of the galvanoscope and the two wires con- nected with it. From the point B another wire is taken, and is moved along the rail on the other side of the joint until nc Fig. 493.— Rail. Joint Test. sign of current can be discerned in the galvanoscope. By the principle of the Wheatstone bridge the resistance of the joint is equal to that of the portion C D of the rail. Twelve inches is a usual distance for the space including the joint. The current passing in the rail from the operation of the road is the current which is used in the experiment. It may be necessary to protect the galvanoscope by resistances. If so, they must be placed as shown. A couple of keys as indicated are convenient also. The CD resistance of the joint is expressed as AD This gives the length of rail equal in resistance to a joint. rieasurement of Insulation Leakage. — An insulated electric 64a ELECTRICIANS' HANDY BOOK. cable represents a condenser or Leyden jar. The insulation is the dielectric, or a part of it. If the cable is an aerial one, the air is also part. If the cable is sheathed, the metal sheathing repre- sents the outer coating. Otherwise the earth or moisture on the cable may be representative of the outer coating. The inclosed conductor represents the inner coating. It can be charged just like any condenser. A definite quantity of coulombs or micro- coulombs of electricity at a definite potential can be charged upon the metal surface of its conductor if both ends are disconnected from everything, leaving the conductor insulated. If so charged and left to itself, the charge will slowly leak out, owing to im- perfect insulation. The resistance of the insulation determines the time. The value of the insulation in megohms, R meg., per mile is given by the following formula. In it C is the capacity in microfarads per mile of the cable, E the potential of charge ac the beginning of a certain number of seconds T, and e the poten- tial at the expiration of that period. Then the resistance is given by „ 26.06 R meg. = Clog^ e Insulation leakage varies with this resistance, whieli iB called the insulation resistance. Insulation Resistance of a Metal -Sheathed Cable. — This is the resistance between the wires and the sheath in a given length of cable. If it contains a quantity of wires, they are bunched at one end for the test, or for special purposes the wires may be tested individually. The diagram, Fig. 496, gives the theory of the connections. One wire is connected to the sheathing o»f the cable, one is connected to the bunched end of the wires; a special switch V is in parallel with the galvanometer G and shunt S. Thus the connections are the same as in a Wheatstone bridge, with the exception that the switch has been introduced, and that the insulation between wires and sheath in the cable takes the place of the unknown resistance of the bridge. A battery B and known resistances O and R complete the system. Before closing the circuit by connecting the wire to the cable ELECTRICAL ENGINEERING MEASUREMENTS, 647 sheath or bunch of wires, the switch K is closed. Then the con- nections are completed, and the battery charges the cable. This sudden rush of current does not affect the galvanometer, as it goes almost entirely through the switch. The switch is now opened, and the galvanometer deflection, after it has stood a few minutes, is noted. To keep the deflections within proper limits, the shunt S may have to be adjusted. The battery must give the voltage used in determining the constant. If the same shunt has been used as was employed in deter- mining the constant, the galvanometer constant divided by the Bi^ Fig. 496.~lNStJLA.TioN Resistance of a Metax-Sheathed Cable. deflection gives the meg-ohms resistance of the insulation of the cable tested. If a different shunt has been employed, multiply the result by the multiplying value of the original shunt and divide by the multiplying value of the shunt used in the test. The first deflection of the galvanometer is not noted. An extra quantity of electricity flows into the cable at first. After stand- ing a minute the reading is generally assumed to be correct. The slow absorption of electricity is termed electrification. Suppose that with a galvanometer shunt of fir/999, the constant of 25,000 was determined, and that using the same constant and of course the same voltage on a cable insulation, the galvanometer deflection was too small to be accurately read. Suppose the gal- 648 ELECTRICIANS' HANDY BOOK. vanometer shunt had to be changed to g/99 and that a deflection of 55 resulted. The insulation resistance would be ^^^ \. iOOO_ 55 100 "" 454.6 meg-ohms. What is wanted is often the meg-ohms of insulation resistance per mile. In such case the meg-ohms found are multiplied by 5,280 and divided by the length in feet of the piece of cable tested. Telephone cables show from 1,500 to 2,500 meg-ohms per mile. Determination of Capacity of a Cable.— When the capacity 1 ^•l-lil^ B Fig. 497.— Capacity Test. Fig. 498.— Determination of Capacity of a Cable. of a cable is to be determined, the ballistic galvanometer is used, and a standard condenser. A galvanometer shunt will ordinarily be needed unless the standard condenser and line to be tested have capacities rather close together. The cut, Fig. 497, shows the determination of the throw due to a standard capacity, which for telephone cables would be about 1/10 microfarad. Seven or eight cells of battery may be used. C is the standard condenser, B the battery, G the ballistic galvanom- eter. On depressing the key the condenser is charged; 15 or 20 seconds may be allowed for this. Then the key is suddenly re- leased, and it springs upward and connects the galvanometer to the terminals of the condenser. The latter discharges its charge ELECTRICAL ENGINEERING MEASUREMENTS. 649 through the galvanometer, and the deflection, which is an in- stantaneous throw only, is noted. Next for the terminals of the condenser are substituted connections to the wire in a cable and to the outer metallic sheathing of the same. The distant ends of the wires are disconnected from the sheath and from the ground. The operations are repeated, and the throw of the galvanometer is again noted. The throw due to the discharge of the cable is divided by that due to the discharge of the condenser, and the quotient is multiplied by the capacity of the condenser. The re- sult is the capacity of the cable. The connections for a capacity-testing apparatus are shown in the cut. Fig. 498. G is the galvanometer, S its shunt, C the standard condenser, V a double-pole switch, K K discharging switches, and B the battery. The switches K K are depressed on their lower connections. This brings the battery in circuit with the condenser. The latter should be arranged so that various capacities can be ob- tained from it. The battery now is allowed to charge the con- denser for 15 or 20 seconds, when the keys K K are released and the throw of the galvanometer is noted. The double-pole switch V is now thrown to the right. This substitutes the cable to be tested for the condenser. The switches K K are again depressed for 15 or 20 seconds and suddenly released, and the throw of the galvanometer again noted. If the shunt had to be used, the multiplying power used for the cable is divided by that used for the condenser, and the figure obtained as described on page 648 is multiplied by this quotient to get the capacity of the cable. The capacity of a cable affects its use in telephony; the greater its capacty, the more poorly will it work. An interesting appli- cation of the test is involved in the determination of a break in the conductor of a cable. It is only applicable in cases where there is high insulation resistance — a meg-ohm at least. The capacity of the broken wire or conductor is determined first from one end of the cable, and then from the other end. It is perfectly evident that the capacities of the two parts will vary as their lengths. A simple proportion will give the lengths. Thus, call X the distance to the break from one end, and K the 650 ELECTRICIANS' HANDY BOOK. capacity of this part, K' the capacity of the other part, and a — x the distance to the "break from the other end. Then we have the proportion: K : K' :: X : a — x K' X =^ K a — K X (K' + K) x — Ka K a -" -K -h K If there is a good wire in the cable, this can be used to avoid the necessity of carrying the instruments from end to end of the line. Deflections are obtained, d for the near section of broken wire, d' for the good wire, and d" for the good wire connected to the distant broken section, which last gives the sum of the deflections due to the good wire's and distant section's capacities. Hence the capacity of the two sections of broken wire is ci" — d' + d. As before let x be the length of the near section and a the total length of the wire. Then the deflections being propor- tional to the capacities, and consequently to the lengths of the wires, we have: d" — d' -\- d : d : : a : X a d ~~ c?" — d' + d If there is low insulation resistance, this test is inapplicable. If the capacity of the cable per mile or other unit of length is known, a determination of the capacity of the near section gives the requisite datum in combination with what is known to calcu- late the location of the break. Thus, call the capacity of a mile of submarine cable K, and that of the broken section K'. It is XT evident that the length of this section is equal to -. As before, K' the deflections may give it directly if those due to capacity K are known. As K' might be due to many miles or only a few, the galvanometer shunt might have to be used, and possibly different galvanometers for extreme cases. This would introduce simple multiplication factors or divisors into the calculation. Galvanoscope Cable and Line Tests. — An uncalibrated gal- vanometer with three or four cells of battery is useful for testing ELECTRICAL ENGINEERING MEASUREMENTS. 651 lines for grounds, crosses, or breaks. One terminal of the battery is grounded, the other end is connected to the end of the line to be tested, which must be on open circuit at the n-ear end. The galvanoscope will be apt to move at the instant the connection is made. Suppose that the distant end of the line is on open circuit also. A permanent deflection of the galvanoscope will indicate a ground. Suppose there is no deflection. Then as the line has shown no ground, the distant end is to be next connected to earth. If the galvanoscope shows a permanent d-eflection, the line is con- tinuous and without ground. But if the galvanoscope shows no deflection on the second test, the line is broken somewhere, and the part beyond the break may be full of grounds. The break prevented them showing on the first test. If on the first test the galvanometer gives a strong throw of the needle, followed by a return to zero, it goes to prove that the line is continuous, as this throw is due to the capacity of the line, and the capacity is greater as the line is longer. But there is nothing accurate about this test. If the observer knows the line and knows the instrument, he may draw a useful conclusion from the observation of the first throw of the needle. Such conclu- sions are on a par with those which an observer draws from the loudness of the ring of his magneto bell. With a galvanometer and battery, all the wires in a cable can be rapidly tested for crosses, by connecting across from one to another seriatim or in succession. Thus one wire can be con- nected through the galvanometer and battery to all the others bunched. The distant ends of the wires in the cable are supposed to be disconnected or on open circuit. If no permanent deflection is shown, one wire is shown to be all right, and without cross connection with any of the others. This wire is bent aside, tagged or marked if desired, and the end of another wire pulled out of the bunched ends, and is tested against the rest of the wires. This time there is one less wire in the bunch than before. If the wire is without cross, it is put aside and another tested. Eventu- ally, only a pair will be left to be tried, one against the other, if no crosses have been found. A cable full of wires can be rapidly gone through. If a ground is found, it will be between the single wire and one or more of those left in the bunch. By 652 ELECTRICIANS' HANDY BOOK. testing one wire after another out of the bunch against the crossed wire, the fault can be located as far as the specific wires are con- cerned. The two or more wires which are crossed can be found and tagged or marked, so as to exclude them from the working wires of the cable. Frequently the testing is done to cables while rolled up on the wooden reels on which they are transported. The cable on a reel is tested and found perfect, and is then drawn into the duct in the conduit. Wires found defective should be tagged with the indicatlQn of their defect, whether grounded, crossed, or broken. Fig. 499.— Cable Testing on Reel for Breaks in Wires. Cables are supposed to be subjected to severe tests by the manu- facturers, so that new cables are often assumed to be perfect, and no test is applied by the purchasing company. There are two possibilities. One is that the cable has been injured in trans- portation; the other is that it may be injured in being drawn into the duct. Tests of Cable on Reels.— When cables are still on reels, both ends are accessible, and they can be conveniently tested. The cut. Fig. 499, shows the connection for finding breaks of contin- uity in individual wires of a cable. The bunched ends of the wires at one end are connected to a battery and galvanoscope. The other end of the wire of the circuit is touched to the other ends of the cable wires, one by one. These ends are opened so as ELECTRICAL ENGINEERING MEASUREMENTS. 653 not to touch each other. As shown in the cut, a bell is used as galvanoscope. A ringing will indicate that there is no break in the wire which is touched. The test for crosses should also be made. Finding Wire Ends in a Cable. — It will be seen that similar tests can be applied to picking out the two ends of a wire in a cable. The distant end of the wire is grounded. The near end of one wire after another is connected through the battery and galvanometer to the ground, until the galvanometer shows the existence of a current. Making Branch Connection in a Cable. — Sometimes it is de- sired to take a branch line from an intermediate point in a cable. The above test can be applied to pick out a wire from the cable whose sheathing has to be opened for the purpose of making the connection. The distant end of a wire is grounded. The wires are loosened or opened. Connection of one wire after another is made with the galvanometer and battery which are grounded. The connection is made by means of a pointed w^ire or needle point, which forms the terminal of the wire from the battery and galvanoscope. This is thrust through the insulation of one w^ire after another in the opened part of the cable, until the indi- cations of a current on the galvanoscope show that the right wire has been found. The Telephone as a Galvanoscope. — The telephone, whose diaphragm gives a sharp click upon making or breaking an active circuit of which it forms a part, is an exceedingly sensitive indi- cator of current. It can be substituted in many cases for a galvanometer, being connected in series with three or four cells of dry battery. Portable sets are made for this purpose, com- prising a pocket battery and small telephone. Where much testing is to be done, a strap or spring should be used to hold the tele- phone against the ear of the observer. Otherwise, where perhaps a hundred wires in a cable have to be tested, the w^ork of holding the telephone by hand will become quite laborious. The tele- phone and battery represent the combination of hand magneto and bell. The test is made by touching and separating the eDd of a wire in the circuit to the telephone terminal, or by otherwise suddenly making and breaking the current. Due regard must 654 ELECTRICIANS' HANDY BOOK, be had to capacity of the wire. A click in the telephone will be produced by this alone if it is at all considerable. The following method of using the telephone test for crosses and grounds in a cable is given in Roebling's pamphlet on tele- phone cables. It may be applied most conveniently to cable on the reel, as both ends are then accessible from the observer's posi- tion. At the near end of the cable the wires are spread a little, and the particular wire under test has a short piece of wire con- nected to it. The rest of the wires are bunched, and by a short piece of bare wire are connected to the sheath of the cable. The arrangement and connections are shown in Fig. 500. Fig. 500.— Cable Testing on Reel for Short Circuits. By a short piece of wire one terminal of the battery is con- nected to the cable sheath. The other terminal of the battery is connected with the telephone. The observer holds in his hand the end of the wire, which is connected to the wire to be tested. He suddenly taps with it one of the binding posts of the tele- phone. This gives it a charge of electricity, and if it is not crossed or grounded, which means connected to the lead sheath of the cable, a click will be heard in the telephone. This tells nothing. But if the wire is in good condition as regards crossing and grounding, it will hold its charge, and on a sec- ond tap being given the telephone will give no sound or a great- ly diminished one, and a third tap will be almost sure to produce no sound whatever. But if the wire is crossed or grounded, a ELECTRICAL ENGINEERING MEASUREMENTS, 655 closed circuit will result from each tap. The circuit will include the telephone, battery, sheath, and wire, and perhaps another wire or wires if there is a cross; consequently, in such a case every tap will give a click on the telephone. If no click is given on the first tap of all, it will indicate that the wire is broken off, probably close to the observer. Even moisture in the cable will impair the insulation enough to give the indications described. The loudness of the click will give some clew to the extent or degree of the trouble. It will be seen that the test for continuity of the wire is a part of the test for crossing and grounding. The telephone can be BAD WIRE X ^ GOOD WIRE Q ->ea FAULT Fig. 501.— Varley's Loop Test. used for it. A continuous clicking, as the telephone terminal is tapped, represents a permanent deflection of the needle of the gal- vanoscope. The Vibrating Magneto Bell as a Qalvanometer. — A vibrating bell can be used in tests where a galvanometer is applicable. Three or four cells in series with it give the current. It is very convenient for continuity tests. Some of the tests just given are not quantitative — which means that no measurement of current, resistance, or other function is executed. An idea of the degree of trouble with a cable can be obtained from the indications of the instruments, but that is all. Experience will teach the observer to place the right amount of dependence, rather little than much, upon differences of degree which have been alluded to. 656 ELECTRICIANS' HANDY BOOK, Varley Loop Test. — This is an application of the Wheatstone bridge. Suppose that there is a bad wire in a line, as shown in Fig. 501. Its distant end is connected to the end of a good wire, and the two are connected into a Wheatstone bridge, as shown. The battery is grounded as shown; the points i and e represent the ends of the bridge, R + X is the resistance of one arm of the bridge; C + Y is that of the other. We have as the equation of the bridge: A R4-X B ~ C + Y The entire resistance of the tv/o wires is equal to X + Y + C. Calling this resistance L, we have: L = X + C + YandL. — X=C + Y. Substituting L — X for C + Y in the first equation, we have: A=^^jt^andX= AL-BR B L— X A + B If the resistance A is eaual to B, we have: X- 2— If L is known, X can be determined; L is found by calculation from the size and length of the wires or from records. If there is only one ground, it can be measured uy bridge connection, the battery terminal being taken from the ground and connected between R and X. Hand flagneto Tests.—The hand magneto is a bipolar gener- ator with shuttle or H-section armature. The latter is rotated by multiplying gear, and the current is taken off by two contact rings. It is made in stock size, and is in itself a rough measuring in- strument. A bell is mounted in the box, whose armature is polarized and is acted on by an electro-magnet. The winding of the magnet is in circuit with the armature winding of the magneto. On turning the handle, the bell will ring if its circuit is closed. The diagram, Fig. 502, shows the connections. The hand magneto is much used in testing insulation. One terminal may be connected to a line to be tested, and the other terminal to a water or gas pipe to give a good ground. On turning the handle, the bell will ring if the insulation of the line ELECTRICAL ENGINEERING MEASUREMENTS. 657 is defective within the limits of the sensitiveness of the instru- ment. The magneto in question with its bell is generally so wound that the bell will ring through 20,000 to 25,000 ohms. By practice and use of the same magneto, the operator using it can roughly approximate to the seriousness of a ground, or to the resistance of the circuit rung through, if he notes the loudness of the bell and the clearness of its ring. It is dangerous to rely too much on such indications. Hand Magneto Test for Ground — If a line or cable is to be tested for grounding, the circuit is opened at both ends. The Fig. 502.— Hand Magneto and Bell, for Testing. practical point is to be sure that the far ends of the line wires are on open circuit. The near end is opened, connected to the magneto, and the other end of the magneto circuit is grounded. If the bell rings on turning the handle, the assumption is that there is a ground. But simple as this test seems, it is not reliable. The alternat- ing current produced by the magneto m-ay, by charging and dis- charging an ungrounded line of some capacity, ring the bell and so lead to false conclusions. This test is of great use where the line is of slight capacity, which is the case with a bare wire on poles. But cables with metal sheathings represent a sort of Leyden jar, and may ring the bell when there is no ground upon them. 658 ELECTRICIANS' HANDY BOOK. Hand Magneto Test for Cross Connections.— To test for a cross connection in a cable, one terminal of the magneto is con- nected to the wire to be tested, and the other to the ends of the remaining wires, which are bunched for the purpose. If the bell can be rung, a cross is present. The test can be applied if de- sired to a single pair of wires, so as to go through the cable wire by wire instead of bunching all except the one. Engineering Tests. — The distinction between engineering and laboratory tests and measurements is definite. Much apparatus is used in laboratory work which it would be quite impossible to employ in outdoor work in the streets of a city. Any tendency to the refinement of what may be called street tests is accomp- anied by a corresponding tendency to apply the finer processes of the laboratory to more and more of the every-day problems which confront the engineer. It is therefore a fair conclusion that the distinction between the two classes will always exist. The object of the engineer should be to use the finest class of measurements in his work, and to constantly appeal to the lab- oratory for final data. CHAPTER XXXVII. ELECTROPLATING. Electroplating:. — The decomposition of a metallic salt by the electric current in such a way that the metal is deposited where desired constitutes electroplating. To deposit metal, a current must pass through the solution by electrolytic conduction. This gives the general case without reference to electrons or to the ionic theory. The latter only affects the theory of the case. Energy Absorbed in Electroplating. — The rate of expenditure of electric energy is expressible in volt-amperes or watts. To deposit metal some current must pass through the solution, be- cause each coulomb precipitates a given and invariable amount of each metal. Until some current passes, no metal will be de- posited. If the current passing through a given bath be multi- plied by the voltage required to force the current through the bath, the rate of energy will be given in watts or volt-amperes. The voltage may be derived from an outside source, such as a battery or dynamo, or part or all of it may be derived from the bath and its electrodes. A Daniell's cell is an example of a plating bath, which deposits copper upon a surface of copper, the reaction between the electrodes and solution producing all the required voltage. If copper electrodes are immersed in a bath of copper sulphate and a current is passed through the solution, all the voltage is derived fromi an external source. In other cases the solution and electrodes may generate part of the voltage only, the rest being supplied from an external source. General Principles. — The general principle is easiest fixed on the mind by reference to the primary battery. If from the terminals of such a battery wires are led to a bath filled with a plating solution, and if the ends of the wires are attached to ob- jects of metal adapted for the purpose, and if the metal objects 659 t)60 ELECTRICIANS' HANDY BOOK. are immersed in the bath, electro-deposition of the metal of the bath will take place, and the object connected to the zinc plate of the battery by the wire will have metal deposited upon it. The one attached to the copper, carbon, or platinum plate of the battery will have no metal deposited on it, and in many cases will be dissolved in the bath, and gradually disappear. Anodes.— The plate on which no metal is deposited is called the anode. Thus, for nickel-plating nickel anodes are a regular article of commerce. They are dissolved in the nickel bath in the course of the plating operation. For each ounce of nickel deposited, an ounce should be dissolved. There are other terms, such as cathode, for the plate on which metal is deposited, which have never come into general use. Reproduction.— Electroplating is used for two purposes. One is to reproduce objects. To do this, a mold is taken from the object. This mold may be of wax, papier mache, fusible metal, or any substance which can be made to give a reversed reproduc- tion of the object. A thick layer of metal may be directly electro- plated on the object. This layer peeled or stripped from the original gives a reversed reproduction. On such the metal is deposited, which on removal obviously gives the direct unre- versed reproduction of the original object. The other purpose is to coat one metal with another, as spoons and other table ware are coated with silver. Current for Electroplating.— A source of heavy current and of low voltage is required for electroplating. If a battery is used, it is a low-resistance battery. The amperes required are generally found by determining the area to be plated, and allowing a definite amperage to each square inch or other unitary area of the articles. A solution is contained in a vessel, which is called the bath. The objects to be plated are immersed in it, and opposite to them are the anodes. The wire from the zinc pole of the battery, if such is used, or from the corresponding pole of the dynamo, is connected to the objects. The other wire is connected to the anodes. As current passes, the metal is deposited. The voltage varies for different solutions. From one to ten volts is a good range. It must be noted that a high voltage does no harm as long as the current is of proper strength, but the voltage must ELECTROPLA TING. 661 be high enough to produce the requisite current and to decom- pose the solution. Regulation of Current.— The strength of current is regulated by adjusting the resistance in the circuit. A simple resistance frame, such as is shown in Fig. 503, is often used for this pur- pose. As the handle is swung in the direction indicated by the arrow, it cuts in less resistance. Simple Plating Apparatus. — Electroplating on the small scale is often done by the amateur with apparatus on the lines of the Daniell battery. Fig. 504. The object to be electroplated or reproduced takes the place of the copper electrode, and is attached by a wire to the zinc electrode. If the object is of a metal electro-negative to zinc or is coated with plumbago, copper will be deposited on it. Fig. 505 shows a circle of porous cups in a circular tank containing zinc plates, all con- nected by a circle of wire. A metal cross rests on the circle, and carries at its center the object to be plated. The large tank contains copper sulphate; the porous cups, water with a little salt. Large Plating Apparatus. — The illustration. Fig. 506, shows a bath for electroplating, around whose upper edge two frames of metal run. The outer frame is a little higher than the other. Long rods rest upon the outer frame, and the anodes are sus- pended from them, and short wires rest on the inner frame. The objects to be plated are suspended from these wires. One frame is connected to one pole, the other to the other pole of the battery or other source of current. The next cut. Fig. 507, shows a plat- ing bath A and battery D . . . . There are two main wires or bus-bars, a & and c d. One has the anodes K K connected to it by transverse wires mm; the other has the objects connected Fig. 503.— Electroplater's Resistance Frame. 662 ELECTRICIANS' HANDY BOOK. to it. One battery pole is connected to one bus-bar, the other to the other. Insulation is applied to the wires where required to prevent short circuits. Metals Deposited — Copper, nickel, silver, and gold are the metals generally deposited. Copper-Plating.— The bath, for objects not attacked by sulphuric acid or copper sulphate, may be a solution of copper sulphate with one-tenth of its volume of sulphuric acid. It should have a density of 1.197, and is used cold. If the bath contains too much copper sulphate, this will form crystals on the surface of the '=:^:^„iifr=Smtii^^ J^^^rTT^^>ii.^ ^ /^^^P^-i?f^ r^^^^^x /i^^U^^fji^-j — ^v^t ^ ^^»^ ^ >\^^^^ 1' ^■'^'- ^ ^^r--^^: ^T^WE" - — u ^? ^^^^ig zzz "^ ^^^^/ 1 1 F^ :e 1 1 l^B ^^ r^ Hm -^ L^^'^m^m^ Fig. 504.— Daniell's Battery Plating Apparatus. Fig. 505.— Large Daniell's Battery Plating Apparatus. anode. Such crystals, perhaps invisible, will prevent the passage of current. This bath is of limited application, as it cannot be used for iron or zinc. It is applicable to wax molds, such as are used in electrotyping. For depositing copper on zinc and similar metals, the following baths are applicable: Copper sulphate, 2 pounds; water, 1 gallon. Add ammonia until the precipitate first formed is just redissolved. This colors the solution blue. . Then add potassium cyanide until the blue color disappears. This bath should be used at a tem- perature of 122° F. to 131° F. (50° C. to 55° C). If zinc is to be plated, the piece is first dipped into a mixture of 4.5 per cent sulphuric acid, and then after washing into a solution ELECTROPLATING. 663 of caustfc soda or of sodium carbonate. It is then ready for plating. EiQ. 506.— Electroplater's Bath. Electroplating such metals as iron or zinc is only to be recom- mended for special purposes. On any water getting at the zinc or iron, galvanic action commences and the metal is attacked. Fig. 507.— Electroplating Apparatus. In Paris copper-plating has been applied to lamp-posts for the streets, they being first varnished, or coated with oil mixed with copper powder. It is not a perfect success. 664 ELECTRICIANS' HANDY BOOK. One case in which iron may be copper-plated with advantage is when the metal is to be silver- or gold-plated, and a preliminary copper-plating is often recomm.ended as a preparation for nickel- plating on iron. Copper and potassium tartrate and copper and ammonium oxa- late are bases of the formulas. Before copper-plating iron, it should be dipped in dilute sul- phuric acid, and then after washing into an alkaline solution, as prescribed for zinc. Nickel-plating.— The following are formulas for nickel-plating baths with sulphates as the base: Ammonium and nickel sulphate. . . 4 parts 1 part Distilled water 100 parts 10 parts Ammonium carbonate (about) .... 3 parts The double sulphate as above is a salt very much used in nickel-plating. The chloride may also be a basis for nickel-plating as in the following formula: Nickel chloride . .■ 298 parts Water 2250 parts Dissolve and add Ammonium chloride 70 parts Water, enough to make 10,000 parts Edward Weston recommends the addition to nickel-plating baths of boric acid— 2 parts of boric acid to 5 parts of nickel chloride or 1 part of boric acid to 3 parts of nickel sulphate. Too much alkali in a nickel bath gives a yellow deposit; too much acid gives a non-adherent coat. The bath must be per- fectly neutral. The bath should have a specific gravity of 1.041 to 1.056. If it is weaker, the bath works slowly; if stronger than specific gravity 1.070, salts crystallize on the anodes. The bath must be constantly watched for changes in its specific gravity. The pieces to be plated must first be polished, the last polishing being given with powdered lime. The pieces are cleaned of grease with a 10 per cent solution of caustic potash. They are sometimes scrubbed with a brush in a mixture of warm water, Spanish white, and sodium carbonate. Sometimes benzine is used to remove grease. ! 1 ELECTROPLATING. 665 If copper is to be nickel-plated, it is first dipped into a 10 per cent solution of nitric acid, and after washing is dipped into a solution of 5 parts of potassium cyanide in 100 parts of water. For iron the first acid dipping bath is a 1 per cent solution of sulphuric acid. The pieces are rubbed with powdered pumice stone, and then dipped in a 20 per cent solution of hydrochloric acid. Iron objects must at once be put into the bath after treatment; otherwise they will rust. A thin plating with copper may precede the nickel-plating. Zinc can be nickel-plated by receiving first a good coating of copper, or it may be amalgamated. The latter tends to make it almost as brittle as glass. On removal from the bath nickel-plated objects are first washed in cold then in hot water, and are dried in wood sawdust. They are polished by regular processes. Nickel anodes must be chemically pure; they are suspended by nickel wires. Their surface area must be a great deal larger than that of the objects to be plated, because the solutions dissolve nickel with difficulty. It is a great object to have the anode dis- solve exactly as fast as the metal is deposited. If this occurs, the solution remains of unvarying strength. The voltage to be used varies. It may start as high as 5 volts, and is to be reduced when the piece appears white, and may eventually run down to 1 volt. The evolution of hydrogen must be kept down as much as possible, although there is always more or less of it. Change of relative position of the anodes and pieces to be plated is often advisable, to prevent the deposition concentrating itself on salient parts. Silver=Plating.— Baths for silver-plating are generally made of potassium-silver cyanide> Pure silver nitrate is the starting point for this preparation. The following are examples of silver- ing solutions: A solution of silver nitrate in water is precipitated by addition of lime water, the silver oxide appearing as a brown powder. The precipitate is washed with care, and is kept in vessels full of water. To prepare a bath for plating, some of the brown oxide is dissolved in solution of potassium cyanide in distilled water. A solution of silver nitrate may be precipitated by solution of €66 ELECTRICIANS' HANDY BOOK. potassium carbonate, or of sodium chloride (salt), and treated as above. 332 parts of silver nitrate are precipitated by hydrocyanic acid. The acid must be made imrbediately before use by adding nitric acid to potassium cyanide in quantity just sufficient to neutralize it. The precipitate is washed and put into 10,000 parts of water, and dissolved by addition of potassium cyanide. One or two thousandths of ammonia added to a bath improves the adherent power and brilliancy of the deposit. A very small quantity of carbon disulphide is sometimes added for the purpose of securing a bright deposit. To obtain an even deposit, the pieces in the bath must be moved about. This is sometimes done mechanically. The anodes are of pure silver, and their surface should be about equal to that of the pieces to be silvered. Iron or lead wire is used to hang them by. Copper wire must not be used, as it would dissolve and in- jure the solution by introducing copper into it. At least 4 inches space must be between the anode and the pieces to be plated. On commencing, a current of 42 to 43 amperes per square yard of surface to be plated is required. A potential of not over 2 to 3 volts is also prescribed, although it is to be remembered that as long as the voltage is sufficient, the amperage is the critical thing in electroplating. After a quarter of an hour in the bath the pieces are taken out and examined to see if they are acquiring a uniform coating. They are then washed in a warm solution of potassium cyanide and replaced in the bath, and left there until the plating is thick enough for the require- ments. Four hours should complete the operation if there is enough current. Every 3600 coulombs or each ampere-hour de- posits 62.4 grains of silver. When the deposition is completed, the pieces are removed from the bath, washed with clean water, and then with water slightly acidulated with sulphuric acid. They are finally brushed and Ijolished by the regular processes. Preparation for Silvering. — Preparation of articles to be silver- plated begins with the removal of grease by boiling for a few seconds in a ten per cent solution of caustic potash. This is fol- lowed by washing in water and then dipping in a ten per cent ( ELECTROPLATING, 667 solution of sulphuric acid and water and washing. Next they are passed through a bath composed of Nitric acid (36°) 100 parts Salt (sodium chloride) 2 parts Calcined lampblack 2 parts After a few seconds they are washed vigorously, and then are passed at once through this bath: Nitric acid (36° ) 600 parts Sulphuric acfd (66°) 80 parts Salt (sodium chloride) 4 parts Again they are vigorously washed and placed in the "quick- ing" bath until they appear white on the surface. This bath is made up of: Water 100 parts Mercuric nitrate 1 part With enough sulphuric acid to dissolve the mercuric nitrate. The pieces are then washed and put into the plating bath. QoId=Plating.— Gold-potassium cyanide is used for the bath. 154 parts of gold chloride are dissolved in 2000 parts of water. A separate solution of 200 parts of potassium cyanide in 8000 parts of water is made. The two solutions are mixed and boiled for half an hour. This bath is employed at the ordinary temperatures. To keep up its strength, gold chloride and potassium cyanide may be added in equal parts as needed. The anode is a plate of gold. A bath too rich in gold gives a blackish or reddish coating. A gray coating slowly formed indicates too much potassium cyanide. Platinum suspension wires are employed for the anode. The anode should not be left in the bath except during the plating. For gilding with a warm solution the following baths may be used: 1. 2. Sodium phosphate (crystallized). 600 parts. 500 parts. Sodium bisulphite 100 parts. 125 parts. Potassium cyanide 10 parts. 5 parts. Gold chloride 12 parts. 12 parts. The first formula is for gold-plating silver, copper, and alloys rich in copper. The second formula is for iron and steel. 668 ELECTRICIANS' HANDY BOOK. The sodium phosphate is dissolved by heat in 8000 parts of water, the gold chloride in 1000 parts of water, and the two solutions are mixed. The remaining salts are dissolved in 1000 parts of water and added to the others. This gives nearly 10,000 parts of solution by weight, or about a one-tenth of one per cent solution of gold cyanide. These baths are employed at temperatures of 122° to 176"" F. (50° to 80° C). A few minutes is time enough to give a coat- ing. A platinum anode is used. If a large area is immersed, the deposit is reddish in color; if the anode is partly withdrawn, the tendency is toward a pale deposit. This bath is best made up new as required. Enriching it by addition of gold salt and potassium cyanide is not recommended. The reason so short a period of plating is required is that gold has the property of giving an exceedingly thin and uniform coating. A very small thickness "covers." Platinum = Plating. — Platinum-plating is not often done. The following is a formula for the solution for plating copper and its alloys : Dissolve 17 parts platinic chloride in 500 parts of distilled water. Dissolve 100 parts ammonium phosphate in 500 parts of distilled water. Mix the solutions. A precipitate will be formed. Little by little a solution of 500 parts sodium phosphate in 1000 parts of water is added and the whole is brought to boil- ing, water lost by evaporation being constantly replaced until, the ammonia being boiled away, the solution becomes acid and loses the yellow color it possessed and becomes colorless. This bath is used hot with a strong current, and its strength must be kept up by additions of the ammonium-platinum phos- phate precipitate, obtained as above described. Another formula is carried out by adding to a solution of platinum chloride a sufficient excess of potassium cyanide to form a clear solution of ammonium-platinum cyanide. A moderate cur- rent is required, or else a black powder will be deposited. The anode in platinum-plating is always platinum. Tin.— The following is a solution for the deposition of tin: Sodium pyrophosphate 10 parts. Water 1000 parts. ELECTROPLATING. 669 In this solution is dissolved 1 part of fused tin chloride (stan- nous chloride, tin protochloride). There is liable to be some difficulty in the solution. If pieces of the tin chloride fall to the bottom, they may become coated with a sort of crust which is difficultly soluble, and which retards the solution of the tin salt. One way is to put the tin salt into a perforated ladle, like a cul- lender, and keep the salt near the surface of the liquid, and agi- tate it until it dissolves. The anode is of tin. The strength of the bath is maintained by adding, by means of the perforated ladle, equal parts of sodium pyrophosphate and tin chloride. Another solution is made thus: Metallic tin is dissolved in hydrochloric acid, and is precipitated by addition of caustic potash solution. The precipitate is mixed with a solution of potassium cyanide and caustic potash until it dissolves. Steeling.— A coating of iron is sometimes deposited on copper electrotypes of engravings in order to harden the surface. The iron thus deposited is so hard and durable that it is sometimes termed steel, although it is not steel at all, but pure iron. The bath may be thus prepared : A solution of 1 part sal-ammoniac (ammonium chloride) in 5 parts of water is made. In it are suspended two plates of iron connected to the poles of a strong battery. After some hours the solution is ready, as some of the iron will be dissolved. The electrotype which is to be steeled is put into the bath after thorough cleaning and washing with caustic potash solution. About 4 volts electromotive force are prescribed. After the steeling the plates are washed in cold water and rubbed with benzine. To preserve them from rusting they are covered with a film of beeswax. Size of Conductors. — The conductors leading from the source of current to the bath should be as thick as convenient. All re- sistance of battery, generator, and conductors absorbs energy, which is wasted. The slight additional expense of large conduc- tors is compensated for by the economy of power. Current Intensity. — The quality of the deposit is greatly modi- fied by the intensity of current per unit area of surface plated. Thus in a copper-plating bath too strong a current will give a 670 ELECTRICIANS' HANDY BOOK. brown deposit almost powdery in quality. To remedy any such tendency when it is observed, the current strength must be dimin- ished. This is easily done by raising the anode so as to decrease its immersed surface. Another way is to move the anode and ob- jects plated away from each other in the bath. If the current is too weak, the anode can be dipped deeper, or an extra one added, or the distance spoken of above can be de- creased. The latter is only advisable when a rather flat surface is being plated, because an irregular piece near the anode will have a thicker deposit formed on its protuberant parts than on its retreating parts. The difference in distance between projections and recesses and the flat anode will vary less proportionately for large than for small distances between anode and object. Too acid a bath gives less resistance, and tends to the develop- ment of too strong a current. Too little acid has the reverse effect. Copper deposited w^ith too weak a current is crystalline and brittle. A general rule is to have the surface of the anodes equal to that of the objects to be plated. • All such rules are only general. Thus, the question of excess of acid only applies to the limited number of baths in which free acid is present. Most baths are of alkaline reaction. The Relative Position of Anode and Surface to be Plated has its effect on the result. They should be as nearly parallel as pos- sible on general considerations. But there is a tendency for the lov/er parts of the objects to receive the thickest coating, as the solution tends in use to become more dense at the bottom of the bath. This tendency is counteracted by changing the position of the pieces, by moving them constantly, which is often done by power, and by agitating or stirring the solution in the bath. When an object is quickly plated, these precautions are unneces- sary. But when objects remain a long time in the bath, streaks are liable to appear on their lower area if they are not moved, or if the liquid is not stirred about. The tendency of metal to be deposited most thickly on parts nearest to the anode leads to the following rule: When a piece in high relief is to be plated, the anode should be as far removed as is possible from the object. The anode can be increased in area ELECTROPLATING. 671 to compensate for the greater distance. Especially is this main- tenance of distance important when the solution is of such nature as to attack the object to be plated. In such a case it may attack the object in its deep parts while metal is being de- posited on its high places. A long distance as above is sup- posed to give a better quality of deposit as regards flexibility. Sometim.es for very high relief auxiliary anodes carried oa supporting wires into the deeper parts of the relief may be used to secure the deposition of metal there. Temperature of Baths. — The temperature of the baths has an Important effect. For some baths heat is prescribed; for others^ no heat is required, but they are to be used at the ordinary tem- peratures. Sometimes it is necessary to prevent the formation of insoluble deposits on the anode. These deposits may become sa thick as to prevent the passage of any current. The best way of heating the bath is to place the vessel in an- other larger one containing water. The water in the outer vessel is heated, so that the whole arrangement constitutes a water-bath. Or the bath may be placed on an iron tray filled with sand, which is heated. This constitutes a sand-bath. The sand enables the vessel constituting the bath to be more evenly heated than if it rested on an iron plate in more or less imperfect contact with it. Material of Vessels.— For dipping baths for sulphuric acid a lead-lined tank may be used. For alkaline dipping sheet-iron or cast-iron vessels are excellent. For nitric or hydrochloric acid earthenware or gutta-percha vessels are best. Glass, enameled earthenware, varnished wood, or gutta-percha-lined wood are good materials for the plating baths. Metal Molds. — Sometimes for reproducing articles molds are required. These gfve the reverse of the article. By depositing by the electric current a thick coating of metal on them, the mold is produced in reverse, which is the true reproduction of the original article. Molds are sometimes made of fusible metaL The following is an alloy suitable for the purpose: Bismuth , 28 parts Tin 10 parts Lead 19 parts This alloy melts a little below the boiling point of water. C>72 ELECTRICIAN^ S' HANDY BOOK. One way of using it for the reproduction of metals or coins is to melt it and pour some into a slight depression in a slab of marble. It will lie there in a flattened globule and will stay liquid for some time. The medal or coin to be reproduced is chilled and dried and is dropped flat upon the globule from a height of two or three inches. After the metal has solidified, the medal is separated by light jarring. An exceedingly delicate mold is thus produced, on which the plating is executed. Only a slight hollow m the marble is required to retain the melted metal. Another alloy is the following: Bismuth 250 parts Tin 125 parts L^ad 160 parts Antimony 30 p^rts This alloy is used in a pasty condition, to which it is brought by proper degree of heating. It is then applied to the object. After cooling, a light yet decided blow will separate the two. The mold is then ready for its deposit. Wax and Stearine Molds.— These are more generally used than metal molds. Simple white wax or stearine may be melted and poured over the surface of the object, which latter has been previously ofled with a little olive oil. The wax must be allowed to cool several hours before any attempt is made to detach it from the original. Another way of using wax is to soften ft by heat and to press the object into it. Other formulas are given, such as the following: Spermaceti 225 parts Beeswax > 50 parts Mutton tallow 50 parts Plumbago may be mixed with these compositions with benefit. White lead in dry powder gives still better results. The wax mixtures serve especially for the reproduction of flat objects, such as medals, coins, or for electrotyping. Plaster flolds.— The object is covered with a thin coating of olive oil. Plaster of Paris mfxed to a cream with water is painted on with a brush, and after perfect contact of plaster and object has been thus assured, the rest of the plaster is poured on. It may be held in place by a band of paper. ELECTROPLATING, 673 Elastic Holds. — For dijOacult pieces the plaster mold can some- tfmes be applied in sections, which are then put together after re- moval. This is not always an easy thing to do. Elastic molds are often used for such cases. These are made by mixing a strong solution of glue with molasses, about four parts of glue solution to one of molasses. By heating together a perfect mixture is obtained. This softens when heated, and on cooling becomes elas- tic, like a very stiff jelly. It is melted and poured over the object to be molded. A box may be used to hold the object and to pre- vent the composition from running off. Sometimes threads are led along the surface of the object, secured by glue, if necessary with long ends. They are drawn away through the composition when it has set, and divide it into sections. This composition can be used on undercut and complicated ob- jects. It springs out of shape on being drawn off, and springs back at once. Gutta>Percha fields. — Gutta-percha softens in hot water, can be pressed upon an object so as to give the most delicate outlines, and is indefinitely durable. Gelatine or glue compositions such as just described may give finer results, but do not form durable molds. Alcohol, acid, and alkaline solutions are without effect on gutta-percha. It is worked by being softened in hot water and pressed upon the object to be copied. When the object is of such a shape that the gutta-percha will not leave it, application of hot water will soften it enough to permit it to be removed, and as it cools it will retain the form given it by the object. In using any of these materials, personal experience counts for a great deal. Preparing riolds.— The molds of non-conducting materials just described have to be given a conducting surface. Such are glue mixture, plaster, or gutta-percha molds. Plumbago is generally used for giving this quality to them. The mold is moistened a little, by steam if of plaster, and the plumbago is applied by a soft brush. It is rubbed on until the surface is bright and me- tallic in appearance, and of uniform luster. Copper powder is often mixed with the plumbago. It can be made by putting lumps of pure zinc into boiling and saturated solution of copper sulphate. The zinc is soon covered with the 674 ELECTRICIANS' HANDY BOOK, copper precipitated. The lumps of zinc may be removed and the copper brushed off, washed, and dried. This powder can be used alone. Sometimes fine iron powder is dusted over the surface after plumbago has been applied. On dipping into a solution of copper sulphate, metallic copper is precipitated by the iron, and helps to give a good surface for plating. It is to be noted that deposition of metal spreads on all sides. If it begins energetically, in one spot, it spreads as well as builds up, and this action tends to produce even results. Varnish.— Red sealing wax dissolved in alcohol is an excellent varnish for coating parts of objects on which no deposit is desired. Thus the sides and backs of the molds of fusible metal must be varnished to prevent deposition where it is not desired, and where it would prevent removal of the metal deposited. Oiling. — For reproductions on metal molds, the surface must be slightly oiled to prevent adherence. Too thick a coating of . oil will prevent the deposition of any metal, and makes the pro- cess inoperative. Placing Molds In the Bath.— The general system is to place the molds vertically as near as may be and opposite to the anode. Sometimes the mold is placed horizontally below the anode. One objection to this arrangement is that if any dirt or scale is de- tached from the anode, it will fall upon the object and impair the result. If metal molds are used, all connections of the anode should be completed before the mold is introduced. If not, there is danger that the mold may be attacked by the solution and oxi- dized. If the mold, connected to the zinc plate of the battery or equivalent wire of the plating dynamo, is introduced after the anode is in place, its introduction into the bath will be the last thing to complete the circuit, and it will be at once covered with> a thin coating of metal. This is enough to prevent the moldi from being attacked. Plating on flolds.— In using non-conducting molds, it is well to begin with a current of low intensity. The deposit begins near the points of attachment of the conductor, and spreads laterally as already described. If hydrogen is disengaged, the coating will ELECTROPLATING. 675 be brittle, and to prevent this generation of hydrogen the current is started at low intensity. A wire may be used as electrode until the mold is pretty well covered with the deposit. If air bubbles are seen in the interstices of the mold, they can be removed with a camel's hair pencil or other soft brush. Backing Up Deposits. — A thin deposit will suffice in reproduc- ing hollow objects, if it is strengthened by fusible brass or spel- ter such as is used for brazing. This can be put into the interior of the reproduction in small pieces with some borax. The blow- pipe is then used to heat the whole to redness. The spelter runs, and is made to spread all over the surface by inclining the mold from side to side, so that it attaches itself to all the interior surface. The copper although thin resists the heat much longer than the spelter. It must be remembered that there is some danger of the spelter attacking and alloying with the copper and thus destroying the reproduction. Too long applied and too high heat will do this, and it will occur the more easily as the copper is thinner. If the copper is about one-tenth of an inch in thickness, there will be little danger of such an accident. Plating on Glass. — A good deal of this work has been done recently, bottles especially having silver deposited upon them m various open-work designs and engraved as desired. Several methods have been used. Originally a varnish or lacquer was painted over the entire surface of the vessel on which the metal was to be deposited. When almost dry, plumbago was dusted over it, and it was polished with a soft brush. It was then wired, connected to the tank wire, placed in the tank, and left there for about eighteen hours' operation of the electro-plating current, or until a sufficient thickness of metal was obtained. The snow- white deposit of silver was polished. The designer then took it in hand, and painted a design with wax on the surface. The article was then immersed in an acid bath, and the silver not covered with wax was dissolved. Diluted nftric acid would an- swer for this operation. The engraver finishes the process by putting in any lines desired. The plumbago with the lacquer formed a black background, which was undesirable. A more recent process has been substi- tuted for the one described. Nitrate of silver solution mixed 676 ELECTRICIANS' HANDY BOOK. with dextrose solution is poured over the article. Silver is de- posited by reduction. Any of the methods of silvering used on astronomical reflectors, or so frequently used now on mirrors, can be employed. In this way an exceedingly thin coating of me- tallic silver is deposited all over the surface of the article. The electroplating is done upon this, and the processes detailed for the plumbago system are applied. Finally, a metallic oxide has been applied with some varnish- like agent following the design. On baking the oxide is reduced, giving a metallic coating for the electroplating. Practical Processes.— It must be kept in mind that in these as in other electroplating processes a description is not sufficient to enable operations to be successfully performed. Every elec- troplater has methods of carrying out processes which he has acquired by practice, and in many instances these methods are kept secret. A piece of plating may present an excellent ap- pearance, yet on use the metal may scale off. A considerable interval of time may be required to show defects. It is there- fore important when a good and satisfactory process is evolved to stick to it, and not to be too anxious to try something new. I CHAPTER XXXVIII. TELEPHONY. Sound.— Sound is due to vibrations of matter, generally or al- ways vibrations of masses of matter. Thus a piano produces sound by the vibrations of its strings. A pair of cymbals dashed together vibrate, and produce sound. A plate of iron acted on by electro-magnetic pulses of attraction vibrates, and produces sound. The latter is the telephone receiver. Pitch.— Sounds vary in pitch. Some are high and some low. The sounds of high pitch are produced by relatively high-fre- quency vibrations, sounds of low pitch by low-frequency vibra- tions. The lowest note in a church organ may be due to 16 vi- brations and the highest to over 1000 vibrations per second. Fundamental Note.— Every piece of metal or other solid has a note which is produced if the whole piece vibrates as a whole, and is called the fundamental note. If two pieces are of equal thickness and of the same material, the larger one will have the lower natural or fundamental note. This follows out the law of strings; the longer string in a piano has the lower note. Overtones.— When a string in a piano is struck by the hammer, a number of notes are produced. The note due to the motion of the whole string in one arc of vibration back and forth is pro- duced, and is called the fundamental note, due to what we may call n vibrations. Besides this the string produces notes of 2n, 37i, and other multiples of the fundamental note. The sound due to 2n vibrations is one octave higher, that due to 4n vibrations is two octaves higher, and many notes of intermediate value are produced every time a piano string vibrates. These high notes are called "overtones." Sounding Plate,— The bottom of an oil can pushed in and al- lowed to spring out produces a sound due to its motion as a whole. In some sense it is a fundamental; at least it represents 677 678 ELECTRICIAXS' HANDY BOOK. the meclianical action of a half-vibration in each of its motions. If by some process a plate of metal can be made to vibrate as a whole, it v^ould produce its fundamental note. Its motions would resemble those of the bottom of a spring-bottom oil can. Then if in addition the plate could be made to vibrate back and forth in a quarter of its area, as if divided by two lines at right angles to each other, one quarter springing up while the adjacent one sprang down, an overtone would be produced. If in addition to this the total surface vibrated in areas of one-eighth a still higher overtone would be produced. The Human Voice when it produces musical sounds is very rich in overtones. When it speaks, a very complicated vibration of various fundamentals irregularly succeeding each other, and complicated by all sorts of higher-pitch vibrations in addition to what may be called the fundamentals, is produced. It is not exactly a case of overtone production, but of simultaneous vibra- tions of a number of pitches produced by the vihrations of the vocal organs. Principle of Telephone Receiver. — In the telephone a plate of iron is acted on by impulses of attraction and release from attrac- tion. These impulses vary in periodicity and intensity exactly as do the vibrations in the human voice. The impulses force the plate into vibrations identical in frequency and relative strength with those in the human voice. The plate in producing these vibrations does not vibrate as a whole, but is forced to divide itself into areas of vibration. The natural vibration period is not the controlling factor. The plate has to correspond to the pulses of current going over the line and through the coil of the telephone. The pulses of current act upon the iron plate through the Intermediation of an electro-magnet. The receiving instrument contains an electro-magnet connected to the transmission line. In front of the pole of this magnet a plate of iron is held very near to the pole. The pole faces its center. The changes in current passing through the coil are reflected in the attraction it exerts on the magnet. The variations in attraction, which may be many yiundred in a second, force the plate into corresponding vibra- tions. The plate vibrates in subdivisions of its area, which sub- TELEPHONY, 679 divisions vary with great rapidity in number, areas, and shapes. A. low note may make the plate vibrate in three or four areas, while simultaneous nigher notes divide it into a quantity of smaller areas wnich vibrate without interfering with the large areas or with each other. Such is the vibration of the telephone plate. It is easier to think of when we picture the complicated vibrations of a string producing a fundamental tone and a lot of overtones. But the telephone plate only gives its natural or fundamental note by chance coincidence. The magnetic attraction forces it into vi- bration often quite unnatural to it, and such as cannot be referred to its natural periodicity of vibration. The Telephone Transmitter. — The above describes the theory of the telephonic receiver. It is connected by a wire with a dis- tant instrument, which is spoken into and is called the trans- mitter. The original transmitter was an instrument which was a duplicate of the receiver. Two telephone receivers connected in an electric circuit can be used as a complete telephone system. The same instrument can act alternately as transmitter and re- ceiver. As transmitter, the above type performs the functions of a dynamo. The voice makes the plate vibrate fn the same forced manner as described for the receiver. The movements of the plate, which acts as an armature of the magnet, induce currents of high frequency of impulse in the circuit, and the distant tele- phone reproduces them in its plate armature as described. The intensity of the vibrations of the plate of the receiver is very much less than that of the transmitter. In the old-time telephones the speaker shouted into the instrument in order to make himself heard at the other end of the wire. Invention of the Microphone. — Soon after the telephone was invented about 1876, the microphone was invented, and the great defect of the telephone was overcome. Shouting into the trans- mitter ceased to be a requisite. The microphone varies the re- sistance of the telephone circuit. A current is kept passing through it as long as it is in use. As the resistance of the circuit changes, the intensity of the current also changes. These changes act upon the plate of the receiver and make it produce sound. ^%u^ 680 ELECTRICIANS' HANDY BOOK. The changes in resistance are produced by the sound waves produced by the voice acting on the microphone. The distant receiver reproduces the sound of the voice. Hughes Microphone. — This is the original microphone, whfch has been modified indefinitely in the many telephone receivers which have appeared from time to time. Referring to the cut, Fig. 508, C fs a board on which are screwed two blocks BB of hard carbon. Holes in the blocks receive the ends of a little rod of carbon A, which rests in its position quite loosely. The apparatus is placed in circuit with a battery as indicated, and with a tele- phone receiver. The least agitation to which the carbon rod is subjected causes the resistance of the microphone to vary. The varia- tion in resistance causes the cur- ^^JiMSMX^_^^^ rent to vary, and a sound is pro- ! _j^ duced in the receiver. A fly walk- ing on the instrument will produce a sound with every footfall. If talked against, the sound of the voice will be reproduced in the re- ,, ceiver more or less perfectly. riG.508.-HTTGHEs'sMicBo- ^he Blake Transmitter. -For PHONE. many years the Blake transmitter was the classic telephone receiver. In it a highly-finished block or button of hard carbon and a bit of platinum are held in contact with each other. One is pressed against a metallic diaphragm, the other is attached to an arm capa- ble of moving back and forth. The primary circuit of an induction coil includes these two buttons, and the primary current, when- ever the transmitter is in use, passes through the buttons from one to another, and therefore depends on their contact for its comple- tion. "When the mouth is placed close to the diaphragm and words are spoken, the vibrations of the diaphragm change the degree of pressure existing between the buttons, these changes exactly corre- sponding in form with the form of the sound waves. This varia- tion of pressure causes changes in resistance, and therefore by Ohm's law in current also, corresponding in form with the sound waves. These changes of current acting on the distant receiver TELEPHONY, 681 cause its magnet coils to vary in excitation also in form corre- sponding to the original sound waves. This throws the diaphragm of the re- ceiver, which is of iron and is the armature of the magnet, into vibrations exactly similar in form to those of the transmitter diaphragm, so that speech is reproduced. The Blake transmitter depends on pressure changes. Whether these affect resistance by direct variations in pres- sure or by changes in the area of con- tact due to pressure is not certain. It is probable that both actions have a part in the phenomenon. The cut. Fig. 508a, shows the Blake transmitter. A is the opening to be spoken into, closed by a plate of iron or other diaphragm E. At the end of spring F is a bit of platinum, which presses against the diaphragm. K is the carbon button carried by a brass block P at the end of a spring G. B B are pillars, to the upper one of which a heavy counter weight C is attached by a spring. The lower pillar carries an adjusting screw N. The pressure between the platinum and carbon button is thus regulated, and the freedom .given by the spring M, which carries the contact button and platinum contact piece, adds to the sensitiveness. The cur- rent passes through the contact by the springs F and G. Loose Carbon Transmitters. — A variation in the Blake transmitter ap- pears in instruments with quite loose carbon contacts. Thus, two disks of carbon, each with a number of depressions in them, may be Fig. 508a.— The Blake Transmitter. Fig. 509— The Clamond Transmitter. 682 ELECTRICIANS' HANDY BOOK. Fig. 510.— The Western Union Trans- mitter. supported face to face, one attached to a diaphragm. The disks do not touch. In each pair of depressions, which face each other as the disks are placed, is a little carbon sphere. These are thrown into motion by the voice, and act exactly like the orig- inal Hughes microphone. Sometimes little carbon cylinders are used. The cut, Fig. 509, shows the Clamond transmitter. Hunning Transmitter. • — The more modern type of receiver which has met with most favor in this country, is of the so-called Hunning type. This in- ventor substituted for the vary- ing contact of a few pieces of a.ccurately-shaped carbon, the varying contact of a quantity of granular carbon or carbon dust. It is on this basis that the mod- ern transmitter is constructed. In the cut. Fig. 510, is shown an exceedingly simple embodi- ment of this idea. To the right of D is a diaphragm, back of which is a second parallel plate B; the space between them is half filled with carbon dust C. The rest explains itself. This transmitter is of importance as involving the use of granulated carbon instead of regularly- shaped pieces. Edison^s Telephone.— This is shown in Fig. 511. E is the mouthpiece and D the metal dia- Fig. 511.— Edison's Telephone. TELEPHONY, 683 phragm. I is a carbon disk with adjusting screw V. A platinum plate B B with an ivory button h is attached to the carbon disk. The ivory button is pressed against the diaphragm. C C is an in- sulating ring. The connections bring the disk into the circuit, and the resistance is varied when the instrument is spoken into. The Solid Back Transmitter.— This modern instrument con- tains essentially the following parts: Two small disks of polished carbon face each other. They are in a cylindrical case of diam- eter slightly larger than their own. Their faces are maintained near together, but not touching each other. One is attached to the diaphragm, which is m modern instruments often made of alu- minium. The wires from the primary of the induction coil go^ one to one disk and the other to the other. The space between the disks and any space left in the cylindrical case fs filled with fine carbon dust. The action is similar to what has been described. The pressure exerted on the carbon powder by the disks is changed by the vibrations of the diaphragm. The powder is also agitated. One or both of these actions produces the changes in resistance, to which the transmitting power of the circuit is due. Which action is the prevailing one, or what degree of efficiency is to be ascribed to each, is uncertain. The Receiver.— Modern telephone receivers are of several types of construction. The straight hand telephone embodies the fol- lowing points in its construction: Within a hard-rubber cylindrical case is a compound perma- nent horseshoe or U-shaped magnet. A more powerful magnet is produced by clamping together several thin steel magnets than where the magnet is made of one piece of steel as thick as the combined thinner ones. A lamellar or compound magnet is there- fore the best. This magnet has its limbs so close that it fits into the standard India-rubber. case. At its end are pole pieces pro- jecting in line with its limbs, and on these are placed coils or spools of fine insulated wire, wound like coils on a horseshoe mag- net oppositely to each other. At its forward end the cylindrical case carries an expansfon, somewhat like the mouth of a trumpet, over whose front a hard- rubber cover with a central aperture is secured by a thread cut 684 ELECTRICIANS' HANDY BOOK. in the rubber. It screws on like the cover of a box. When in position it holds a disk of iron across the mouth of the tube very- close to the magnet poles. The disk closes the aperture in the cover. It is essential that the distance from disk to magnet poles shall be invariable. If the magnet were secured by its distant end, changes of temperature would constantly cause this dis- tance to vary as the metal of the magnet expanded and con- tracted. In the older receivers this defect was present. The magnet was secured by its distant end to the case; sometimes it was fastened at both ends. In the latter case changes of tem- perature were liable to produce damaging strains. In the modern instrument the magnet is fastened by its forward end only. The four or five inches of steel extends back into the case, and is free to expand or contract without affecting the ad- justment. The critical distance between pole faces and metallic disk is invariable. The bobbins are wound with very fine wire. One of the early troubles with receivers was the breaking of this wire. In the modern instruments it is protected absolutely from all strain. Through the bottom of the handle, closed with a solid disk of India-rubber, pass two binding screws. Within the case these connect to two heavy pieces of insulated wire, which by being twisted together or other simple arrangement, are held fast, so that the upper ends cannot be moved by any manipulation of the binding screws. The upper ends are connected to the ter- minals of the windings of the bobbins. This secures the fine "Wire from all strain. The telephone receiver is secure from all possibility of a broken circuit. The working parts of a modern telephone receiver are shown in Fig. 512. The two limbs of the magnet are seen held parallel to each other, with their upper ends connected by a block of cast iron. On their forward ends are the two bobbins. The plate of Iron is held by a screw cover across the opening of the cup, within which the coils are seen. The screw cover forms the part of the case which Is held against the ear of the person receiving the message. The Telephone Induction Coil.— The telephone transmitter is TELEPHONY, 685 placed in the circuit whictr includes the primary circuit of an induction coil and the exciting battery. This circuit is of far lower resistance than that of the long telephone line and of the coil or coils on the transmitter would be. In this fact lies one great incentive to its use. The sound waves vary the resistance of the microphone or transmitter, for the modern transmitter is invariably a microphone. If the transmitter is in a circuit of low resistance, its variations in resistance will be larger propor- tions of the total resistance of the circuit than if the total resistance of the circuit were high. A variation of 1/100 ohm on a 5-ohm circuit would be a variation of 1/500 of the total Fig. 512.— Telephone Receiver, resistance. By Ohm's law such a variation would cause the cur- rent to vary 1/500 fn intensity. But if the circuit were of 300 ohms, the 1/100 ohm variation would only be 1/30,000 of the total resistance, and would only produce that variation in the current. The use of an induction coil secures this feature. The primary of the coil and the battery for actuating it need only have a comparatively small resistance. Acting as the primary of the induction coil upon the secondary, the variations in current due to the microphone action of the transmitter induce potential changes in the secondary. This im- presses a much higher set of voltages upon its circuit, and a diminished current varying in intensity in proportion to the changes in the primary goes over the line. The receiver is wound for this current with many turns of wire, so that the action of the current on the magnetic field of the receiver is in- 686 ELECTRICIANS' HANDY BOOK. creased or accentuated. The slight current multiplied by the large number of turns of wire in the receiver gives a tangible number of ampere turns. The induction coil effects two things. It brings about a relative- ly hfgh variation in the current changes, due to microphonic ac- tion, and it enables a much smaller wire to be used for trans- mission. There are other things involved, into which this book will not go, affecting the capacity of the line, the relative qual- ities for clear transmission of a small wire with small current or of a large wire with correspondingly large current, and other similar points. In Fig. 513 B is a battery, T represents a transmitter, and P the I Fig. 513.— Induction Coil in Telephone Circuit. primary of an induction coil, whose secondary is indicated by S. The telephone receiver R is brought into circuit with the second- ary of the induction coil by the line wire L L'. This illustrates the place of an induction coil in a telephone circuit. The extension of the telephone is made much easier by the use of small wires. A lead-covered cable, not much over three inches in diameter at its bulkiest parts, such as joints, can accom- modate wires enough for fifty or more metallic circuits. In the country almost invisible wires can be carried overhead through long spans at very slight expense. The induction coil makes these practicable in service. Dimensions of Teleplione Induction Coils.— The dimensions of TELEPHONY, 687 induction coils include the turns of wire in primary and second- ary, the size and length of Vires, and the consequent resistances. The iron core made of soft iron wires fs not generally stated. The best dimensions are determined by trial rather than by calculation. Coils are tested over various lengths of line with transmitters of the class eventually to be used in the service. In general, it has been found that a coil good for one distance was good for another. With the old Blake transmitter in this country an induction Figs. 514 and 515.— Telephone Induction Coil. coil of one-half ohm primary and 250 ohms secondary was used. An extreme case of a low-resistance coil has been used on long- distance lines in this country. This one had a primary coil re- sistance of 0.3 ohm and a secondary coil resistance of 14 ohms. The ratio or resistance in the first case was 1 to 500, in the second 1 to 46. The last-described coil had a very large core. The following are the dimensions of a typical modern coil for ordinary work: Core about 5 inches long and 9/16 inch in diam- eter, composed of 500 strands of No. 24 American gauge soft Swedes iron wire. The core is contained in a thin tube of fiber with square wooden heads or flanges at the ends. The primary coil is wound on the tube. It is composed of No. 20 wire, and two ess ELECTRICIANS' HANDY BOOK. layers are wound in 200 turns. Paper is wound over it some layers deep, and the secondary is wound on this. It consists of two lines of No. 34 wire making 1,400 turns. The resistance of the primary coil is 0.38 ohm, of the secondary 75 ohms. This gives a resistance ratio of about 1 to 19 and of turns 1 to 7 only. Large wires are connected to the windings, and secured so as to prevent any strain coming on the windings. Figs. 514 and 515 give a sectional view and side view of a modern cofl with its primary coil and secondary coil wound on a core, consisting of a bundle of iron wire. Fig. 516.— Bracket Telephone. Fig. 517. — Induction Coil in Bracket Telephone. Induction Coils in Bracket Telephones. — Coils are sometimes placed in the bases of swinging bracket telephones. Fig. 516 shows such a telephone, and Fig. 517 shows the section of the chamber at its base, within which the induction coil is placed. Effect of the Telephone Induction Coil. — The universal use of induction coils shows that they are valuable in telephony. The ratio of reduction of current is not so great as it would seem that it might be. They exercise an effect on the current also. The microphone current is uniform in direction, but of varying intensity. By the induction coil this current is changed to an alternating one. The direction during the increase of microphone resistance is in TELEPHONY. 689 one direction, and during the decrease in the other. When the microphone is inactive, a steady current passes in the microphone circuit if the receiver is off its hook, while in the secondary induc- tion coil circuit under this condition no current whatever passes. In modern central battery practice they can no longer be con- sidered to have much effect in reducing the size of line wire, ow- ing to the absence of any battery at the customer's telephone ap- paratus. The Telephone Magneto.— The bell magneto has already been spoken of in this book. Although not part of a telephone sys- FiGS. 518 AND 519.- Calling Magneto. tem strictly speaking, so many have been and still are in use as calling apparatus that some description must be given of them here. The magneto used for calling the central office consists of a field composed of several U-shaped permanent magnets, between whose poles a single-coil armature is rotated by turning a handle. On the shaft of the handle is a cogwheel which actuates a much smaller one on the shaft of the armature, so as to give it a suflS- ciently high speed of rotation. Two sprocket wheels and chain are sometimes used for the same purpose instead of gear wheels. Figs. 518 and 519 show a magneto generator, and Fig. 520 shows its armature core. Some device is usually applied to cut the armature out of cir- 690 ELECTRICIAN^ 8' HANDY BOOK. Fig. 520,— Armature Core of Mag- neto Bell. cuit when not in use. Sometimes the shaft of the large gear wheel is free to move in the direction of its length a short dis- tance. In one position which it takes when at rest, it makes con- tact with the wire and short-circuits the armature. When the lat- ter is to act, then the turn- ing of the handle automati- cally shifts the shaft a short distance, and breaks the cir- cuit, so that all the current passes through the armature. The shifting of the axle may be effected by the use of a cylindrical cam on which a projection on the shaft rides up. This cam may be formed on the hub of the large gear wheel, which wheel is so mounted that it cannot move along the line of the shaft, while the shaft can move back and forth through the aperture in its hub. In Fig. 521 this apparatus is illustrated. The large gear wheel C carries the cam on which the pin P rides up, shifting the shaft to the right and breaking the contact be- tween its end and tho spring O. The magnets in this generator, made by the "Western Telephone Con- struction Company, are made of magnet steel, in cross section % inch by % inch, and are bent into shape cold. The air gap, which is the distance from the arma- ture surface to the surface of the magnet poles, may be as small as 1/100 inch. A cast-iron core turned so that it fits with this slight clearance between the poles is wound with insulated wire. The pole pieces of the magnet are attached to the ends of the magnets, and are bored out to form a chamber for the armature to revolve in. (t^ Fig. 521.— Automatic Magneto Switch. TELEPHONY. 691 Many different magnetos have been constructed, differing only in detail. The current produced is alternating and of the sine type approximately. The bells which are mounted on the face of the magneto case are rung by a hammer operated by an electro- magnet with polarized relay. The armature is in improved constructions made of laminated type, built up of thin disks held upon a shaft. The pole pieces are sometimes laminated also. The armature of the 10,000-ohm magneto is wound with No. 35 or 36 American wire gauge silk-covered wire. The classification of magnetos is based on the resistance of a line through which they can ring a bell. The figure such as 10,000 ohms above ex- presses line resistance, and has no direct reference to the di- mensions of the magneto. The resistance of the armature of the above magneto may vary from 400 to 550 ohms. The magnets of the bells are wound with No. 31 American wire gauge wire to a resistance of 75 to 100 ohms. Silk-insulated wire is used for winding. These dimensions apply to magnetos used on series work. But sometimes calling bells are connected across a circuit like lamps in parallel. Such arrangement is called in telephone practice bridging work. For this work a high inductance in the bell magnets fs required to prevent the rapidly alternating speaking current from going through the coils. It must be shunted through the receiver in parallel with the bell magnet coils. The generator for bridging work should have a stronger field than that of the one just described, with a longer armature wound with No. 33 wire to about 350 ohms. The bell magnets are wound to as high a resfstance as 1,000 ohms with No. 33 single silk-covered wire, or to 1,200 to 1,600 ohms with No. 38 wire. The thing prin- cipally wanted is not resistance, but inductance, so that they shall act as choke coils for the speaking current. In central stations large magnetos or alternators are drfven by power and kept constantly in action. Current for ringing is taken from them by the operatives as required. Polarized Bell.— This is the bell which is rung by the magneto. It is shown fn Figs. 522 and 523. The electro-magnet has an armature pivoted below it, and to the center of which the clapper 692 ELECTRICIANS' HANDY BOOK, of the bells is attached. The armature is a bar of steel, and is magnetized so as to have a north pole at one end and a south pole at the other. When an alternating current from the mag- neto passes through the windings of the magnets, their strengths change with each alternation of current, so that the polarized armature swings first one way and then another, thus keeping the clapper in motion, so as to ring the bells placed within its range of motion. Telephone Systems. — The general installation of a telephone system includes these elements: A microphone, termed the transmitter, is the apparatus spoken against. This is in a cir- cuit through which a current flows when the transmitter is in Figs. 522 and 523.— Polarized Armature Bell, use. On this circuit is a source of potential which maintains the current. Generally, this circuit includes the primary circuit of an induction coil. There is a secondary circuit of the induc- tion coil, which is in circuit with the receiving instrument. The receiver, as it is called, is a modification and only a slight one of the Bell telephone of twenty-five years ago. To effect electrical connection between different customers, there is a central station. Wires from the customers' houses go to the central station, and by means of one or more switchboards communication between any two customers is brought about in a few seconds. Finally, call- ing apparatus fs included. In the houses of customers this gives an audible signal, generally the ringing of a bell. In the central TULEPZIONY. 693 office a shutter is dropped, making a click and exposing the num- ber of the customer, or else in more modern practice an in- candescent lamp at the customer's number on the switchboard is lighted when a call is made. In many systems at the present day there is a battery in every customer's house. In more modern practice all the current is supplied from a storage battery at the central station. Protec- tive devices to secure the system from lightning and damage from crosses with other wires are among details which form in mod- ern practice essential parts of the system. House Connections. — The house connections for a telephone Fig. 524.-:-House Telephone Connection. Hook-S witch Depressed. instrument with private battery have been variously carried out from time to time. A diagram of a typical system is given in the cuts. Figs. 524 and 524a. LL' are the lines, M the magneto, C the call bell, T the transmitter, B the customer's battery, P and S the primary and secondary of the induction coil, R the re- ceiver, and the hook-switch on which the transmitter is hung when it is out of use is seen on the left of the coil. The first position shown is that in which the hook-switch is down. This is brought about in practice by hanging the receiver on the hook-switch. The illustrations show the circuit, including the receiver, induction coil, battery, and transmitter, and the con- necting line &, all of which are thrown out of circuit when the hook-switch is depressed and connects with the stud 3. 694 ELECTRICIANS' HANDY BOOK. An alternating current sent over the line from the central sta- tion goes through the magneto armature coil, then through the bell, ringing the latter, and by way of the hook-switch and stud 3 and connecting wire a back to the other line L'. In this position, if the handle of the generator is turned, an alternating current will be sent over the line and will ring the bells at the central station or will operate any form of signal ap- paratus employed there. The receiver is not shown on the hook, in order to make the diagram clearer. It is supposed to be hung upon the hook-switch. On hearing the call the customer unhooks the receiver, and the hook-switch springs up. It opens the circuit at 3 and closes the n Line L >2 3 H Line L' Fig. 624a.— House Telephone Connection. Hook-Switch Raised. circuits at 1 and 2. The cut, Fig. 524a, shows the connections thus brought about. What before were inactive wires become active. The magneto and its bell are cut out of circuit. Tracing the connections on this diagram, it will be seen that the receiver is now in circuit with the line L', and through the secondary S of the induction coil with the line L. A message can be received by it. The transmitter is in circuit with the battery and primary of the induction coil. The circuit contain- ing these three is closed through the contact 2 and the hook- switch. The hook-switch acts as a conductor for both primary and secondary currents from the induction coil, and as conductor for the talking current from the distant instrument. In the posi- TELEPHONY, 695 tion shown, the primary coil of the induction coil is on closed circuit, and a direct current goes through the transmitter. "When the transmitter is spoken into, the primary current varies as described, and the secondary current induced goes through the receiver to the line L L', and through the hook-switch, connection 1, line d, to the other line L. Series Telephone Circuit.— 'This is shown in Fig. 525. The ~l_i_u_ Fig. 525.— Series Telephone Circuit. Fig. 526.— Bridged Telephone Circuit. line wires connect at 1 and 3. When the hook-switch is depressed, the bell B is in circuit with the line and the magneto M is short- circuited. When the customer operates his generator this short circuit is automatically opened. When the hook-switch is de- pressed the receiver R is also cut out of the circuit. The connec- tions are now adapted for calling up by the bell B only. On unhooking the receiver R, the hook-switch L springs up, opens the bell circuit, and closes both the circuit of the transmit- ter T with the primary P of the induction coil in circuit with it and the circuit of the receiver R with the secondary S of the in- 696 ELECTRICIANS' HANDY BOOK. duction coil in circuit with it. The connections are now ready to transmit and receive. In this cut and in Fig. 526 central binding posts 2 are shown. These are fcr connecting to the ground for the lightning arresters. Bridged Telephone Circuit. — Circuits of this description are characterized by the fact that the bell is connected across the lines permanently. A bridged circuit is given in Fig. 526. The bell C is permanently connected between the lines 1 and 2. It3 magnets are wound of high resistance and have high inductance. "Whether the hook-switch H is up or down, the bell circuit is in the same connection, being quite independent. But the resistance and reactance of its magnets make it an effectual barrier to tele- phonic currents; for them it is a choke coil. The magneto M is in a second bridged circuit, normally open, but closed w^hen the handle is turned. These are parts of the calling circuit. The talking circuit with the receiver R in circuit with the secondary S of the induction coil is a third bridge circuit, open when the hook-switch is depressed. V/hen the telephone is taken off the hook, it rises and closes the talking circuit and also the local transmitting circuit. In the latter the primary P of the induction coil, the transmitter T, and battery B are included in series. A switch is shown at Ic at the magneto. This is supposed to be operated by hand to close the magneto circuit when the central is to be rung up. An automatic closing device similar to that already described for the magneto is also used. The Hook-Switch. — Considerable thought has been expended on the best construction of the hook-switch. Platinum connecting points or studs are the best, and it is an object to have a little sliding action as they open and close with the rise and descent of the switch. This tends to keep the contacts in good condition and free from dust. The contact action is only due to gravity when the receiver is hung up, and to a spring when the receiver is removed and the hook-switch springs up. If the contacts do slide, they should slide only on a conducting surface, not on an in- sulating surface and then on a conducting one. Sliding contacts bring about cutting as one of their objectionable features. If platinum contacts are used, sliding contacts are not neces- TELEPHONY. 697 sary; for such metals as brass they are requisite. To prevent cutting, it is a good plan* to make the two surfaces of dissimilar metals, just as in steam engine and heavy machine practice. The use of brass surfaces with German-silver springs, sliding as they make contact on the brass, is considered good practice. Sometimes the lever or hook-switch arm forms no part of the circuit, but it generally does. The journal or pivot screw should not be de- pended on as part of the cfrcuit, but should be reinforced by a flexible wire twisted into a spiral spring, with its ends soldered one to the base, and one to the switch arm. It is well to pass the ends of the wire through holes and solder it in them after burring or riveting its ends. Common Battery Systems. — The most advanced system of tele- phone installation has no local batteries in the house s«ts of tele- Fig. 527.— Common Battery Metallic Cikcuit System. phoning apparatus. In very many installations the local battery is still employed, and the circuits hitherto shown in this book have embodied it. The simplest representation of a metallic-circuit common bat- tery system is shown in Fig. 527. B is the battery at the central station. This is always a storage battery. At P are plug switches. The line drawn through the center of each switch indicates insu- lation of the two sides of the plug from one another. When the plugs are inserted, the two line transmitters T and receivers R and the battery are all thrown into circuit. It will be understood that the hous-e connections are here omitted, the receiver and transmitter merely indicating them. They follow the general lines of those used with the local battery. The system is shown in this cut as applied to several sub- scribers. The low resistance of the battery prevents any notice- able amount of current being deflected from one circuit into the other. 698 ELECTRICIANS' HANDY BOOK, Sometimes choke coils are used between the central battery and the subscribers' lines. These coils permit the passage of di- rect current from the battery, which gives the basis for the trans- mitters to work on. The choke coils cut off all chance of inter- communication between independent circuits. Stone's Common Battery System.— The diagram. Fig. 528, shows two circuits supplied from a single battery B. The coils used are of but slight resistance compared to that of the rest of the circuit, but are of considerable impedance. The battery main- tains a direct current through any of the circuits it is plugged into or connected with. The transmitter when talked into causes this current to vary, and a speaking current is thus produced, restricted practically to its own circuit by the inductance of the coils. This inductance resists the passage of an alternating or Fig. 528.— Stone's Common Battery System. undulatory current, such as that of the speaking type produced by the microphonic action of the transmitter. This system is due to John S. Stone. Dean's Common Battery System. — A most ingenious applica- tion of the choke coil enables both lines of a metallic circuit to be used in parallel for sending current to the transmitting circuit with a ground return. The diagram. Fig. 529, gives the general features of the system. The central station battery B is grounded. It connects from its other terminal to the center of a choke coil I whose winding Is connected across the two leads of the metallic circuit. At the subscriber's end of the circuit another choke coil, I, is connected across. From its center a connection fs taken to a local closed circuit, including the primary of an induction coil and a trans- TELEPHONY. 699 mitter. A ground connection is taken from a point of thfs cir- cuit opposite to the other connection and between transmitter and primary of the induction coil. From the battery when all connections are made a direct cur- rent goes through the two branches of the choke coil I to both leads of the metallic circuit. It goes through both of these in parallel and through both parts of the choke coil I to its central connection. Thence it goes through both branches of the closed primary of the coil in the transmitter circuit to the ground. In this closed circuit the current divides. Part, goes through the primary p of the induction coil, but without effect, as it is a con- stant current. Part goes through the transmitter T. These variations in current through p, due to the voice acting Fig. 529.— Dean's Common Battery System. on the transmitter, induce a speaking current in the metallic cir- cuit, which includes the secondary s of the induction coil, and the receiver R at each station. The inductance of the choke coils pre- vents any of the talking current going through them, so that the circuit for talking purposes is a true metallic one. By the use of choke coils on the same principles it has been pro- posed to have local storage batteries in«the subscribers' sets, and to charge them from the central station. The use of choke coils enables the two lines of the metallic circuit to be used fn parallel for the charging current with a ground return. The parallel cir- cuit thus given is of one-half the normal resistance of the line. The current flowing in the same direction in both leads at the subscriber's station divides between the storage battery on one side and the transmitter with a special resistance coil on the other. The resistance coil and transmitter with their resistances in series shunt most of the current through the storage battery. 700 ELECTRICIANS' HANDY BOOK. Party Lines. — The expense of a telephone distribution system is materially diminished by grouping private stations which are near to each other in groups of four or more, and serving them all with a single circuit from the central office. The first thing Involved is the calling up of any one subscriber of the groups from the central station without calling up the others. Polarized Bells for Party Lines. — In some systems polarized bells are used. The general principle of these may be given in a few words. One type of polarized bell is one whose magnet armature is a permanent magnet, and which is attracted to the electro-magnet by current in one direction and repelled by that in another. If a current in the direction which attracts the armature is sent through the magnet coils, the armature will move toward the magnet poles and the striker will strike the bell. If the current ceases, it will be drawn back by the spring. An intermittent current in one direction will keep the striker in vibration, and the bell will ring continuously. If the current is in the other direction, it will repel the armature only and no ringing will be produced. The bells in two of the subscribers' houses are connected to one lead of the circuit only, and are grounded from the other terminal. These bells are oppositely polarized. By sending a ringing current in one or the other direction, either bell can be rung as desired. On the other lead of the main circuit two more bells, also op- positely polarized, are connected and grounded, each at a sub- scriber's house. By using this lead and sending ringing currents of opposite directions, either of these subscribers can be called up. The four bells by this arrangement can be individually rung from the central station. Eight bells can be individually rung by calling upon variation in current strength as well as polarity. Four polarized bells wound to high resistance are connected exactly as described, on the four most distant stations. These can be rung by a light current one at a time as desired. At the nearer stations four polarized bells are connected in series, two oppositely polarized on each lead. These are wound TELEPHONY. 701 to low resistance, and are not actuated by the slight current which rings the further bells. But if a strong enough current to ring one of them is sent, a relay situated beyond them is actuated and grounds the line at that point. The bell at the nearer station is then rung through the ground connection, which connection cuts out the bells beyond the house in question. In any case, only one of these would be rung, on account of its polarity and of the direction of the current. Suppose that a line is fitted with four polarized bells as de- scribed on page 700 for four separate subscribers' stations. Using both lines in parallel, as if they were one, two oppositely polar- ized bells can be connected thereto and grounded. They can be operated exactly as the two bells on either lead of the line. The full metallic circuit can be utilized for two more bells. This gives eight subscribers on a single circuit. In practice this system is operated by ordinary bells actuated from a local battery. The bell and battery are connected in a local circuit opened and closed by polarized and neutral relays, differently connected as regards their polarization, there being two relays for each station and bell. The armatures of both relays at a house must be released from attraction and rest against their back-stops to cause their bell to ring. The operator by sending current in one or the other direction over one or the other of the leads, or over both in parallel, can ring any of the eight bells. One arrangement is to utilize six of the single-wire and through circuit connec- tions for subscribers' signals, and to use the remaining two for locking the hook-switch, so that the central office cannot be called when the line is in use by another subscriber. Harmonic Signal for Party Lines. — The armature of an elec- tro-magnet can be mounted with a flat straight spring in place of a pivot. Such an armature if pulled to one side and released will swing back and forth and vibrate at a frequency depending on its weight, the length and the stiffness of the spring. If a series of impulses are imparted to it, coinciding in frequency with its own natural frequency, it will be caused to vibrate. If the impulses are irregular or have no correspondence with the periodicity of the armature movements, they will give it some 702 ELECTRICIANS' HANDY BOOK. motion perhaps, but not with the same energy as if they harmon- ized with each other. If through a magnet facing the armature impulses were sent, they would have little effect on the armature unless their frequency corresponded with that of the magnet. If their frequency was one-half or one-quarter or other integral fraction of that of the armature they would affect it, but would have most effect if of its exact frequency. Such series of impulses of current would start it into vibration. A contact point must be Fig. 530.— Harmonic Bell Signal. provided with which the armature will make contact as it vibrates. By this contact a bell circuit is to be closed. For each closing the bell will ring. As long as the armature vibrates, the bell will ring in unison. The cut, Fig. 530, shows the general idea of such a harmonic signal. The magnet C receives the current broken into impulses of definite number per second. When this number corresponds with the natural number of vibrations of the armature B, carried by the flat spring screwed on the top of the block h, the armature will vibrate, and only then. If it vibrates, it will close the bell circuit at the point n, whert^ TELEPHONY. 705 there are two contacts, one on B and one on D. When this con- tact is closed, the bell rings. The armature B will not be thrown into vibration by a broken current whose impulses do not correspond with its own natural period of vibration. By having armatures of different rates of vibration at different subscribers' houses, any subscriber can be called by a broken current of frequency corresponding to that of his armature. In some systems the vibrations are used to close a bell circuit as above, in some to open a shunt in parallel with the bell, and which when closed prevents it from ringing by taking the cur- rent. In some the bell-hammer and armature are one. The harmonic system is very little used in American tele- phone practice. A practical limitation exists to the number of subscribers that can be served by one wire, because the amount of service ex- acted by four to six subscribers is about all that one line can take care of. If harmonic calls were used, only four to six rates of current impulses would be needed at the central station. Distributing Boards. — A central telephone station may have six thousand or more individual circuits entering it. Every one of these has to be taken to its place, where a number is assigned it on the main switchboard, which in all large offices is of the multiple type. The mass of wires back of the main switch is complicated, and if it had to have its connections changed and shifted about, endless confusion would result. To avoid the necessity for changing the wires at this point, a special arrangement called a distributing board is used. It pro- vides two faces or boards, separated a little from each other. On one face are secured all the wires of the circuits which enter the building. These connections are supposed never to be dis- turbed under ordinary conditions. A multiple switchboard has a number of identical panels. Each panel has plug sockets for all tlie circuits that enter the building, with perhaps one or two thousand others to provide for future extensions. Circuits from the multiple switchboard equal in number to the sockets on one panel of the board run to the other side of the T04 ELECTRICIANS' HANDY LOOK. distributing board. As the panels of the multiple switchboard all are connected, No. 1 to No. 1 and so on all the length of the board, it follows that every connection on the distributing board connects with every panel of the multiple switchboard. It does more than this. Taking No. 1 connection from the distributing board, this wire connects with every No. 1 plug on the switchboard. There may be fifty or more panels, on each panel a single No. 1 plug and all connected to one circuit. This circuit goes to No. 1 connection on the distributing board. The same is done for every socket on each panel; a connection from all of each given number runs to a corresponding number on the distributing board. These connections are normally never disturbed. The space of some feet in depth intervening between the front and back of the distributing board is bridged across by wires, one for every active connection on the switchboard. These wire connections are subject to frequent change. If it is necessary to change a subscriber's number, the wire from his connection to the distributing board is connected to the other face of the board, to the connection leading to the desired sets of numbers on the multiple switchboard. The shape the distributing board takes is a sort of open rectangular rack. Several feet intervene between the two faces, and within this space the connections are made. There is little that is distinctive about them. Each one has to have front and rear connections corresponding in number, and on the face next to the switchboard they must correspond in designation with the sockets on each panel of the multiple switchboard. A wire circuit enters the building, and is connected to the rear of the distributing board. It may be decided to connect it to the set of sockets numbered 75 on the multiple switchboard. By short wire leads within the distributing board the con- nection from the incoming wire to the No. 75 connection on the other face of the distributing board is made. This one connec- tion puts the subscriber whose wire circuit is thus disposed of in connection with every plug bearing the number 75 on the multiple switchboard, as well as with the calling plug for the operator. TELEPHONY, 705 Fig. 531 gives a cross section and view of the side of a dis- tributing board. At C a cable from the street is supposed to enter. Its end is opened, and wires tv are taken from the cable head H and carried to the near face of the board. Wires from the other face run to the plug connections S S on the switch- board C. Wires called bridle or jumper wires connect the front and rear connections of the distributing board with each other. The board illustrated is the Hibbard board. The frame is open work built of iron pipe, forming a sort of trellis. Fig. 531.— Distributing Board. It has been aptly said that the object of the distributing board Is to concentrate the changes of connections into a definite local- ity. The sLort wire connections are of No. 20 to 22 wire tinned, rubber-covered, and twisted in pairs to give the elements for continuing the metallic circuit. Lightning arresters are often included in the connections The best and generally accepted practice is to solder all the ju^nper wire connections. Repeating Coils. — The repeating coil used in telephone practice is an induction coil. Its core is made of a bundle of annealed iron wire. Its windings are generally of the same number of turns and of the same size of wire for both primary and second- 706 ELECTRICIANS' HANDY BOOK. ary circuits or windings. It is used to cause the speaking cur- rent in one line to be transferred to another. It has four binding posts — a pair for each circuit. Thus a ground telephone circuit may extend to a certain point in the district and there terminate. Its end may be connected to one terminal or binding post of the coil, whose other corres- ponding binding post is grounded by another wire connection. To the other pair of terminals are connected the ends of a metal- lic telephone circuit. Any conversation on the grounded line will be transferred to the metallic circuit by induction, and the reverse action will also take place. Thus, a circuit may be part grounded and part metallic. One principal use of this combination is to avoid interference from other lines, and not have the expense of a full metallic cir- cuit system. Where there is no danger of interference a ground circuit is used, with one winding of the repeating coil in series at its outer end. For the part where interference is feared a metal- lic circuit is put in, with the other winding of coil in series. By another repeating coil a ground circuit may be brought into the circuit again if the area of disturbance and interference is passed. There are other uses of the repeating coil in central station prac- tice. The Multiple Switchboard Is used in central telephone ex- changes to effect the connection of one subscriber with another. If there were but one or tv/o hundred subscribers in a district, the connections between them could be effected by a single board. On the board the terminals of all the subscribers could be placed and each one numbered. A flexible wire with proper end con- nections could connect any subscriber's terminal to any other. If there were some thousand subscribers in a district, it would be impossible for a single operator to answer all the calls which would come in from them. Therefore, it would have to be de- termined what number of subscribers could be attended to by a single operator. Each operator would have a calling-up board or set of connections to a limited number of subscribers. These subscribers would be able to call up this operator and no other. Calling-up connections for the entire number of subscribers TELEPHONY. 707 are arranged along the full length of the switchboard. The number that one operator can take care of are arranged within reach of the arm as the operator sits on a chair or high stool. For each subscriber there is only a single calling-up connec- tion. This portion of the switchboard is single, not multiple. Each operator must without leaving the chair be able to con- nect any one of the limited number of calling-up connections, which may vary in different cases, to any one of the subscribers in the whole district, who may be several thousand. In front of each operator is a panel of the board, with a connection on it for every one of the subscribers, and all these connections within reach of the arm. Corresponding in width with the panel is the row of calling-up connections. If there are fifty operators, there are fifty panels. Every connection of a given number on one panel is repeated there fifty times along the series of panels. This multiplication of panels constitutes the multiple feature of the switchboard. 12345 12 3 46 13345 12345 12345 6789 10 6T89 10 6789 10 6789 10 6789 10 11 12 13 14 15 11 12 13 14 15 11 12 13 14 15 11 12 13 14 15 11 12 13 14 15 16 17 18 19 20 16 17 18 19 20 16 17 18 19 20 16 17 18 19 20 16 17 18 19 20 21 22 23 24 25 21 22 23 24 25 21 22 23 24 25 21 22 23 24 25 21 22 23 24 25 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2 ) 21 22 23 24 25 The diagram shows the relation of panels to calling-up con- nections, and also indicates the multiple connections for identi- cal numbers of the series of panels. Each panel is shown with twenty-five subscribers' connections, and five calling-up connec- tions are shown for each panel. This is as if each operator was only called up by five subscribers, and as if there were only twenty-five subscribers in the district. In reality there might be several thousand connections on each panel, and fifty to two hundred calling-up connections for each panel. Each panel in- dicates one operator; as above shown, there would be five. The total number of subscribers divided by the number of panels gives the number of calling-up connections to one operator. The number of panels fixes the number of operators, and under each panel are the number of calling-up connections, each one for a 708 ELECTRICIAXS' HANDY BOOK, designated subscriber, which that particular operator must answer. The number of calling-up connections which one operator can attend to depends on the number of subscribers. When there are a large number of subscribers connected to a central station, each one will call for more connections in a day than if there were only a few. If a subscriber has six thousand co-subscribers at a station, he will call up more times in a day than if he only had one thousand or five hundred. Therefore, as the number of subscribers in the district served by the central station is larger, the calling-up connections assigned to each operator must be fewer in number. The number of operators required on a multiple switchboard does not increase in simple ratio. Doubling the number of sub- scribers exacts more than double the number of operators. The rate of increase approximates to the geometric ratio. The above simple description merely gives the outlines of the theory of the multiple switchboard. As it comes in practice when there are several thousand • subscribers to be included in every panel, and w^here the panels have to be consequently very numerous, the complication becomes enormous. There are a number of modifications designed to bring about more efficient working. Operation of Switchboard. — Calling-up connections of the sub- scribers on a multiple switchboard are operated by the sub- scriber. When the handle of his magneto is turned, or when the receiver is removed from the hook-switch as the case may be, a current is sent over the line. At the caliing-up connection on the switchboard in the central station this current operates some kind of annunciator to indicate the number of his station to the operator. The Mechanical Annunciator is a falling shutter or drop, seen in so many forms in ordinary house-bell annunciators, in hotel annunciators, and the like. A little shutter hinged at the base is held up in a vertical position by a catch or hook which holds its upper edge. The hook is operated by an electro-magnet. When the magnet is excited by a current passing through it, it attracts an armature TELEPHONY. 709 to which the hook is attached. This raises the hook, and the shutter at once drops. On its inner surface is painted or other- wise marked the number of the subscriber's station to which its circuit is connected. The current sent over the circuit by the distant subscriber drops the shutter and discloses his number. The operator replaces the shutter by hand as soon as the magnet releases its armature, so that the retaining hook drops to its lowest position. The drop is now ready for another call. Such annunciators are used in great number. An advance is made by having an electric system of replacing the shutter. It is considered that automatic setting of the annuncia^tors effects a saving of time and energy for the operator, who is often worked to the full extent of her power during the busy hours of the day. Where mechanical annunciators are used, the ten- dency is to use self-restoring drops. Lamp Annunciator. — The most advanced practice on switch- boards is to substitute incandescent lamps for mechanical drops. Eight- to twenty-volt lamps are used, one for each calling-up con- nection. A simple low-voltage lamp represents the maximum of simplicity and takes the place of the mechanism of the drop, inevitably more or less complicated. Lamps are cheaper than the modern self-restoring drops. They operate v/hen current passes, and cease when it ceases, thus presenting the self-restor- ing feature of the most improved drops without complication of the latter. Lamp signals are rapidly coming into use on the larger and more important switchboards. At first very low voltage lamps were used. These proved quite unreliable; they were very sensitive to slight changes in voltage, were hard to make, and burnt out very easily. Ten- to twenty-volt lamps are now frequently used. To produce the lighting current a storage battery at the central station is used. This gives an almost constant voltage. By having the lamps of reasonably high voltage, a drop or rise of a fraction of a volt has a much less effect on the duration of a lamp than when they are of only two to four volts potential. One-half of a volt on a two-volt lamp is twenty-five per cent of its voltage. On a twenty-volt lamp it is only two and a half per cent. 110 ELECTRICIANS' HANDY BOOK, In one system the removal of the receiver from the hook- switch throws the lamp into circuit with the station storage battery. The lighting current goes through the whole line, and through the transmitter at the subscriber's station. Several objections have been cited in reference to this sys- tem. The valid one is that if a cross occurs between lines, a very low resistance circuit may be produced, through which current will reach the lamp. The low resistance will operate to burn out the lamp. A very obvious way to dispose of this trouble is to put the lamp on a relay circuit. The calling cur- rent closes the relay, and the lamp is lighted from the storage Fig. 533.— Spring Jack. battery through the unchanged resistance of the local circuit. Apparently more complicated than the straight circuit system, the relay system avoids the necessity of adjusting the resistances of long and short circuits so as to give each of the lamps the proper current. Spring Jacks. — Connections to subscribers' lines on multiple switchboards are made by the agency of plugs thrust into spring jacks. Some boards have between one and two hundred thou- sand spring jacks. Fig. 532 illustrates the principle of construc- tion of one kind. The spring jack is screwed to the back of. the board, and a tube in its front projects through it. When the plug is withdrawn, the spring f rests on the contact screw p, and a closed circuit is made through the spring jack. When the plug is pushed into place, it pushes the spring up from the TELEPHONY. 711 contact screw, opening the circuit and connecting its own lead thereto. The front of a telephone switchboard appears full of holes, regularly spaced, and along the middle level appears a straight row of such holes extending its whole length. These are the spring jacks. On a five-panel board each subscriber will have Line Jacks Line Drops Subscribers Lines Clearing-out Drop Tig. 533.— Switchboard Connections. six spring jacks; five are on the face of the board and one is in the horizontal row. Various constructions of spring jacks are in use. They may act to give a simple metallic contact, or if the plug is in two divisions insulated from one another, a double connection may be made by plugging the hole. A flexible wire ("flexible cord") is connected to the plug. 712 ELECTRICIANS' HANDY BOOK. Switchboard Connections.— The diagram, Fig. 533, illustrates the work of a switchboard. Two subscribers' lines are shown entering an exchange, each including in its circuit an annunci- ator drop. When the subscriber by magneto or otherwise sends a current over his line, his special annunciator drop falls and discloses his number. Spring jacks such as have just been illus- trated are indicated, one for each line. It will be seen how in their closed position, Vv'hich is when the plugs are not inserted, the spring jacks act to complete ihe circuit through the annunciators to the earth at G. In the lov/er part of the diagram are shown the sv/itchboard connections. R and T are the receiver and transmitter used by the operator; B is the local battery, which v/ith the trans- mitter T is in circuit with the primary p of an induction coil. P P' are plugs to go into the spring jacks at the top of the cut. Assume that a subscriber desiring to use the telephone has by means of his magneto sent a current through his line. It drops the shutter of an annunciator, disclosing his number. The operator inserts the plug P' into the subscriber's spring jack. Each spring jack, it will be understood, has its number, that of the subscriber to whom it belongs. When the plug is inserted in the socket with the same number as that shown when the shutter drops, the subscriber is cut off from the ground at G, and is connected in circuit with the operator's receiver R and secondary S of his Induction coil to the ground at G'. Immediately on inserting the plug, the key K", which has been hitherto open, is closed, completing the circuit described. The operator by the transmitter T asks the number desired, and the subscriber tells it. The operator receives it by the receiver R. The plug P is inserted into the spring jack of the subscriber who is to be called up, and the key K is depressed. By work- ing the magneto M the bell of the second subscriber, the on? who is to be called up, is rung. When the second subscriber has answered, his answer being received by the operator's re- ceiver R, the keys K and K" are opened, and the subscribers ar^ in circuit with each other, and can speak together. The oper- ator can listen by closing or depressing the switch K". When the subscribers are through they both probably ring TELEPHONY. 713 off, although it would be just as good if only one did so. This sends a current through the coil of the magnet of the annunci- ator, called the clearing-out drop. Its shutter drops, showing that the conversation is finished. The plugs are pulled out of the spring jacks, and the lines are again ready for work. The proportion of plugs for a given number of subscribers is a matter for consideration. One pair of plugs for every ten subscribers is a proportion which in many cases is found advantageous. Lamp Signal System.— A simple presentation of the lamp signal annunciator system is given in Fig. 534. The sub- scriber's apparatus is shown in the upper part of the cut at B, and the central office con- nections in the lower part of the cut at C. At C, g and h in- dicate choke coils, I the annun- ciator lamp, and i a battery; g, Ji, and I are all in one metallic circuit. There is one such cir- cuit with numbered lamp for each subscriber. When the receiver is on the hook-switch, the circuit including the annunciator lamp I is closed through the high resistance of about 1000 ohms of the calling bell e. This cuts down the current so that the lamp I shows no light. When the receiver is taken off the hook-switch, this springs up and closes a circuit as it does so by coming in contact with two terminals above it, as shown at C. This short-circuits the bell coils. In the short circuit is included the secondary of the subscribers' induction coil. But this short circuit m.ay aggregate less than 50 ohms Fig. 534.— Lamp Signal Switch- board Connection. 714 ELECTRICIANS' HANDY BOOK. resistance, and the lamp is lighted. This calls the central sta- tion operator, who effects the desired connection, when told it by the calling subscriber. The local battery at the subscriber's house is a storage battery. When the hook-switch is released, this battery operates the trans- mitter. When the receiver is hung on the hook-sv/itch, the secondary battery is in closed circuit with the coils of the calling bell magnet, and receives a slight charging current of about 1/50 ampere. This keeps it in good condition for use. This description is of the simplest kind of lamp signaling system. There are many modifications, involving more compli- cated connections. If the local battery becomes too weak, the subscriber's trans- mitter will work with current from the central station battery, and the local battery will act as a sort of equalizer. In general practice, the calling lamp is on a relay circuit, and the relay closes its circuit when the receiver is taken off its- switch, by short-circuiting somewhat as described above. Conduction Interference. — Electric conductors such as line wires are sometimes subject to much trouble from induction and other electric disturbances. This is especially true of telephone lices. The telephone receiver is a wonderfully sensitive de- tector of any current change. It tells nothing if a constant cur- rent is passing through it, but reveals sudden changes in intensity of the current passing through it by producing a sound. Grounded circuits are peculiarly subject to disturbance. A grounded tele- phone circuit may be rendered quite useless by the presence in its vicinity of an electric trolley. The latter use the rails as their return circuit, and some of the current leaks into the ground and into grounded circuits in the vicinity. A grounded telephone line will sometimes sound in accord with the motions of the car motors. A part of the return current will go through the line, following the law of divided circuits. Induction Interference. — The above is a disturbance by con- duction. Sometimes induction from neighboring irregular cur- rents will affect a line. Insulation is without effect on induction, so whether the wire is insulated or bare, it will suffer disturb- ance as far as telephonic uses are concerned. A neighboring TELEPHONY. 715 telegraph line will act on a telephone line, so that its signals will be heard in the receivers. Such induction is usually treated as electro-magnetic. Experi- ments go to show that it is electrostatic. If a telephone receiver is placed in the center of a line, and one at each end, the end receivers will give a sound when a disturbing circuit acts on the line, while the central instrument will be mute. Even if the line is cut in the center, the two halves will give current changes which will make the end telephones sound. The use of metallic circuits does away with much of this f^ C . X I \x -^ Figs. 535, 536 and 537.— TEiiEPHONB Line Induction. trouble. If the disturbing line lay parallel with and between the two leads of the metallic circuit and equidistant from both, it would affect both leads equally and in the same direction, so that the two effects would neutralize each other. But in practice the disturbing line never occupies just such a position. One or the other lead is nearer to the source of disturbance than the other, and a disturbance results, which may be very annoying, particularly in telephone service. In Fig. 535 the heavy line indicates a circuit of varying cur* rent. The telephone circuit is seen parallel with it, with ona side nearer to it than the other. The nearer side of the tele- phone circuit will have the stronger potential impressed on it, and the result is indicated by the relative length of the arrows. 716 ELECTRICIANS' HANDY BOOK. The induced current will be due to the difference of the elecuio- motive force on the two leads of the telephone circuit. In Fig. 536 the effects of an inducing wire equally distant from both the telephone leads is shown. Equal electromotive force is impressed on both leads and in the same direction. Therefore no current is produced, and the telephones are un- affected. In Fig. 537 transposition is illustrated. The wires are un- equally affected because of their different distances from the source of disturbance. If the result is followed out on the dia- gram, it will be seen that the net result is the impressment of equal electromotive forces on both leads of the wire and in the same direction, so that they neutralize each other. In induction the polarity of the electromotive force induced constantly changes. The arrows in these diagrams illustrate the condition at one instant of the induction. When a^ number of lines are carried on one set of poles, the transpositions of the lines must not be the same for alL If the identical transposition were given' to all, there would be mutual induction. This induction is avoided by transposing the leads of the different circuits at intervals or at points varying for each pair of leads. Thus, two pairs of lines may be transposed at intervals of one mile for each case. To overcome mutual in- duction the places of transposition may vary, so that there would always be one-half mile between them. Other pairs could be transposed every half mile, and could also be varied in their places of transposition. Transposition on pole lines is effected by transposition insu- lators. These have two grooves. The wire is cut, and each end is turned about the insulator in its own groove. The same is done for the other wire of the circuit, and by short wires the rear end of one lead is connected to the forward end of the other, and the remaining ends are cross-connected in like man- ner. Twisting the leads of a circuit is much used. This secures comparative immunity from induction. In cables containing a number of pairs of wires twisting is extensively applied, and has been found to prevent induction. TELEPHONY, 111 Induction troubles are felt most on telephone circuits. Ordi- nary telegraph, power, or lighting circuits are relatively or com- pletely free from them. The cable construction companies endeavor to supply non-inductive cables, and have much success in their construction. Fig. 537a.— Pole Connections for Subscriber's Circuit. Subscriber's Pole Connections.— The method of taking a sub^ scriber's connection from a pole line is shown in Fig. 537a. A double-grooved insulator, such as referred to in the preceding paragraph, receives the ends of the line wire, which is cut at this point. From the ends a branch circuit is taken, as shown in the cut, two single-grooved insulators being provided, which take the strain off the main line insulator. Improvements. — No branch of electrical engineering is more subject to development and improvement than telephony. The utmost that can be done in these few pages is to give the outlines of what is a very complicated subject, much of who^e theory is largely unformulated. Automatic exchange systems dispensing in part or in whole with the central station operators are coming to the front, and if they ever reach full development, may exer- cise profound influence on the future of the business, by introduc- ing a different ratio of expenses to number of subscribers. CHAPTER XXXIX. BELL. WIRING. Bell Wiring is a class of work in which bad insulation leads to endless trouble. Size of Wire. — Cheapness often induces the use of undersized "wire. A small current will ring a bell, and the lengths of wire in a house are so short that the question of resistance hardly needs to be considered. Undersized wire is objectionable be- cause of its weakness. Wire stapled to joists under a floor, and led back of lath and plaster, seems out of all danger, but thin wire in house work will break and give much trouble. Circuits sometimes need changes; an extra bell, or more likely an extra p«ish button, is to be put in. Thin wire is far less easy to con- nect, because it is liable to break and give the work of extra sjjlicing to restore it. Wire in houses is sometimes cut by tacks or nails. Heavy wire has at least a better chance of escap- ing this accident than thin wire has. Nos. 16 and 18 American wire gauge are standard sizes. When No. 20 or even finer wire is used, the standard of the work is greatly lowered. The wire should be double-coated and paraf- fined. This makes it slippery, which is a great advantage, be- cause the coating is not so much injured by pulling around corners as a wire without paraffin in its coating would be. This is an incidental advantage. The first object of the paraffin is to improve the insulation and to exclude dampness. In putting wires into a finished house, they have often to be led up or down between studding and back of the lath and plaster. The processes used are called "fishing." In executing this, the greatest care should be taken to avoid dragging the wire around a sharp corner. If it is unavoidable, the paraffining helps to save the insulation. 718 BELL WIRING. 719 Fishing. — To run a wire behind lath and plastering, a space between studding must be found by sounding with a hammer. There is a slight difference, which to the practised ear discloses the hollow chamber or space. To lead a wire through it, a hole is bored through lath and plaster. A piece of Tery flexible string is used for the ''fishing.'* Well-waxed sail twine is ex- cellent. Sometimes fishing line is used. Waxing is advisable for it also. To its end a weight is attached, for which purpose a few inches of No. 19 double jack chain is recommended. The flexible chain can be pushed through the hole, and doubling down will go through a small space. Often studs on a brick wall are only an inch thick, so that the chain is excellent for such places. A half dozen spherical lead bullets, bored and strung like beads, are better than the chain. With the weight at its end the cord is fed through the hole and goes down until it reaches the desired point, provided all is clear. With a plumb line or by the sound of the weight on the end of the cord the line is located, and a hole is bored through the wall or surbase or wherever it may be to meet it. A piece of wire with a short hook is inserted, and the cord is hooked by it and drawn out; the bell wire is attached to it, and is drawn back by the cord. This principle takes care of all vertical and often of inclined runs of wire. The wire can be drawn downward from the other end by the same cord. Work Under Floors. — In running wire under the floor, a steel spring or flat wire % by 1/64 inch, with a hook at its end called a snake or fishing wire, is used. This can be pushed quite a long distance horizontally, and the string or the jack chain at its end, which has been dropped through a hole in the floor at some distant point, can be caught by a hook at its end and drawn back. If the beams run in the right direction or with the wire, it facilitates floor work greatly. If they run in the other direction, which is across the line to be followed, the wure must be taken through the beams one by one. A beam is located, and a hole is bored from the floor diagonally down from a point above its center. Floor beams are about a foot apart. If the string is dropped through a corresponding hole at the next beam, it is readily fished up to the surface of the beam in question. A 720 ELECTRICIANS' HANDY BOOK, second diagonal hole is then bored, through the beam, so as to form an inverted V, and the cord is passed down it, to be fished up from the next beam. The process and result is shown in the cut. Fig. 538, The hole must be nicely closed by putty cir plugs. For very particular work the holes must be kept as small as is consistent with gettting the cord and fishing wire through, them. This kind of work w^ould not be allowed for fine-finished hard- w^ood floors, unless possibly a joiner would undertake to close the holes so neatly that the plugging would be unnoticed. Many cases .would occur where this method would not be allowed. Fig. 538.— Fishing Bell Wire Under Floor. All sorts of expedients may be adopted. Houses differ from each other. Some have clear spaces running from plate to sill. If they can be found, a heavier weight, called a mouse, may be dropped at the end of a string, and thus one fishing will take a wire from base to top stcry. More will be learned by a few weeks with a competent man than by any description. Moldings may be removed and the wires put back of them, grooves being raced out for the wires, or a corner may be planed off the lower inside corner of the molding to give room for the wire. Racing is cutting a narrow groove in a floor or other wooden surface with a tool called a racing tool. It consists of a handle into which blades with hooked ends can be inserted. The groove BELL WIRING. 721 made is big enough to hold a wire. Sometimes wires can be laid in such grooves secured in place with tacks, but under ex- isting conditions of house construction and furnishing this is not so often allowable as formerly, when floors were of soft wood and were fully carpeted. Leading the Wires. — Exposed wires are used in some places, and are selected of color to match the paint or woodwork on which they lie. These can be stapled. The greatest care must be taken to keep them away from electric light wires. The distance between two parallel lines of bell wires should be half an inch, two wires never being put under one staple. Occasionally it may be necessary to adopt gutta-percha-covered wire for damp places, but this is not often the case. To splice wires, strip four inches of each and mak-e the regular telegraph lineman's splice, as shown on page 508. If a very good job is to be executed, solder each joint, using no acid, but only rosin or some non-corro- sive flux. The joints may be taped, but this is not usually necep- sary. If the joints are not well soldered, so that the solder fails to cover the copper, paper should be wrapped around them before taping. Grounding Wires. — Some inches of the end are stripped of in- sulation and brightened by scraping or otherwise. They are wound around a gas or water pipe, the part being scraped or sand-papered. The place should be soldered. It is good practice to solder all grounded ends of the same system to gas pipes alone or to water pipes alone, and not to solder some to water pipes and others to gas pipes. In case of disconnection of a pipe sys- tem, the grounds will still be good. Thus bells could be rung during repairs to plumbing or gas pipes. The removal of the gas or water meter removes the water or gas system from the ground in great measure. Soldering. — For soldering joints between wires a rather hard solder, one containing more than half its weight of tin, should be used. The soldering iron may be filed to the shape of a wedge with a groove filed across it, about Vs inch deep. The groove In the hot and well-tinned soldering iron is filled with solder. The twisted joining of the wires is dusted over with powdered rosin or other non-corrosive flux, and the groove full of melted solder 722 ELECTRICIANS' HANDY BOOK. is applied to its under side. The iron is rocked back and forth and ultimately turned completely around the wire, or else the wire itself is turned around while in the groove. Soldering joints is not universal, but it adds to the quality of an installa- tion. In all cases before joining wires use emery paper or some equivalent on the ends, so as to brighten them and remove cop- per oxide and dirt and secure good electrical connection. Solder will not take hold of a dirty surface. Wires. — Annunciator wire is double cotton-covered wire, with the cover saturated with melted paraffin. Office wire is a grade better in the quality of its cover. Sometimes wires are carried through tubes. As this brings them close together, wire of thor- oughly good quality of insulation should be used in this case. Distinguishing Colors may be used for different wires. This is a regular practice in other branches of work, and in bell work wires covered with different colored insulations can be used to distinguish the runs of wire. Otherwise, more or less frequent tagging of the wires can be adopted to make them readily traced through the house. As the tendency is now to use exposed plumb- ing, bell wiring should be done as much on the exposed order of work as possible. CHAPTER XL. ELECTRIC HEATING. Electric Cooking and Domestic Heating is possible because the current need only be turned on a few minutes before it is needed, and can be at once turned off. If it were kept on by the day, the expense would be prohibitive. Various utensils re- quire a certain period of heating before cooking can be begun with them. For an electric stove or griddle a period of 5 to 8 minutes is given; for a broiler, 12 to 14 minutes; for an oven, 20 minutes. The cooking operations proper are about the same as for coal fires. For boiling water, 15 to 20 minutes, and for heat- ing fiatirons, 8 to 12 minutes are required. The cost of electric cooking in one experiment was found to be about five times that of coal cooking. All such figures are approximations only, as circumstances vary so greatly. Power Required for Cooking. — A small broiler, 6 by 8 inches in area, will require from 340 to 400 watts, a 1%-pint kettle a little less; a 16-quart kettle, 1140 watts. A full electric range of 6 square feet area consumed 1650 watts per square foot of surface. Efficiency of between 80 and 90 per cent can be attained in boiling water. Electric Furnaces may be divided into two classes. In one the voltaic arc is the heating agency; in the other class incan- descence is the principal source of heat. With many materials both arc and incandescence may operate simultaneously. The illustration, Fig. 539, gives a cross section of a simple electric furnace. The square box or case may be of iron. It is lined with some insulating refractory substance, such as lime or mag- nesia. Carbon rods pass through holes in the box, and are in- sulated therefrom as shown. To operate such a furnace a strong 7x'3 724 ELECTRICIANS' HANDY BOOK. alternating current is required. The carbons are connected in the circuit, and an arc is started across the interval between them. This may be done by pushing the carbons together, and thus closing the circuit. They are then drawn apart, and the arc ^^^ ^ Pig. 53S,— Open Elzctric Furnace, Fig. 540.— Closed Electhio Furnace. is thus ''struck" or formed. Material to be operated on is placed in the cavity, and as it reaches the level of the arc becomes heated by it. Although this describes arc heating, it may often happen that w^hen the material reaches the level of the carbons it conducts the current, and the furnace operates by incandescence. YxQ, 541.— SiEMENS'S Electric "itJRN^CE. In Fig. 540 is shown an advance on the last. It is a covered furnace adapted to receive a vessel to be heated by the arc. In this apparatus, where the substance to be heated and the arc are distant from each other, there is no question of incandescence. The heat is due to the arc. The furnace shown in Fig. 541 is virtually a lined crucible through whose sides two electrodes project. The electrodes are mounted so that they can be run in and out, thus varying the ELECTRIC HEATING, 725 length of the arc. Worm gear is provided for this purpose. One electrode is a carbon tube, the other is of metal and hollow, and water circulates through it, introduced by a pipe placed on its axis and reaching nearly to its end. In the furnace shown in Fig. 542 a brick structure filled with Fig. 542.— Cowles's Horizontal Furnace, Fig. 543.— Cowles's VerticaIj Furnace. Fig. 544.— Mechanical- ly Operated Electric Furnace. non-conducting material, such as sand, B, holds a retort A. At the left end is a carbon electrode C, and at the other end is a carbon crucible which acts as the other electrode. The crucible D is perforated at d to permit the escape of any gas generated in the reactions. The furnace of Fig. 543 has a hopper, through which material to be acted upon is introduced. From the bottom of the hopper a 726 ELECTRICIANS' HANDY BOOK. tubular electrode extends downward, and a second one rises from the bottom, sb as nearly to meet the other. Material can also be introduced from outside the upper tubular electrode. Gases which escape are condensed or cooled in a condenser, indicated to the right of the furnace. As the charge melts it runs down the central opening of the lower electrode and is withdrawn. In Fig. 544 is given a section and plan of a more complicated furnace. In this structure the upper electrode can be moved not only up and down, but its end can be swung about over the area Fig. 545.— Verttcal Carbon Furnace. Fig. 546.— The Electric Blowpipe. of the crucible below it, so that all parts of the charge can be sub- jected to its action. The crucible is below the end of the carbon electrode, and forms itself the lower electrode. It is carried on trunnions, so that it can be turned down for pouring out its contents. Another simple form of furnace is shown in Fig. 545. The crucible with its lining forms one electrode, and a carbon rod de- scends into its center from above, constituting the other electrode. Material enough may be added to cover the charge acted on and to supply new material as the materials melt down. The electric furnace is a very simple thing. The factor abso- lutely necessary for it is plenty of electric power. The furnace ELECTRIC HEATING. 727 used in the manufacture of carborundum has sometimes Deeu little more than a pile of coke covering the charge and held in place by a loose brick wall. Carbon elec- trodes entered the ends, and the current acting by incan- descence heated the charge to white heat. The Electric Arc Blow- pipe. — The voltaic arc is re- pelled when a magnet pole is brought near it. This prin- ciple has been applied to producing an electric blow- pipe, in which the arc driven to one side, as shown in Fig. 546, is used like a blowpipe flame for local heating. Fig. 547.— Electric Arc Heating. Direct Heating by the Electric Arc is carried out by making the object to be heated one of the electrodes of the arc. Thus, a boiler is shown in the cul, Fig. 547, as under treatment by the arc. One conductor is connected to it, the other is connected to a carbon rod carried in a holder and held over the point to be heated. An arc is caused to form, and is brought where de- sired by moving the carbon pencil over the spot. A colored glass screen pro- tects the eyes of the operative. The carbon holder has a handle with shield to protect the hand, something like the hilt Pig. 548,— Diagram of Arc Heating. 728 ELECTRICIAXS' HANDY BOOK. of a fencing foil. A general diagram of the connections is given in Fig. 548. There is a storage battery, with a connection box U, by which the number of its cells supplying current to the arc can ^s^nnnmnn :xii=3:I^In> Figs. 540 and 550.— Electric Soldering Irons. be increased or diminished. W is a resistance frame, and the car- bon holder is at K. V is a voltmeter, and A is an ammeter. Be- low is seen the carbon holder with its protecting shield. The Electric Soldering Iron, Figs. 549 and 550, uses less than red heat. The first figure shows the external view of one, and the lower figure is a section. The copper bolt is surrounded by a coil of wire insulated by fire- proof insulation, which on the passage of a sufficient current keeps the bolt at the proper tem- perature for soldering. Electric Welding, the princi- ples of which are shown in Fig. 551, uses the heat of direct in- candescence. An induction coil or transformer has two coils, a high- and a low-tension one. An alternating current is passed through the high-tension coil, which induces in the low-tension coil a much more intense cur- rent than itself, but impresses a much lower voltage on the same circuit. In the cut the high-tension coil is the inner one lying flat on the paper, and the simple bar of iron outside it is the Fig 531.— Electhio Welding. ELECTRIC HEATING, 729 low-tension coil. Wires are seen leading to the high-tension coil. These are connected to the source of supply. Two heavy coils of iron wire surround both coils and act as core. The pieces to be welded are held in the clamps as shown, and are rapidly V=V-A-^=^ FiQ. 552.— Electric Incubator. heated by the induced current. By the screw they are forced to- gether so as to weld. Almost any conducting metal can be welded by this process. Very remarkable results have been attained with various metals and shapes. Fig. 553.— Electric Kadiator. The Electric Incubator, Fig. 552, is a curiosity in electric heat- ing. A basket holds eggs and has a cover which contains a coil of wire, through which a current of electricity passes. By a ther- mometer the temperature is watched, and regulated by a resist- ance coil. The young chickens are kept in a coop which contains a heater to represent the mother hen. Both are shown in the cut, each with a thermometer on top. 730 ELECTRICIANS' HANDY BOOK. Electric Radiator. — Many forms are made, consisting of long wire conductors which may be covered with asbestos insulation. The cut, Fig. 553, gives a simple form in which the conductor is carried up and down over studs on a frame. Iron and steel wire are good materials for the conductor. Their principal use is for heating electric cars. Economy of Electric Heating. — When electric power is pro- duced by a steam plant, the loss of energy is very great. By a law underlying the operation of heat engines, of which the steam engine is the most conspicuous example, by far the greater part of the potential energy of the fuel is wasted. From 90 per cent upward of the heat of the coal burned is lost. The law is termed the second law of thermodynamics. Under these circumstances the efficiency of electric heating is necessarily low. On the other hand, when water power is used for its production, it may be very efficient. Its economy when produced by a steam-driven plant is low on its face, but is relatively high when intermittent heating is in question. The current can be cut' off so readily that long periods of useless expenditure of fuel, inevitable in many cases of heat- ing with, coal, are avoided. The economy thus brought about compensates for the low efficiency explained above. The electric heating of trolley cars is possible because the power is rather advantageously produced and the repairs of stoves are avoided. CHAPTER XLI. WIRELESS TELEGRAPHY. Wave Transmission of Signals.— The ocean is thrown into waves by the motion of the air or winds. The particles of water in making the waves constantly move in vertical circles, round and round. The diameter of the circles is several times the height of a wave. The particles of water do not move forward or backward except through a limited range, and a wave on the deep sea does not transfer or carry water along with it. There is no displacement. A man at one side of a pond of still water could send a message to another by waves, if there was any good way of detecting them. The constant reflection and repe- tition of the waves would occasion trouble. But on a large stretch of water a sharp impulse given might send waves of water, which could be detected at a considerable distance. Such waves could be used to transmit messages. The air is thrown into waves of another type by the vibrations of bodies, and transmits sound. As air is much lighter than water, air waves travel at much higher speed than do water waves. About a thousand feet a second is the rate of sound transmission by air waves. The luminiferous ether is thrown into waves by various kinds of dis- turbances, electrical among others. Ether waves are transmitted with a velocity which would take them around the earth nearly eight times in a second. Air waves are the medium for propa- gating sound, such as the human voice. By wireless telegraphy ether waves are produced at one place and detected at another, and are made to transmit intelligence by the Morse code or some equivalent. The waves used are called Hertz waves, from the cele- brated Prof. H. Hertz, an early demonstrator in this field. Their existence was predicated on Clerk Maxwell's celebrated electro- magnetic theory of light. 731 732 ELECTRICIANS' HANDY BOOK.: ' If a discharge is produced between the terminals of an induc- tion coil, a spark as it has long been called is produced. In reality this is an enormous multitude of discharges or sparks beating back and forth with decreasing intensity, but uniform frequency. The time occupied by the multiple discharge is very short, but the dur- ation of a single element of it is in the second order of duration, and is almost infinitesimal. The discharge beating back and forth is called an oscillatory discharge. The time in fractions of a second of a discharge is calculated by the formula. T = 27r 'x/KL; K being capacity and L inductance of the circuit. The oscillation in Hertz oscillators, as the special circuits for these experiments are called, varies from 10,000,000 to 300,000,000 per second. If it were possible to increase them to a sufficient fre- quency, light would be the result. The trifling light given by the spark is due to the heat of the discharge, not to its oscillations. Hertz, Receiver. — The oscillator transmits waves. As a receive:: Hertz used a broken circle of copper wire. The diameter of the circle v/as about 16 inches. It terminated in little metal balls or knobs, whose distance apart was adjustable. When the oscillator was discharged, a minute spark passed across from ball to ball ol the detector or receiver, when everything was in adjustment The receiver only operated at a short distance. At more thau the length of a room the effect was too attenuated to produce r.. spark in the detector. Braniy's Coherer. — This investigator found that loose meta.\ filings were astonishingly ?^ensitive to ether waves of slow fre- quency, such as produced by oscillatory electric discharges. A tube containing loose metal filings and of relatively high resist- ance had its resistaDce greatly diminished by being held near the place where such ether waves could reach it. The ether waves make the loose filings take up a new condition and act in a de- gree like solid metal. As the molecules of every solid metal co- here, the tube of filings is appropriately termed a coherer. When once cauj^ied to cohere, the filings remained so until disturbed by agitation or otherwise. If such a tube is placed in circuit with a battery and relay, ether waves will by reducing resistance close the relay. If they cease, then tapping the tube will increase the resistance and open the relay. WIRELESS TELEGRAPHY. 733 Wireless Telegraphy is based on the production of an oscil- latory discharge at a transmitting station, its transmission by ether waves through space, and the detection of the waves due to the discharge at the distant station. Originally only the coherer v/as used. It is still in extensive use, although many other receiving instruments have been invented. It is now sharing the work with other devices more rapid in action. Transmitting Apparatus. — The principle of the Marconi trans- mitting apparatus is shown in the cut, Fig. 554. One or more vertical wires, W, are supported by a mast or other support. The lower ends, if there are several wires, are joined together and are connected to one ball, cl, of a spark gap. The other ball, d, is connected to the earth. From d and d wires c' c' are carried to an induction coil c. The ])rimary of the coil with key & is in circuit with a battery a. On depressing the key, an oscillatory discharge takes place across the gap, and by charging and discharging affects the whole length of the vertical wire. Ether waves go off from the wire through space, with a general tend- ency to follow the curvature of the earth. They travel best over water, so that the ocean is peculiarly adapted for the use of wireless telegraphy. Marconi, beginning with vertical wires only twenty feet long, sent signals a mile. He found that increasing the length of the wire increased the distance of transmission, and the rule of the distance varying with the square of the length of the wire was at one time suggested, but has been abandoned. Receiving Apparatus. — At the distant receiving station the system of antenna is established, whose lower end is grounded with the primary of an induction coil in series. The secondary of the induction coil is in series with a battery and relay mag- net. In parallel with the battery and relay magnet is the coherer. Fig. 554.- Principle of Transmitiikg Apparatus. 734 ELECTRICIANS' HANDY BOOK. Connection of Stations. — Referring to Fig. 555, 1 is the trans- mitting station with its antenna Ai, spark gap h &, induction coil secondary Si and primary P^, sending key K, and battery B,. On working the key, sparks pass between h and &, affecting the antennae. Ether waves fiy through space and are caught by the antennae A., of the receiving station 2. The disturbance sends a momentary current through the primary P.. of the in- duction coil. Its secondary S2 then sends a current through ^ZB2 Fig. 555.— Wireless Telegraphy CoxxectionS. the coherer c. This reduces the resistance of the coherer, and a current goes through it due to the battery B. and through the relay magnet R, operating a Morse receiver on local circuit. There is a hammer which actuated by an electric magnet and make and break constantly taps the coherer, so that the coherer only retains its conductivity while acted on by ether waves. The instant they cease, the tapping restores its resist- ance. Long and short signals for the Morse code are sent by holding down the key K at the transmitting station for long or short periods. They are received by the receiving station and printed on a tape. WIRELESS TELEGRAPHY. 735 Marconi's Coherers. — Marconi's coherer, shown in Fig. 556, is a tube 1^ inches long and 1/12 inch internal diameter. A chamber is made in the center by introducing two silver plugs with their ends 1/30 inch apart. A mixture of 9 per cent nickel filings and 10 per cent silver filings is contained in this space with a minute quantity of mercury. Hysteresis and Other Receivers. — The coherer used as a re- ceiver operates a relay circuit, and prints the message on a tape in Morse characters or their equivalent. This has the advantage of giving a fixed record. There are a number of receivers which do not give a record, some of which are based on magnetic lag. The hysteresis of iron is modified by ether waves impinging on it. In one of Marconi's receivers an end- FiG. 556. — Marconi's Coheeer. less iron wire is stretched around two pulleys, and passes through the core or axis of a double coil of insulated wire. The arrangement represents an induction coil with moving core. The primary receives the impulses from the receiving antennae as already described. The secondary connects with a telephone receiver. The impulses modify the hysteresis of the moving core, and sound is produced in the telephone. In another construction the core of the induction coil is fixed, and an electro-magnet rotates in front of it. The ether waves modify the hysteresis as in the case just cited, and* the message is received by a telephone. There are a number of other constructions of receiving in- struments in which a telephone is used as a receiver, the great sensitiveness of the telephone receiver causing them to be operative. The action of the ether waves on these classes 'of instruments is so slight that the instruments can only be 736 ELECTRICIANS' HANDY BOOK. used with a telephone receiver, and cannot actuate a printing recorder. On the other hand, the acoustic instruments are faster, other things being equal. Detectors. — V/hen an impulse from a distant station is re- ceived on the receiving antennas or aerials of a station, it has to be rectified or otherwise affected in order to be heard on a telephone. A number of instruments have been in- vented to do this, and the name detector has been given them. The currents dealt with are so exceedingly minute that the construction of rectifiers is based to an extent on principles which would not be applicable to large currents. The name detector is also applied to coherers and other apparatus and appliances, and even to simple spark-gaps, in which there is no rectification of current. Italian Navy Coherer. — This consists of a small glass tube with two iron plugs, between which plugs is a drop of mer- cury. The mercury only makes good contact with the iron when the ether waves act upon it. It is self-decohering. It resembles the Branly instrument, and is used in the same relation to the aerials. The Liodge-Muirhead Coherer. — This is also self-decohering or restoring. A small steel wheel with a sharp edge is rotated by clockwork, and just touches the surface of a globule of mercury, which is held in a cup on the top of a brass pillar. A thin film of oil is maintained over the surface of the mer- cury. The oil insulates the wheel from the mercury so that no current can pass, unless the apparatus is excited by a dis- charge. The apparatus is put in a local circuit, as in the case of any coherer. When electrical oscillations are set up in the receiving aerials, the insulation of the oil breaks down and a local current can pass and actuate the receiving instrument. The constant rotation of the wheel decoheres it when the Impulses cease, its rotation representing in a sense the tapping of the Branly coherer. No relay is required with this instru- ment; the resistance is so low that it can operate a recording apparatus directly. The Stone Coherer has steel plugs with loosely packed carbon granules between their end faces. It is self-decohering.' ^VIRELESS TELEGRAPHY. 737 It is lacking in sensitiveness, but is well adapted to be used in, portable outfits, as it is not easily put out of order or ad- i%\stment. The Fleming Valve Detector. — A small incandescent lamp hivs a cylinder of copper surrounding its filament; it is kept incandescent by an independent circuit. Besides this the nega- tive terminal of the filament is connected to one terminal of the telephone receiver. The other telephone terminal con- nects to the copper tube by a platinum wire sealed into the Fig. 557. — Fessexdex Detector. lamp bulb. When acted on by the discharge from the aerials the copper cylinder is alternately charged with positive and negative electricity. Electrons are constantly being emitted . from the filament; they are repelled when the cylinder is at negative potential, but when at positive potential they are at- tracted by it. This attraction and reception of electrons con- stitute an electric current, and from what has been said it is obvious that it can only pass in one direction. Hence it is a true rectifier, and gives a current whicb can operate a tele- phone. Electrolytic Detectors. — There are a number of these based on the same principle. The following gives the general fea- 738 ELECTRICIANS' HANDY BOOK. tures of the Fessenden detector. An exceedingly fine platinum wire is sealed into a glass tube so that only its minute end is exposed. This is immersed in a solution of 1 part sulphuric acid to 5 parts of water. A globule of mercury or a plate of silver lies below the platinum also in the acid. This consti- tutes a small electrolytic cell; silver or mercury and platinum have each their own terminal, and it is connected in series with the receiving telephone and a source of current. The platinum point is at once polarized. When the oscillations from the receiving aerials are passed across such a cell they depolarize the platinum electrode. This enables a current to pass, for as long as the platinum was polarized it cut off the passage of current. Thus the successive impulses of discharge depolarize the platinum and current passes; when they cease the polarization takes place again and current ceases. The telephone answers to the changes and gives a note, and can receive a message. It is believed that the action may be more complicated than the above; resistance, charge and counter- electromotive force it is believed may play a part. Heterodyne Detector. — This device is also due to Fessenden. It is based on a very simple and obvious principle. If the discharge from the aerials is received in a telephone it is of such high frequency that no sound is produced. Now if simul- taneously with it a second coil on the telephone received a local current of slightly less or greater frequency, inter- ference would result and the beats could be made to occur at any desired frequency, so that the telephone would react to them. Thus suppose the discharge had a frequency of 50,000 per second. This would be without effect on a telephone. Suppose that the second current had a frequency of 49,000 or of 51,000 per second; the result would be the production of 1,000 beats per second. This is a good telephone pitch. Thus as long as the discharge lasted the telephone would sound by the beats of the interfering waves. When the current alone passed the telephone would not speak. This instrument has very high power as regards .receiving from great distances. Goldschmidt's Tone Wheel is a mechanical rectifier. It is simply a toothed wheel, and the discharge goes through it by WIRELESS TELEGRAPHY. 739 brushes making contact with its teeth. It is obvious that it is a simple matter theoretically to adjust its speed of rotation so that there is contact for each wave. When the brushes are between the teeth no current can pass; therefore only half the current can pass, and the adjustment is such that this is all in one direction. The other portions of the waves are cut off. This receiver has been used successfully on transatlantic work. The adjustment to a given high frequency is far from simple practically. But if a slight slip is allowed to the wheel, it is obvious that it will work in a manner analogous to that of the Fessenden device just described. It will gen- ^---i—.j^ Receiver Circuits -> Fig. 558. — Heterodyne Detector. erate a wave of good telephone frequency when the difference in frequency between it and the discharge is correctly deter- mined. Crystal or Contact Detectors. — There are a number of these used most extensively in practice. They are rectifiers, based on the principle that many natural or artificial minerals will only pass current in one direction. Suppose a crystal of carborundum, artificial aluminum silicide, is placed in the circuit of a current. In one direction the current may be 100 times stronger than in the reverse direction. Silicon, galena, iron pyrite, zincite, and molybdenite are some of the sub- stances which have been adopted for practical work. There 740 ELECTRICIANS' EAXDY BOOK. are all sorts of details in the connections of these detectors, but the principle is simplicity itself. They give directly a one direction current for the telephone when acted on by the oscil- lating discharge from the aerials. Spark Gaps. — There are many kinds of spark gaps. One of the simplest of the commercial ones consists of a pair of hemi- spheres, often made of zinc, and protected from leakage by Fig. 559. — Single Crystal Detector. Fig. 560. — Double Crystal Detector. discs extending from their peripheries. To adjust them, their distance apart is varied, and to prevent damage to the appa- ratus a pair of terminals at a fixed distance are arranged directly below them, between which a spark may pass, if the tension rises to too high a point. The protecting discs are not found on all this class of gaps. The Disc Discharger is a rotating disc, with teeth. On each side are electrodes, and sparks pass as the teeth go by the electrodes. The apparatus is quite complicated, the above giving merely the basis of its construction. The Telefunken WIRELESS TELEGRAPHY, 741 quenched spark is produced by a series of discs, arranged cyl- inder fashion, insulated from one another by mica. The discs are of copper faced with silver plates, and the discharge goes through the series as a set of sparks, which may be eight in series. T SERIAL ^^ Z^ L AERIAL sm«IBKr ^ V VE.RTICWL. aEPIRU V acBmL U MBA EUL A RETRIAL FfJN AERIAL Fig.. 561. — Aerials or Antenna. Aerials or Antennae. — There are some six prominent classes of aerials. The illustration shows them and is self-explana- tory. The arrowheads indicate the grounding. In some large stations most elaborate and complete grounding is provided by burying zinc plates and connecting the leads thereto. The higher an aerial is the more power will it have as regards 742 ELECTRICIANS' HANDY BOOK. distance or range of action. The same applies to size; the larger it is the more powerful will it be. The all-important thing is that there should be plenty of metal surface, and stranded conductors are used to insure this. Bronze or copper wire is generally employed. The distance apart of the con- stituent wires is important to avoid self-induction in the hori- zontal leads; a distance or interval of five or of ten feet is standard practice. The highest power of transmission is in a direction parallel to the horizontal component of the aerials. By carrying out this feature to the greatest possible extent, directional systems are produced, which tell the direction of Fig. 562. — Magnetic or Inductive Fig. 563. — Electric or Capacity CorPLiNG. Coupling. the transmitting system. This is, among other cases, pecu- liarly applicable to airplane navigation over the ocean. Couplings are the connection between the two or more ele- ments into which a receiving or a transmitting circuit can be resolved. Of these two parts one is in each case the aerial or antenna, the other is the circuit containing receiving or transmitting apparatus. These divisions are called also sys- tems, and coupling is defined as the arrangement of two systems, so that oscillations in one of the systems always causes oscillations in the other. Conductive Coupling is when the two systems are in actual! metallic or conductive connection. It is also called DirectI coupling and Galvanic Coupling; a pure conductive or galvanic coupling must have no mutual inductance between the two systems. Magnetic or Inductive Coupling is so arranged that the only connection between the two systems is by mutual induction. WIRELESS TELEGRAPHY. 743 This connection is established by means of two coils with par- allel axes, one in each system, which coils react on one an- other when there is any change in the magnetic field of either one. It is shown in the diagram. Electric or Capacity Coupling is obtained by having one or more condensers common to the two systems. An electric dis- turbance in the one system reacts upon the other by the agency of the condenser or condensers. This is also shown in dia- grams. Combined Couplings are produced by having two of these couplings in the same system. Loose and Close Couplings. — These are terms of a descrip- tive nature indicating that couplings are so arranged that in the one case (loose coupling) a slight reaction only occurs between the two systems, and in the other case (close coup- ling) a strong reaction occurs between the two. The strength of the reaction is increased in a magnetic coupling by bringing the inductive elements of the two systems closer together, and vice versa. In a galvanic coupling, the larger the common portion of the two systems the closer will their coupling be. Marconi Sending Plant. — The connections used in this plant are given in the cut. A is an alternating current generator. K is the sending key; B is an induction coil with iron core, to tune the system to the frequency of the current. C is a step- up transformer, transforming the voltage of the alternator up to 15,000 or 20,000 volts. This produces an inductive coup- ling. Di, D2 are choking coils, without any cores, designed to protect the transformer from high frequency oscillations. Next comes the oscillation circuit, which includes the rotary spark-discharger, Ej which regulates the frequency of the system from 50 to 300 periods per second. It also contains a variable condenser, indicated by the two parallel lines with a diagonal arrow, and a variable inductance, F, which is used to tune the circuit to the desired frequency. Another trans- former makes inductive coupling with the aerials, which in their turn are provided with a tuning inductance, H", and an earth arrester, 0. The latter, shown in the cut, is essentially a short spark-gap, too short to insulate the antennae in send- 744 ELECTRICIAXS' HAXDY BOOK, ing, on account of the high potential of the circuit, but which resists the passage of the oscillations as received from the distant station from going to the earth. Thus it acts as a _ rH, Spark Tuning *^^ >r Discharger ^Inductance *^V^ Fig. 564. — Marcoxi's Sexdixg System. sort of valve. The sparks pass from the outer edge of tKe upper plate to the surface of the lower and larger plate. The sparks are about 0.01 inch long, their length being determined by the thickness of the mica. Fig. 565. — ^]\Iaecoxi's Eakth Abrester. Marconi Receiving Plant. — This portion of the apparatus is connected in shunt to the earth arrester, G, and after what has been said the cut will be almost self-explanatory. H is the WIRELESS TELEGRAPnY, 745 aerial variable inductance, C is a variable condenser, D a vari- able inductance, and £7 is a transformer, giving a coupling to the next circuit, X, which circuit in its turn is coupled, also Earth Arrester Fig. 566. — :Maecoxi's PvECeiving System. inductively, to the listening or receiving circuit, F. F and J are adjustable condensers, and I is the detector. The listening telephone is not shown; it may be connected in parallel with 746 ELECTRICIANS' HANDY BOOK. the detector. 5 is a choking coil which takes off static charges from the antennse. Enough has been given to enable the reader to form a general conception of wireless telegraphy. The subject is one of great extent, its development is going on rapidly, and it is in a constant state of change. The mathematics of the subject are enough to constitute a treatise. It cannot be adequately treated in as few pages as we are at liberty to devote to it here. The same applies to wireless telephony. Wireless Telephony. — A number of systems of wireless telephony have been invented, based on various principles, and so numerous that it is beyond the scope of this book to do more than give an outline of some of the basic principles underlying the typical ones. Suppose that a stream of radiations is emitted from the antennae of one station, such radiations to be constant in am- plitude and in frequency, the frequency to be several thou- sands per second. Such emanations received by a receiving station will produce no sound in a telephone, although the diaphragm will respond to them, as the pitch will transcend the range of audibility of sounds. A microphone is connected in the sending system in such a way as to change the resist- ance or the capacity of the antennse, directly or by magnetic or static coupling, as such microphone is acted on by the human voice. At the outset there are two different ways of affecting the emanations; the amplitude of the waves ema- nated may be made to vary, or their wave length may be altered by the action of the microphone. The emanations are changed by the action of the microphone from a steady stream to a varying one, whose variations are proportional to the vibrations impressed on the microphone by the voice; therefore when this varying stream of emanations is received on the distant antennse of a receiving station, the variations im- pressed on the telephone connected to the antennse reproduce the words spoken at the transmitting station. The microphone may be used in a large variety of ways. It may change capacity or resistance of the antennse, it may vary the excitation of the generator supplying the basic cur- WIRELESS TELEGRAPHY. 747 rent, when such is used; if an arc is used to give the basic current, it may affect the action of the arc in a number of ways; and the action may be such as to simultaneously affect the amplitude and the frequency of the current. Sending Arrangement for Wireless Telephony. — Referring to the cut M indicates a microphone, connected in series with the antennae. A continuous arc is maintained as shown to the left of the diagram, between a copper and a carbon elec- FiG. 567. — Sending Arrangement for Wireless Telephony. trode in a magnetic field and in an atmosphere of hydrogen gas. This is the Poulsen arc. The resistance of the micro- phone varying with the action of the speaker^s voice, impresses the steady oscillations due to the arc with the varying oscilla- tions due to the voice, so that a set of speaking undulations is sent out through space, to be picked up by a distant station. The coupling as shown at L, is an inductive one. The micro- phone could be placed in the arc-circuit, it could be used to change the action of the arc, and could be made to produce a speaking emanation in several other positions or connections than the one indicated in the cut. Thus C indicates a con- 748 ELECTRICIAXS' HAXDY denser which determines the capacity of the arc circuit. The microphone can be so constructed as to change the capacity of the circuit it is connected in, and thus to statically affect the undulations of the emanations and produce a speaking current. Receiving Arrangement for Wireless Telephony. — In general principles this is simple; there is no question of variations Fig. oG8. — Eeceivixg Aeraxgemext for Wireless Telephoxy. such as are met with at the transmission end; the cut gives the outline of the arrangement, which, of course, is subject to various changes. The antennae receive the undulations; by inductance at L they are impressed on the coupled circuit as indicated in the cut; O is a condenser, and in parallel with it is a circuit con- taining a detector, T, and a telephone, B. The listener at the telephone hears the distant talker at the transmitting station. CHAPTER XLII. METALLIC FILAMENT INCANDESCENT LAMPS. Metallic Filament Lamps. — Various metals have been em- ployed in incandescent lamps. Platinum was one of the first metals tried, but it was too fusible. Tantalum met with some degree of success, but tungsten, a metal of the iron group, is the survivor of all. One great advantage of metallic fila- ments is that the illuminating power of the lamp is not affected by changes in voltage to anything like the same extent as in the case of the carbon filament lamp. A change of one per cent in the voltage supplied to ^ carbon lamp effects a change of six per cent in the light. If a carbon filament is charged with a metal, or metallized, as it is termed, the corresponding change is less than five per cent. It is still less in the case of a full metallic filament. Tvingsten Filaments. — Tungsten is a metal of the iron group, very infusible and abundant in nature. Several ways of making filaments of this metal have been employed. The starting point is powdered tungsten. It may be mixed with a moist binding material, squirted through a die, and heated in a reducing atmosphere, such as hydrogen. Another process is to heat it to a high temperature and weld the powder together by pressure or percussion. In the amalgam process 40 per cent of the powdered metal is mixed with an amalgam of equal parts of cadmium and mercury, the mixture is forced through a die, and heated to expel the cadmium and mercury. The wire thus produced is quite fiexible and ductile. The electro-plating process is based on coating the wire with gold, silver or some other metal, which protects the tungsten from oxidation during treatment. Finally the thorium process, fol- lowing the lines of the Julian Pintsch invention, is now much 750 ELECTRICIANS' HANDY BOOK. used. The ordinary tungsten filament is very brittle and crys- talline. To apply this process the tungsten oxide is reduced to the metallic state by ignition in hydrogen. The powder thus obtained is mixed with thorium oxide; two per cent is specified in the patent. A very little binder is added and mixed with it, the soft mixture is squirted through an aperture or die, is dried and passed slowly through a tube. A current of hydrogen gas is maintained through the tube, and in the tube is a coil of tungsten wire, kept at a high heat. The fila- ment is passed very slowly through the length of the tube, and as it emerges therefrom is of high tensile strength and has lost its brittle, crystalline nature. It is of 164 kilogram tensile strength per square millimeter. The thorium seems to lengthen the crystals. Sometimes as much as 50 per cent of thorium oxide is em- ployed. An addition of 20 per cent operates to increase the resistance of the resulting filament 50 per cent. More than 50 per cent of thorium oxide gives a high resistance material which is brittle, but which makes it possible to employ a short filament. One square millimeter of surface of the tungsten filament gives 0.5 candle at 0.55 watt to the given area. The carbon filament gives 0.182 candle at 0.63 watt. The light given by the carbon filament varies with the 6.3 power of the voltage; that of the tungsten filament with the 3.6 power of the voltage. The dies for drawing tungsten wire for filaments are now made of high-speed steel or of diamonds. Tungsten is the most refractory of all metals, but it melts at a temperature far below that at which carbon fuses, and far above the temperature at which carbon volatilizes. In general the limiting temperature at which metal filaments can be used depends on their melting points, and that of carbon on its volatilization point. Once a metal filament lamp bulb begins to blacken the process is very rapid, although it is gen- erally a long while before it begins. Carbon on the 'other hand is apt to begin to obscure the bulb earlier and to take a much longer time to do it. Gas-filled Incandescent Lamps. — In the early days of elec* METALLIC FILAMENT INCANDESCENT LAMPS. 751 trie lighting attempts were made to use a glass bulb or tube filled with nitrogen gas to hold the incandescent carbon rod or filament. This was discarded in favor of the vacuum, which for many years was in universal use for carbon filament lamps. To-day it is used for both metallic and for carbon filament lamps. For the latter the vacuum is the only available way of preventing the burning of the carbon. Tungsten filament lamps now are frequently filled with gas. For large lamps nitrogen gas is used; for small lamps argon is sometimes em- ployed; a little nitrogen is mixed with it to help it to resist a discharge. Argon has a lower conductivity for heat than nitrogen, and therefore prevents the bulbs getting over- heated. A tungsten filament in a vacuum-bulb, if heated too intensely, or sometimes by proper heating long prolonged, throws off a discharge of metallic vapor, which plates the bulb and obscures the light. If the bulb is filled with gas this dis- charge is in great part prevented. Any which does occur is carried up to the top of the bulb by the gas currents in the bulb, set up by convection, and deposited there where it does little harm. Owing to the cooling effect of the gas, a greater current is needed to operate a gas filled lamp, but on the other hand the filament can be heated to a much higher tem- perature than in the vacuum. This restores the condition of economy. A thin filament exposes a proportionately larger cooling surface than a thick one. Hence to avoid too great cooling effect from the gas filling, the filament is made thicker and shorter, and a heavier current is required than for the vacuum lamp filament. Leading-in Wires in Incandescent Lamps. — Instead of plat- inum, for many years the only metal used for leading-in wires, a compound wire, dumet wdre, is used to pass through the glass of the incandescent lamp. As the glass has to be sol- dered or melted around and must adhere to the wires, they must be of as nearly as possible the same coefficient of expan- sion as glass. As platinum has such a coeflacient it was for many years the only metal used for this purpose. Dumet wire, presumably meaning two-metal wire, consists of a core of an alloy of 45 per cent nickel and 55 per cent iron; within a 752 ELECTinCIAXS' HAXDY BOOK, sleeve or coating of copper surrounding it. Copper expands much more than the iron-nickel alloy, and much more than glass; the high coefficient of the one works against the low coefficient of the other, with the result that the expansion of the compound wire is almost the same as that of glass, and it solders perfectly into the bulb. The filament is attached to it by spot welding or some equivalent method. The Auer Process of Attaching Metallic Filaments to the leading-in w^ires is as follows: The filament is secured a short distance below the ends of the leading-in wires; one end of the filament to one of the leading-in wires. An arc is then started through the space between the ends of the leading-in wires. The ends instantly fuse, forming beads or little balls of the metal, and these balls run down the ends of the leading-in u'ires and melt around the ends of the filament. The arc is at once short-circuited when this takes place. Molybdenum Supports of Filaments. — The upper supports attached to the glass stem inside a tungsten lamp are now largely made of molybdenum. As the wire of the filament is put in they are made to spring down a little under the strain, so as to keep the filament tight if it lengthens. Duration of the Timgsten Filament's Efficiency. — After a thousand hours' burning it w^as found that a tungsten lamp had only lost 3 per cent to 4 per cent of its candle power. Photometering Gas-filled Lamps. — The spinning or rapid rotation of an incandescent lamp on the photometer, so as to obtain an average candle power, is liable to give inaccurate results in the photometry of gas-filled lamps. The convection currents of the gas filling of the bulb are undoubtedly dis- turbed, so as to introduce conditions not present in the lamp w^hen in regular use. For each lamp, however, there is a speed of rotation, which gives correct results on the photom- eter; this has to be known or determined to employ the rotating apparatus and get the average candle power directly and simply. This correct speed exists for most gas-filled lamps. Overshooting. — This term describes the quick attainment of its full brightness by a tungsten filament lamp. The ccld METALLIC FILAMENT IXCAXDESCEXT LAMPfi, 753 filament is of much lower resistance than is the hot one. Hence at the first instant of turning on the current a heavier current than the lighting current passes and brings the fila- ment up to white heat with what may be called undue rapidity. A tungsten filament in series with a carbon one is more quickly brought to incandescence. In the action of overshooting the momentary candle power is higher by about 20 per cent than the normal. This high power only lasts for an instant. Helion Filaments. — This is a filament of silicon, with a slender carbon filament as a base for its production. The carbon filament is heated or flashed in an atmosphere of silicon fluoride or some gas containing silicon and decomposable by heat. The filament thus treated becomes coated with silicon and is of very high efficiency, running as high as 1 watt to the candle power, and resists a very high temperature before softening. 3Ietallic Carbon Filaments is the name of such as have been subjected to a special heating treatment. After the regular flashing in a hydrocarbon the filament is exposed for several minutes to a very high temperature, nearly that of the evap- orating point of carbon, which causes the shell of the filament to acquire metallic properties, such as low electric resistance, positive temperature coefficient of resistance, metallic lustre and low vapor tension. It can be operated at a temperature high enough to give an efficiency of 2.5 to 2.6 watts to the candle power. Osmium Filament Lamps. — This metal of the platinum group w^as one of the first metals experimented with as a material for filaments. Its conductivity is inconveniently high. In Germany three were put in series on a 110 volt circuit, giving about 27 volts for each lamp. Their efficiency was high, li/^ watts to the candle power. They were almost untransportable on account of the high fragility of the fila- ment. Although one pound of the metal will make 30,000 filaments, it is estimated that there is not enough of the metal known to exist to supply incandesecnt filaments for one year of consumption. 754 ELECTRICIANS' HANDY BOOK. Tantalum Filament Lamps. — This metal, with a fusion point of about 2,250° to 2,300<^ C. (4,018° to 4,108° F.) has been experimented with as a filament. Its specific gravity is from 16.8 to 17. To prepare it its oxide is treated in an electric furnace in vacuo, and melted, giving a button of the metal, which is drawn into wire of extreme fineness. Thus prepared it compares with steel in strength, but gradually becomes brittle in service. At first it lengthens so as to hang some- what loosely in the bulb, but it next grows shorter and strains at its supports. Its efiiciency is 1.8 watts per candle power. A typical filament for a 22 candle power lamp is 2 5.6 inches long. 0.002 inch in diameter, weighing so little, that although the metal is expensive, the question of cost was taken as eliminated. CHAPTER XLIII. VACUUM TUBE LAMPS, FRAME ARC LAMPS, NOTES ON ILLUMINATION AND PHOTOMETRY. Vacuum Tube Lamps. — These lamps have been extensively experimented with and now are reduced, as far as practical results are concerned, to two kinds: One is the McFarlan Moore lamp, now used for color comparisons, and in other ways to a rather limited extent, and the other is the Cooper-Hewitt mercury vapor arc light, a vacuum tube charged with mercury vapor, and which is being extensively used for factory and out- door lighting installations. The McFarlan Moore lamp is really a Geissler tube. In the Cooper-Hewitt lamp the heavy mercury vapor gives characteristic effects, which differentiate it from the Geissler tube, as usually understood, although it is, properly speaking, a mercurial vapor Geissler tube. The Cooper-Hewitt Mercury Vapor Lamp consists of a tube, exhausted of air or other gas, and containing some mercury. It follows that it is filled with a small amount of gaseous mercury, the amount of which will vary with the temperature. The tube is held in a vertical or inclined position, never in a horizontal one. The reason for this will be understood from the description. At the lower end is a pocket of mercury, and into this mercury a wire terminal extends or is in contact with it. The wire goes through the end of the tube and is sealed in the ^lass. This is the positive electrode. At the upper end of the tube, through which a wire also extends, sealed in the glass, there is an electrode, cup-shaped, made of thin iron, and connected to the wire. If a current of elec- tricity is caused to pass through the tube a bright light is produced. Owing to the inclined position of the tube any mercury which condenses on the walls of the tube runs back to the positive electrode. Thus the tube is never obscured with globules of mercury, and the electrodes are kept in 755 ioQ ELECTBICIASB' HAXDY BOOK. proper condition. As the lamp is designed to operate upon an incandescent lighting circuit, it is obvious that the first prob- lem is the starting of the arc. Starting the Cooper-Hewitt Arc. — Each tube is provided v\ith an auxiliary apparatus, which steadies the light and also operates in the lighting of the lamp, or what is the same thing, the starting of the arc. This apparatus is generally mounted above the tube. If for a direct current system it contains a resistance coil in series with an electric magnet. If Shiffen starting Resistance Tube- Fig. 569. — Cooper-Hewitt Mercury Vapor Lamp axd Connections. for an alternating current system a choke coil occupies the place of the resistance coil. Near the poles of the magnet there is mounted a mercury switch called a **shifter.'' This is a small receptacle of glass, carried on pivots, and with an armature attached to it. It contains mercury in contact with two wires sealed in the glass. In its natural position the mercury connects the wires, thereby short-circuiting the light- ing tube or lamp proper. With the apparatus connected and the mercury switch in its natural position, if the current is turned on, it would go through the resistance coil or choke coil and then through the magnet coils, and so out to the line again. But suppose the mercury coil is free to oscillate, TACrUM TUBE AXD FRAME ARC LAMPS. 757 or rotate in its bearings, which it is arranged to do. Then when the current is turned on, it starts through the coils, as just described, excites a magnet, which attracts its arma- ture, causing the mercury switch, "the shifter," to rotate and assume a new position, opening its circuit and bringing the lighting tube into the circuit. This breaks the main circuit, producing an induced discharge of such high potential that it jumps the length of the lighting tube, starts the mercury- arc, and thereby puts the lamp into operation. Cooper-HeA\itt Tube. — The tube is made of lead glass, and of various lengths, as given in the table. It must be chem- ically clean in its interior, and before receiving its charge and before exhaustion, is cleaned with acid and potassium bichro- mate. The iron electrode must also be free from all extrane- ous matter. It is about three-quarter inch in diameter, with rather thin walls. To its interior, at the lower or mercury end there is attached a patch of about a square inch area of carborundum. This supplies a multitude of points, and such are favorable to an efectric discharge. The outside of the same end is coated with tin-foil. This along with the mer- cury provides a sort of Leyden jar, or condenser, to intensify the starting discharge; as the tin-foil is connected to the posi- tive electrode, while the mercury inside the tube is in contact with the negative electrode, the condition of a static condenser or Leyden jar is produced. The tin-foil is called the ''starting band."' Visual Acuity. — This term is now much used in discussing artificial illumination. It means simply sharpness of vision or the relative ability to see things. There is a certain amount of chromatic aberration in the optical apparatus of the human eye, and the theory is maintained by the advocates of the monochromatic type of lamp, that visual acuity is greatly in- creased when the eye has only to focus light of a small range of wave-lengths — approximately monochromatic light. Thus it is found that for many purposes the light given by the mer- cury arc enables the eye to discern differences and details which quite escape it in even the best sunlight. The large area of the Cooper-Hewitt tube takes it out of 758 ELECTRICIAN^ 8' HANDY BOOK, \ the law of inverse squares. A bank of tubes placed side by- side or otherwise distributed produce a light of approximately uniform strength all through the lighted space. The same applies to a great extent to a single tube, if the space lighted by it is not too laYge, or if the tube is not too remote from the observer. This fact is one of its great features of superiority over the ordinary artificial light of small area. Its operation produces an approximation to the uniform strength of sun- light. The latter, owing to the distance of the sun, is also outside the law of diminution of intensity with distance. Tak- ing the Cooper-Hewitt lamp as a line, for small distances its light should vary inversely as the distance. Potential Drop in the Cooper-Hewitt Lamp per unit of length of arc or of tube depends upon the pressure of the mercury vapor. The pressure in the glass tube lamp is about one-eighth inch of mercury. Its intrinsic brilliancy is so low that it can be looked at directly without disagreeable effect upon the eye, and approximates to the '*cold" light of the fire- fly. The light given per inch length* of tube is small, and the total i-s made sufficient by the length of the tube. In the quartz lamp, with short tube of heat-resisting quartz, the pres- sure of the mercury vapor may exceed that of the atmosphere, and the drop of potential per unit of length may be thirty times as great as that of the glass tube lamp. A 700 candle power tube on a 110 volt circuit will pass a current of 3.5 amperes, giving 885 watts total expenditure of energy, at 0.55 watt to the candle power. The tube is supposed to last 2,000 to 3,000 hours, although there are records of as high as 5,000 hours* burning. Data of Cooper-Hewdtt Glass Tube Mercury Vapor Lamp Length of Tube over all 55% inches Length of Portion of Tube giving Light.- 50 ** Diameter of Tube 1 inch Mean Spherical Candle Power .-. . 670 to 850 candles Watts per Candle Power 0.51 to 0.52 Power Factor about 50% Total Watts 350 to 430 Average Life of a Tube about 7,900 hours VACUUM TUBE AND FRAME ARC LAMPS, 759 Lengths of Tubes, Average mean spherical Candle Power. Average Watts. Watts per Candle Power. 21 inches 200 102 38'5 385 0.64 45 *' 700 0.55 50 '' 800 0.48 (Alternating Current) 50 inches 800 400 0.50 Cooper-Hewitt Quartz Tube Lamps. — The use of the oxygen blowpipe, especially with acetylene or blau gas, makes it pos- sible to work with quartz as under the ordinary air blast flame with glass. White hot quartz in a pasty condition can be plunged into water without cracking, and in general is im- mune to all ordinary heat, much more so than glass. It is used for mercury vapor lamps of high intensity, where a glass tube would succumb to the intensity of the heat. Quartz also permits the passage of the ultra-violet rays of light; for some purposes this is desired, as in the treatment of oils. Quartz softens not far from 1385° C. (2,500° — F.), or rather does not soften below that temperature. The potential drop in a mer- cury arc varying with the pressure of the vapor, and the intrinsic brilliancy and temperature varying with the same factor, the quartz tube lamp is made to operate at a pressure above that of the atmosphere, the potential drop per unit of length of arc is thirty times as great as in the glass tube lamp, and the heat is very great; it is thought that it is one of the highest heats produced by man. Construction of Quartz Tube Lamp — The tube is about one- half inch in diameter and five inches long. At each end is a bulb or enlargement. A quantity of mercury is contained in the tube and the space above it is filled with mercury vapor. This suggests a vacuum, but in operation the heat destroys the vacuum as long as the arc exists, by volatilizing the mercury. The tube is carried in a pivoted support. If the tube is kept in a horizontal position the mercury will lie in a thin column or thread along its entire length. It has sealed-in electrodes or terminals at each end, and the thread of mercury connects these in the position assumed. If a current is turned on the TCO ELECTRICIANS' HAS BY BOOK. mercury will short-circuit the vapor in the tube. When the current is turned on, if the tube has its position changed so as to break the short-circuit, then the mercury-vapor arc starts and is maintained, as long as the tube holds its open circuit position. There are two systems for effecting the lighting. In one the tube is kept in the closed circuit position when Fig. 570. — Cooper-Hewitt Quartz Tube Lamp. there is no current passing. When the current is turned on in order to light it, an electro-magnet in circuit with it is excited, attracting its armature, which is connected mechan- ically to the tube. The end of the tube is drawn up, the mer- cury thread is broken, and the mercury vapor arc is started and continues as long as the current passes. In the cut h is the tube fastened to a plate, 5, the latter carried by a pivot at the end of the supporting bar, c. m is TACri'M TUBE A\D FRAME ARC LAMPS. 701 the electro magnet. When it attracts its armature it draws up the rod, /, swinging the left end of the tube upwards and breaking the mercury thread. A dash-pot, d, prevents any shock. 7*1 and r2 are two resistance coils, the former provided with a sliding resistance connection. The rod, p, operates to bring the dash-pot into operation and is in parallelism with Fig. 571. — Cooper-Hewitt Quartz Tube Lamp. the support, c. There is a permanent magnet, so arranged as to operate a catch to hold the tube in the open circuit position, if any reversal of current should operate to drop the tube and short-circuit the vacuum. A reflector is indicated immediately above the tube. In the next cut the other arrangement is shown. Here there are two electro magnets, m and ms. The tube is normally on open circuit so that no current can pass through it. The cir- 762 ELECTRICIANS' HANDY BOOK. cuit of the regulating and lighting apparatus is closed through the two magnets, the magnet, m, being in shunt. When the current is turned on, the magnet, ms, is excited and attracts its armature, drawing the tube into such a position as to close the circuit by the mercury flowing along its bottom. At once the other magnet, m, attracts its armature, which operates a switch to cut the shunt magnet out of circuit. The tube drops back to its open circuit position, the arc is started and the lamp works as long as current passes. Resistance coils, r^, ^2, are indicated, and the magnet, m, operates as an inductance if such is required. Action of the Quartz Tube Lamp. — In this is involved the starting characteristic and the stationary characteristic. Starting Characteristic. — When the current starts through the cold tube the voltage of the circuit is taken up in greater part by the resistance coil in series with the lamp. On a 220 volt circuit a cold lamp will show a potential drop of only 25 volts. As the heat inside the tube increases, the voltage in- creases quite slowly; at the end of eight to ten minutes it reaches its maximum, about 165 volts. The current starts at about 11 amperes, slowly decreasing as the voltage increases, until it reaches, in about 11 minutes, its normal value of about 3 % amperes. The appearance of the arc changes during the lighting. At first the whole tube is filled with the arc, re- sembling the glass tube lamp in operation. As it attains its working temperature the arc concentrates in the middle of the tube in a thin dazzling path, losing its pale bluish color and becoming whiter. Stationary Characteristic. — The heat capacity of the quartz tube is considerable, and this introduces a lag between the electric current changes and the burner voltage. This lag brings about the peculiarities of the starting characteristic. The voltage and current changes are due to the changes in vapor pressure varying with the heat of the tube. When the lamp is regularly operating, after its proper amperage is reached, a variation in voltage hardly affects the current at all. If at 8 volts the current is 3^ amperes, then at 165 volts it will be only 3% volts. Recurring to the starting char- VACUUM TUBE AND FRAME ARC LAMPS, 763 acteristic, some of its features are involved in a sudden change of voltage. If it is an increase, the current at first increases> the heat of the tube gradually rises, cutting down the current, which soon is reduced to its original value, although the in- creased voltage is maintained. All that has been said refers to a 22 volt lamp. If the voltage on the line is suddenly decreased, the same changes take place, but in reverse order. The same phenomena attend the operations of lamps at other voltages. Ultra- Violet Rays in Quartz Tube Lamps. — The quartz tube lamp is very rich in these highly actinic rays. They are some- times desirable, in some chemical processes, for instance, but for ordinary lighting purposes they are not wished for. At close range they act like a strong sun or even more intensely^ affecting the skin just like sunburn. The interposition of a glass shade cuts off a quantity of thie ultra-violet rays so as to make the light comparatively innocuous, but depriving it of its actinic quality, something which in many cases is desirable or required. If used in the open air for lighting large areas. a hemispherical globe is used, more to protect it than for the elimination of ultra-violet rays, for the air acts to eliminate them. To get the full actinic effect of the light, it should be placed close to the surface to be acted on. Data of Quartz Tvibe Lamps. — The following table gives some of the principal data of the quartz tube lamps. It will be noticed that there is a large loss of voltage between lamp and line. For the 550 volt circuit, this loss runs as high as thirty-five per cent of the line voltage. Nominal supply or line voltage 110 220 550 Full range of supply voltage.. 100-125 200-250 450-625 Average current 3.8 3.3 2.0 Maximum burner voltage 90 170 345 Candle power mean hemispher- ical with clear glass globe.. 1,000 2,400 3,500 Commercial efficiency, watts per mean hemispherical can- dle at nominal line voltage. 0.42 0.30 0.31 For each voltage a special lamp is used. 764 ELECTBICIAXS' HAXDY BOOK. The lamp gives 1,000 to 2,600 candles at ^^ watt per can- dle power, which referred to the voltage of the tube reduces to about 0.40 watt. The 3Ierciiry Arc Lamp for Alternating Cui^rents has two positive terminals connected to the ends of a transformer. The negative lead is connected to the middle of the same transformer through an inductance and a ballast resistance. The current alternates between the two electrodes on the positive end, so as to be always in the same direction. The lamp acts as a rectifier, changing the alternating current into a direct one. The McFarlan Moore Vacuum Tube Lamp. — This is a Geissler tube, charged with any desired gas, according to the character of light desired, operated by alternating current. The actual vacuum is 0.19 millimeter of mercury, and the vari- ation in the vacuum must not exceed 0.01 millimeter, which is 0.00001 atmosphere. In operation the vacuum tends to increase and this is provided against by a valve operated by the primary current, which actuates the transformer, which operates the tubes. The tubes are made in section, each one 8 feet 6 inches long, and soldered or melted together. This operation is done with remarkable rapidity, two minutes suf- ficing for a joint. The tube when joined up may be 22 feet long. The glass is 1/1 6th inch thick. The tubes operate on an alternating current circuit. The current comes from the secondary of a transformer, and is normally about 24 amperes. The vacuum rapidly increases and as it does . so the current also increases, and small changes in the vacuum produce enormous changes in the re- sistance. Thus at 0.11 millimeter vacuum a current of 24 amperes may pass; at the end of a minute the vacuum will have increased and a current of 2^ amperes will pass. At this point a valve will operate and will admit a little gas to restore the proper conditions. A candle power of 10.5 can- dles per lineal foot is assigned to the tube with an expenditure of 1.50 to 2.84 watts per candle power. The automatic gas valve admits gas when the vacuum in- creases. A glass tube connects the gas supply to the interior VACUUM TUBE AXD FRAME ARC LAMPS. 765 of the large lighting tube. This gas tube is closed by a dia- phragm of carbon. On top of the diaphragm is mercury. Above the valve is a solenoid and armature. The latter car- ries a glass tube immersed in the mercury. If the vacuum increases and more current enters the lighting tube, a greater fribui^d ^rm desired lengths of ZOO ft Fig. 572. — IMcFarlax Moore Vacuum Tube La:n[p with Automatic Air Valve. current passes through the primary of the transformer, and consequently through the solenoid. It draws up its armature, raising the tube from the mercury and exposing the carbon. Gas flows through the pores of the carbon, the current de- creases and the solenoid armature descends. This raises the level of the mercury, which covers the carbon so that no gas can pass through it. This series of operations is repeated 766 ELECTRICIANS' HANDY BOOK. over and over again, so as to maintain a constant vacuum within very narrow limits. For proper working of the Moore tubes the current should have sharp peaks to its alternations. In a Moore tube there is a point of maximum conductivity, above or below which pressure the resistance increases. In practice the tube is operated at a lower vacuum. It follows that, if the vacuum is increased, the resistance will fall and a greater current will pass. This increase is what operates the valve, as just described. To fill the tube with nitrogen and so get the pink color, the air supplying the tube is made to pass over phosphorus. The maximum frequency in the current operating a Moore tube is 60 cycles per second. Every foot is supposed to give a candle power of ten candles. Neon Lamps. — Vacuum tube lamps filled with the rare gas neon have been experimented with by Georges Claude. A very slight potential is required to cause a vacuum tube con- taining it to glow. In wide tubes the potential fall per centi- meter is only one volt. If helium is present it is slowly ab- sorbed by the electrodes. The presence of the latter gas is always to be expected with neon and the other rare gases of the atmosphere. Flame Arc Lamps. — These are arc lamps of the open type, whose carbons are impregnated with a chemical salt. The invention is assigned to Bremer in 1898. In the ordinary carbon arc the greater part of the light is due to the ignited ends of the carbons; in the fiame arcs about three-quarters of the light is due to the arc proper. The carbons may be ar- ranged vertically, one over the other, as in the regular arc lamp. An inverted shallow fire clay bowl, fixed a short dis- tance above the arc, operates to retain the gases of combustion and to retard the combustion of the positive carbon, which nor- mally burns twice as fast as the negative one. The positive carbon is in this case placed uppermost; if the positive car- bon is to be placed lowest, then it must be thicker than the upper one. A probably preferable arrangement is to use in- clined carbons, arranged like the letter F. The angle of incli- VACUUM TUBE AND FRAME ARC LAMPS. 767 nation may be about 15°. With the inclined carbons a "blow down magnet," as it is called, is used. The coils of the magnet lie on each side of the carbons, well out of the path of light. f^^ Fig. 573. — Flame Akc Lamp Showing Inverted Gas Bowl. y///,vM Fig. 574. — Flame Arc Lamp with Ixclixed Carbons. The core is of circular contour, lying in a horizontal plane. The poles are carried down close to the arc, one on each side. They act on the arc to drive it down to its proper position. 768 ELECTRICIANS' HAXDY BOOK. The length of the arc is 15 or 16 millimeters, seven or eight times as long as the carbon arc. A current of 4 am- peres and a potential difference of 68 volts is standard. The arc is so long that a special striking appliance is needed to start it when the current is turned on. Fig. 575. — Filame Aec La:mp with Ixclixed Carbons. 'r^rWrirj Fig. 57G. — Flame Arc Lamp with Inclined Carbons. VACUUM TUBE AND FRAME ARC LAMPS, 7G9 Blondel's Carbons have an outer zone of ordinary electrode carbon. Then comes an intermediate zone of carbon impreg- nated with salts, and the core is of softer carbon, also im- pregnated. A usual impregnating material is calcium fluoride. By the employment of different salts, different tints are ob- tained. To get a long period of operation, the carbons may be two feet in length. Magnetite Arc Lamp. — This lamp, the invention of Alfred Steinmetz, has a lower electrode of copper, which is the posi- tive one. The upper one is made of a mixture of magnetic oxide of iron and of titanium oxide, along with salts of chromium, titanium or of other metals. Ulrich's Integrating Sphere. — This is used to determine mean spherical candle power. It is a hollow sphere, with whitened interior walls, and it is provided with a window of milk or opal glass for observation. A standard lamp is mounted within it, in such a manner that its direct rays are screened off from the window. The brightness of the win- dow is proportional to the light within the sphere, as it im- pinges on the walls thereof. The window is photometered. Then the light to be tested is introduced or lighted within the sphere and the window is again photometered. The re- sult is the relative mean spherical candle power of the two lamps. It is not of high scientific accuracy, but is extensively used. For arc lanips it is five feet in diameter; for incan- descent lamps it is two feet in diameter. The factor or coefficient for conversion of mean horizontal candle power of an incandescent lamp into mean spherical candle power varies from 0.8 to 0.95. Lux and Lumen. — Referring to page 564 et seq. of this book, we see that the intensity of light given by a source of inconsiderable area varies inversely as the square of the dis- tance. To make this true the source of light must be so -small that it can be considered a point. In photometry a candle flame is accepted as such, although it has size and is by no means a point. The intensity of light produced by a candle at a distance of one meter is called a lux. This unit is also called a meter-candle. If this intensity is referred to 770 ELECTRICIANS' HANDY BOOK, a square foot we have a unit of quantity of light called a lumen. It is a conglomerate unit, a mixture of the English and metric systems, and is used in engineering work and in cases where the illumination of areas is to be considered. Another lumen, and a more consistent one, is the foot- candle lumen. This is the light given by a candle at a dis- tance of one foot from the illuminated surface and distributed over one square foot area. Steinmetz defines the lumen as the unit of light-flux, the light-flux passing through unit surface at unit light-flux den- sity. One candle intensity of light at uniform distribution gives 4jt lumens. This is because the area of a sphere is equal to the square of the radius multiplied by 4jt, or to four great circles, and 4jt is the area of the surface of a sphere of unit radius. Another way of expressing it is to say that one mean spherical power is equal to 4jt lumens. 1 The unit of flux is the candle-lumen, of the total 4;t light-flux from a source of one mean spherical candle power. One lumen per square meter is sometimes called the lux or meter-candle. Utilization Efficiency. — This is a figure derived from the circumstances of each particular case, depending on the color and reflecting power of the walls. Suppose that it is required to light a room with a total of C lumens per square feet of area. The net lumens needed would then be the product of C by the area of the room, which area we may call A, so that A C would be the net lumens needed. But owing to the fact that there is inevitably a loss of efficiency, as no wall can have a 100% reflecting coefl^cient, the utilization efficiency is less than unity in value, and the total lumens required are A C expressed by the fraction, , in which E is the coefficient E of utilization efficiency. Then if this expression be multi- plied by the reciprocal of the lumens of a single lamp, we will have the number of lamps required. Calling the lumens of a VACUUM TUBE AND FRAME ARC LAMPS, 771 single lamp, L, and carrying out the above multiplication, we have: Lamps required = ^~p- X-^— ^^ E L EL. As this efficiency coefficient is entirely practical, there is no scientific accuracy in such calculations, since the data have to be determined by practical observation for each case. The color of the walls, their texture, roughness or smoothness and similar factors determine the coefficient. Intrinsic Brilliancy. — If we know the era of a surface- giving light, and divide its candle power by its area, the quotient is its luminosity per unit area and is called its in- trinsic brilliancy. The intrinsic brilliancy of various sources of light is given in the table, principally from Ives and Luckiesch: Carbon Arc (Crater) 84,000 Flaming Arc 5,000 Nernst Glower 3,010 Tungsten Incandescent, Nitrogen filled. . . . 2,200 Tungsten Incandescent (1% Watt C.P.) . . . 1,600 Carbon Incandescent (2i^ Watt C.P.).... 400 Welsbach Mantle 31 Cooper-Hewitt Tube Lamp 14.9 Kerosene Flame 9 The general practical rule is that the most acceptable lights are those of lowest intrinsic brilliancy. The Student Lamp, with kerosene oil as its fuel, is often referred to as the best reading light, and kerosene is at the foot of the list. Yet the sun is of very high brilliancy and is the perfect luminary, in spite of this fact. Its perfection is largely on account of its distance, taking it out of any law of ratio of light to distance; its distance causes perfect uniformity of illumina- tion. In artificial light an approximation to uniformity is obtained by using a number of lights and by having the walls of our rooms so treated that they reflect a quantity of light. A room with light-colored walls is bettet* illuminated by a given artificial light than one with dark walls. CHAPTER XLIV. ELECTRIC FURNACES. Electric Furaaces. — These furnaces are divisible into two principal classes, those which heat by incandescence and those in which the arc is the heating agent. Except for special uses, it is fair to say that the arc furnaces are in more general use. In the steel and iron industry there is every advantage in keeping the metal from contact with a more or less impure fuel, and this is one great advantage of the electric furnace. In general terms it may be said that the furnace is limited in size by the possible size of the electrodes. One great point in these is that they should be of as poor heat-conducting quality as possible. As they pass through the walls of the furnace without the possibility of avoiding some leakage of the gases around them, there is a tendency for them to burn or waste away outside of the furnace. This is guarded against by sheathing them with metal or by some other simple arrangement. To insure even consumption of the carbons the current is alternating and it has a power factor of 0.80 to 0.8 3. A seven and a half ton furnace may have 12-inch graphite electrodes, with a cur- rent of 10,000 amperes and a potential of 8 volts. Taking it as operated at two phases, we have for the power absorbed: 2 X 10,000 X 80 ^ ^^^ , ., 8x1 000 ~ 2,000 kilowatts. Electric Steel Furnaces. — The diagrams illustrate the prin- cipal types of electric furnaces used in the manufacture of steel. The Heroult furnace is simply a closed hearth fur- nace, through whose roof the carbon electrodes project, 772 ELECTRIC FURNACES. 773 being fed down as fast as they are consumed. Alternating current is supplied from a transformer, which gives the proper voltage as determined for each case; arcs are struck and continue to exist as long as the current is supplied. The steel is melted absolutely out of contact with anything ex- cept the furnace walls and bottom and any flux or other ad- Elsctric connecfions to common ' re f urn of the fwo phases m:.,;' Silica Brick Dolomite ^ Magnesife Brick Carbon Mixture^ Chrome Brick Fig. 577. — Heroult Furnace. Fire Brick dition purposely made. It is as pure a type of arc heating as any. The Girod Furnace, operated by a monophase current, is shown next. The arc goes from upper electrode to the charge proper, the latter being in electric contact with the- lower electrodes, which extend through the bottom of the furnace. The next cut shows the same furnace for three phase current. In it the arcs are produced between the upper carbons, while the lower ones, connected as indicated, operate as the neutral point. 774 ELECTRICIAN^ S' HANDY BOOK. Girod Furnaces in Bethlehem, Pa. — The following is the description of the Girod steel furnaces being installed at the Bethlehem Steel Works. They are cylindrical in shape, 15 feet in diameter and 5 feet high, with shells of %-inch steel plates. The whole structure is mounted on rockers, so that it can be tilted to pour out the charge. The bottom of the hearth is 14 inches thick, of dolomite or magnesite. The metal bath on top of this is 14 inches deep. Fourteen water- /7U Bottom. £ (ect nodes- Fig. 578. — Girod Furnace. cooled steel electrodes connect the bath electrically with the bottom shell of the furnace. The top or roof is of 9-inch silica brick, separated from the hearth by asbestos. Three electrodes enter through the top; each one is 18 inches in diameter, with a working length of 6 feet. They are carried in copper holders, which are water-cooled. They are raised and lowered by a motor, so as to keep them at a standard distance of 4 inches above the surface of the melted steel. The furnace weighs about 90 tons. It operates at 1,500 kilowatts, 3 phase and 25 cycles at 65 to 8 5 volts, the carbon electrodes taking two phases, with the shell of the furnace connected in the neutral point. 2,500 kilowatt, 6,600 volt 2 5 cycle generators are employed. There are three oil-cooled transformers for each furnace protected by reactance coils. ELECTRIC FURNACES. 775 Of 700 kilowatts leaving the transformers, 620 are absorbed through the electrodes. The Kjelliu Furnace is a purely inductive one. In the dia- gram is shown a heavy core of rectangular shape, around one of whose sides a coil of wire is placed. The opposite side is surrounded by a circular trough-like hearth for the steel. If an alternating current is passed through the coil, the current induced in the metal in the trough heats it to any desired degree within the capacity of the apparatus. A rotary seg- Trons/br/ner Core Fig. 579. — Kjeixin Furnace. ment at the top of the core is supplied with^ direct current, and rotates under the effect of the alternating current and is designed to increase the efficiency of the electric operation. One of the problems is the starting of this type _o^ 'furnace before the charge has become conductive. One method is to introduce an iron ring extending around the hearth, which heats up under the induced currents and brings the charge to fusion, when the charge makes a connection of itself and the operation goes on normally. The Roechling Furnace heats its charge by conduction through the metal itself. The cut is self-explanatory. The current induced in the secondary passes through the length of the charge as it lies in the hearth, the metal in the hearth 776 ELECTRICIAN'S' HANDY BOOK, forming a portion of the secondary of the induction coil which is constituted by the construction of the furnace. A Swedish Shaft Furnace is next shown. This is a single arc furnace of the cupola or blast furnace type. A Hering Ore Reducing Furnace is the next illustrated. This heats by conduction, and as the charge is reduced to the Ti'Onsformer Cot^ Lead 'Llectrode Fig. 580. — Roechling Furnace. Fig. 581. — Swedish Shaft Furnace. metallic and fused state it is withdrawn from time to time through the tap hole. The cut of the Hering combined ore reducing and steel furnace is also self-explanatory. The left-hand furnace is a simple melting furnace and next to it a melting furnace of identical construction, into which the metal from the re- ducing furnace flows and is subjected to any additions of alloys or to any desired refining or treating process desired. Aluminum Furnace. — The Heroult furnace, shown on page 778, for the production of metallic aluminum, has a massive ELECTRIC FURNACES. 777 iron casing. It is, in recent 'practice, made hollow so as to be water-cooled. At E it is connected to one of the electric leads. It has a refractory lining, B, and at C is a mass of aluminum or copper, which is in electric contact with the shell or casing of the furnace. H represents the carbon anodes. The material operated on is purified aluminum oxide, mixed with cryolite; the latter is sodium-aluminum 'f'/€C/rv(fes Fig. 582. — Hering Ore Beduc- ING Furnace. Fig. 583. — Hering Combined Ore Reducing and Steel Furnace. fluoride; it is easily fusible and when fused is a solvent for the oxide. After the furnace has received its charge, an arc is started by bringing anode, H, and cathode, C, into momen- tary contact; on drawing them apart the arc forms and melts the central portions of the charge. The current is unidirec- tional, so there is not only the intense heating effect of the electric arc to melt the cryolite and cause the solution of the oxide, but electrolytic decomposition of the dissolved oxide is started and metallic aluminum is produced and runs down through the charge and adds itself to the metal in the bot- tom of the furnace. As the charge is exhausted more is 778 ELECTRICIANS' HANDY BOOK. added through the top as required. The metal produced is drawn off at the tap-hole, F, Carborundum and Graphite. — Carborundum is a chemical compound of carbon and silicon, C Si. It is a powerful abrasive and has largely supplanted natural emery. It is used in powdered form and as grinding wheels and other shapes. The furnace in which it is made is an extremely simple one. It is a receptacle of trough-like form, about 15 Fig. 584. — Heroult Aluminum Furnace. feet long and 7 feet high and wide. It is made of bricks, laid up dry. C and C are carbon electrodes connected at B and B to the electric circuit. It operates by incandescence mainly. For this furnace a charge of ten tons is required, consisting of about the following mixture: Coke, 34%; Sand, 54%; Sawdust, 10%; Salt, 2%. In putting in the charge, as the level of the carbon electrodes is reached, a core of coke 9 feet long and 2 feet thick is placed horizontally, so as to connect the electrodes electrically and give a path to start the current. When the furnace is in operation and intensely hot all of the charge is highly conductive. When cold it is ELECTRIC FURXACES. 779 non-conductive. On turning on the current the heat slowly grows and for two hours large quantities of carbon monoxide are evolved and burn all over the outside as the gas escapes through the crevices of the brickwork and elsewhere. In twelve hours the whole furnace and charge is red-hot. After a run of twelve hours the current is turned off and the mass is allowed to cool. Four thousand amperes at 185 volts are used; the current is alternating. The mass is annular or ring-like in distribution. The outer layer is insufficiently converted and is saved and goes into another charge, so as to get a second treatment. Next comes a sixteen-inch layer or cylinder of true carborundum. This is the desired product. r^-^r /«^''->''l'^->?~^j'^-\":' ci^ '^'j.^'^y^-f<^. mmmmmmm: Fig. 585. — Atcheson's Carborundum and Graphite Furnace, The central core is graphite, also a merchantable product. The carborundum is ground and made into the various prod- ucts as required. Each group of electrodes comprise 6 carbons, each one 3 inches in diameter and 2 feet long. About 50% of the theo- retical yield of the carborundum is obtained. Calcium Carbide. — This substance is interesting, as it rep- resents the first success ever attained in the synthesis of carbon and hydrogen, in which its production is the first step. Its formula is Ca C2. A three phase small shaft furnace is shown in the cut. At ^ is a cast-iron disc, whose upper surface is protected by a layer of graphite. This disc can be lowered to the bottom of the furnace by the screw, J3, operated by the hand wheel on top of the shaft, C. The arc is started with the plate just in contact with the carbons; con- 780 ELECTRICIAN^ 8' HANDY BOOK. tact is at once broken and purified lime and coke, which make up the charge, are slowly poured in after the arc is started. The heat is intense and it causes the combination of carbon and the calcium of the lime, and as fast as this takes place the disc is lowered, and more charge is introduced. Thus when the disc reaches the bottom the shaft of the furnace is filled with calcium carbide. The product is about 90% pure. Fig. 586. — Calcium Carbide Furnace. Sodimn Furnaces. — In the Castner process, which was about the first electrolytic method successfully used for the pro- duction of sodium, fused sodium hydrate is electrolyzed at a temperature of 325° C. (617° F.). This is a low red heat. Referring to the cut the vessel containing the melted hydrate is shown. Through its bottom the cathode B enters. A cer- tain amount of solid hydrate forms around the neck of the vessel at the bottom and makes a tight joint around the cathode. The anodes, D, which are arranged in a circle, surround the upper end of the cathode. A tube, F, has a cylinder of wire gauze, E, at its lower end, intervening be- ELECTRIC EURNACES. 781 tween anodes and cathode. The anodes, cathode and pot or melting vessel are made of iron. As the hydrate decomposes under the effect of the current metallic sodium goes to the Fig. 587. — Castner's Sodium Furnace. Fig. 588. — Borciier's Sodium Furnace. cathode and collects in the tube, F, floating on top of the fused hydrate. From time to time it is removed by ladling it out. A horsepower of electric energy will produce about one-half ton of sodium per annum. 782 ELECTRICIAN^ S' HAXDY BOOK. In the Borcher process sodium chloride is decomposed, giving chlorine and metallic sodium. The cut shows two vessels connected at their bottoms by a flanged joint with an intervening metallic ring. A, and clamps, D and E; it is kept tight by solid sodium chloride which forms about it. The metallic ring is cooled so as to effect this result. The vessel, C, is of refractory clay, and the vessel, B, is of iron; the latter forms the cathode, and the anode is seen passing through the top part of the vessel, C, When the current passes it decomposes the salt, the sodium collects in B, float- ing on the surface of the melted salt, and as it accumulates it overflows through the outlet, K. Chlorine is evolved in C; this is drawn off through the pipe, H. As the salt is de- composed more is added through the opening, R; it falls into a perforated vessel, and melts. The lower end of the tubular vessel in question is immersed in the melted salt, so as to trap off the chlorine. In the Castner process there .is a certain amount of oxygen and of hydrogen produced. The oxygen is saved, constituting one of the useful products; in the Borcher process the chlo- rine is also saved as a valuable side-product. Fixation of Atmospheric Nitrogen. — This is effected in elec- tric arc furnaces by passing a stream of air through the greatly enlarged arc. In one kind of furnace the arc is flattened out and greatly extended by a strong magnetic fleld; in another the air treated in the furnace is made to expand or lengthen the arc. The chemical reaction gives nitric oxide, N O, which in the presence of water gives nitric acid, thus: 2 N O + 3 O + H20=2 H N O3 The reaction is given in its simplest form; it is really made up of intermediate steps. Electrodes in these furnaces are made of metal; different metals are used, the selection being made of such as wear away slowly. The potential difference is very high, it may be several thousand volts, so that very good insulation is required- As there is no solid substance in the furnace, no slag or melted metal, the furnace lining lasts much longer ELECTRIC FURNACES. 783 than in an ordinary metallurgical furnace. The stopping or starting of such furnaces is not much more of an operation than in the case of an ordinary arc lamp. In the Brickland Eyde furnace the electrodes are quite close at their ends, about 0.3 inch separating them, but a strong magnetic field which surrounds them repels and flat- tens out the arc to comparatively large dimensions. One arc after another is described as driven out radially in great semicircles. A 50 cycle alternating current at 10,000 volts, reduced to 5,000 volts, is employed. As the arc is acted on Air Tube Air /?eacf/o/7 Chamber f/re Br/c/c Oas £xif Fig. 589. — Cross-Sectiox of the ScHOE^HERB Xitrogits^ Fixation Furnace. by the magnetic field its potential rises across the electrodes, and air is blown through the chamber in which the arcs are produced. The nitric oxide is withdrawn as fast as it is pro- duced, because it would dissociate again if left in the fur- nace too long. Its temperature as it is withdrawn is about 1,250° C. (2,282° F.) One kilowatt-hour gives 67 to 70 grams of nitric acid. The Schoenherr furnace is a tubular structure; the arc produced in the lower end of a vertical tube is lengthened out by the inrushing air to a length of 23 to 25 feet. The tube in which the arc is produced is of iron, and constitutes one of the electrodes. The air is made to take a tangential or spiral path, so that the arc is kept in the axis of the tube and its sides are not attacked. An alternating current at 784 ELECTRICIANS' HANDY BOOK. 4,500 to 5,000 volts is employed. The gases at the outlet have a temperature of 850° C. (1,562° F.) The largest size furnace uses 800 kilowatts and gives 65 grams of nitric acid per kilowatt-hour. The cut shows the cross section of the Schoenherr tubular furnace. The air is preheated in the annular passages be- £/ecfroc/e Fig. 590. — ^The Pauling Nitrogen Fixation Furnace. fore entering the central tube containing the long arc. Seven- teen per cent of the heat is lost by radiation, and about 30% by the water used for cooling the electrodes. The diagram of the Pauling furnace illustrates the shape of the electrodes and the contour of the arcs formed be- tween them as they are drawn out by the inrush of air. The ends of the arcs are described as moving along the surfaces of the diverging electrodes in a succession of little leaps, the arc being momentarily held at each point until the air blast drives it on. The electrodes last a long time under the conditions obtained, as their surfaces are comparatively large. In furnaces using electrodes, from whose ends the arc starts, the wear is much greater. A modification of the Pauling furnace is shown in the next diagram; it is due to Phaeler and Heckenbleckner. Movable ELECTRIC FURNACES. 785 electrodes, called kindling blades, are used to start the arc. The electrodes are cooled at the base only. There are two air supplies; one at high pressure cools the starting blades; the greater portion of air is admitted at low pressure after it has been preheated. Some of the gases from the outlet of the furnace are blown in through cooling ducts. These gases /nst/Zoftor Pressure Pressure ^fr /i/r Fig. 591. — ^Phaeleb and Heckenblecker Nitrogen Fixation Furnace. are identical in composition with the freshly treated gases, and they operate to reduce the temperature so as to prevent dissociation of the newly formed nitric oxide. The J. S. Island furnace has V-shaped rings, as shown in the illustration, the central one rotating. Air enters through holes in the angles of the stationary outside rings. The cen- tral ring rotates rapidly, carrying with it, as it were, the arc, which presents the appearance of a ring of fire, although it 786 ELECTRICIANS' HANDY BOOK, is only a single arc. The air holes in this construction are said to burn out rather rapidly. The Moscicki furnace is shown in diagram. One electrode is seen passing up from the base of the furnace in its axis; CcrsOu/^/et (Pas Ouf/ej^ Sf-ea/7? A/r/n/et Fig. 592. — J. S. IsLArs'D Nitrogen FlXATIOX FURXACE. Co/7 Fig. 593. — Moscicki Nitrogex Fixation Furnace. the other electrode is formed by the casing, indicated by the heavy black lines. Arcs spring from the conical top of the central electrode in a radial direction, and the magnetic field due to the coil seen surrounding the furnace keeps the arc or arcs in rotation. The casing is cooled by water-jacketing. The power factor in these furnaces is from 65% to 70%. The nitric oxide is absorbed by water in a tower or proper mixing structure. Air Supply for Nitrogen Fixation. — In all these furnaces the air should be as pure as possible, dust particles, oil or moisture and the other impurities in air impairing the re- sults obtained. By direct experiment it has been found that the yield on moist days is inferior to that obtained on dry days. CHAPTER XLV. ELECTRIC ^BLEACHING. Electric Bleaching involves the decomposition of a solution of sodium chloride by electrolysis, under conditions which insure the liberation of chlorine as the bleaching agent or the formation of hypochlorite, a powerful bleaching salt. If a solution of sodium chloride is decomposed by a strong electric current at high voltage, sodium hydrate goes to the cathode and chlorine to the anode. This accounts for the sodium chloride and for one hydroxyl group of a water mole- cule for each molecule of sodium chloride. The remaining atom of hydrogen escapes from the surface of the cathode as a gas. The reaction is the following: Na CI + HoO=Na O H + H + CI. To decompose sodium chloride requires a potential differ- ence of 4.2 volts. One horsepower hour gives 232.92 grams or 73.5 litres of chlorine. Of course, in practice there is a considerable loss, and such figures as the above are not realized. In the original Kellner process the sodium chloride solu- tion is kept in circulation by a rotary pump, betvx^een elec- trodes wound with platinum-iridium wire. As the current passes, the solution grows stronger, and when its strength is sufficient the current may be stopped and the solution of sodium hypochlorite is withdrawn. In the Haas and Oettel process a stoneware vat has verti- cal electrodes which stct as partitions and divide it into com- partments. There is an orifice in the bottom of each compart- ment and an overflow at the top. The vat is called the electrolyzer. It is placed in a stoneware tank, considerably 787 788 ELECTRICIAXS' HANDY BOOK. larger than itself, so that there is a space between the walls of the two vessels. Salt solution poured into the outer vessel rises through the electrolyzer, nearly to the level of the overflow. The current is turned on, the solution froths up as it is decomposed, and overflows into the outer vessel. So- lution enters at the bottom to make up for what passes out, and thus a constant circulation, is kept up, and the operation is continued until a proper strength of the hypochlorite solu- tion has been attained. There is a cooling coil in the outer vessel. The electrodes are made of graphitic material and last eighteen months. The Kellner process, improved by Foester, Siemens and Halske and others, has its anode at the bottom of the de- composition vessel. The gaseous hydrogen, the cause of the foaming of the solution, is liberated from the cathode near the top, so that the liquid in the bottom is undisturbed. Here in operation the chlorine is liberated and, reacting with the sodium hydrate also formed at the anode, gives hypo- chlorous acid. This is a powerful bleaching agent. At the anode at the bottom of the vessel the reaction starts thus: Na CI + H20=Na O H -f H + CI The hydrogen escapes as a gas; the chlorine dissolves in the solution, which, as it diffuses upward, brings the chlorine into contact with the sodium hydrate formed at the cathode at the top of the vessel. 2 Na O H + 2 Cl=Na CI + H2O + Na O CI The hypochlorous acid acted on by the excess of sodium hydrate present forms sodium hypochlorite, and this in its turn gives hypochlorous acid: Na O CI + 2 CI + H20=Na Gl + 2 H O CI Some lime salt and some organic matter should be present. This coats the cathode, so that there can be no reduction to sodium chloride by contact with the metal of the cathode. ELECTRIC BLEACHING. 789 The electrodes are of platinum wire gauze. The Leclanche bleaching apparatus has carbon anodes and iron cathodes. The carbons are enclosed in a sort of in- verted bell, whose sides are asbestos diaphragms, closed at the top by metal tops, whence the gaseous chlorine escapes through a tube connected to the apex. Sodium hydrate is withdrawn from the compartment outside the asbestos, where it forms at the cathodes. The chlorine is introduced into the vessel receiving the sodium hydrate and the liquid is agitated by a paddle wheel so as to mix gas and liquid thoroughly and produce sodium hypochlorite, ready for use as a bleach- ing liquor. It is an object to remove the hypochloric solution from contact with the anode, in the case of the apparatus first described, as soon as possible; action of the anode on sodium iiypochlorite oxidizes it to sodium chlorate, Na CI O3, which is of no value as a bleaching agent. CHAPTER XLVI. PUPIN'S COILS. LIFTING MAGNETS. Piipin Coils. — Long distance telephony has been made pos- sible, it is not too much to say, by what is called Pupinization, the installation on the line of inductances. The idea was conceived by Prof. M. I. Pupin about as far back as 1899, and it took some ten years to bring it to perfection, and it involved elaborate mathematical investigation to develop the application of the invention. A constant, called the Pupin constant, enters into the calculations. Long lines of telephone circuits of high or moderate ca- pacity act to damp the telephonic waves. Capacity, induct- ance, resistance, length of line and leakage all enter as factors in the transmission. The leakage is not especially important. If the inductance is small the enunciation is loud and indistinct, and the overtones and hissing sounds are damped. On the other hand, high inductance damps all the sound-producing waves to more nearly the same degree, and the volume or loudness of the sound is reduced. In the Pupin invention self-inductance is added at intervals all along the line. This is done by connecting coils of high in- ductance at determined distances to the transmission line. The cores of the loading coils, as the self-inductance ele- ments are called, are ring shaped for land wires, - and are made of fine iron wire of specially selected iron, whose con- stants insure the best results. The wire is of 0.004 inch diameter; the windings are insulated from each other to reduce hysteresis. A core receives two windings, one for each side of the transmission circuit, in the directions shown in the diagram. Each coil, it will be seen, magnetizes with the same polarity, so that there is no mutual induction added 790 PUPINS COILS, LIFTING MAGNETS. 791 to the self-induction. The coils are immersed in an insulat- ing compound, after being wound • with tape, and are put into iron cases. The installation of the coil in its case on a telephone pole is shown in the cut. On aerial lines there may be a coil for every 6 to 9 miles; on subterranean lines they are about five Fig. 594. — Pupin's Coil for Long Distance Telephone Lines. times more closely installed, from 1 14 to 2 miles apart, and in sea cables a sea mile is a standard distance. The dispo- sition of a coil in a submarine cable is shown in the cut. It only involves a slight swelling or enlargement in diameter of the cable at the proper intervals. The inductance of each coil may vary %;uite widely, accord- ing to requirements. A range from 0.02 to 0.2 henry is given as covering the maximum. The coils add greatly to the length of the line. On the line from New York to San Francisco, a metallic circuit of 6,8 00 miles, the coils contain 13,6 00 .miles of copper wire, which is added to the length of line. A. circuit provided with Pupin coils is said to be loaded. 792 ELECTRICIANS' HANDY BOOK. On a loaded circuit the insulation must be particularly good; the best types of insulators are used on the poles and pole arms, and the bridle wires have special insulation to pre- vent leakage. These are the wires connecting the coils to the line, and at first much trouble was experienced with them. The cable contains four wires. They are twisted in pairs, and then the two pairs are twisted around each other. This arrangement gives two through telephone circuits and one Fig. 595. — Mounting of Putin's Coil on Telephone Pole. phantom circuit. Two wires make up one side of a phantom circuit, so that the phantom current goes in the same direc- tion on both wires of the actual circuit. On the other hand, these wires have opposite currents on them, as regards the actual circuit. Eight telegraphic messages and three tele- phone messages can be sent simultaneously on the four wires of the cable. The cable or group of four twisted wires, as described, is called a quad. The twisting* is essential to enable them to be phantomed. PUPINS COILS. LIFTING MAGNETS. 793 Some of the factors of the U. S. transcontinental line are given here. The electric constants are referred to the mile: New York to San Francisco: Length miles, 3,4 00; re- sistance ohms, 4.95 per m. ; inductance henries, 0.0365 per m.; capacity micro-f*ds, 0.0091 per m.; attenuation, 0.0013, Fig. 596. — Pupin's Coil on Cable. Lifting Magnets. — These have attained quite an extensive application in the iron and steel industry. They are simply specially constructed electro magnets, used instead of hooks and chain slings to lift iron. They are made of great power, a range of from V2 to 1 V2 tons being standard. Fig. 597. — Lifting Magnets in Pairs. In use they are attached to the end of a lifting chain tackle. When iron or steel is to be lifted the tackle is paid out until the magnet is in contact, or virtually so, with the iron to be lifted. The iron adheres and can be lifted about and released as desired. To lift the iron the magnet is excited by a strong 794 ELECTRICTAXS' HA^^DY BOOK. current. On reversing the current the iron drops. A small magnet will lift 15 times its own weight; a magnet 2 to 3 feet in diameter will lift 8 to 12 times its weight, and a 3 to 5 foot one will lift only 5 to 6 times its weight. As many Fig. 598. — Case of Phenix Lifting Magnet. as six layers of plate iron can be lifted at once. By manipu- lation of the commutator they can be dropped off one at a time. Mica sheets are used to insulate the coils and they are protected by an insulating compound, sometimes introduced under a vacuum. This eliminates sweating and makes them waterproof. ■~--D -" ry Fig. 599. — Tngranic Lifting IMagnet. The Phenix Lifting Magnet has two concentric annular coils carried in a casing AC\ one in the space F, the other in E. The poles N and S and the casing are made of high permea- bility cast-steel. The annular well D separates the coils. The PUPIXS COILS. LIFTING MAGNETS. 795 coils are wound on brass forms so as to insure rigidity, as there is a tendency in the turns of the coils to move slightly as the current is switched on and off. Plates H hold the coils in place and protect them from injury; these coils are made of phosphor bronze. The Ingranic Lifting Magnet has a magnet steel body. A, with lifting lugs cast integral with it. The terminal box is also part of the casting. Pole shoes, D, are bolted to the frame or body. ^ is a flanged disk of corrugated manganese steel which protects the coils. The coil, of copper strip, is Fig. 600. — "Wilton Kramer" Lifting Magnet Lifting Pig Iron. wound on the spool C, Mica insulation is used as needed and waterproof insulation is introduced under vacuum. In shutting off the current a resistance coil is first con- nected across the terminals of the magnet windings, then the current is reversed and shut off. The object of doing this is to prevent a discharge current from piercing the insula- tion. Such a discharge would otherwise take place on ac- count of the high inductance of the magnet. Lifting magnets are on record as reducing the expense of unloading pig iron to one-eighth of the expense when done 796 ELECTRICIAN^ 8' HANDY BOOK. by hand. Two 62-incli magnets and two men with hoisting apparatus unloaded 1,900 tons in 10% hours, with an average lift of 1.53 ton. To do this by hand would require 48 hours' work of twenty-eight men. In another case, 55 tons of pig iron were lifted in half an hour by one 5 2 -inch magnet, with a capacity of 0.84 ton. The reversal of the current is to effect the detachment of the load by demagnetizing the massive core. SUPPLEMENTARY INDEX. PAGE Aerials 741 Acuity, Visual 757 Aluminum Furnace 776 .\ntennae 741 Atmospheric Nitrogen, Fix- ation of 782 Auer Process of Attaching- Lamp Filaments 752 Bleaching-,, Electric 787 Blondel's Carbons 769 Brilliancy, Intrinsic 771 Calcium Carbide 779 Carbide, Calcium 779 Carborundum and Graphite 778 Coherer, Italian Navy 736 Coherer, Lodge-Muirhead. .. 736 Coherer, Marconi's 735 Coherer, Stone 736 Coils, Pupin 790 Contact Detectors 739 Cooper-Hewitt Arc, Start- ing 756 Cooper-Hewitt Lamp, Data of 758 Cooper-Hewitt Lamp, Po- tential Drop in 758 Cooper-Hewntt Mercury Va- por Lamp 755 Cooper-Hewitt Quartz Tube Lamp 759 Cooper-Hewitt Tube 757 Coupling, Capacity 743 Coupling, Conductive 742 Coupling, Electric 743 Coupling, Inductive 742 Coupling, Magnetic 742 Couplings 742 Couplings, Close 743 Couplings, Combined 743 Couplings, Loose 743 Crystal Detectors 739 Detector, Fleming Valve. . . 737 Detector, Heterodyne ...737-738 Detectors 736 Detectors, Contact 739 Detectors, Crvstal 739 Detectors, Electrolytic. .737-738 PAGE Detectors, Fessenden 737-73& Discharger, Disc 740 Electric Bleaching 787 Electric Furnaces 772 Electric Steel Furnaces.... 772 Efficiency, Utilization 770 Fessenden Detectors. .. .737-739 Filament, Metallic, Lamps.. 749 Filament, Osmium 753 Filaments, Auer Process for 752 Filaments, Helion 753 Filaments, Carbon Metallic 753 Filaments, Duration of Ef- ficiency of Tungsten 752 Filaments, Molybdenum Sup- ports for 752 Filaments, Tantalum 754 Filaments, Tungsten 749 Fixation o f Atmospheric Nitrogen 782 Flame Arc Lamps 766 Fleming Valve Detector.... 737 Furnace, Aluminum 776 Furnace, Girod 77a Furnace, Girod, in Bethle- hem 774 Furnace, Hering Ore Re- ducing 776 Furnace, Kjellin 775 Furnace, Roechling 775 Furnace, Swedish Ore 776 Furnaces, Electric 772 Furnaces, Electric Steel. . . 772 Furnaces, Sodium 780 Girod Furnace 773 Girod Furnace in Bethle- hem 774 Goldschmidt's Tone Wheel.. 739 Graphite, Carborundum and 778 Helion Filaments 753 Hering Ore Reducing Fur- nace 776 Hysteresis Receivers 735 Incandescent Lamps, Lead- ing in Wires 751 797 798 SUPPLEMENTARY INDEX, PAGE Incandescent Lamps, Gas Filled . 750 Ingranic Lifting- Magnet.. 7«5 Kjellin Furnace 775 Lamp, Action of Quartz Tube 762 Lamp, Cooper-Hewitt Mer- cury Vapor 755 Lamp, Cooper-Hewitt Quartz Tube 759 Lamp, McFarland Moore 764 Lamp, Mercury Arc for Al- ternating- Current 764 Lamps, Data of Quartz Tube 763 Lamps, Flame Arc 766 Lamps, Gas-Filled Incan- descent 750 Lamps, Magnetite Arc 769 Lamps, Metallic Filament. . 749 Lamps, Neon 766 Lamps, Photometering Gas- Filled 752 Lamps, Tantalum Filament 754 Lamps, TJltra-Violet Rays in Quartz Tube 763 Lamps, Vacuum Tube 755 Leading- in Wires 751 Lifting Magnets 793 Lumen, Lux and 769 Lux and Lumen 769 McFarland Moore Lamp 764 Marconi Receiving- Plant.. 744 Marconi Sending Plant.... 744 Marconi's Coherer 735 Magnet, Phenix Lifting... 794 Magnet, Ingranic Lifting. .. 795 Magnets, Lifting 793 Magnetite Arc Lamps 769 Mercury Vapor Lamp, Cooper-Hewitt 755 Molybdenum Supports for Filaments 752 Neon Lamps 766 PAGE Nitrogen, Atmospheric, Fix- ation of 782 Osmium Filaments 753 Overshooting- 752 Phenix Lifting- Magnet.... 794 Photometering- Gas - filled Lamps 752 Pupin Coils 790 Quartz Tube Lamp, Con- struction of 759 Receivers, Hysteresis .735 Receiving Plant, Marconi.. 744 Receiving Arrangement for Wireless Telephony 748 Roechling Furnace 775 Sending Arrangement for Wireless Telephony .... 747 Sending Plant, Marconi.... 744 Sodium Furnaces 780 Spark Gaps 740 Starting- Characteristic . . . 762 Starting the Cooper-Hewitt Arc 756 Stationary Characteristic... 762 Tantalum Filaments 754 Telephonv, Wireless 746 Tone Wheel, Goldschmidt's 739 Tungsten, Duration of Ef- ficiency of Filament of . . . . 752 Tungsten Filaments 749 Ulrich's Integrating- Sphere 769 Utilization Efficiency 770 Vacuum Tube Lamps 755 Visual Acuity 757 "Wireless Telephony 746 Wireless Telephony, Re- ceiving- Arrangement for 748 Wireless Telephony, Send- ing- Arrangement for.... 747 Consult Supplementary Index on Page 797 INDEX. Access of Air 439 Accidents to Motors 415 Acid Depolarizers, Sulphuric and Isitric 110 Acid Storage Batteries, Zinc. 140 Action of a Circuit 66 Action of Arc Lamp on Con- stant-Potential , Circuit 553 Action of a Storage Battery. 125 Action of Conductor 51 Action of Current on the Mag- net 202 Action of Currents, Mutual... 199 Action of Drum Armature.... 233 Action of Inclosed-Arc 546 Action of Mag-net Poles on Each Other 195 Adjusting Lamps 559 Adjusting Weight 552 Adjustment, Clutch Stop 560 Air-Blast Cooling 386 Air Blast Effect of 538 Air Gap and Sparking 310 Air Pump, The Mercury 528 Air Switches 447 Air-Vane Damping 461 Alarm, Ground 457 Alternating-Current Arc 544 Alternating - Current Arc Lamps, Distribution of Light of 544 Alternating - Current Arc Lamps, Efficiency of 545 Alternating-Current Arc, Pow- er-Factor in 544 Alternating-Current Armature Winding, Principle of 350 Alternating- Current Circuit, Counter and Forward E.M.F, in 334 Alternating - Current Circuit, E.M.F. in 333 Alternating Current, Dirrct Current from 300 Alternating Current Distribu- tion 504 Alternating-Current Dynamo, Elementary Idea of . .*! 219 Alternating Current, Genera- tion of : 348 Alternating-Current Ground In- dicator 456 Alternating-Current Lightning Arrester, Low-Equivalent. . .' 521 Alternating-Current Potential Regulator 454 Alternating Current, Y Connec- tions for 506 Alternating E.M.F 316 Alternating E.M.F. and Cur- rent, Cause of Form of 319 Alternating E.M.F. and Cur- rent Curves 320 Alternating E.M.F. and Cur- rent, F'orm of 318 Alternating E.M.F. and Cur- rent, Production of 318 Alternating Quantities, Multi- plication of 344 Alternating Quantities, Sum- mation of 342 Alternator Brushes 435 Alternator, Inductor 356 Alternators, Field Magnets of 436 Alternators in Step 405 Alternators, Trouble in Rotors of 435 Alternator Winding, Six-Wire Connection of Three-Phase . 359 Aluminium Negative Plate. . . . 101 Amalgamation of Primary Bat- tery 96 American Storage Battery.... 134 American Wire Gauge 85 Ammeters 467 Ammeter, Shunted 469 Ammeter, Total-Current Sole- noid 467 Ammeter, Transformer 470 Ammonium-Chloride Batteries. 119 Ampere, Analogy for the 53 Amperes and Coulombs, Cur- rent 52 Ampere's Rule 203 Ampere's Rule Adapted to In- duction 213 Ampere's Theory of Magnetism 200 Ampere's Theory of Terrestrial Magnetism 201 Ampere Turns 176 Ampere Turns and Lines of Force, Relation between.... 183 Analogies of Drop of Poten- tial 60 Analogy for the Ampere 53 Angular Measurement 32 Annealing 181 Annunciator Lamp, Telephone. 709 Annunciator, The Mechanical. 708 Annular Chambered Magnet... 191 Anodes 660 Anode and Cathode, Position of 670 Antennae and Connections.... 735 Anti-Parallel Systen^s 488 Apparatus, Large Plating. . . . 661 Apparatus, Photometer 567 Apparatus, Receiving, Tele- phone 733 799 800 ELECTRICIANS' HANDY BOOK. Apparatus, Simple Plating. . . . 661 Apparatus, Transmitting, Tele- phone 733 Appliances and Generators in Circuits 65 Appliances, Circuits without. . 64 Application of Circular Mil System 84 Application of Lenz's Law... 214 Arc, Alternating-Current 544 Arc and Incandescent Lamp Circuits 472 Arc Blowpipe, The Electric... 727 Arc, The Direct-Current Open. 540 Arc, Direct Heating by the Electric 727 Arc. Distribution of Light in Direct-Current Open 541 Arc, Heat of 536 Arc, Hissing 543 Arc Lamp, Direct Photometer- ing of 572 Arc Lamps, Features of Series or Constant-Current System for 475 Arc Lamp on Constant-Poten- tial Circuit. Action of 553 Arc Lamp, Photometry of . . . , 576 Arc Lamp, Resistance Coil in. 553 Arc Lamps, Commercial Rating of 542 Arc Lamps, Constant-Current or Series 551 Arc Lamps, Constant-Potential 555 Arc Lamps, Distribution of Light of Alternating-Current 544 Arc Lamps, Eflaciency of Alter- nating-Current 545 Arc Lamps, Noise in 545 Arc, Length of 545 Arc Length, Voltage Drop and 538 Arc Light, Candle-Power in, Watts per 579 Arc Light Carbons 530 Arc Light, Efficiency of 537 Arc Light, Quality of 580 Arc Lights, Distribution of Light from 582 Ave. Power Consumed in 538 Arc. Power-Factor in Alternat- ins^-Current 544 Arc Proper, Light Given by.. 543 Arc ProDer, Resistance of . . . . 537 Ave, Striking the 535 Ave, The Voltaic 535 Arcs, Resistance of Longer... 543 Arcs, Resistance of Short.... 543 Area of a Circular Mil, Exam- ples of 84 Arithmetic and Mathematics. 17-40 Armature, Action of Drum... 233 Armature and Core 221 Armature, Balancing of 4*^8 Armature out of Center 429 Armature, Centering: of 428- Armature, Closed-Coil Direct- Current 223 Armature, Commutator Connec- tions of Ring 227 Armature Connections, Drum. 350 Armature Core, Action of Field Poles on 265 Armature Core, Increasing E.M.F. by Adding 221 Armature Cores, Eddy Cur- rents in 271 Armature, Current in Ring. . . 229- Armature, Disk 303 Armature, The Drum 231 Armature, End Leakage of Lines of Force in 272 Armature, Mounting of Ring.. 230 Armature, Multipolar Ring. . . 231 Armature, Open-Wound Four- Part Ring 230- Armature, The Pacinotti 225 Armature, Pole Single-Phase.. 354 Armature Polarity Due to Windings 265 Armature Reaction Diagrams. 266 Armature in Rotary Field .... 364 Armature Rotation, Reversal of 404 Armature Running 428 Armature Shaft, End Motion in 423 Armature, Short Circuits in. . 423 Armature, Single-Phase 348 Armature, Sixteen-Conductor Bipolar 235 Armature, Squirrel Cage 367 Armature, Spindle or H 222 Armature Winding, General Formulas for Drum 246 Armature Winding^ Nomencla- ture for Drum 246 Armature Winding:, Principle of Alternaling-Current .... 350 Armature Windings, Break in. 422 Armature Windings, Drum... 233 Armature Windings and Fi*ame, Short Circuits between 435 Armature. Winding a Drum.. 241 Armature Winding, Simple Sys- em of 234 Armature Windings, Laying out Drum " 243 Armatures, Cores of Ring. . . . 228 Armatures, Modern Types of Closed-Coil 226 Armatures, Open Coll 222 Armatures, Pole 301 Armatures, Various 193 Armntures with Formed Coil, Winding 301 Arranirements of Batterips. . . . 122 Arrestpr, Comb or Saw-Tooth, Li Calculations, Voltage 92 Calibration 524 Calorimeter. Here's 9^^"* Camacho Cascp'^p Batterv.... 107 Candle, .TablorhVoff " 561 Candle. Standard. Ensfli^i"'. . . . 567 Candle-Power in Arc Light, Watts per 579 Candle-Power in Incandescent Lamps, Watts per 580 INDEX, 803 Candle-Power of Incandescent Lamps 57G Cand]e-I*ower, Spherical 574 Capacity 47, 388 Capacity, Composition of Re- sistance, Inductance, and. . 343 Capacity, Examples of 49 Capacity of Cable, Determina- tion of 648 Capacity of Condensers 44 Capacity, Ohmic Equivalent of - Reactance of 337 Capacity, Reactance of ...... . 338 Capacity, Specific Inductive. . . 48 Capacity, Storage 130 Carboning a Lamp 561 Carbonization 524 Carbon-Feed Lamps 549 Carbon Holder, Globe and 547 Carbon Holders 550 Carbon Negative Plates 103 Carbon Transmitters, Loose... 681 Carbons, Arc Light 539 Carbons, Duration of 545 Carbons, Inclosed-Arc.547, 559, 560 Carbons, Positive and Nega- tive 535, 560 Carbons, Quality of 538 Carbons, Wearing of 539 Cardew Voltmeter 463 Car, Economical Running of. 597 Car, Jerking 601 Car Heating 601 Car, Leaving the 598 Car Motor, Construction of . . . 588 Car Motors, Horse-Power of. . 586 Car, Reversing the. . ." 598 Cascade Battery, Camacho. . . . 107 Cathode, Position of Anode and 670 Causes of Lag and Lead 341 Cell 96 Cell, Modification of Gravity. 116 Cell, The Primarv Battery 93 Cells, Counter, E.M.F.... 165 Cells. Insulation of Storage Batterv 161 Cells of Storage Battery. End 164 Cells, Short-Circuiting of Sin- gle 152 Cells, Storage Battery 161 Center, Armature out of 429 Centering of Armature 428 Change, Graphic Representa- tion of Rate of 323 Change, Rate of 323 Changing Voltage 400 Characteristic Curves. See Curves, Characteristic. Charge, E.M.F. and the Static 58 Charge, First 150 Charge of Storage Battery, The 145 Charging 43 Charging, English Rule for... 151 Charging from Lighting Cir- cuits, Connections for 157 Cnarging Storage Battery. . . . 166 Chemical Action of Storage Batteries 131 Chemical Decomposidon, Cur- rent Strength and 90 Chemical Decomposition, E.M. F. in 91 Chemical Decomposition, En- ergy in 91 Chemical Equivalents, Atomic Weights and 89 Chloride Battery 136 Choke Coils 505 Choking of Transformer 376 Circle, Generating 322 Circle, Interpretation of the Generati^"* .'^'"*'^. Circuit, Action of a 66 Circuit and luduc.ance, Turns of a 334 Circuit. Action of Arc Lamp on Constant Potential 553 Circuit-Breaker 150 Circuit-Breaker, Magnetic Re- lease Underload 453 Circuit-Breaker. Mechanical Re- lease Fnderloaf^ . . . 453 Circuit-Breaker, Reverse Cur- rent 4o4 Circuit-Breaker, Switch Boxes and, on Car 590 . Circuit-Breakers as Switches.. 454 Circuit-Breakers, Con'-bin'^d . . 454 Circuit-Breakers, Overload.... 451 Circuit-Breakers, T^ndor!oa:l. . 4r)2 Circuit, Bridged Telephone. . . 696 Circuit, Condensers in a 64 Circuit, Constant Current. ... 78 Circuit, Constant Potential... 7S Circuit, Constitution of a . . . . 63 Circuit, Counter and P'orward E.M.F. in Alternating-Cur- rent 334 Circuit, Discharge of Storage Batterv on Open 144 Circuit, E.M.F. in Alternating- Current 333 Circuit, Energy and the Mag- netic 172 Circuit, Exciting Series Coils from Main 261 Circuit, Nature of the Mag- netic 174 Circuit, Outer 68 Circuit, Permeance of a ^'ag- nctic 184 Circuit, Polaritv of the 159 Circuit, Qualities of a 331 Circuit, Series Telephone 695 Circuit, Short 69 Circuit, Short Circuits in Out- er 432 804 ELECTRICIANS' HANDY BOOK. Circuit, The Electric 63 Circuit, Tlie Magnetic 172 Circuit. Tliree Elements in a . 74 Circuit, Three Factors of Mag- netic 175 Circuit without Resistance . . 71 Circuits. Arc and Incandes- cent Lamp 472 Circuits, Appliances and Gene- rators in 65 Circuits. Calculation of Resist- ance of Parallel 81 Circuits. Connections for Charging from Lighting. ... 157 Circuits, Independent 492 Circuits, Open and Closed. ... 64 Circuits without Appliances. . 64 Circular Developments 238 Circular Functions, Numerical Value of 34 Circular Mil, Area of 84 Circular :Mil System 83 Circular Mil Sj'stem, Applica- tion of 84 Classification 483 Clerk Maxwell's Rule 213 Closed Circuits, Open and.... 64 Closed-Coil Armatures, Modern Types of 226 Closed - Coil Direct - Current Armature 223 Closet Svs^em 484 Clouds. E.M.F. in Thunder... 58 Clutch. The 548 Clutch Stop Adjustment 560 Coherer, Branly's 732 Coherer. Marconi's 736 Coil and Electro-Magnet, Mag- netizing by 197 Coil, Damping 460 Coil, Dimensions of Telephone Induction 686 Coil, Disconnecting or Open- ing Shunt 260 Coil. Effect of Independent Ex- citation of Shunt 260 Coil, Effect of Telephone In- duction 688 Coils, Arrangement of Resist- ance 627 Coils, Choke 505 Coils, Direction of Current In- duced in 217 Coils from !Main Circuit, Excit- ing Series 261 Coils, Heating of Field 420 Coils in Arc Lamp Rpsistance. 553 Coils in Bracket Telephones, Induction 688 Coils in Comnound Dynamos, Excitation of Fi'^ld." 260 Coils. Modern Arrangements of Resistance 6*^8 Coils, Pancake 381 Coils, Practical Notes on Re- sistance 631 Coils, Proportional 636 Coils, Reactance or Economy. 545 Coils, Repeating 705 Coils, Resistance 626 Coils, Resistance, Spools for. . 631 Coils, Separate Excitation of Shunt 261 Coils, Telephone Induction ... 684 Coils, Winding Armatures with Formed 301 Collecting or Slip Rings . . 219 Collector Rings 419 Colors, , Distinguishing, of Wires 722 Commercial Rating of Arc Lamps 542 Commutator Bars. Bad Con- tacts between Winding and. 419 Commutator Bars, Loose 420 Commutator Connections. .238, 245 Commutator Connections, De- velopment of 240 Commutator Connections of Ring Armature 227 Commutator Construction 303 Commutator, Filing 433 Commutator, Material of .... 419 Commutator, Oval 420 Commutator, Position of 305 Commutator, Sandoanering. . . 433 Commutator, Smoothing .... 433 Commutator, Sparking of 424 Commutator Surface, Gummy or Sticky 420 Commutator Surface, Lubricat- ing '. 420 Commutator, Temperature of. . 419 Commutator. Turning down... 432 Compensated Voltmeter 470 Compensating Resistance 61^ Compensator. Inductance .... 471 Compensator, Ohmic 471 Compensator, Starting 36R Compensators 409, 470 Composition of Resistance, In- ductance and CaDacitv 343 Compound Dvnamos. Excita- tion of Field Coils in 260 Comnound Dvnamos, Parallel Coupling of 402 Compound Dynamo. Wrong Connections in 4R2 Compound Winding 256 Compound Winding Calcula- tion, Examnle of 259 Compound Winding, Long- Shunt 256 Cornnound Winding, Short- Shunt 256 Co^nonnd Wound Dynamos, Self-R emulation of 257 Compounding, Over- 259 INDEX. 805 Concentric Magnets in Lamps. 550 Condenser, Earthing a 44 Condenser, Single Surface 45 Condensers 42 Condensers, Capacity of 44 Condensers in a Circuit 64 Conditions for Inducing Elec- tric Energy 206 Conditions of Sensitiveness... 636 Conductance 69 Conductance and Cross-Sec- tional Area of Conductors.. 83 Conductibility 69 Conduction, Electrolytic 73 Conduction Interference 714 Conductivity 69 Conductor, Action of a 51 Conductor, Calculation for Conical 486 Conductor in a Field of Force, Motion of 169 Conductor, Inductance React- ance in Subdivided 336. Conductor, Lines of Force Pro- duced by a Curved 169 Conductors, Conductance and Cross-Sectional Area of . . . . 83 Conductors, Cylindrical and Conical 485 Conductors, Eddy Currents in. 272 Conductors, Electrolytic 66 Conductors and Non-Conduc- tors 50 Conductors, Size of, for Plat- ing 669 Conical Conductor, Calculation for 486 Conical Conductors, Cylindri- cal and 485 Connection, Delta or Mesh.361, 507 Connection in Cable, Making Branch 653 Connection of Stations 734 Connection, Y or Star 359 Connections, Antennae and, in Wireless Telegraphy 735 Connections, Auxiliary Feeder 495 Connections, Booster 406 Connections, Booster and Stor- age Battery 410 Connections, Commutator. 238, 245 Connections, Development of Commutator 240 Connections, Drum Armature. 350 Connections, Examining, on Cars 599 Connections for Alternating- Current, Y 506 Connections for Charging from Lighting Circuits 157 Connections in Compound Dynamos, Wrong 432 Connections, ]Making Storage Battery 7. 161 Connections, Line 362 Connections, Multipolar Dyna- mo 264 Connections of Ring Armature, Commutator 227 Connections, Switchboard .... 712 Conservation of Electricity.... 57 Constancy of Magnetism 199 Constant-Current Circuit 78 Constant-Current or Series Arc Lamps 551 Constant-Current System 472 Constant-Current System for Arc Lamps, Features of Series or 475 Constant-Current Transform- ers 7 387 Constant, Determination of the 620 Constant, Hysteretic 186 Constant of Galvanometer.... 620 Constant-Potential Arc Lamps 555 Constant- Potential Circuit.... 78 Constant-Potential Systems... 473 Constants, Useful 35 Constitution of a Circuit 63 Contacts between Winding and Commutator Bars. Bad 419 Control, Crocker-Wheeler Speed 414 Controller Points 593 * Controller, Rheostat 596 Controller, Series-Parallel . . . 594 Controller Troubles on Cars.. 600 Controllers 592 Construction, Advantages of Multipolar 252 Construction, Multipolar 350 Conventional Representations of Machines 264 Converter, Functions of a Ro- tary 394 Converter in Three-Wire Sys- tem, Rotary 393 Converter, Rotary 390 Converter, Starting a Rotary.. 394 Cooking and Domestic Heat- ing, Electric 723 Cooking, Power Required for. 723 Cooling, Air-Blast 386 Cooling, Oil 384 Cooling, Water ■ 385 Copper Loss, The 584 Copper-Plating 662 Copper, Saving in 497 Copper Storage Batteries 139 Core, Action of Field Poles on Armature 265 Core, Armature and 221 Core Grooves 301 Core. Increasing E. M. F. by Adding Arm>ature 221 Core, Permeance of Ring 228 Core Transformers 380 Cores. Eddy Currents in Arma- ture 271 806 IL^rjEOTRICIANS' HAXDY BOOK. Cores of Ring Armatures 228 Cotton Waste 439 Coulomb, Hj'drogen i^iberated by the 88 Coulomb, Water Decomposed by the 88 Coulombs, Current, Amperes and . 52 Counter Electromotive Force. See Counter E. M. F. Counter E. M. F 333, 530 Counter and For\Yard E. M. F. 173 Coimter and Forward E. M. F. in Alternating-Current Cir- cuit 334 Counter E. M. F. Cells 105 Counter E. M. F., R. I. Drop and 79 Counter E. M. F., Drop Cal- culation, Example of. .... . 82 Couple 96 Coupling of Compound Dyna- mos. Parallel 402 Coupling of Shunt Dvnamos, Parallel 402 Critical Point of Shunt- Wound Dynamo 281 Crocker-Wheeler Speed Con- trol 414 (^rompton-Howell Battery . . . 135 Crushers 41^ Current from Alternating Cur- rent. Direct 390 Current, Amperes and Coul- ombs 52 Current and Rate Units 50 Current Armature, Closed-Coil Direct 223 Current. Cause of Form of Alternating E. M. F'. and.. 310 Current Circuit, Constant.... 78 Current. Critical ' 279 Current Curves, Alternating E. :\r. F. and 320 Cuvrent Curves, Drawing E. M. F. and 321 Current Curve. E. ^l. F. and. 310 Current. Direction of a 55 Cuvmiit Distribution, Alternat- ing 504 Curvont Dvnamo, Elementary Trioa of ^Altprnating 219 Cuvv^nt Dvnamo. Elementary Td«a of Direct 220 Ciirvorit in Electroplating, Rpfmlation of 001 Cuvrput fnr Electroplating. . . 0^0 Current, Etbor and 107 Current, Fields of Force and T-ines of Force Due to. . . . 55 Current, Form of Alternating E. Af. F. find 318 Current, Generation of Alter- nating 348 Current, Induced, Develop- ment of 240 Current Induced in Coils, Di- rection of 217 Current Intensity in Plating. . 669 Current Measurement with Potentiometer 040 Current on the Magnet, Ac- tion of a 202 Current, Production of 52 Current. Production of Alter- nating E. M. F. and 318 Current in Rotary Transform- er, Relations of Voltage and 392 Current, Reversing Direction of 410 Current, Speed of a 54 Current, Standard Series Lighting 477 Current Strength 53 Current Strength and Chemi- cal Decomposition 90 Current, Three-Phase 347 Current, Time Required to Pro- duce a 52 Current to Drop. Relation of. 491 Current, Two-Phase 346 Current, Wattless 246 Current, Y Connections for Alternating 506 Currents in Armature Cores, Eddy 271 Currents in Conductors, Eddy. 272 Currents in Core Disks. Eddy. 271 Currents, Foucault or Eddv. 215, 271, 429 Currents, Mutual Action of. . 199 Currents in Pole Pieces. Eddy. 272 Curve, Arrival 54 Curve, Drooping; Characteristic 275 Curve, E. M. F'. and Current. 316 Curve of Shunt Dynamo, Total Characteristic 203 Curve, of Shunt Dynamo, Total Current Characteristic .... 282 Curve, Sine 321 Curve,. Vector Diagram of Sine 325 Curved Conductor. Lines of Force Produced bv a 169 Curves, Alternating E. M. F. and Current 320 Curves, B and H 178 Curves, Characteristic. 258, 273, 276 Curves, Determination of .... 182 Curves, Drawing Cliaracter- istic 277, 321 Curves, General Notes on Characteristic 279 Curves, Hysteresis 18.") Curves, Internal Characteristic. 277 Curves. Interpretation of B and H 179 Curves. Interpretation of Characteristic 276 i INDEX. 807 Curves, Ohm-Volt 284 Curves, I'ermeability 180 Curves. lV)wer 345 Curves, Shunt-Wound Dynamo Characteristic 280 Curves, Types of Character- istic 273 Cutting Lines of Force. 205, 209, 224 Cut-Outs 5G0 Cut-Out, Automatic 3 5() Cut-Out, Film 478 Cut-Outs, Board and 59G Cut-Outs. Overload and Under- load 448 Cycle 316 Damping Air-Yane 461 Damping Coil 460 Daniell Battery Ill, 112 Daniell Battery, Sand Type of 114 Dash Pots 550 Data with Storage Batteries, Manufacturer's 144 Deflection, Direction of .... 637 Degree System 321 Delta or Mesh Connection.361, 507 Demagnetizing Turns 268 Density, Field 176 Density of a Field of Force.. 172 Densities of Field, Varying. . 267 Depolarizers. Sulphuric and Nitric Acid 110 Depolarizing Mixtures in Pog- gendorfC's Battery 109 Deposits. Backing Up 675 Deprez-D'Arsonval Galvanom- eter 611 Detection of the Field of F'orce 168 Determination of Curves.... 182 Determination of Discharge of Storage Battery 145 Dean's Common Battery Sys- tem 698 Decade Plan, The 629 Decomposition, Current Strength and Chemical .... 90 Decomposition, E. M. F. in Chemical 91 Decomposition, Energv in Chemical 91 Developments, Circular 238 Diagram of Sine Curve, Vector 325 Diagrams, Armature Reaction. 26() Dielectrics 48 Difference of Potential, E. M. F. and 61 Diffractive Photometer 574 Dimensions of Telephone In- duction Coil 686 Dip Batteries 108 ^Direct-Current Armature, Closed-Coil 223 Direct-Current Dynamo, Ele- mentary Idea of 220 Direct-Current from Alternat- ing Current 390 Direct-Current Ground Indi- cator 456 Direct-Current Motor and Torque 285 Direct-Current Open Arc..'... 540 Direct-Current Open Arc, Dis- tribution of Light in 541 Direction of Current Induced in Coils 217 Direction of Current, Pteversing 416 Direction of Deflection 637 Discharge of Storage Batteries 144 Discharge of Storage Battery, Determination of . 145 Discharge of Storage Battery on Open Circuit 144 Disconnecting or Opening Shunt Coil 260 Disintegration of Storage Bat- tery 153 Disk Armature 303 Disk, Bunsen 564 Disk, Leeson 565 Disk Winding of Transform- ers 387 Disk Windings 357 Disk-Wound Transformers . . . 380 Disks, Mounting Photometer. . 565 Disks, Eddv Currents in Core. 271 Distortion, Field 265 Distributing Boards 703 Distribution, Alternating-Cur- rent 504 Distribution, Calculations for Series 475 Distribution, Disadvantages of Parallel 482 Distribution. Elementary Case of Parallel 482 Distribution, Limitations of Series 474 Distribution, Objections to Series 481 Distribution of Light 534 Distribution of Light of Alter- nating-Current Arc Lamps. . 544 Distribution of Light from Arc Lights 582 Distribution of Light in Di- rect-Current Open Arc... 541 Distribution of Light from In- candescent Lamps 583 Distribution. Parallel 481 Distribution, Series 472 Driving Points 593 Drop and Arc Length, Volt- age 538 Drop and Fall of Potential.. 79 Drop Calculation, Example of Counter E. M. F 82 808 ELECTRICIANS' HANDY BOOK, Drop Calculations, Examples of R. I 82 Drop in Parallel System, Po- tential 482 Drop of Potential 60 Drop of Potential, Analogies of 60 Drop, Relation of Current to. . 491 Drop, Voltage 536 Drum Armature 231 Drum Armature, Action of . . . 233 Drum Armature Connections. 350 Drum Armature, Winding a. . 241 Drum Armature Winding, Gen- eral Formulas for .". 246 Drum Armature Winding, Nomenclature for 246 Drum Armature Windings.233, 243 Dry Batteries 121 T)uties, Motorman's 597 Dynamic and Static Electri- city 56 Dynamo, Action of Separately- Excited 262 Dynamo and Motor, Inter- changeability of 218 Dynamo and Motor, Reversibil- ity of 285 Dynamo, Balancing 501 Dynamo Characteristic Curves, Shunt-Wound 280 Dynamo, Critical Point of Shunt-Wound 281 Dynamo Connections, Multi- polar 264 Dynamo-Electric Generators. .. 218 Dynamo, Elementary Idea of Alternating-Current 219 Dynam.o, Elementary Idea of Direct-Current 220 Dynamo Frames, Earthing . . 431 Dynamo, Modern Multipolar.. 250 Dynamo or Motor, Tempera- ture of 439 Dynamo, Separate-Circuit . . . 263 Dynamo, Starting a 428 Dynamo, Telephone Receiver, a 212 Dynamo, Three-Brush 499 Dynamo, Total Characteristic Curve of Shunt 283 Dynamo, Total Current Char- acteristic Curve in Shunt. 282 Dynamos as Boosters, Motor. 408 Dynamos and Magnetos, Reg- ulation of Separately-Ex- cited 262 Dynamos, Excitation of Field Coils in Compound 260 Dynamos, Field Magnet for Multipolar 310 Dynamos, Field Winding of. . 252 Dynamos, Parallel Coupling of 402 Dynamos, Self-Regulation of Compound-Wound 257 Dvnamos, Separatelv- and Self- Excited 263 Dynamos, Series Winding of. 252 Dvnamos, Shunt Winding of. . 254 Dynamos, Soft Steel in 181 Dynamos, Varieties of 219 Dynamos, AVrong Connections in Compound 432 Dynamometer, The 39 Dynamometer, Siemens's .... 622 Economy Coil, Reactance or. . . 545 Economy, Feeder 496 Edison's Meter 513 Edison's Storage Battery.... 140 Edison's Telephone 682 Eddy Currents, Foucault or, 215, 271, 429 Eddy Currents in Armature Cores 27:f Eddy Currents in Conductors. 272 Eddy Currents in Core Disks. 271 Eddy Currents in Pole Pieces. 272 Effective Values 328 Effective Values, Formulas for 330 Efficiency of Alternating-Cur- rent Arc Lamps 545 Efficiency of Arc Light 537 Efficiency of Electric Heating 723 Electric Arc, see Arc. Electric Circuit, The 63 Electric-Car Motor, The 584 Electric Cooking and Domestic Heating 723 Electric Energy, Conditions for Inducing 20'3 Electric Generators, Dynamo. 218 Electric Heating. Economy of. 730 Electricity, Conservation of. . 57 Electricity, Dynamic and Static 56 Electricity, Ether Waves Pro- duced by 50 Electricity, Quantity of, Mean- ing of 44 Electric Resonance 339 Electric Quantity, Storage of. 41 Electrodes, Moving 103 Electro-Chemical Equivalents. 89 Electro-Chemistry, Summary of 90 Electrolyte and Tests, Impuri- ties in 155 Electrolyte, Preparing the . . 155 Electrolyte, Specific Gravity Variation of Storage Bat- terv 146 Electrolvtic Conduction . . .(d6, 73 Electro-Magnet 188 Electro-Magnet, Magnetizing by Coil and 197 Electro-Magnet, Spiral 189 Electro-Magnet, Tractive Force of the 188 Electro^Maenets, F-Shaned... . 189 Electro-Magnetic Induction . . 205 IJMJJ^JL. Electro-Magnetic Tractive I'ower 192 Electrometer, Thomson or Kelvin Absolute 617 Electromotive Force. See E. M. F. Electromotive Force, Counter. See Coimter E. M. F. Electroplating 659 Electroplating, Current for... 660 Electroplating, Energy Absorb- ed in 659 Electroplating, General Prin- ciples of 659 Electroplating, Regulation of Current in 661 E M F 55 e! M.' F.*, Alternating* '.'.'/.'.'.'. 316 E. M. F. in Alternating-Cur- rent Circuit 333 E. M. F'. in Chemical Decompo- sition 91 E. M. F. the Cause of Current 59 E. M. F. and Current. Cause of F'orm of Alternating. . . . 319 E. M. F. and Current Curves. 316 E. M. F. and Current Curves, Alternating 320 E. M. F. and Current Curves, Drawing 321 E. M. F. and Current, Form of Alternating 318 E. M. F. and Current, Produc- tion of Alternating 318 E. M. F. and Difference of Potential 61 E. M. F. Drop Calculation, Example of 82 E. :M. F*. and Energy 56 E. M. F. Forward 173 E. M. F. by Increasing Turns, Increasing 221 E. M. F. Increasing, by Adding Armature Core 221 E. :\r. F. and the Static Charge 58 E. M. F., Counter. See Count- er E. M. F. E. M. F., Production of 56 E. M. F. in Thunder Clouds . . 58 E. M. F., Variations in Im- pressed 216 End Cells of Storage Battery. 164 End Motion in Armature Shaft 423 Energy Absorbed in Electro- plating 659 Energy in Chemical Decompo- sition 91 Energy, Conditions for In- ducing Electric 206 Energy Due to Hysteresis, Loss of 186 Energy, E. M. F. and 56 Energy and the Field of Force 174 Energy of the Field of Force, I'otentiai 173 Encrr;y and the Magnetic Cir- cuit 172 Energy Relations 209 Energy, Resistance and 70 ' English Rule for Charging. . . 151 E. P. S. Battery 135 Equalizer in Three-Wire Sys- tem, Storage Battery 50 Equivalents, Atomic Weights and Chemical 89 Equivalents, Electro-Chemical. 89 Equivalent of Reactance of Capacity, Ohmic 337 Equivalent of Reactance of In- ductance, Ohmic 336 Ether and Current 167 Ether, Luminiferous 40 Ether Waves Produced by Elec- tricity 50 Evolution, Gas 149 Excitation of Field Coils in Compound Dvnamos 260 Excitation of Shunt Coil, Ef- fect of Independent 260 Excitation of Shunt Coil, Sep- arate 261 External Resistance, Internal and 871 Factor, Form 329 Factor in Alternating-Current Arc, Power 544 Factor, Power 330 Factors of Magnetic Circuit, Three 175 Fall of Potential, Drop and.. 79 Faraday's Law 212 Faure's Battery 130 Faure-Sellon-Volckmar Battery, The 131 Feeder Connections, Auxiliary 495 Feeder Economy 496 Feeders 483. 493 Fiber Suspension 605 Field Coils, Heating of 429 Field Coils in Compound Dyna- mos, Excitation of 260 Field Densitv 172, 176, 268 Field Distortion 265 Field of Force, The 167 Field of P'orce, Building up the 173 Field of Force and Lines of Force Due to Current 55 Field of Force, Detection of the 174 F'ieid of Force, Energy and the 174 Field of Force, Iron and the. 175 Field of Force, Motion of Con- ductor in 169 Field of Force, Potential En- ergy of the 17S 810 ELECTRICIANS' HANDY BOOK. Field of Force. Varying . . . > 207 Field Ma.^•nets of Alternators. 436 Field Magnet for Multipolar Dynamos 310 Field Magnets, Laminated. . . 314 Field Poles 222 Field Poles. Action of on Arm- ature Core 265 Field. The Rotary .. .363. 364, 366 Field, Stray 184 Field, Varying Densities of. . 267 Field Winding, Break in 430 Field Winding of Dynamos. 252, 315 Field Winding, Short Circuits in 430 Field, Wrong Polarity of . . . . 417 Fields of Force in Practice. . 209 Fields, Windings for Multi- polar 237 Figure of Merit 622 Film Cut-Out 478 F'ilament. Occlusion of Gases by Incandescent Lamp .... 525 Filaments, Metallic 530 Filaments, Oxide 530 Filaments, Squirted 524 Filaments. Tamidine 523 Filing. Commutator 433 Fishing for Bell Wires 719 Fiye and Seyen-Wire Systems. 502 Flashing Incandescent Lamp Filaments 524 Flashing. Lowering of Resist- ance of Incandescent Lamp Filaments by 525 Flashing. Making Joints in Incandescent Lamps 525 Fleming's Rule 212 Floating Storage Battery. .166, 410 Forbes' Meter 514 Force. Cutting Lines of . .209, 224 Force of the Electro-Magnet, Tractive 188 Force, Electromotiye. See E. ]\r. F. Force. The Field of. etc. See Field of Force. F'orce. Leakage of Lines of. . 183 Force, Magnetic 176 Force, Memoria Technica for Lines of 171 Force.. Relation Between Amp- ere' Turns and Lines of.... 183 Force, Spreading of Lines of. 188 Force, Threading. Interlinking and Cutting Lines of 205 Force. TTtility of Conception of Lines of * 171 Form Factor 3'^9 Forming Storage Batteries. . 129 Form of Alternating E. M. F. and Current 318. 319 Formula, Bipolar Winding by. 247 F'ormula, Multipolar Winding by 248 Formulas, Bipolar Winding. . 247 Formulas for Drum Armature Windins-. General 246 Forward E. M. F.. Counter and 173 Foucault or Eddy Currents, 215, 271, 429 Foucault's Photometer 572 Frame, Short Circuits Between Armature Windings and . . 435 Frequency 316 Frequency, Lengtli of Waye and 319 Fuller's Mercury-Bichromate Battery 106 Function of a Storage Battery 128 Functions, Trigonometric .... 33 Fundamental Note 677 Furnaces, Electric 723 Fuses, Safety 441, 4-18 Fuses, Safety, in Cars 599 Galyanic Pile 97 Galvanometer 604 Galyanometer, Astatic 605 Galvanometer, Ballistic 610 Galvanometer, Constant of . . . 620 Galvanometer, Deprez-D'Arson- val 611 'Galvanometer, Magneto Bell as 655 Galvanometer. Reflecting. .606. 609 Galyanometer Resistance .... 622 Galyanometer, Shunts 618 Galvanometer. Shunt to the. . 636 Galvanometer. Simple 604 Galyanometer, Sine 616 Galvanometer, Suspension Fiber for 605 Galvanometer, Tangent 615 Galvanometer, Thomson or Kelvin 609 Galvanoscope 636 Galvanoscope Cable and Line Tests 650 Galvanoscope, Telephone as. . 653 Gas Battery, Grove's 126 Gas Evolution 149 Gassing 149 Gassing, Indications from . . 156 Gauge. American Wire 85 Gauges, Wire 85 Gauss. The 178 Generating Circle 322 Genorating Circle, Interpreta- tion of the 323 Generation of Alternating Cur- rent 348 G^n^^ra^or, ]\t^noneto 2r»0 Gen'^rator Without INTotiou . . 210 Generator and INIotor Con- nected 28.*) Generator, Separately-Excited. 261 INDEX. 811 Generators in Circuits, Appli- ances and 05 Generators, Dynamo-Electric. . 218 Gibbs' Batterj^ 105 Globe and Carbon Holder... 547 ^«k)be. The Inclosing 559 Gold-Plating 607 Gonld Storage Battery 132 Graduation of Voltmeter Scales 463 Gramme Ring. The 226 Graphic Representation of Rate of Change 323 Gravity Battery 114 Gravity Cell Modification of. 116 Gravity Variation of Electro- lyte, Specific 146 Greek Letters 34 Grenefs Battery 108 Ground Alarm 457 Ground, Bad 508 Ground, Hand Magneto Test for 657 Ground Indicator, Alternating- Current 456 Ground Indicator, Direct-Cur- rent 456 Grounding Wires 721 Grouping of Windings 350 Grove's Battery 102 Grove's Gas Battery 126 Gutta-Percha Molds 673 II Armature. Spindle or .... 222 H and B Curves 178 H and B Curves, Interpreta- tion of 179 H and B Syonyms for 178 Hand Magneto Test for Ground 657 Hand Magneto Tests in Gen- eral 656 Hare's Calorimeter 99 Harmonic Signals for Party Lines \ 701 Heat of Arc 536 Heat in Transformers 3S4 Heating, Cp : 601 Heating, Conditions Causing. 584, 586 Heating, Economy of Electric. 730 Heating by the Electric Arc, 'Direct 727 Heating. Electric Cooking and Domestic 723 Hpating of Field Coils 429 Heating of Motors 585 Heating of Windings of Stator, Local 436 Helios-Upton Battery 133 Henry. Inductance and the.. 332 Hertz Receiver 732 High Potential. Ad^':^ntages of. 476 High-Resistance Measurements 643 High-Voltage Determinations with I'otentiometer 638 High- Voltage Parallel Systems 503 Hissing Arc 543 Holder, Globe and Carbon... 547 Holders, Brush 305 Holders, Brushes and Brush. 420 Hook-Switch, The 696 Horse-Power of Car Motors. . 586 Horse-l*ower Lines 274 Hot Resistance 596 Hot-Wire Voltmeter, The Stan- ley 466 Hot-Wire Voltmeters 466 House Connections, Telephone 693 Hushes Microphone 680 Human Voice, The 678 Hunning Transmitter 682 Hydrosen Liberated by the Coulomb 88 Hydrogen to Oxygen, Propor- tion of 88 Hydrometers for Storage Bat- teries 147 Hysteretic Constant 186 Hysteresis . . . •. 185 Hysteresis Curves 185 Hysteresis, Loss of Energy Due to r. 186 Hysteresis and Other Receiv- ers 736 Idle Wire 229 Illuminating Power, Standards of 563 Impedance 339 Impressed E. M. P.. Varia- tions in 216 Impurities in Electrolyte and tests 155 Incandescence, Open-Air .... 562 Incandescent Lamp. The 523 Incandescent Lamp Circuits, Arc and 472 Incandescent Lamps, Candle- Powers of 576 Incandescent Lamps, Distribu- tion of Light from 583 Incandescent Lamps, Watts per Candle-Power in 580 Incandescent Lighting 523 Incandescent Lighting, ''Muni- cipal" Series 470 Incandescent Lighting, Series. 477 Inclosed-Arc, Action ^of 546 Inclosed-Arc Carbons ....547, 559 Inclosed-Arc Lamp r)45 Inclosed-Arc Lamp. Carbons for .-,00 Inclosing GMbe, The 559 Incubator. Th^ Electric 729 Independent Circuits 492 Independent Excitation of Shunt Coil, Effect of 260 812 ELECTRICIANS' HANDY BOOK. Indications From Gassing. . . . 156 Indicator, Alternating-Current Ground 456 Indicator, Direct-Current Ground 456 Individual Transformers .... 505 Individual Voltages of Lamps. 490 Inducing Electric Energy, Con- ditions for 206 Inductance 332 Inductance and Capacity, Com- position of Resistance 343 Inductance Compensator, The. 471 Inductance and the Henry. . . 332 Inductance. Ohmic Equivalent of Rfactance of 336 Inductance, Reactance of .... 335 Inductance, Reactance in Sub- divided Conductor 336 Inductance, Turns of a Cir- . cuit and 334 Induction 205 Induction, Ampere's Rule Adapted to 213 Induction Coils in Bracket Telephones 688 Induction Coil, Dim'ensions of Telephone 686 Induction Coil, Effect of Tele- phone 688 Induction Coil, The Telephone 684 Induction, Electro-Magnetic... 205 Induction, Examples of 211 Induction Interference 714 Induction. Laws of 212 Induction Motor, The. 363, 366, 369 Induction Motor, Lenz's Law and the 369 Induction Motor, Polyphase . . 436 Induction Motor Rotors 436 Induction Motor, Three-Phase. 365 Induction, Two Systems of. . . . 210 Inductive Capacity, Specific. 48 Inductivity 48 Inductor Alternator 356 Iron and the Field of Force.. 175 Iron Negative Plates 101 Iron. Saturation of 177 Inspection of Transformers. . 438 Insulation Resistance of Cable 646 Insulation of Storage Battery Cells 161 Insulation Tests. Line 643 Insulation for Transformers. . 300 Insulation of Windings 441 Insulator of Magnetism, ISo.. 177 Insulators 511, 603 Intensity 53 Internal and External Resist- ance 71 Interchansreabilitv of Dynamo and Motor 218 Interchangeability of Parts. . 439 Interference, Conduction .... 714 Interference. Induction ..... 714 Interlinking and Cutting i^ines of Force. Tlireading 205 Interpretation of B and H Curves 170 Invention of Microphone .... 679 Inverted ^Addition and Sub- traction 23 Jablochkofe Candle, The 561 Jacket Tvpe Transformers, Shell or^ 378 Joints in Line Wire 508 Kelvin Absolute Electrometer, Thomson or 617 Kelvin Galvanometer, Thom- son or 609 Kirchoff's Laws 83 Lag 325 Lag, Angle of Lead and .... 326 Lag and Lead, Basis of .... 327 Lag and Lead, Causes of .... 341 Lalande and Chaperon Bat- tery. ^Modifications of .... 119 Laminated Field Magnets . . . 314 Lamp Annunciator 709 Lamp, Carboning a 561 .Lamp Circuits, Arc and Incan- descent 472 Lamp on Constant Potential Circuit, Arc 553 Lamp, Direct Photometering of Arc 572 Lamp, Efficiency of Nernst... 533 Lamp, The Incandescent .... 523 Lamp, Inclosed-Arc, Carbons for 560 Lamp, Making .Joints by Flashing in. Incandescent. . . 525 Lamps, Metallic Filaments in Incandescent 530 Lamp, The Nernst 530-533 Lamp, Pasted Joints in, Incan- descent 526 Lamp, Photometrv of Arc... 576 Lamp. Putting. Into Service. . 560 Lamp Signal Svstem 713 Lamp. The Sun 562 Lamp, The Wallace 562 Lamps, Adjusting 559 Lamps, Candle-Powers of In- candescent 576 Lamps, Carbon-Feed 549 Lamps, Commercial Rating of. Arc 542 Lamps, Constant-Current or Series Arc 551 Lamps, Constant-Potential Arc 555 Lamps, Direct-Current. Nernst 532 Lamps. Distribution of Light from Alternating-Current Arc 544 INDEX. 813 Lamps, Distribution of Liglit from Incandescent 583 Lamps, Efficiency of Alternat- ing-Current Arc 545 Lamps, Electroplated and Other Joints in Incandescent 526 Lamps, Features of Series or Constant-Current System for Arc 475 Lamps, Inclosed-Arc 543 Lamps, Individual Voltages of 490 Lamps, Making Incandescent. . 526 Lamps Without Mechanism . . 561 Lamps, Noise in Arc 545 Lamps, Relief 478 Lamps, Vacuum Nernst 533 Lamps, Watts per Candle- Power in Incandescent 580 Lap Winding 239, 249 Lap Winding, Multipolar . . . 246 Lap Winding, W^ave and 238 Law, Faraday's 212 Law 'and the Induction Motor, Lenz's 369 Law, Kirchoff's 83 Law, Lenz's 213 Law, Ohm's 74 Law, Right-handed Screw . . 204 Laws of Induction 212 Lead 325 Lead, Basis of Lag and 327 Lead, Causes of Lag and.... 341 Lead and Lag. Angle of 326 Leading-in Wires 526 Leakage of Lines of Force. . . . 183 Leakage, Measurement of Re- sistance 645 Leeron Disk, The 565 Length, Voltage Drop and Arc 538 Length of Wave 318 Lenz's Law 213 Lenz's Law and the Induction Motor 369 Letters, Greek 34 Light of Alternating-Current Arc Lamps, Distribution of. 544 Light from Arc Lights, Dis- tribution of 582 Light in Direct-Current Open Arc, Distribution of 541 Light, Distribution of 534 Lights, Distribution of Light from Arc 582 Light, Efficiency of Arc 537 Light Given by Arc Proper. . 543 Light, Mechanical Equivalent of 577 Light, Quality of Arc 580 Light, Watts per Candle- Power in Arc 579 Lighting Circuits, Connections for Charging from 157 Lightin g Current. Standard Lighting, Incandescent 523 Lighting, "Municipal" Series Incandescent 479 Lighting, Series Incandescent. 477 Lightning Arrester, Double- Pole 522 Lightning Arrester, Low- Equivalent Alternating-Cur- rent 521 Lightning Arrester, Tank . . . 522 Lightning Arrester, Westing- house 521 Lightning Arresters ....518, 591 Limitations of Series Distri- bution 474 Line Connections 362 Line Insulation Tests 643 Line of Ohms 278 Line Resistance Tests 643 Line Tests, Galvanoscope Cable and 650 Line Wire, Joints in 508 Lines of Force in Armature, End Leakage of 272 Lines of Force, Cutting. .209, 224 Lines of Force Due to Current, Field of Force and 5.^ Lines of Force, Leakage of. . 183 Lines of Force about a Mag- net, Illustrating 188 Lines of Force, Memoria Technica for 171 Lines of Force Produced bv a Curved Conductor ....*... 169 Lines of Force, Relation Be- tween Ampere Turns and. . 183 Lines of Force, Spreading of. 188 Lines of Force, Threading, Interlinking and Cutting. . 205 Lines of Force, Utility of Conception of 171 Lines, Harmonic Signals for Party 701 Lines, Horse-Power 274 Lines, Party 700 Lines, Polarized Bells for Party 700 Load, Slow Speed Without.. 418 Local Action of Primary Bat- tery 96 Logarithms 81 Longer Arcs, Resistance of . . . 543 Long-Shunt Compound Wind- ing 256 Long-Sliunt Windings, Action of Short-Shimt and 257 Loop Svstem 483 Loop Test, Varley 656 Loose Carbon Transmitters... 681 Loss of Energy Due to Hyster- esis 186 Loss of Magnetic Polarity... 417 Low-Equivalent Alternating- 814 rJLEOTRICIAXS' HAXDY BOOK. Luminescence 530 Liiminiferoiis Ether 40 Lnminometer. The ri"2 Lummer-Brodhun Screen, The. 565 Machine, Starting a 427 Machine, Stopping a 416 Machines, Conventional Repre- sentations of 2G4 Machines in Series, Shiint- WoLind 403 Magnet, Action of a Current on the 202 Magnet, Annular Chambered. . 191 Mas-net.- Blow-Out 596 Ma«-net, Effect of 538 Mas-net, The Electro- 188 Ma2:net- Illustrating Lines of Force About a 188 Magnet. Magnetizing by Coil and Electro- 197 Magnet for Multipolar Dyna- mos, Field 310 Magnet, The Natural 194 Magnet, The Permanent .... 194 Magnet Poles on Each Other, Action of 19.5 Magnet, Spiral Electro- .... 189 Ma2:net, Tractive Force of the Electro- 18S Magnetic Blow-Out Arrester. 518 Magnetic Circuit, The 172 Magnetic Circuit, Energy and the 172 Magnetic Circuit, Nature of the 174 Magnetic Circuit, Permeance of a 184 Magnetic Circuit, Three Fact- ors of 175 Magnetic Force 176 Magnetic Induction, Electro-. 205 Magnetic Needle in Rotating Field, The 864 Magnetic Polarity, Loss of. . 417 Magnetic Poles, Attraction and Remilsion of 202 ^Tagnetic Release Starting Box 398 Macrnetic Release I'nderload Circuit Brpaker 452 Maarnetic Tractive Power, Electro- 192 Ma9-netism, Ampere's Theory of 200 Masrne^-ism. Ampere's Theory of Terrestrial 201 Magnetism, Constancy of. ... 190 Magnetism, No Insulator of.. 1 "^7 Magnetism, Residual 1^5 Magnetized 199 Magnetizing hv Coil and Elec- tro-Majrnet 197 ISIao-npto BpH as Galvanometer, Ti>e 655 Magneto Generator, The .... 250 Masneto Test for Ground, Hand 657 Magneto, The Telephone .... 689 Magneto Test for Cross Con- nections, Hand 658 Magnetos. Regulation of Sep- eratelv-Excited Dynamos and \ 262 :Magueto Tests, Hand 656 ^Magnets of Alternators, Fiekl 436 ^Magnets, Concentric, in Lamps 550 Mas'nets bv Double Touch, Making .' 196 Magnets, Examples of Perman- ent 198 Magnets, Laminated Field... 314 Magnets, Making U-Shaped. . 196 Magnets,. Memoria Technica of 201 Magnets, Multipolar 192 Magnets, Preservation of 198 Magnets by Single Touch, Making .^. 195 Ma^rnets, Steel for 197 Magnets, U-Shaped Electro-. 189 Main Circuit, Exciting Series Coils from 261 Main, and Leads, Feeders .... 483 Marconi's Coherers 736 Material of Commutator .... 419 Material of Vessels for Plating 671 Measurement. Angular 32 Measurement. Ballistic 613 Measurements, High Resist- ance 643 Measurement with Potentiom- eter, Current 640 Measurement of Resistance Leakage 645 Measurement of Resistance, Voltmeter 642 Mechanical Equivalent of Light 577 Mechanical Release Underload Circuit Brea^ker 453 ^leidinger's Battery 114 Mercurv-Bichromate Battery, Fuller's 106 :\rerit. Figure of 622 Memoria Technica of Mag- nets 201 Memoria Technica for Lines of Force 171 :Mesh Connection, Delta or.... 361 AFetal Brushes. Trimming.... 307 AFetals DPDOsited in Plating. 662 Meter Bridge, The 634 ureter, Edison's 513 ureter, Forbes' 514 ureter, Shallenberger's 5i7 dieter. Thomson's 515 Meters, Electric 5i 3 Method. Null 634 ?Ii-rofarad 49 INDEX, 815 Microphone, Hughes 680 Microphone, Invention of .... 679 Mil System, Application of Circular 84 Mil System, Circular S3 Mil, Examples of Area of Cir- cular . .'. S-t Mil, Area of Circular 84 ^loisture in Transformers.... 4o8 Molds in Bath, Placing 674 Molds, Elastic 673 Molds, Gutta-Percha 673 Molds, Metal 671 Molds, Plaster 672 Molds, Plating on 674 Molds, I'reparing 673 Motor and Booster 501 Motor Connected, Generator and 285 Motor, Construction of Car.. 588 Motor Dynamos as Boosters.. 408 Motor, The Electric-Car 5S4 Motor Heating, Cause of . . . . 584 Motor, The Induction 363 Motor, Interchangeability of Dynamo and .'. 218 Motor, Lenz's Law and the Induction 369 Motor Without Load, Speed liegulation of 418 Motor, Operation of Synch- ronous 370 Motor, Refusal of, to Start . . 417 Motor, Reversihility of Dyna- mo and . . . 285 Motor Rotors, Induction 436 Motor, Self-Starting Single- Phase ... 435 Motor, Seif-Startiiig * Synch- ronous 373 Motor, Single-Phase Synchron- ous 371 Motor, Synchronous Polyphase. 372 Motor, Temperature of Dyna- mo or 439 Motor, Three-Phase Induction 365 Motor and Torque, Direct- Current 285 Motor Transformer 400, 401 Motor Troubles 600 Motors, Accidents to 415 Motors, Construction of In- duction 369 Motors, Determining the Heat- ing of 585 Motors, Horse-Power of Car.. 5K(j Motors, Idle 418 Motors, Induction 366 ^Motors, Polyphase Induction.. 436 Motors, Starting and Stop- ping 307, 418 Motors. Synchronous ....369, 436 Motornian's Duties 597 Mounting of Ring Armature. 230 Moving Electrodes 103 '•Municipal" Series Incandes- cent Lighting 479 ]Multii)le Series 67 :Multiple Series System 478 Multiple Switchboard, The . . 706 ^Multipolar Construction. .252, 350 Multipolar Construction, Ad- vantages of 252 Multipolar Dynamo Connec- tions 264 Multipolar Dynamo, Modern. 250 Multipolar Dynamos, Field Magnet for 310 Multipolar Fields, Windings for 237, 315 Multipolar Lap Windings 246 Multipolar Magnets 192 Multipolar Ring Armature. . . 231 Multipolar Winding by Formu- la 248 Multipolar Windings 245 Mutual Action of Currents... 109 Nature of the Magnetic Cir- cuit 174 Natural Magnet, The 194 Negative Carbon, Positive and, 535, 560 Negative and Positive Plates. 95 Negative Plate, Aluminium . . 101 Negative Plate, Carbon 103 Negative Plates, Iron 101 Nernst Lamp • 530, 533 Nernst Lamp, Direct Current. 532 Nernst Lamp Ballast 532 Nernst Lamp Glower 530 Nernst Lamp Glower Terminals 531 Nernst Lamp Heaters 531 Nernst Lamp Vacuum 533 Neutral Points ,. 268 Neutral Wire in Three-Wire System 497-499 Neutral Wire in Y System... 362 Nickel-Plating 664 Nitric Acid Depolarizers, Sulphuric and 110 Noise in Arc Lamps 545 Nomenclature for Drum Arma- ture Winding 246 Nomenclature of I*rimarv Bat- tery 95 Non Conductors, Conductors and o . . 50 Notation. Exponential 29 N^te, Fundamental 677 Null Method 634 Nuiuerical Value of Circular Functions 34 Occlusion of Gases by Fila- ment 52.'5 Ohm, The 71 816 ELECTRICIAKS' HAXDY BOOK, Ohm, British Association Standard 627 Ohm's Law 74 Ohm's Law, Examples of . . . . 75 Ohm's Law, Five Forms of . . 76 Ohmic Equivalent of React- ance of Capacity 337 Ohmic Compensator, The 471 Ohmic Equivalent of React- ance of Inductance 336 Ohms, Line of 278 Ohm-Volt Curves 284 Oil 560 Oil Coolinsj 384 Oil for Filling Transformers, 389, 437 Oiling 440, 674 Oil Switches 447 Open Arc, The Direct-Current 540 Open Arc, Distribution of Light in Direct-Current. . . 541 Open-Air Incandescence 562 Open Circuit, Discharge of Storage Battery on 144 Open and Closed Circuits.... 64 Open-Coil Armatures 222 Open-Wound Four-Part Arma- ture 230 Opening Shunt Coil, Discon- necting or 260 Operation, Two-Phase 436 Opposition, Quadrature and. . 327 Outer Circuit 68 Outer Circuit, Short Circuits in 432 Overtones ....'. 677 Over-Compounding 259 Overload Circuit Breakers. . . . 451 Overload and Underload Cir- cuit Breakers 448 Oxide Filaments 530 Oxygen, Proportion of Hvdro- gen. to 88 Pacinotti Armature, The 225 Pancake Coils 381 Panels, Switchboard 445 Parallel Circuits, Calculation of Resistance of 81 Parallel-Circuit System 554 Parallel Coupling of Com- pound Dynamos 402 Parallel Coupling of Dynamos 402 Parallel Coupling of Shunt Dynamos 402 Parallel Distribution 481 Parallel Distribution, Disad- vantages of 481> Parallel Distribution, Ele- mentary Case of 482 Parallel, Series and 08 Parallel and Shunt 67 Parallel System, Potential Drop in 482 Parallel Svstems, Anti- 488 Parallel Systems, High-Voltage 503 Parts, Interchangeability of.. 439 Parts of Primary Battery.... 93 Party Lines 700 Party Lines, Harmonic Sig- nals for 701 Party Lines, Polarized Bells for 700 Partz's Battery 109 Pasted Plates 135, 139 Percentage 20 Permeability 175, 177 Permeability Curves 180 Permeance 175 Permeance of a Magnetic Cir- cuit 184 Permeance of Ring Core 228 Phase 325 Photometer Apparatus 567 Photometer, Bar 563 Photometer Bar, Calculating Scale of 568 Photometer, Bouguer's 570 Photometer, Diffractive 574 Photometer, Foucault's 572 Photometer Observations .... 569 Photometer, Principle of 563 Photometer, Pupillary 573 Photometer. Shadow 570 Photometering of Arc Lamp, Direct 572 Photomteric Screens 564 Photometric Standards 570 Photometric Standards, Table of 570 Photometry of Arc Lamp.... 576 Pile, Galvanic 97 Pile, Volta's 97 Pile, Zamboni's 99 Pilot Wires, Pressure Lines or 470 Pitch 6''7 Plante's Battery 12S Plate, Cadmium 156 Plate, SoundinsT 677 Plates, Buckling of 153 Plates, Negative and Positive. 95 Plates, Pasted 135, 139 Plates, Suspended 138 Plating Apparatus 661 Plating on Molds 674 Platinum-Pin fir g 668 Poggendorff's Battery, Depolar- izing :Mixtures in 109 Poggendorff's Batterv, Exciting Solutions in . . . .' 109 Poggendorff's Battery, Modi- fications of 106 Points. Controller 503 Points, Driving 503 Points, Neutral 2^8 Polarity of the Circuit. TItp.. 159 Polarity Due to Windings, Armature 265 Polarity of Field, Wrong. . . . 417 INDEX. 817 Polarity, Loss of Magnetic... 417 Polarity, Tests for 404 Polarization of Primary Bat- tery 96 Polarized 199 Polarized Bell 691 Polarized Bells for Party Lines 700 Pole Armatures 301 Pole Pieces, Eddy Currents in 272 Pole Single-Phase Armature.. 354 Poles, Attraction and Repul- sion of Magnetic 202 Poles, Development of Field.. 240 Poles on Each Other, Action of Magnet 195 Poles, Held 222 Polyphase Induction Motors.. 436 Polyphase Synchronous Mot- ors 372 Position of Anode and Cathode 670 Positive and Negative Carbons, 535, 500 Positive and Negative Plates... 95 Potassium Bichromate Solu- tions 110 Potential, Advantages of High 476 Potential, Analogies of Drop of 60 Potential Circuit, Arc Lamp on Constant 553 Potential Circuit, Constant . . 78 Potential, Drop of 60 Potential, Drop and Fall of.. 79 Potential Drop in Parallel System 482 Potential, E. M. F. and Differ- ence of 61 Potential Energy of the Field of Force 173 Potential Methods, Uniform.. 492 Potential Regulator, Alternat- ing-Current 454 Potentiometer, The 637 Potentiometer, Current Meas- urement with the 640 Potentiometer, High - Voltage Determinations with 638 Potentiometer, Principle of. . . 637 Power 77 Power Calculations, Examples of 79 Power Consumed in Arc 538 Power Curves 345 Power, Electro-Magnetic Tract- ive 192 Power, Examples of 77 Power Factor 330 Power Factor in Alternating- Current Arc 544 Power Required for Cooking. 723 Power, Standards of Illumin- ating 563 Powders of Ten or Exponential Notation 29 Practical Processes 676 Preparations for Silver-Plat- ing 666 Preparing Storage Battery Electrolyte 155 Preservation of Magnets 198 Pressure, Brush 420 Pressure Lines or Pilot Wires. 470 Prevention of Sulphating. . . . 151 Primary Battery, The 125 Primary Battery, Amalgama- tion in 96 Primary Battery Cell. The.. 93 Primary Battery, Exhaustion of 96 Primary Battery, Local Action of 96 Primary Battery, Nomenclat- ure of 95 Primary Battery, Parts of . . 93 Primary Battery, Polarization of 96 Production of Alternating E. M. F. and Current 318 Production of Current 52 Production of Electromotive Force 56 Prony Brake, The 38 Proportional Coils 636 Proportion of Hydrogen to Oxygen 88 Protectors, Lightning 518 Pump, The INIercury Air .... 528 Pupillary Photometer 573 Quadrature and Opposition... 327 Quality of Arc Light 580 Quality of Carbons 538 Qualities of a Circuit 331 Quantity of Electricity, Mean- ing of 44 Quantity of Electricity, The Storage of 47 Quantity, Electric 41 Quantity, Unit of 45 Quantities, Multiplication of Alternating 244 Quantities, Summation of Al- ternating 342 Racing 720 Radian System of Angular Measurement 32 Radiator, Electric 730 Radiguet Battery 108 Radius Vector and Resultant. 324 Rail-Joint Test 644 Rate of Change 323 Rate of Change, Graphic Rep- resentation of 323 Rate Units, Current and .... 50 Rating of Arc Lamps, Com- mercial 542 Ratio of Transformation .... 379 Reactance 332 ^ 818 ELECTRICIAXS' HAXDY BOOK. Reactance of Capacity 337,338 Reactance Coil or Economy Coil 545 il^actance of Inductance. .335, 336 Reactance in Subdivided Con- ductor, Induciance 336 Reaction Diagrams, Armature 2G6 Receiver, The Telephone 683 Receiver, a Dynamo, Telephone 212 Receiver, Hertz 732 Beceiver, Principle of Tele- phone 678 ileceivers. Hysteresis and Other Wireless Telegraphy. . 736 Receiving Apparaus 733 Rectifier, The 304 Reflecting Galvanometer ..606-609 Regeneration 126 Regulation of Voltage, Auto- matic 492 Regulation of Current in Elec- troplating 661 Jlegulator. Alternating - Cur- rent Potential 454 Regulators or Boosters 4U6 Relation Between Ampere Turns and Lines of Force . . 183 Relations, Energy 209 Relief Lamps 478 Reluctance and Reluctivity... 178 Repeating Coils 705 Reproduction 660 Repulsion of Magnetic Poles, Attraction and 202 Residual Magnetism 185 Resistance 69 Resistance of Arc Proper.... 537 Resistance Boxes 626, 630 Resistance of Cable, Insulation 64 Resistance, Circuit Without. . 71 Resistance Coil in Arc Lamp. 553 Resistance Coils 626-631 Resistance Coils, Modern Ar- rangement 628 Resistance Coils, Practical Notes on 631 Resistance, Compensating . . . 619 Resistance and Energy 70 Resistance by Flashing, Low- ering of 525 Resistance, Galvanometer . . . 622 Resistance, Hot 596 Resistance, Inductance and Ca- pacity, Composition of 343 Resistance, Internal and Ex- ternal 71 Resistance Leakage, Measure- ment of 645 Resistance of Longer Arcs... 543 Resistance of Parallel Circuits, Calculation of 81 Resistance of Short Arcs .... 543 Resistance, Spurious 270 Resistance, Starting 367 Resistance of Storage Batteries 132 Resistance Tests, Line 643 Resistance, Voltmeter and Am- meter Determination of ... 642 Resistance, Voltmeter Measure- ment of 642 Resistance Wire 627 Resonance, Electric 339 Resultant, Radius Vector and 324 Reverse Current Circuit Breaker . . 454 Reverser, Car " 596 Reversing the Car . 598 Reversing Direction of Current 416 Revolving Field, Rotary and. 366 Rheostat 624 Rheostat Controller 596 R. I. Drop Calculations, Ex- amples of 82 R. I. Drop and Counter E. M. F 79 Ring Armature, Commutator Connections of 227 Ring Armature, Current in. . 229 Ring Armature, Mounting of. 230 Ring Armature, Multipolar. . . 231 Ring Armature, Open-Wound Four-Part 230 Ring Armatures. Cores of . . . 228 Ring Core, Permeance of 228 Ring Oiling 440 Ring, The Gramme 226 Rings, Collecting or Slip 219 Rings, Collector 419 Rotary Converter 390 Rotary Converter, Functions of a 394 Rotary Converter, Starting a. 394 Rotary Converter in Three- Wire System 393 Rotary and Revolving F'ield.. 366 Rotary Field, Armature in. . . . 364 Rotary Field, The 363 Rotary Transformer, The 391 Rotary Transformers. Rela- tions of Voltage and Cur- rent in 392 Rotary Transformer, Fse of. . 391 Rotating Field, The Magnetic Needle in 3G4 Rotation, Reversal of Arma- ture 404 Rotor and Stator 355 Rotors of Alternators, Trouble in 435 Rotors, Induction Motor 436 Rule, Ampere's 203 Rule for Charging Storage Bat- teries. Enslish 151 Pule, Clerk Maxwell's 213 Rule, Fleming's 212 Safety Fuses 441, 448 INDEX, 819 Sand Type of Daniell's Bat- tery 114 Sandpapering Commutator.... 4:>2 Saturation of Iron 175, 177 Saw-Tootli Arrester, Comb or. 518 Scale of Photometer Bar, CaJ- culating 5G8 Scale, Translucent, Galvanome- ter 4G3 Scales, Graduation of "Volt- meter 4G8 Screen, The Lummer-Brodhun. 5(55 Screens, Photometric 5G4 Sediment in Storage Batteries 153 Sensibility, Regulation of, in Galvanometer , GIO Sensitiveness, Conditior.s of in Galvanometer 636 Self-Excited Dynamos, Sepa- rately- and c 263 Self -Regulation of Compound- wound Dynamos 257 Self-Starting Single-Phase Mo- tor 435 Separately-Excited Generators 261-263 Series 67 Series Arc Lamps, Constant- Current or 551 Series or Constant-Current Sys- tem for Arc Lamps, Features of 475 Series Distribution 473 Series Distribution, Calcula- tions for 475 Series Distribution, Limitations of 474 Series Distribution, Objections to 481 Series Incandescent Lighting. . 477 Series Incandescent Lighting "Municipal" 479 Series liighting Current, Stan- dard 477 Series-Multiple 67 Series-Multiple System 479 Series and I'arallel 68 Series-l'arallel Controller .... 594 Series, Shunt-^yound Machines in 403 Series Telephone Circuit 695 Series Winding, Action of.... 254 Series Winding of Dynamos... 252 Service, Putting Lamp into... 560 Service, Taking Storage Battery out of 160 Seven-Wire Systems, Five and 502 Shadow Photometer 570 Shaft, End Motion in Arma- ture 423 Shallenherger's Meter 517 Shock. Treatment of Electric. 442 Shell or Jacket Type Trans- formers 378, 379 Short Arcs, Resistance of . . . . 543 Short-Circuitiug of Single Cells 152 Short Circuits 69, 433 Short Circuits in Armature... 423 Siiort Circuits between Arma- ture Windings and Frame. . 435 Short Circuits in Field Wind- ing 430 Short Circuits in Outer Circuit 432 Short Circuits in Transformers 438 Short-Shunt Compound Wind- ing 256 Short-Shimt and Long-Shunt Windings, Action of 257 Shunt Coil, Disconnecting or Opening 260 Shimt Coil, Effect of Indepen- dent Excitation- of 260 Shunt Coil, Separate Excita- tion of 261 Shunt to the (:^alvanometer . . . 636 Shunt, Galvanometer 618 Shunt Dynamo, Total Charac- teristic Curve of 283 Shunt Dynamo, Total Current Characteristic Curve in.... 282 Shunt Dynamos, Parallel Cou- pling of 402 Shunt Winding, Action of . . . . 255 Shunt Winding of Dynamos. . 254 Shunted Ammeter . ^ 469 Shunt-Wound Dvnamo, Critical Point of 281 Shunt-Wound Dynamo Charac- teric Curves 280 Shunt-Wound Machines in Se- ries 403 Siemens's Dynamometer .... 622 Siemens's Plan for Resistance Boxes 627 Signal System, Lamp 713 Signals, Emergency and Dan- ger 441 Signals for Party Lines, Har- monic 701 Signals, Wave Transmission of 731 Silver-Plating 665 Silver-Plating, Preparation for 666 Silver Voltameter 90 Simple Batteries 93 Sine-Curve. The 321 Sine-Curve, Vector Diagram of 325 Sine Galvanometer, or Com- pass, The 616" Single Cells, Short-Circuiting of 152 Single - Dynamo Three - Wire System 49? Single- I^hase Armature 348 Single-Phase Armature, Pole. . 354 Single-Phase Motor, Self-Start- ing 435 Single-Phase Synchronous Mo- tor 371 820 ELECTRICIANS' HANDY BOOK. Single Surface Condenser 45 Single Touch, Making Magnets by 195 Siphon Battery, Baudet 107 Six-Wire Connection of Three- Phase Alternator Winding. . 359 Size of Conductors for Plating 669 Slip Rings, Collecting or 219 Slow Speed without Load.... 418 Smee's Battery 103 Smoothing Commutator 433 Soldering 441, 721 Soldering Iron, The Electric. 728 Solenoid Ammeters, Total-Cur- rent 467 Solid Back Transmitter, The. 683 Solutions in Poggendorff Bat- tery, Exciting 109 Solutions, Potassium Bichro- m.ate 110 Sound 677 Sounding Plate 677 Sparking, Air Gap and 310 Sparking of Commutator 424 Specific Inductive Capacity. ... 48 Speed Control, Crocker-Wheel- er 414 Speed of a Current 54 Speed without Load, Slow. . . . 418 Speed Regulation of Motor without Load 418 Spherical Candle Power 574 Spindle or H Armature 222 Spiral Electro-Magnet 189 Spreading of Lines of Force. . 188 Spring Jacks 710 Squirted Filaments 524 Standard English Candle, The. 567 Standard Series Lighting Cur- rent 477 Standard Voltage and Allow- able Temperature 584 Standards of Illuminating Power 563, 570 Stanley Hot-Wire Voltmeter, The 466 State, Stationary 543 Static Charge, E.M.F. and the 58 Static Electricity, Dynamic and 56 Stations, Connection of 734 Stator, Local Heating of Wind- . ings of 436 Stator, Rotor and 355 Star Connection, Y or 359 Starting Box 397, 308 Starting Compensator 368 Starting a Dynamo 427, 428 Starting ]Motors 307 Starting Resistances 367 Starting pnd Stopning Motors 418 Starting Torque 367 Start, Refusal of Motor to. 41 7, 508 Steel in Dynamos, Soft 181 Stoel for Magnets 197 Steeling 669 Step-Down Transformers, Step- Up and 378 Step-Down and Step-Up Trans- formation 400 Step, Alternators in 405 Stone's Common Battery Sys- tem 698 Stop Adjustment, Clutch 560 Stop» Emergency 600 Stopping Motors, Starting and, 416, 418 Storage Battery, Action of a.. 125 Storage Battery, American... 134 Storage Battery Cells 161 Storage Battery, The Charge of 145 Storage Battery, Charging.... 166 Storage Battery Connections, Booster and 410 Storage Battery Connections, Making 161 Storage Battery, Determination of Discharge of 145 Storage Battery, Disintegration of 153 Storage Battery, Edison's 140 Storage Battery, End Cells of. 164 Storage Battery Equalizer in Three-Wire Systems 501 Storage Battery, Floating. . . . 166 Storage Battery, Function of a 128 Storage Battery, Gould 132 Storage Battery on Open Cir- cuit, Discharge of 144 Storage Battery, Overcharge of 151 Storage Battery, Requirements of a 128 Storage Battery, Setting Up a. 153 Storage Battery, Specific Grav- ity Variation of Electrolyte in 146 Storage Battery, Suspended Plates in 138 Storage Battery, Taking out of Service 160 Storage Batteries, Chemical Ac- tion of 131 Storage Batteries, Copper 139 Storage Batteries, The Dis- charge of 144 Storage Batteries, Forming... 129 Storage Batteries, Manufac- turer's Data with 144 Storage Batteries, Notes on.. 163 Storage Batteries, Resistance of 132 Storage Battorips, Sediment in 153 Storage Bntteries in Three- Wire System 500 Storage Batteries, Varintion of Electrolyte, Specific Grav- ity 146 Storage Batteries, Zinc Acid. 140 INDEX, 821 Storage Capacity 130 Storage of Electric Quantity. 41 Storage of Quantity of Elec- tricity, The 47 Stray Field 184 Strength and Chemical Decom- position, Current 90 Strength, Current 53 Striking the Arc 535 Subdivided Conductor, Induct- ance Reactance in 336 Subscribers' I'ole Connection. 717 Sulphating, Prevention of . . . . 151 Sulphuric and Nitric Acid De- polarizers 110 Summary of Electro-Chemistry 00 Summation of Alternating Quantities 342 Sun Lamp, The 562 Switchboard 445 Switchboard Connections .... 712 Switchboard, The :Multiple. . . 706 Switchboard, Operation of . . . . 708 Switchboard Panels 445 Switch Boxes and Circuit Breaker on Car 590 Switches, Air 447 Switches, Circuit Breakers as. 454 Switches, Oil 447 System, Anti-Parallel 488 System for Arc Lamps, Fea- tures of Series or Constant- Current 475 System, Closet 484 System, Common Battery . .697-698 System, Constant-Current .... 472 System, Constant-Potential . . 473 System, Degree 321 System, Five-and Seven-Wire. 502 Svstem, High-Voltage Parallel 503 System, Lamp Signal 713 System, Loop 483 System, Multiple-Series 478 System, Neutral Wire in Y. . . . 362 System, Parallel-Circuit 554 System, Potential Drop in Par- allel 482 System, Rotary Converter in Three-Wire 393 System. Series-Multiple 479 System, Single-Dynamo Three- Wire 499 System, Storage Battery Equal- izer in Three-Wire 500, 501 System, Telephone 692 System, Three-Wire 497 System, Tree 484 System, Two-Dynamo Three- Wire 499 Synchronous Motors. 369, 373, 405, 436 Table, Traction 588 Tab.es, Winding 236 Tamidine Filaments 523 Tangent Galvanometer, The... 615 Tangential Brushes 306 Telegraphy, Wireless 733 Telephone Circuit, Bridged.... 696 Telephone Circuit, Series .695 Telephone, Edison's 682 Telephone as Galvanoscope, The 653 Telephone Induction Coil, The. 684 Telephone Induction Coil, Di- mensions of 686 Telephone Induction Coil, Ef- fect of 688 Telephone, Induction' Coils in Bracket 688 Telephone, Magneto, The 689 Telephone Receiver a Dynamo. 212 Telephone Receiver, Principle of 678 Telephone Systems 692 Telephone Transmitter. The . . . 679 Temperature of Commutator . . 411> Temperature of Dynamo or Mo- tor 439 Temperature of Plating Baths 671 Temperature. Standard Volt- age and Allowable 584 Terminology of Analytical Ge- ometry 278 Terrestrial Magnetism, Am- pere's Theory of 201 Test for Cross Connections, Hand Magneto 658 Test For Ground, Hand Mag- neto 657 Test, Rail-Joint 044 Test of Resistance, Voltmeter and Ammeter 642 Test, Varlev Loop 05(> Tests of Cable on Reels 652 Tests, Engineering 658 Tests, Galvanoscope Cable and Line 650 Tests, Hand Magneto 656 Tests, Impurities in Electrolyte and 155 Tests, Line Insulation 643 Tests, Line Resistance 643 Tests, Polarity 404 Theory of Magnetism, Am- pere's 200 Theory of Terrestrial Magnet- ism, Ampere's 201 Thomson or Kelvin Absolute Electrometer 617 Thomsion or Kelvin Galvano- meter 609 Thomson's Meter 515 Threading, Interlinking and Cutting Lines of Force 205 Three-Brush Dynamo 499 Three-Phase Alternator Wind- ing, Six-Wire Connection of 359 822 ELECTRICIAXS' HANDY BOOK. Three-Phase Current 347 Tliree-Phase lucUiccion Motor. 365 Threo-Pliase Winding 358 Three-AVire System 497 aiiree-Wire Sj'stem, Neutral Wire in 497-499 Three-Wire System, Rotary Converter in 393 Three - Wire System, Single- Dynamo 499 Three-AVire System, Storage Bitteries in oOO Three-Wire System, Storage Battery Equalizer in 501 Three-Wire System, Two-Dyna- mo 499 flhimder Clouds, Electromotive Force in .... : 58 Tin Plating 6()S Torque 36 Torque, Direct-Current Motor and 285 Torque. Starting 367 Total-Current Solenoid Amme- ter 467 Traction Table 588 Tractive Force of the Electro- Magnet 1S8 Tractive Power, Electro-Mag- netic 192 Transfer Bus-Bar 495 Transformation, Ratio of 379 Transformation, Step-Down and Step-Up 400 Transformations. Algebraic. . . 18 Transformer, Action of the . . . 383 Transformer, Action of Motor. 400 Transformer. Ammeter 470 Transformer. The Auto- 381 Transformer. Choking of 376 Transformer Construction, Ba- sis of 375 Transformer, Construction of Rotary 391 Transformer, The Limitation of a 377 Transformer, Motor 400 Transformer, Object of 376 Transformer Practice, Motor. . 401 Transformer, The Principle of a 377 Transformer. Relations of Volt- aee and Current in Rotaiy. 392 Transformer. I'se of Rotary.. 391 Transformers, Breakdowns in. 437 Transformers, Care of 437 Transformers, Constant - Cur- rent 387 Transformers, Core 380 Transformers, Disk Winding of 387 Transformers, Disk-Wound. . . . 380 Transformers, Economy of Mo- tor 401 Tranrjformers, Heat in 384 Transformers, Individual .... 505 . Transformers, Inspection of. . . 433 Transformers, Insulation for.. 390 Transformers, Moisture in. . . . 438 Transformers, Oil for 389 Transformers, Oil for Filling. 437 Trnnsformers, Operation of . . 396 Tniusfomc-rs, Shell or Jacket Type 378. 379 Transformers, Step-Up and Step-Down 378 Transformers, Short Circuits in 438 Translucent Scale for Galvano- meter 609 Transmission of Signals, Wave 731 Transmitter, The Blake 680 Transmitter, Hunning 682 Transmitter, Loose Carbon... 681 Transmitter, The Solid Back. 683 Transmitter, The Telephone. . 679 Transmitting Apparatus 733 Tree System 484 Trigonom.etric F'unctions 33 Tripping Platform 5-19 Troubles, Motor 600 Tudor Ba ctei y 138 Turns, Action of Demagnetiz- ing 2«9 Turns, Ampere 176 Turns of a Circuit and Induct- ance 334 Turns, Dead 270 Turns, Demagnetizing 268 Turns, Increasing E.M.F. by Increasing 221 Turns and Lines of Force. Re- lation between Ampere 183 Twelve - Conductor Bipolar Armature 235 Two-Dvnamo Three-Wire Sys- tem 499 Two-Phase Current 346 Two-Phase Operation 436 Two-Phase Windings 357 U-Shaped Electro-Magnets 189 U-Shaped Magnets. Making... 196 Underload Circuit Breaker .448-453 Uniform Potential Methods... 492 T'nit of Quantity 45 Units. Current and Rate 50 Useful Constants 35 Vacuum 527. 528 Values, Average 327-328 Values, Effective 328-330 Varley Loop T?st 656 Varying Field of Force 207 Vector Diagram of Sine Curve 325 Voice, The Human 678 Voltage 62 Voltage and Allowable Tem- perature, Standard 584 IXDEX. 823 Voltage, Automatic Regulation of 492 'S'oltage, Changing 400 Voltage Calculations 02 Voltage and Current in Rotary Transformers, Relations of. 392 Voltage Drop '. 530 ^'oltage Drcp and Arc Lengtli 53S Voltages of Lamps, Individual 490 Voltaic Arc, The 53o Voltameter, Silver 90 Volta's Pile 97 Voltmeter, The 458 "\'oitmeter end Ammeter Deter- mination ot Resistance .... 642 Voltmeter, Cardew's 4P3 Voltmeter, Compensated 470 Voltmeter, Empire 461 Voltmeter, Measurement of Re- sistarce 642 Voltmeter Scales, Graduation of 403 Voltmeter, The Stanley Hot- wire 466 Voltmeter, Weston's 4r)9 Voltmeters, General Notes on. 463 Voltmeters, Hot-Wire 466 Wadell-Entz Battery 140 Wallace Lamp, The 562 Waste, Cotton 439 Water Cooling 385 Water Decomposed bv the Cou- lomb "^ 88 Wattless Current 246 Wattmeter 470, 513 Watts per Candle-Power in Arc Light 579 Watts per Candle-Power in In- candescent Lamp 580 Wave. T he 316-3:1 8 Wave Form, Inliuence of 544 Wave and Frequeucv, Length of 319 Wave and Lap Winding 238 Wave Transmission of Signals 731 Wave Winding . . . . 239 Waves Produced bv Electricity, Ether 50 Wearing of Carbons 539 Weights and Chemical Equiva- lents. Atomic 89 Welding. Electric 728 Weston's Voltmeter and Am- meter 459 Wheatstone Bridge, or Bridsje Box 632-634 M^heels. Flat 597 Wheels, Skidding 598 Wheels, Sliding • 598 Wire Ends in Cable, Finding. . 653 Wire Gauges 85 Wire, Idle 229 Wire, Joints in Line 508 Wire, Resistance 627 Wire in Y System, Neutral... 362 Wires for Bell Circuits 722 Wires, Grounding 721 Wires, Leading Bell 719-721 Wires, Leading-in 526 Wireless Telegraphv 733 Wiring, Bell 718 Wollaston's Battery 99 Winding, Armature Polaritv Due to /. 265 Winding, Bipolar ...243, 244, 247 Winding, Break in Armature.. 422 ^^'inding, Break in Field 430 Winding Calculation, Example of Compound 259 Winding and Commutator Bars. Bad Contacts between 419 Winding, Compound, Long Shunt and Short Shunt. 256, 257 Winding, Disk 357 Winding, a Drum Armature, 241-243, 246 Winding of Dynamos, Field. . . 252 Winding and Frame. Short Circuits between Armature. 435 Winding, Grouping of 350 Winding, Insulation of 441 Winding, Lap 238, 239, 249 Winding, Multipolar, 237, 245, 246, 315 Winding, Principle of Alter- nating-Current Armature. . . 350 Winding, Series 252-254 Winding, Short Circuits in Field 430 Winding, Shunt 254. 255 Winding, S'x-Wire Connection of Three-Phase Alternator.. 359 Winding, Three-Phase 358 Winding Tables 236 Winkling, Two-Phase 3')7 Winding, Wave 238, 239 Windings of Stator, Local Heating of 430 Y Connections for Alternating- Current 506 Y or Star Connection 359 Y System, Neutral Wire in... 362 Zamboni's Pile 99 Zinc A(ii Storage Batteries. . 140 1919—1920 CATALOGUE o/LATESTaradBEST Practical and Mechanical Books Including Automobile and Aviation Books PRACTICAL BOOKS FOR PRACTICAL MEN Any of these books will be sent prepaid to any part of the world, on receipt of price. Remit by Draft, Postal Order, Express Order or Registered Letter. Published and For Sale by Ihe JNorman W . Henley Publishing Co. 2 West 45th Street New York, U. S. A. INDEX TO SUBJECTS Accidents 21 Air Brake 25, Id Arithmetic 15, 29, 38 Automobiles ....3, 4, 5, 6, 7 Automobile Charts .... 7 Aviation 8 Batteries 18 Bevel Gears 22 Brazing and Soldering. 9 Cams 22 Charts 7, 8, 9 Chemistry 21 Civil Engineering 29 Coke 10 Compressed Air 10 Concrete 10, 11, 12, 13 Cosmetics 34 Dictionaries 14 Dies— Metal Work ..13, 14 D r a w i n g — Sketching Paper 14, 15 Electric Bells 16 Electricity.. 15, 16. 17, 18, 19 Encyclopedia 29 Factory Management, etc ^ 19 Ford Automobile 6 Fuel 20 Flying Machines 8 Gas Engines and Gas, 20, 21, 22 Gearing and Cams .... 22 Hydraulics 22 Ice and Refrigeration.. 22 Inventions — Patents ... 23 Knots 2Z Lathe Work 23, 24 Link Motion 25 Liquid Air 24 Locomotive Engineering, 24, 25, 2(>, 27 Machine Shop Practice, 27, 29, 30, 31 Manual Training 32 Marine Engineering . , , 32 Mechanical Magazine . . 28 Mechanical Movements. 30 Metal Turning 2Z Metal Work Dies 13, 14 Mining 33 Motor Cycles 6, 7 Patents and Inventions. 23 Pattern Making 33 Perfumery 34 Plumbing 34 Receipt Book 35, 40 Refrigeration and Ice.. 22 Repairing Automobiles.. 6 Rubber ZS Saws ZS 'Screw Cutting ZG Sheet Metal Work ...13, 14 Smoke Prevention 20 Soldering 9 Starting Systems 5 Steam Engineering. 36, Z7f 38 Stea-i Heating and Ven- tilation 38 Steel 38, 39 Storage Batteries 18 Switch Boards 17, 19 Tractor 22, 39 Turbines 39 Ventilation 38 Waterproofing 13 \\' elding 5 \^iring 17, 18 Wireless Telephones .. 19 t^^Any of these books will be sent prepaid to any part of the world, on receipt of price. REMIT by Draft, Postal Money Order, Express Money Order, or by Registered Mail. 2 GOOD, USEFUL BOOKS AUTOMOBILES— MOTORCYCLES The Modern Gasoline Automobile, Its Design, Construction, Operation. By Victor W. Page, M.S.A.E. This is the most complete^ practical, and up-to-date treatise on gasoline automobiles and their component parts ever published. In the new revised and enlarged 1919 edition, all phases of automobile construc- tion, operation and maintenance are fully and completely described and in language anyone can understand. Every part of all types of automobiles, from light cyclecars to heavy motor trucks and tractors, are described in a thorough manner; not only the automobile, but every item of its equipment, accessories, tools needed, supplies and spare parts necessary for its upkeep, are fully discussed. It is clearly and concisely written by an expert familiar with every branch of the automobile industry and the originator of the practical system of self-education on technical subjects; it is a liberal education in the automobile art, useful to all who motor for either business or pleasure. Anyone reading the incomparable treatise is in touch with all improvements that have been made in motor car construction. All latest de- velopments, such as high speed aluminum motors and mul- tiple valve and sleeve valve engines, are considered in detail. The latest ignition, carburetor and lubrication prac- tice is outlined. New forms of change speed gears, and £nal power transmission systems, and all latest chassis im- provements, are shown and described. This book is used as a text in all leading automobile schools, and is conceded to be the standard treatise. The chapter on Starting and Lighting Systems has been greatly enlarged, and many automobile engineering features that have long puzzled lay- men are explained so clearly that the underlying principles can be understood by anyone. This book was first pub- lished six years ago, and so much new matter has been added to the book that it is nearly twice its original size. The only treatise covering various forms of war automobiles and recent developments in motor truck design, as well as pleasure cars. This book Is not too technical for the layman nor too elementary for the more expert. It is an incom- parable work of reference for home or school. 6x9. Cloth, 1,000 pages, nearly 1,000 illustrations, 12 folding plates. Price, $3.50 Questions and Answers Relating to Modern Auto- mobile Construction, Driving and Repair, By Victor W. Page. A self -educator on automobiling with- out an equal. This practical treatise consists of a series of thirty-seven lessons, covering' with over 2,000 questions and their answers — the automobile, its construction, operation and repair. The subject matter is absolutely correct and explained in simple language. If you can't answer all of the following questions, you need this work. The answers to these and 2,000 more are to be found in its pages. Give the name of all important ^ parts of an automobile and describe their functions. Describe action of latest types of kerosene carburetors. What is the difference between a **double" ignition system^ and a ''dual" ignition system? Name parts of an induction coil. How are valves timed? What is an electric motor starter and how does it work? What are advantages of worm drive gearing? Name all important types of ball and roller bearings. What is a **three-quarter" floating axle? What is a two-speed axle? What is the Vulcan electric gear shift? Name the causes of lost power in automobiles. Describe all noises due to deranged mechanism and give causes. How can you adjwst a carburetor by the color of the exhaust gases? What causes **popping" in the carburetor? What tools and supplies are needed to equip a car? How do you drive various makes of cars? What is a differential lock and where is it used? Name different systems of wire wheel construction. What is a ^'positive" drive differential? etc., etc. Answers every question a'sked relating to the modern automobile. A popu- lar work at a popular price, 554x7H» Cloth, 650 pages, 392 illustrations, 3 folding plates. Revised Edition just published. Price, $2,00 How to Run an Automobile. By Victor W. Page. This treatise gives concise instruc- tions for starting and running all makes of gasoline auto- mobiles, how to care for thera, and gives distinctive features of control. Describes every step for shifting gears, con- trolling engine, etc. Among the chapters contained are: I. Automobile Parts and Their Functions. II. General Starting and Driving Instructions. III. Typical 1919 Con- trol Systems — Care of Automobiles. Thoroughly illustrated. 178 pages, 72 illustrations. Price, $1.00 The Automobilist's Pocket Companion and Ex- pense Record. By Victor W. Page. This book is not only valuable as a convenient cost record, but contains much information of value to motorists. Includes a condensed digest of auto laws of all States, a lubrication schedule, hints for care of storage battery and care of tires, location of road troubles, anti- freezing solutions, horsepower table, driving hints and many useful tables and recipes of interest to all motorists. Not a technical book in any sense of the word, just a collection of practical facts in simple language for the everyday motorist. Convenient pocket size. Price, $1.00 Gasoline and Kerosene Carburetors, Construction, Installation and Adjustment. By Capt. V. W. Page. All leading types of carburetors are described in detail, special attention being given to the forms devised to use the cheaper fuels such as kerosene. Carburetion troubles, fuel system troubles, carburetor repairs and instal- lation, electric primers and economizers, hot spot manifolds and all modern carburetor developrnents are considered in a thorough manner. Methods of adjusting all types of car- buretors are fully discussed as well as suggestions for secur- ing maximum fuel economy and obtaining highest engine power. 250 pages, 89 illustrations. Price, $1.50 Starting, Lighting and Ignition Systems. By Victor W. Page. A practical treatise on modern starting and ignition system practice. This practical volume has been written with special reference to the requirements of the nontechnical reader desiring easily understood explanatory matter relating to all types of automobile ignition, starting and lighting systems. It can be understood by anyone, even without electrical knowledge, because elementary electrical principles are considered before any attempt is made to dis- cuss features of the various systems. These basic principles are clearly stated and illustrated with simple diagrams. All the leading systems of starting, lighting and ignition have been described and illustrated with the cooperation of the experts employed by the manufacturers. Wiring diagrams are shown in both technical and nontechnical ^ forms. AH symbols are fully explained. Complete data is given for locating troubles in all systems, the various steps being con- sidered in a logical, systematic manner, that can be easily followed by those without expert electrical knowledge. All ignition systems receive full consideration, starting with the simplest battery and coil forms found on early cars to the modern short-contact timer and magneto methods used with the latest eight and twelve-cylinder motors. Every ignition, starting or lighting system component is considered individ- ually, and full directions are given for making all repairs. This book is unusually complete, as it also includes descrip- tions of various accessories operated by electric current, such as electrical gear shifts, brake actuation, signaling devices, vulcanizers, etc. Nearly 500 pages. 297 Specially made en- gravings. New Edition. Trjlce, $2.00 Automobile Welding with the Oxy-Acetylene Flame. By M. Keith Dunham. Explains in a simple manner ap- paratus to be used, its care, -and how to construct necessary shop equipment. Proceeds then to the actual welding of all automobile parts, in a manner understandable by everyone. Gives principles never to he forgotten. This book is of ut- most value, since the perplexing problems arising when metal is heated to a melting point are fully explained and the proper methods to overcome them shown. 167 pages, fully illustrated. Price, $1.25 Automobile Repairing Made Easy. By Victor W. Page. A thoroughly practical book contain- ing complete directions for making repairs to all parts of the motor car mechanism. Written in a thorough but non- technical manner. Gives olans for workshop construction, suggestions for equipment, power needed, machinery and tools necessary to carry on business successfully. Tells how to overhaul and repair all parts of all automobiles. The inf rmation given is founded on practical experience, every- thing is explained so simply that motorists and students can acquire a full working knowledge of automobile repairing. Other works dealing with repairing cover only certain parts of the car — this work starts with the engine, then considers carburetion, ig^nition, cooling and lubrication systems. The clutch, change speed gearing and transmission system are considered in detail. Contains instructions for repairing all types of axles, steering gears and other chassis parts. Manv tables, short cuts in tiguring and rules of practice are given for the mechanic. Exolains fully valve and mag- neto timing, "tuning'* engines, systematic location of trouble, repair of ball and roller bearing, shop kinks, first aid to injured and a multitude of subjects of interest to all in the garage £nd repair business. All illustrations are especially- made for this book, and are actual photographs or reproduc- tions of engineering drawings. This book also contains Special Instructions on Electric Starting, Lighting and Igni- tion Systems, Tire Repairing and Rebuilding, Autogenous Welding, Brazing and Soldering, Heat Treatment of Steel, Latest Timing Practice, Eight and Twelve-Cylinder Motors, etc., etc. You will never "Get Stuck" on a Job if you own this book. 1,000 specially made engravings on 500 plates. 1,056 pages (5>^ x 8). 11 folding plates. Price, $3.50 The Model T Ford Car,Tts Construction, Opera- tion and Repair, Including the Ford Farm Tractor. By Victor W. Page. This Is the most complete and pra«- tical instruction book ever published on the Ford car. A high grade, cloth bound book, printed on the best paper, illustrated by specially made drawings and photographs. All parts of the Ford Model T Car are described and illustrated in a comprehensive manner — nothing is left for the reader to guess at. The construction is fully treated and operating principle made clear to everyone. Complete instructions for driving and repairing are given. Every detail is treated in a non-technical yet thorough manner. To' the 1919 R»visea Edition matter has been included on the Ford Truck and Tractor Conversion Sets and Genuine Ford^ Tractor. All parts are described. All repair processes illustrated and fully explained. Written so all can understand — no theory, no guesswork. New Edition. 106 illustrations, 310 pages, 2 large folding plates. ^ Price, $1.00 Motorcycles, Side Cars and Cyclecars, Their Construction, Management and Repair. By Victor W. Page. Describes fully all leading types of machines, their design, construction, maintenance, operation and repair. 550 pages. 350 specially made illustrations, 5 folding plates. New Edition. Price, $2.00 Automobile Charts By VICTOR W. PAGE, M.S.A.E. THE POPULAR AUTOMOBILE SERIES UNIFORM SIZE— 24' X 38'— PRICE 35 CENTS EACH Location of Gasoline Engine Troubles Made Easy. This chart shows clearly all parts of a typical four-cylinder gasoline engine of the four-cycle type. It simplifies location of all engine troubles. No details omitted. Price, 35 cents Location of Carburetion Troubles Made Easy. It shows clearly how to find carburetion troubles and names all defects liable to exist in the various parts. Instructions are given for carburetor adjustment. Price, 35 cents' Location of Ignition System Troubles Made Easy. In this chart all parts of a typical double ignition system using battery and magneto current are shown, and suggestions are given for readily finding ignition troubles and eliminating them when found. Price, 35 cents Location of Cooling and Lubricating Troubles. This is a combination chart showing all components of the ap- proved form of water cooling group as well as a modern engine lubrication system. It^ shows all points where defects exist that may result in engine overheating, both in cooling and oiling systems. Price, 35 cents Lubrication of the Motor Car Chassis. This chart presents the plan view of a typical six-cylinder chassis of standard design and outlines all important bearing points re- quiring lubrication, and is a valuable guide to the correct lubrication of any modern car. A practical chart for all interested in motor car maintenance. Price, 35 cents' While each chart is complete in itself, the set covers all maintenance instructions for the entire automobile. Sold singly. Securely wrapped. Location of Starting and Lighting System Faults. The most complete chart yet devised, showing all parts of the modern automobile starting, lighting and ignition systems, giving in- structions for systematic location of all faults in wiring, lamps, motor or generator, switches and all other units. Invaluable to motorists, chauffeurs and repairmen. Size 24 x 38 inches Price, 35 cents Location of Ford Engine Troubles Made Easy. Chart showing clear sectional views depicting all portions of the Ford power plant and auxiliary groups. It outUnes clearly all parts of the engine, fuel supply systems, ignition group and cooling system, that are apt to give trouble, detailing all derange- ments that are liable to make an engine lose power, start hard, or work irregularly. This chart simplifies location of all engine faults. Size 25 X 38 inches. Price, 35 cents Location of Motorcycle Troubles Made Easy. This chart simplifies location of all power-plant troubles and will prove of value to all who have to do with the operation, repair or sale of motorcycles. No details omitted. Size 30 x 20 inches. Price, 35 cents AVIATION A B C of Aviation. By Capt. V. W. Page. This book describes the basic prin- ciples of aviation, tells how a balloon or dirigible is made and why it floats in the air. Describes how an airplane flies. It shows in detail the different parts of an airplane, what they are and what they do. Describes all types of airplanes and how they differ in construction; as well as detailing the advantages and disadvantages of different types of aircraft. It includes a complete dictionary of aviation terms and clear drawings of leading airplanes. ^ The reader will find simple instructions for unpacking, setting up and rigging airplanes. A full description of airplane control principles is given and methods of flying are discussed at length. This Book answers every question one can ask about mod- ern aircraft, their construction and operation. A self educa- tor on aviation without an equal. 275 pages, 130 specially made illustrations with 7 plates. Price, $2.50 Aviation Engines — Design; Construction; Repair. By Lieut. Victor W. Page, Aviation Section, S.C.U.S.R. This treatise, written by a recognized authority on all of the practical aspects of internal combustion engine construc- tion, maintenance and repair, fills the need as no other book does. The matter is logically arranged; all descriptive mat- ter is simply expressed and copiously illustrated, so that any- one can understand airplane engine operation and repair even if without previous mechanical training. This work is m- valuable for anyone desiring to become an aviator or aviation mechanician. The latest rotary types, such as the Gnome Monosoupape, and LeRhone, are fully explained, as well as the recently developed Vee and radial types. The subjects of carburetion, ignition, cooling and lubrication also are covered in a thorough manner. The chapters on repair and maintenance are dis- tinctive and found in no other book on this subject. Not a technical book, but^ a practical, easily understood work of reference for all interested in aeronautical science. 576 pages, 253 illustrations. , Price, Net, $3.00 Glossary of Aviation Terms — English-French; French-English. A complete glossary of practically all terms used in aviation^ having lists in both French and English with equivalents in. either language compiled by Lieuts. Victor W. Page, A.S., S.C.U.S.R., and Paul Montariol, of the French Flying- Corps. Price, Net, $1.00 Aviation Chart — Location of Airplane Power Plant Troubles Made Easy. v By Lieut. Victor W. Page, A.S., S.C.U.S.R. A large chart outlining all parts of a typical airplane power plant, showing the points -^here trouble is apt to occur and suggesting' remedies for the common defects. Intended especially for aviators and aviation mechanics on school and field duty. Price. 35 centa 8 BRAZING AND SOLDERING Brazing and Soldering. By James F. Hobart. The only book that shows you just how to handle any job of brazing or soldering that comes along; it tells you what mixture to use, how to make a furnace if you need one.^ Full of valuable kinks. The fifth edition of this book has just been published, and to it much new matter and a large number of tested formulas for all kinds of solders and fluxes have been added. Price, 35c. CHARTS Aviation Chart — Location of Airplane Power Plant Troubles Made Easy. By Lieut. Victor W. Page, A.S., S.C.U.S.R. A large chart outlining all parts of a typical airplane power plant, showing the points where trouble is apt to occur and suggesting remedies for the common defects. Intended especially for aviators and aviation mechanics on school and field duty. price, 35 cents Modern Submarine Chart— With 200 Parts Num- bered and Named. A cross-section view, showing clearly and distinctly all the interior of a submarine of the latest type. No details omitted — everything is accurate and to scale. This chart is really an encyclopedia of a submarine. Price, 35 cents Box Car Chart. A chart showing the anatomy of a box car, having every part of the car numbered and its proper name given in a reference list. Price, 25 cents Gondola Car Chart. A chart showing the anatomy of a gondola car, having every part of the car numbered and its proper reference name given in a reference list. Price, 25 cents Passenger Car Chart. A chart showing the anatomy of a passenger car, having every part of the car numbered and its proper name given in a reference list. Price, 25 cents Steel Hopper Bottom Coal Car. A chart showing the anatomy of a steel hopper bottom coal car, having every part of the car numbered and its proper name given in a reference list. Price, 25 cents Tractive Power Chart. A chart whereby you can find the tractive power or drawbar pull of any locomotive without making a figure. Shows what cylinders are equal, how driving wheels and steam pressure affect the power. What sized engine you need to exert a given drawbar pull or anything you desire in this line. Price, 50 cents 9 Horse-power Chart. Shows the horse-power of any stationary engine without calculation. No matter what the cylinder diameter of stroke, the steam pressure or cut-off, the revolutions, or whether condensing or non-condensing, it's all there. Easy to use, accurate and saves time and calculations. Especially useful to engineers and designers. Price, 50 cents Boiler Room Chart. By George L. Fowler. A chart — size 14 x 28 inches — showing in isometric perspective the mechanisms belonging in a modern boiler room. This chart is really a dictionary of the boiler room — the names of more than 200 parti, being given. Price, 25 cents COKE Coke — Modern Coking Practice, Including An- alysis of Materials and Products. By T. E. Christopher and T. H. Byrom. This, the standard work on the subject, has just been revised and is now issued in two volumes. It is a practical work for those en- gaged in Coke manufacture and the recovery of By-products. Fully illustrated with folding plates. It has been the aim of the authors, in preparing this book, to produce one which shall be of use and benefit to those who are associated with, or interested in, the modern developments of the industry. Among the chapters contained in Volume I are: Introduc- tion; Classification of Fuels; Impurities of Coals; Coal Washing; .Sampling and Valuation of Coals, etc.; Chlorific Power of Fuels; History of Coke Manufacture; Develop- ments in Coke Oven Design; Recent Types of Coke Ovens; Mechanical Appliances at Coke Ovens; Chemical and Physi- cal Examination of Coke. Volume II covers By-products. Each volume is fully illustrated, with folding plates. Price, $3.00 per valume COMPRESSED AIR Compressed Air in all Its Applications. By Gardner D.^ Hiscox. This is the most complete book on the subject of air that has ever been issued, and its thirty-five chapters include about every phase of the subject one can think of.^ It may be called an encyclopedia of compressed air. It is written by an expert, who, in its 665 pages, has dealt with the subject in a comprehensive manner, no phase of it being omitted. Over 500 illustrations. Fifth Edition,, revised and enlarged. Cloth bound, $6.00. Half Morocco, Price, $7.50 CONCRETE Concrete Wall Forms. By A. A. Houghton. A new automatic wall clamp, is illus- trated with working drawings. Other types of wall forms, clamps, separators, etc., are also illustrated and explained. Price, 60 cents 10 Concrete Floors and Sidewalks. By A. A. Houghton. The molds for molding squares, hex- agonal and many other styles of mosaic floor and sidewalk blocks are fully illustrated and explained. Price, 60 cents Practical Concrete Silo Construction. By A. A. Houghton. Complete working drawings and speci- fications are given for several styles of concrete silos, with illustrations of molds for monolithic and block silos. The tables, data, and information presented in this book are of the utmost value in planning and constructing all forms of concrete silos. Price, 60 cents Molding Concrete Bath Tubs, Aquariums and Natatoriums. By A. A. Houghton. Simple molds and instruction are given for molding different styles of concrete bath tubs, swimming pools, etc. Price, 60 cents Molding Concrete Chimneys, Slate and Roof Tiles. By A. A. Houghton. The manufacture of all types of con- crete slate and roof tile is fully treated. Valuable data on all forms of reinforced concrete roofs are contained within its pages. The construction of concrete chimneys by block and monolithic systems is fully illustrated and described. A number of ornamental designs of chimney construction with molds are shown in this valuable treatise. 60 cents Molding and Curing Ornamental Concrete. B}'' A. A. Houghton. The proper proportions of cement and aggregates for various finishes, also the methods of thor- oughly mixing and placing in the molds, are fully treated. An exhaustive treatise on this subject that every concrete worker will find of daily use and value. Price, 60 cents Concrete Monuments, Mausoleums and Burial Vaults. By A. A. Houghton. The molding of concrete monuments to imitate the most expensive cut stone is explained in this treatise, with working drawings of easily built molds. Cutting inscriptions and designs is also fully treated. 60 cents Concrete Bridges, Culverts and Sewers. By A. A. Houghton. A number of ornamental concrete bridges with illustrations of molds are given. A collapsible center of core for bridges, culverts and sewers is fully illus- trated with detailed instructions for building. 60 cents Constructing Concrete Porches. By A. A. Houghton. A number of designs with working drawings of molds are fully explained so any one can easily construct different styles of ornamental concrete porches without the- purchase^ of expensive molds. Price, 60 cents 11 Molding Concrete Flower Pots, Boxes, Jardi- nieres, Etc. By A. A. Houghton. The molds for producing many original designs of flower pots, urns, flower boxes, jardinieres, etc , are fully illustrated and explained, so the worker can easily construct and operate same. Price, 60 cents Molding Concrete Fountains and Lawn Orna- ments. By A. A. Houghton. The molding of a number of designs of lawn seats, curbing, hitching posts, pergolas, sun dials and other forms of ornamental concrete, for the ornamentation of lawns and gardens, is fully illustrated and described. 60c. Concrete on the Farm and in the Shop. By H. CoLviN Campbell. This is a new book from cover to cover, illustrating and describing in plain, simple language many of the numerous appliances of concrete within the range of the home worker. Among the subjects treated are: Principles of reinforcing; methods of protecting concrete so as to insure proper hardening; home-made mixers; mixing by hand and machine; form construction, described and illustrated by drawings and photographs; construction of concrete walls and fences; concrete fence posts; concrete gate posts; corner posts; clothes line posts; grape arbor posts; tanks; troughs; cisterns: hog wallows; feeding floors and barnyard pavements; foundations; well curbs and plat- forms; indoor floors; sidewalks; steps; concrete hotbeds and cold frames; concrete slab roofs; w^alls for buildings; repairing leaks in tanks and cisterns; ,and all topics associated with these subjects as bearing upon securin