REESE LIBRARY
OF THK
UNIVERSITY OF CALIFORNIA.
No.
ELECTRICAL DESIGNS
COMPRISING INSTRUCTIONS FOR CONSTRUCTING
SMALL MOTORS, TESTING INSTRUMENTS
AND OTHER APPARATUS; WITH
WORKING DRAWINGS FOR
EACH DESIGN.
REPRINTED FROM
THE AMERICAN- ELECTRICIAN
NEW YORK:
AMERICAN ELECTRICIAN COMPANY
I QO I
3
COPYRIGHT, 1901,
BY
AMERICAN ELECTRICIAN COMPANY
PREFACE.
The chapters of this book originally formed articles written for the
AMERICAN ELECTRICIAN by the designers of the apparatus described,
which, in many cases, had been actually built and used prior to the pub-
lication of the description. The designs were all prepared with a view
to reducing to the simplest degree the tools and facilities necessary for
the construction of the apparatus. The designs for such of the small
motors as have not been built have been calculated with ample allow-
ance for variations in the quality of iron and steel, so that any discrep-
ancy between the performance predicted for these machines and that
actually realized will be in favor of the motor, if the material is of a grade
at all allowable in good foundry practice.
9409?
TABLE OF CONTENTS
CHAPTER I PAGE
ONE-SIXTH HORSE-POWER MOTOR, WITH DRUM ARMATURE CECIL P. POOLE. . i
Windings for ns-volt circuit and battery current.
CHAPTER II
ONE-SIXTH HORSE- POWER MOTOR, WITH RING ARMATURE, do. ..8
Windings for us-volt circuit and battery current.
CHAPTER III
ONE-FOURTH HORSE- POWER MOTOR, WITH DRUM ARMATURE,. ... do. . . 15
Windings for 115 and 23o-volt circuits and battery current.
CHAPTER IV
ONE-FOURTH HORSE-POWER MOTOR, WITH RING ARMATURE, do. . . 24
Windings for 115 and 23O-volt circuits and battery current.
CHAPTER V
ONE-HALF HORSE-POWER MOTOR, WITH DRUM ARMATURE, do. .. 32
Windings for 115 and 23O-volt circuits.
CHAPTER VI
ONE HORSE- POWER BIPOLAR MOTOR, WITH DRUM ARMATURE do. . . 44
Windings for 115 and 230- volt circuits.
CHAPTER VII
ONE HORSE- POWER FOUR-POLAR MOTOR, WITH DRUM ARMATURE, do. .. 57
Designs for cast-iron and cast-steel magnets and windings
c or 115 and 230- volt circuits.
CHAPTER VIII
TWO HORSE-POWER FOUR-POLAR MOTOR, WITH DRUM ARMATURE, do. ..69
Designs for cast-iron and cast-steel field magnets and wind-
ings for 115, 230 and soo-volt circuits.
CHAPTER IX]
THREE HORSE-POWER MOTOR, WITH DRUM ARMATURE, do. . . 79
Design for cast-iron field magnet ring with wrought-iron
cores, and windings for 115, 230 and 500 volts.
VI TABLE OF CONTENTS
CHAPTER X
ONE-KILOWATT COMBINED ALTERNATING AND DIRECT-CURRENT
MACHINE, J. C. BROCKSMITH. . 87
Design for a one-kilowatt machine which may be used as a
direct-current generator or motor; a single-phase, two-phase
or three-phase alternating-current generator or synchronous
motor; a rotary converter changing single-phase, two-phase
or three-phase alternating currents to direct current; an in-
verted rotary converter changing direct current into single-
phase, two-phase or three-phase alternating currents; a
phase transformer to effect any change in alternating cur
rents within the range of three phases.
CHAPTER XI
TWO-KILOWATT COMBINED ALTERNATING AND DIRECT-CURRENT
MACHINE, do ..loo
Design for a two-kilowatt machine which may be used as
above.
CHAPTER XII
FOUR-KILOWATT COMBINED ALTERNATING AND DIRECT-CURRENT
MACHINE, do. ,.107
Design for a four-kilowatt machine which may be used as
above.
CHAPTER XIII
SINGLE-PHASE RECTIFIER, do. . .116
A machine for changing single-phase alternating current
into direct current.
CHAPTER XIV
UNIVERSAL ALTERNATOR FOR LABORATORY PURPOSES PROF. H. C. CARHART. .125
A revolving-field machine from which may be taken single-
phase, two-phase or three-phase alternating currents.
CHAPTER XV
ONE-QUARTER HORSE- POWER SINGLE-PHASE INDUCTION MOTOR,. P. M. HELDT. .131
CHAPTER XVI
SIMPLE TRANSFORMER IN FOUR SIZES, CECIL P. POOLE. .140
Core-type transformer with a sub-divided primary winding
to work on a 200, 400 or looo-volt circuit, and sub-divided
secondary from which may be taken 18, 32, 50 or 100 volts.
CHAPTER XVII
THE CONSTRUCTION OF A REACTIVE COIL, do. . .147
A specific design with instructions for adapting it to other
conditions.
TABLE OF CONTENTS vii
CHAPTER XVIII PAGE
THE CONSTRUCTION AND CALCULATION OF RHEOSTATS, P. M. HBLDT. . 154
Rules and formulas governing the design of dynamo and
motor rheostats.
CHAPTER XIX
SIMPLE VOLTMETERS, AMMETERS AND WATTMETERS, CHAS. T. CHILD. . 162
Instructions for making magnetic-vane, permanent-magnet
and galvanometer-type ammeters, a hot-wire voltcaeter, and
dynamometer-type and Aron-type wattmeters.
CHAPTER XX
D'ARSONVAL GALVANOMETER, , .Eow pE. SHELDON. .174
CHAPTER XXI
SENSITIVE MIRROR GALVANOMETER JAS. F. HOBART. . 180
CHAPTER XXII
THOMSON ASTATIC GALVANOMETER H. S. WEBB. . 185
CHAPTER XXIII
CHEAP TESTING SET, , JAS. F. HOBART. .194
CHAPTER XXIV
CONSTRUCTION AND USE OK A PHOTOMETER, PROF. A. J. ROWLAND. .198
CHAPTER XXV
CONSTRUCTION OF A SIMPLE STORAGE BATTERY, CECIL P. POOLE. .209
CHAPTER XXVI
CONSTRUCTION OF A CONSTANT- POTENTIAL ARC LAMP, do. , .214
CHAPTER XXVII
AN EXPERIMENTAL NERNST LAMP, -W. S. FRANKLIN. .220
CHAPTER XXVIII
CONSTRUCTION OF AN INDUCTION COIL, GEO. T. HANCHETT. . 223
CHAPTER XXIX
CONSTRUCTION OF A TESLA-THOMSON HIGH-FREQUENCY COIL,.. . . A. F. McKissiCK. .230
CHAPTER XXX
CONDENSER FOR EXTREMELY HIGH POTENTIALS, GEO. T. HANCHETT. .234
CHAPTER XXXI
CONSTRUCTION OF A WIMSHURST INFLUENCE MACHINE ... do. . .237
viii TABLE OF CONTENTS
CHAPTER XXXII PAGE
I ELEPHONE TRANSMITTER AND RECEIVER. E. E. CLEMENT. . 243
CHAPTER XXXIII
CONSTRUCTION OF A DRY BATTERY CELL, .TOWNSEND WOLCOTT O =>25O
CHAPTER XXXIV
SOME HANDY COMMUTATOR. TOOLS, ALTON D. ADAMS.. 255
CHAPTER I.
ONK-6IXTH HORSE-POWER MOTOR WITH DRUM ARMATURE.
In preparing this design and those which follow, it has been assumed
that any one who is sufficiently interested in the subject to undertake the
construction of a motor or dynamo will be sufficiently familiar with
electro-mechanics to exercise individual judgment in the matter of fitting
the various parts, and also in the design and construction of journal boxes,
brush holders, terminal blocks and such other parts as are not of vital
importance in the electrical design of the machines. Detailed descrip-
tions of these parts will, therefore, not be given ; the reader may easily
**
FIG. I. ELEVATION OF FIELD MAGNET.
inform himself concerning these, if necessary, by inspecting a finished
machine of almost any type, or by reference to any good text-book.
The accompanying sketches are intended to serve as working draw-
ings in the construction of a 1-6 horse-power motor, for operation upon a
no-volt direct-current circuit. In Figs, i and 2, M is the field magnet,
consisting of a bar of wrought iron three inches wide and one inch thick,
ELECTRICAL DESIGNS
bent into the shape shown; the inner surface of each limb is machined
smooth a distance of three inches, forming shallow mortises to receive
the pole-pieces, P P, which are secured by Ji-in. cap screws passing-
through the magnet limbs. The pole-pieces, P P, are of gray cast iron,
and should be finished on all sides to remove the scale as well as to im-
prove the appearance of the completed machine. The magnet, M 9 might
be made to look neater by touching up its sides on a coarse emery wheel ;
it should be well annealed after bending and finishing.
Two holes, h, h, are bored through the pole-pieces, after these are
fitted to the magnet, but before the armature chamber is bored out.
These holes are 17-64 in. diameter, and they must be 3^4 i ns - apart,
center to center, and equ'distant from the center of the armature cham-
ber; if the magnet limbs conform strictly to the measurement given
from face to face of the finished part of the limbs, the centers of the
holes, h, h, will each be 3-16 in, from the joint between the magnet
and the pole pieces. In these holes are to be inserted ^4-in. iron or
steel rods 7*4 i ns - long, threaded at each end a distance of J4 in.
o
o
o
FIG. 2. PLAN OF FIELD MAGNET AND JOURNAL YOKES.
Fig. 2, which gives a plan view of the magnet and the journal yokes,
Y, Y, shows the function of these rods ; they support the yokes and carry
distance-pieces, c, c, d, d y made of brass tubing just large enough to slip
over the rods, and having J/g-in. walls. The pieces, c, c, are 1^3 ins. long,
and d, d, are 2% ins. long. The yokes are held in place by brass nuts, not
shown in Fig. 2.
ONE-SIXTH HORSE-POWER MOTOR 3
The journal yokes, Y, Y, are alike. They are of cast brass, J/ in.
thick, with a stiffening rib l /$ in. thick, on each side of the journal box.
The inner end of one box should be trued up to receive the brush arm or
quadrant. The yokes may be much more easily and accurately fitted if a
steel template is used. This may be cheaply provided by taking a piece
of flat steel, I in. wide and 4^/2 ins. long, scribing a straight line approxi-
mately down the center, and drilling three holes as shown by T, Fig. 2,
the center one 11-16 in. and the others ;4 in. in diameter. After the box
is bored, mount it on a mandrel and turn down the inner end to fit the
center hole in the template T, and at the same time face up the ends of
the yoke where they are to touch the distance-pieces ; put the template on
the end of the box and scribe the positions of the j4 m - holes on the ends
of the yoke. This template should also be used to fix the distance apart
of the holes, h, h (Fig. i). The boxes are bored out 9-16 in. in diameter
and fitted with bushings of J4 m - bore and I in. long; oil grooves should
be cut at each end of the box and provision made for taking out the oil.
Oil cups may be used to feed the bearings.
After the yokes are fitted the frame may be centered in a lathe as fol-
lows, for boring out the pole-pieces. Take a piece of J^-in. steel rod it
ins. long, and make the shaft ^(Fig.3) ; the distance from e to g is 3^4 ins.*
and the diameter there is % in. ; from g to i is 3 11-16 ins. and the diam-
eter y in. ; from i to / is I 1-16 ins. and the diameter J / 2 in. ; from / to k is
3 ins., and the diameter is J4 m - Turn the ends of the shaft down to a
point, like that of a lathe center; put it in the boxes, bolt the yokes in
place, and then put the frame on the lathe carriage, adjusting it until the
sharp ends of the shaft are in exact line with the lathe centers. Bolt the
motor down in this position, remove the yokes and shaft, and bore out
the pole-pieces. The ends of the shaft should afterwards be squared off,
care being taken to cut exactly l /2 in. off each end, leaving the shaft 10
ins. long.
The armature (Fig. 3) is built up of iron discs 3 ins. in diameter and
not more than 1-32 in. thick; there are twelve slots, each 5-16 in. wide
and 7-16 in. deep. These may be punched in each disc separately, if a
stamping press is available, or they may be milled after the discs are as-
sembled on the shaft. If the slots are milled the discs should be taken
off the shaft afterwards and the burrs dressed off, care being taken to re-
assemble them exactly as they were when the slots were milled ; this may
be accomplished by taking a very slight cut with a metal saw along the
top of one tooth, using the mark as a guide to get the proper slots to-
gether. In order to get them in exact alignment, a rectangular bar of
metal should be made to fit snugly in one slot before taking the discs off;
ELECTRICAL DESIGNS
when they are put back this bar is inserted in the slot to which it was
fitted and the nut is set up hard. End plates, WW, of brass, 2 ins. in diam-
eter and 3-16 in. thick, serve to prevent the end discs from buckling when
they are compressed. A nut (not shown) fitted to the thread which be-
gins at g on the shaft, serves to clamp the discs, which are held at the
other end by the shoulder, i\ no key is necessary to prevent the discs
from turning on the shaft in so small a machine, but it is essential that
they should be clamped as tightly as a fairly strong man can clamp them,
using a six-inch wrench on the nut. The shaft may be held in a pipe vise
between i and / when setting up the nut ; the nut should be made of very
hard bronze metal in preference to steel, as the latter attracts magnetic
lines of force and is liable to heat.
The commutator may be made as shown in the sketch, or according
to any other modern plan, a number of which were described in the
1
t s
\2
ja
!
3
Am.Elec.
FIG. 3. DETAILS OF ARMATURE AND INSULATING TROUGH.
"American Electrician" for July, 1896. The only essential features are
the space along the shaft which must not exceed % in., the width of face,
which should not be less than y 2 in., and the number of segments, which
must be 12. The commutator here shown is intended to be secured to
the shaft by a small steel set-screw through the hub or boss at the front ;
the end of this hub, /, must be 1^4 ms - from the end of the shaft. Ex-
treme care must be taken to insulate the segments from the shell as well
as from each other; mica is the only reliable material for this purpose.
Carbon brushes 3/2 in. wide and J/J i n - thick should be used.
The armature core is next prepared for winding. Cut four discs of
heavy drilling (so-called twilled muslin), 2^-4 ins. in diameter, with a Y&
ONE-SIXTH HORSE-POWER MOTOR
in. hole in the center ; varnish the ends of the armature core with shellac
and varnish two of the cloth discs, each on one side ; thread them on the
shaft, one at each end, with the varnished sides next to the core, and press
them tightly on the core. While the varnish is hardening cut 24 pieces
of drilling the shape of / (Fig. 3) ; cut two slits % in long in each end, 7-16
in. from each side and 5-16 in. from each other; varnish the strips on one
side, and when nearly dry bend them along the dotted lines so as to form
troughs, with the varnish inside the trough. Varnish the outside of each
trough and the walls of the slots in the core ; put two troughs in each slot
and turn the flaps, u, v, w, flat against the end of the core, applying
enough fresh shellac to hold them down. Then put on the two remaining
end discs of cloth, first varnishing the sides next to the armature ; after
they are in place varnish the outsides and put the core in an oven to bake,
being careful that the oven is not hot enough to scorch the cloth. A
temperature of 130 degs. Fah. is sufficient. After baking, tape the shaft
-n
FIG. 4. MAGNET-COIL HEAD.
Line
FIG. 5. DIAGRAM OF STARTING SWITCH.
thoroughly from to /, and from the other end of the core to where the
commutator will come.
The coils consist of 48 turns of No. 24 double cotton-covered
wire each, wound 8 turns wide and 6 deep in the slots, but spread out as
flat as possible across the heads. Wind coil No. I in slots, A A' \ coil No.
2 in BB'\ No. 3 in C C\ No. 4 in >>'; No. 5 inEE', and No. 6 in FF.
Coil No. 7 goes in A' A, on top of coil No. i, but beginning on the oppo-
site side of the core, as indicated by the lettering; No. 8 in B' B ; No. 9 in
C C; No. 10 in D' D ; No. n in E' E, and No. 12 in F' F. After winding
each coil bring the finishing end across to the slot where the starting end
enters and twist the two lightly together. When all the coils are on un-
6 ELECTRICAL DESIGNS
twist the coil ends and twist the last end of each coil to the starting end
of the coil in the slot next to it on the right ; these twisted ends go each to
a commutator segment, in regular order.
The field magnet is easily made ready to wind by taping the hori-
zontal part of the magnet, two layers deep, with varnished muslin and
putting on two fibre heads. One of these heads is shown by H (Fig. 4).
It is in two pieces, the seams being at the ends, and is cut from J^-in. sheet
fibre. The two halves may be clamped together on the core by means
of a small brass wire drawn around the outer edge, laying in a shallow
groove, the ends being twisted and cut close. The pole pieces should be
removed before taping and putting on the heads, to facilitate these opera-
Am.Elec.
FIG. 6. COMPLETED MOTOR FROM THE PULLEY END.
tions as well as the winding of the coil. One fibre head has a notch, n,
half way of its inner long side, to enter the field wire. The coil consists
cf No. 28 wire, B. & S. gauge, 34 layers deep and 170 turns long, making
5,780 turns in all. The field winding is connected in shunt to the brushes,
and it would be a good plan to provide a starting switch and resistance
lamp connected up as shown diagrammatically by Fig. 5, where F is the
field coil, B B the brushes, L a 32-candle power, loo-volt incandescent
lamp, 5 the starting switch, m a magnet, and S W a. double-pole snap
switch. This arrangement could be mounted on the base of the motor.
Fig. 6 shows the complete motor on a wooden base, Q, without the pul-
ONE-SIXTH HORSE-POWER MOTOR 7
ley ; the latter may be any diameter between ty ins. and 23/2 ins., with
a i-in. crown face or J/>-in. grooved face. The- motor is secured to the
base by flat head machine screws from below, entering the ends of the
wrought iron and countersunk in the under side of the wood. This ma-
chine will stand a momentary overload of 100 per cent., and will work up
to ^4 horse-power for half an hour at a time.
WINDINGS FOR BATTERY SERVICE.
In order to adapt this motor for use in connection with a battery the
following windings, etc., must be substituted for those specified above:
The armature to be wound with six coils of No. 12 wire, each having
twelve turns (three wide and four deep in a slot). The field wire will
be No. 19, wound 17 layers deep and 83 turns in length. The commu-
tator will have six segments, and should have a brush surface Y^ in. wide ;
copper brushes ^ X l /2 in. should be used, the contact faces being cut to
such a bevel as to present an area of l /2 in. square at least. Connect
the field winding in shunt with the armature, instead of in series as is
usually done. This winding is for 6 volts. The machine thus wound
will stand an armature current of 25 to 30 amperes.
CHAPTER II.
ONE-SIXTH HORSE-POWER MOTOR WITH RING ARMATURE.
In Figs 7 and 8 M is a wrought iron magnet core, P P cast-iron pole-
pieces, C the armature core, and Y the journal yoke. The magnet core,
M, is made from a % in. X 4j in. bar of commercial wrought iron bent
to the shape shown. The faces of the arms are machined to a depth of
1-16 in., where the pole-pieces, P P, are attached, so as to form a magnetic
joint of as low reluctance as possible. The pole-pieces are secured to the
magnet arms by j4-in. cap screws passing through smooth holes in the
arms and tapped into the pole-pieces; the latter are of grey cast-iron,
and should be finished on all sides sufficiently to remove the scale. The
magnet, M, might be improved in appearance by touching up its sides
with a coarse emery wheel ; it should be thoroughly annealed after bend-
ing and finishing. It will be noticed by reference to Fig. 8 that the ends
of the magnet arms project slightly beyond the outer faces of the pole-
pieces ; this is done in order to furnish a guide for the flanges of the jour-
nal yoke arms. After fitting the pole-pieces to the magnet arms the com-
plete magnet frame is bolted to the lathe* carriage in position for boring
out the pole-pieces ; before this is done it is necessary to drill a hole
through the back of the magnet to allow the boring bar to pass through,
and also to form a seat for the rear bearing. This hole is ^4 m - i n diam-
eter, and the magnet frame must not be allowed to move from its original
position on the lathe carriage from the time the hole is drilled until all
the circular tooling on it is accomplished.
After drilling the hole in the back of the magnet adjust the boring
bar and bore the armature chamber out, 4 11-16 ins. in diameter. Next
adjust the boring tool so that it will scribe on the ends of the magnet arms
arcs of a circle 6 ins. in diameter; then cut away the wrought iron inside
the scribed marks, down flush with the pole-pieces, as shown in Fig. 7,
forming recesses for the flanges of the journal yoke. The yoke and box
are cast in one piece of brass or other non-magnetic composition ; the
shell of the box is i% ins. long, and projects l /^ in. beyond the inner face
ONE-SIXTH HORSE-POWER MOTOR
Am. Elee.
FIG. 7. ELEVATION OF FIELD MAGNET AT THE COMMUTATOR END.
J
FIG, 8. PLAN OF FIELD MAGNET AND DETAIL OF JOURNAL BOX.
io ELECTRICAL DESIGNS
cf the yoke ; the outer diameter of the shell is ^4 in., and it is bored out to
7-16 in. inner diameter and bushed to % m - The yoke and arm portions
are 3-16 in. thick, with a J /$ in. stiffening rib on each side of the box, and
the arms taper from i in. wide at the flanges to about y 2 in. near the box.
The flanges are 2 ins. long, ^4 m - .wide and % in. thick after facing; the
arms, beyond the bends, are sufficiently long to make the distance from
the face of the pole-piece to the inner face of the yoke 2 ins. After boring
the box it is mounted on a stiff mandrel and the surfaces of the flanges
that go next the magnet are faced up true ; next, the outer edges of the
flanges are skimmed off until the yoke fits snugly between the curved
edges of the recesses previously cut in the ends of the wrought iron mag-
net. Care must be taken in making the pattern for the yoke that the
inner edges will not project inward beyond the bore of the pole-pieces.
The yoke is fastened to the pole-pieces by screws, as indicated in Fig. 7.
The rear bearing, /, is a little peculiar in construction. The box por-
tion is similar to that part of the yoke, but it is cast with a flange, /, I in.
from the farthest end of the shell, which is 1^2 ins. long. A collar, n, is
fitted to screw onto the outer end of the shell, which is threaded for
that purpose. The shell is turned down outside to fit snugly in the hole
drilled in the back of the magnet, and when it is inserted in the hole the
collar, n, is put on and screwed up tight. This box, like the front one,
is bushed to j4 in- bore. The drawing shows the flange, /, and collar, M,
countersunk in the metal of the magnet ; this will not be necessary if the
magnet is smoothed up with an emery wheel, as above suggested, the
object in countersinking being to provide smooth, true bearing surfaces
for the flange and collar.
The armature core, spider and shaft are shown, partly in cross-sec-
tion, by Fig. 9. The core is built up of charcoal iron (not steel) rings,
4*/2 ins. outside diameter and 2 l / 2 ins. inside, not more than 1-32 in. thick ;
these are assembled on a brass drum, shown by Fig. n, which should be
25/& ins. outside diameter before finishing, so that it may be turned down
to exactly fit the inner circle of the armature rings ; the wall of the drum
is y% in. thick after finishing, and there are four equidistant projecting
lugs, /, }/ 2 in. long, on each end by which the drum is secured to the spi-
der (see Figs. 9 and io). The rings forming the core, C (Fig. 9), are
compressed and held on the drum, r, by two brass washers, w, w, 3-16 in.
thick and 3^ ins. outer diameter, which screw onto the ends of the drum.
The core, when compressed, is ift ins. long, and has 20 slots J4 in- wide
and 7-16 in. deep; the washers, w, w, must be set up as tight as the
threads will stand.
The spider, s (Figs. 9 and io), is made of brass, and consists of a hub
ONE-SIXTH HORSE-POWER MOTOR
II
in. diameter, 2 ins. long and 3/2
in. bore) and four arms having T-
shaped ends, the wide part or heads of which project beyond the arms
and hub at each' end, the length of these heads being 2j^ ins. and their
width y ins. The heads of the spider arms are turned off to fit very
closely inside the drum, r, which is mounted on the spider in such a posi-
tion as to bring the spider arms in alignment with the lugs, /, of the drum;
FIG. 9. AXIAL SECTION OF ARMATURE
" * CORE.
FIG. 10. END OF ARMATURE CORE.
An. Elec.
FIG. II. ARMATURE CORE DRUM.
screws through the spider arms into the lugs hold the drum and spider
together.
The shaft, S, is & l / 2 ins. long; the portion /is ij/s ins. long and *4 m -
diameter; k is i in. long and ^8 in. diameter; the part passing through
the core is 3 ins. long and J/2 in. in diameter ; m is ^ in. long and y% in.
diameter ; and p is 3 in. long and *4 in. diameter. The spider, s, may be
secured to the shaft by a key or a set-screw ; the set-screw is sufficient in
12 ELECTRICAL DESIGNS
so small a machine. The commutator (not shown) must not be more than
24 in- over all, along the shaft ; it must have y 2 in. brush surface and 20
segments ; other details may be made to suit the will of the builder. The
front end of the commutator must be not less than 3-16 in. from the
shoulder where / and k join.
The armature is next prepared for winding by removing the drum
and core from the spider and insulating the ends and interior of the core
and the walls of the slots. Cut four rings of heavy drilling of a size to cover
the washers, w w, and the ends of the drum, r ; varnish two of them on one
side with shellac, and apply them to the ends of the armature body.
While these are hardening cut forty strips of drilling ij ins. wide and
2^4 ins. long; in each end of each of these cut two slits 54 m - l n g parallel
with the sides, and located 7-16 in. from each side of the strip. Varnish
these on one side, and when nearly dry fold them into troughs to fit the
slots, two troughs to a slot, one within the other; fold them so that the
varnish will be on the inside of the trough.
When these are dry varnish the slots and the outsides of the troughs
and put the latter in the slots, bending the ends flat against the core and
securing them there with a little fresh varnish. Then varnish the ends
of the core (two cloth rings being on them), and one side of the two re-
maining rings of drilling ; put these rings on top of the first ones, varnish
them on the outside, and put the core in an oven to bake. The armature
coils consist of No. 24 double cotton-covered wire, wound six turns wide
and twelve layers deep. Before winding them four strips of wood 3 ins.
long. y% in. wide and ^ in. thick should be screwed to the inner wall of
the brass drum, in line with the lugs, /, so as to preserve the spaces for the
four arms of the spider. A double thickness of drilling should also be
applied to the interior of the drum to insulate the coils from it. The
connections are the simple Gramme ring arrangement.
The field winding is necessarily divided into two coils, on account of
the rear bearing passing through the magnet. Kach coil consists of No.
28 double cotton-covered wire, wound 17 layers deep and 181 turns or
more long; the two coils are connected in series with each other and in
shunt to the brushes. Heads of hard fibre J/ in. thick should be used to
protect the ends of each coil ; one of these is shown by H (Fig. 12), but the.
width should be 7-16 in. instead cf ^4 in. as marked. 11
It is in two pieces, the seams being at the ends, and is cut from y% in. '
sheet fibre. The two halves may be clamped together on the core by
means of a small brass wire drawn around the outer edge, laying in a
shallow groove, the ends being twisted and cut close. The pole-pieces
should be removed before taping and putting on the heads, to facilitate
OXE-SIXTH HORSE-POWER MOTOR 13
these operations as well as the winding of the coil. One fibre head has a
notch, n, half way of its inner long side to enter the field wire. The pole-
pieces should, of course, be removed before winding the field coil, and
the magnet core should be wrapped with two layers of varnished drilling
where the coils are to go. The entering end of each coil should be re-
mote from the journal, and this means that the magnet must be turned
end for end after one coil is wound, or else the two coils must be wound
in opposite directions in order that the free ends at the center of the mag-
net may be connected together. It is advisable to provide a starting
switch similar to the one shown diagrammatically by Fig. 13, where F.
is the field coil; J3 B the brushes; S the starting switch lever: L a 3 2 -can-
dle-power no-volt lamp; M a magnet, and SW a double-pole snap
switch.
The motor is intended to be mounted on a wooden base-board 8 ins.
X 8 ins., a cleat 3 ins. wide and 7-16 in. thick being put under the pole-
H
Am.Elec.
FIG. 12. MAGNET COIL HEAD.
.Cine
FIG. 13. DIAGRAM OF STARTING SWITCH.
pieces so as to clear the field coil. Bolts from beneath, tapped into the
magnet and countersunk in the under side of the base, should be used to
hold the motor on the base. The pulley may be any diameter from i in.
to 3 ins. by I in. face, if crowned, or }4 in. if grooved.
WINDINGS FOR BATTERY CURRENT.
The armature will be wound with 10 coils of No. 12 wire, each coil
having 16 turns and occupying two (adjacent) slots. The field wire will
be No. 1 8, wound 8 layers deep and 40 turns long in each coil; the two
coils containing 640 turns in all. The commutator must have 10 segments
14 ELECTRICAL DESIGNS
and a brush surface $ in. wide; copper brushes ^ X /4 in. should be
used, the contact faces being cut to such a bevel as to present a surface
at least }4 in. square each. The field winding is to be connected in shunt
to the brushes, instead of in series as is usually the practice in battery
motor construction. This winding is for 6 volts at the terminals; the
current required will depend upon the work done; the machine is cap-
able of standing an armature current of 25 to 30 amperes.
CHAPTER III.
ONE-FOURTH HORSE-POWER MOTOR WITH DRUM ARMATURE.
Fig. 14 represents the field magnet, and Fig. 15 one of the journal
yokes. The magnet is of the familiar single-coil type employed by West-
inghouse, Jenney and others. The core is of round Norway iron, 2 ins,
in diameter and 9 ins. long over all. The ends are turned tapering, as in-
dicated by dotted lines, to insure intimate contact with the yokes ; the
taper is from the full diameter to 1^4 ins., and begins 2 ins. from each end.
The pole-pieces are of cast-iron. Fig. 16 gives a plan view and a face
view of one pole-piece, from which all the essential dimensions may be
obtained. The arms which support the journal yokes are cast solid with
the pole-pieces, and their horizontal thickness tapers from }A in. at the
pole-piece to y\ in. where the yoke is bolted on.
In fitting the magnet frame together the best procedure is to bore the
tapered holes in the lower part of each pole-piece and turn the ends of the
magnet core to the same taper, but just a trifle large ; then dress each
taper down very gradually with a fine file (the core being run in a lathe)
until the pole-piece can be pushed on by hand far enough to bring the end
of the core within 1-32 in. of the back surface of the cast-iron. The pole-
pieces and ends of the core should be punch-marked, so as to insure finally
mounting each pole-piece on the end to which it was fitted. After dress-
ing down the ends of the core as above described, drill and tap in each
end a hole for a X~ m - machine screw, the purpose of which will be ap-
parent by glancing at the right-hand end of the magnet in Fig. 14, where
C is a four-armed claw or spider with a hole through the center where
the arms intersect. The arms are 3-16 in. thick, measured at right angles
to the bolt, and taper from 3-16 in. to }$ in. thick measured parallel with
it. One of these spiders is used at each end, though the drawing shows
it at only one end of the machine.
After drawing one pole-piece home solid by means of its spider and
bolt, slip the other pole-piece on loosely and clamp the pole-pieces lightly
i6
ELECTRICAL DESIGNS
FIG. 15. PLAN OF JOURNAL YOKE.
dxierfcgi) Electrician.
FIG. 14. ELEVATION OF FIELD MAGNET WITH JOURNAL YOKE IN POSITION.
ONE-FOURTH HORSE-POWER MOTOR
FIG. l6. PLAN AND FACE VIEWS OF ONE FIELD-MAGNET POLE-PIECE.
i8
ELECTRICAL DESIGNS
between two iron plates with planed surfaces, applied between the journal
arms, so as to keep the four horns of the pole-pieces in alignment ; then
force the second pole-piece home by means of its bolt and spider, and
clamp the horns hard between the iron plates. The bottom surfaces of
the cast-iron pieces should then be trued up on a planer or shaper and the
clamps taken off the pole-piece horns.
The next operation is boring the armature chamber and the seats
for the journal yokes. The armature chamber bore is 4 3-16 ins; the
seats for the journal yokes, marked "finished part" in Fig. 16, are bored
or cut to 4^ ins. diameter, and this must be done before the position oi
the machine is disturbed after boring the armature chamber. This com-
pletes the machine work on the magnet, except the bolt holes.
The journal yoke may be made of brass or any composition metal.
The bar is 3-16 in. thick and i in. wide, except near the ends, where it
FIG. 17. SHAFT AND CROSS SECTION OF ARMATURE CORE.
flares to correspond with the width of the arms. At each end is a right-
angled lug, y in. thick after machining; these lugs fit the seats in the
ends of the iron arms, and the yokes should be fitted to the magnet imme-
diately after finishing the machine work on the latter, and before it is
taken apart to put on the coil. The box portion is i]/ 2 ins. long over all,
3-16 in. of its length being on the inside of the yoke and iJ/6 ins. on the
outside. As shown by the plan view of the yoke in Fig. 15, there are
stiffening webs starting flush near the ends of the yoke and attaining a
width of 24 m - at the box; these are j in. thick. The box is j in. in
outer diameter, and bored to 17-32 in. inside; it is bushed to ^ in. diam-
eter. These latter dimensions, excepting the final inside diameter of the
ONE-FOURTH HORSE-POWER MOTOR
bushing, may be varied to suit individual ideas, as may also the design
of the box. The only essential measurements are those of the yoke-bar,
the length of the box and the bore of the journal bushing. The journal
yokes are held in place by l / in. cap screws passing through the iron arms
and tapping into the lugs of the yokes. .
Figs. 17, 1 8 and 19 show the shaft and armature core (the latter in
cross-section), an armature disc, and the shell and head. The discs are of
charcoal iron, 4 ins. outside diameter with a iJ/^-in. hole in the center and
a ^s-in. key-seat, annealed after punching and key-seating; there are
eighteen slots 3/ in wide and 3/ in. deep. The shell and one head are
cast in one piece (of brass), and consist of a barrel ij4 m s. outside diam-
eter (when finished) and 2 ins. long, with a head, s, at one end, 3^ ins. in
diameter and tapered in thickness from 3/4 in. near the center to 1-16 in.
at the periphery; at the opposite end of the barrel is a cross-bar J/& in.
thick, cast with the barrel and cf the shape, shown, being 3/ in. wide where
it joins the barrel and Y\ in. at the
center. A j^j-in. hole is drilled in
the center of this cross-bar and an-
other in the center of the head, s, at
the other end of the barrel; the shell
is mounted on a mandrel, the barrel
turned down to fit the hole in the
armature discs, and both sides of the
head faced off smooth. A ^6-in key-
scat 3-16 ins. deep is cut in the bar-
rel so as to come in the center of one
end of the cross-bar, as shown; a ^j-in.
X /^-in. feather, or parallel key, is
laid in the key-seat, and the discs
threaded on the barrel and compressed
against the head by the collar, /i,
and two bolts (not shown) passing through the collar and inside the bar-
rel, and tapping into the head at the other end. This collar, h, is of brass,
3^ ins. in diameter and tapering from 3-16 to 1-16 in. in thickness when
finished. The opening in the center should fit the outline of the cross-bar
en the end cf the barrel at least closely enough to prevent the collar from
shifting under stress of centrifugal force ; the collar must be finished up
smooth on both sides. A disc of insulation should be put on next to the
brass head before the iron discs are put on, and another insulating disc
should go between the last iron disc and the clamping collar, h.
If the slots are cut in the core with a milling machine the discs must
^^^
American Electrician]
FIG. l3. AN ARMATURE DISC.
20 ELECTRICAL DE^IUN'S
all come off the barrel to have the burrs removed, and also be reannealed ;
the key-seat will insure their returning in the original angular position.
It is much better to have discs with the slots punched before the first an-
nealing. The shaft is lofy ms - long over all ; J / 2 in. in diameter in the
largest part, 7-16 in. where the commutator goes and ^ in. in the jour-
nals. A i-i6-in X /^-in. collar, c, is shown back of the armature, the pur-
pose of which is merely to "locate" the armature shell ; it is not absolutely
necessary, however, and may be left off if desired. The easiest way to
provide for it is to make the shaft of Y&-m. stock, leaving the original
metal to form the collar when turning the shaft to proper diameter. The
armature shell may be keyed to the shaft or pinned obliquely through the
thick part of the head; it must be positively secured by some such means.
The commutator shell must be bored to fit the ^-j6-in. portion of the
shaft, and must not exceed 1*4 ins. along the shaft. The lugs where the
wires are attached to the segments may project toward the armature %
American Electrician
FIG. IQ. ARMATURE CORE DRUM AND HEAD.
in. or so. There must be eighteen segments, and a diameter of 2 ins. is
recommended. The quadrant carrying the brush-holders should be fitted
to the inner end of the journal box, and carbon brushes not smaller than
J4 in. X /2 in. (one on each side) on the contact surface should be used.
If the machine be used as a dynamo (it will maintain five or six no-volt
lamps) metal brushes of the same surface should be used to reduce the re-
sistance of the brush contact.
The field coil contains 37 layers of No. 28 double cotton-covered
wire. After the magnet is fitted as described in the beginning of the
article it is taken apart and two circular magnet heads of fibre ^ in.
thick and 3J4 m - outer diameter are put on with a driving fit, care
being taken that the distance along the core from outside to outside
of the heads corresponds with the distance between the pole-pieces when
ONE-FOURTH HORSE-POWER MOTOR 21
the whole is assembled. A groove must be cut on the inner face of
one head from the center to the outer edge in order to lead out the
starting end of the field wire, and this must be covered with two
layers of oil paper to prevent short-circuiting the successive layers of the
coil. The core must be insulated with three layers of shellaced muslin
between the heads and the field wire put on evenly, care being taken not
to "spread" the heads ; if the winding is carefully done the coil will be 216
turns in length. The number of turns in length is not a vital matter, but
the depth must be 37 layers. The ampere turns are the same no matter
what the length of the coil, but it should be as long as possible in order to
reduce the heat loss.
After winding the coil and securing the ends one pole-piece is put on
solid and the other one slipped on until it begins to bind, when the journal
yokes must be inserted between their arms and the bolts put in as far as
possible without jamming. Then by tightening up the journal yoke bolts
and the pole-piece bolt together, being particular never to draw the yoke
bolt hard against the arm, the frame will come together in its original po-
sition. As an additional precaution it may be set on a true plane surface,
and if the base of the loose pole-piece gets out of alignment tap the horn
lightly until the frame is true on the bottom. The magnet frame must
be provided with a non-magnetic base ; hard wood is as good as anything,
the frame being secured by flat-head brass machine screws from below,
two in each casting, countersunk in the wood.
The armature winding is divided into eighteen coils, each having 45
turns of No. 22 double cotton-covered wire, 9 turns wide and 5 turns deep
in the slot. The slots must be insulated with troughs of muslin and
mica, or preferably flexible micanite, 0.03 in. thick. The troughs are
easily made by cutting the material into strips 2 l / 2 ins. long by i*/& ins.
wide, and slitting the ends so as to permit the projecting portion of the
trough to be folded back flat against the core. Before putting in the
troughs a disc of heavy drilling 3^4 ins. in diameter should be secured
to each end of the core by means of varnish, and the outer faces varnished
and allowed to nearly dry. Then put in the troughs and put on two more
muslin discs, varnishing the whole, and bake until thoroughly dry. In-
stead of winding each coil in diametrically opposite slots, take slots lack-
ing one of being precisely opposite.
A good plan is to make a sketch of an armature disc and number the
slots from left to right successively around the periphery. Then wind
the coils as follows, the coil numbers indicating the order in which the
coils are put on, not the order in which they are connected to the com-
mutator.
22 ELECTRICAL DESIGNS
COIL NO. i 2 34 5 6 7 8 9 10 n 12 13 14 15 16 17 is
STARTS IN SLOT NO.--I IO 13 4 7 l6 2 II 1$ 6 14 5 l8 9 3 12 8 17
ENDS IN SLOT NO. 9 l8 3 12 15 6 IO I 5 14 5 13 8 17 II 2 l6 17
Each pair of coils must be covered with muslin where they cross the
heads before the next pair is put on, and before coil No. 8 is wound on top
of coil No. I in slot No. I the bottom coil must be insulated by a strip of
micanite laid in the slot ; this is true of every bottom coil.
After the winding is on, and before connecting up to the commuta-
tor, the band wires should be put on. Use No. 19 B. W. G. soft tinned-
iron wire, known by hardware dealers as "white stove-pipe wire," for the
bands, and put them on under as heavy pressure as possible without en-
dangering the armature shaft. Two bands of eight turns each, y 2 in. from
each end of the core, will suffice. A strip of mica between two strips of
fullerboard must go under each band, and the bands should be soldered
at intervals, not all the way around. Four tin clips located equidistantly,
with a dab of solder at each, will give ample security.
The technical data for this machine are as follows :
TERMINAL E. M. F. , IIO VOLTS.
Armature current, normal 1.9 amps.
" " maximum 2.3 "
" resistance, warm 3.33 ohms
Field current at no volts . . . . .25 amp.
" resistance, warm 440 ohms
C7 a ^?loss in field 27^ watts
C*R loss in armature 12-}- "
Hysteresis loss in armature 20 "
Magnetic flux per square inch :
In field core 90,000 lines
In pole-pieces 39,000 "
In air-gap 25,250 "
In armature teeth 68,000 "
In armature core 56,000 "
Co-efficient of leakage 1.4
Electrical efficiency 84 percent.
Commercial efficiency (friction 10 p. c. estimated). ... 65 f *
Revolutions per minute 2,ooo
If it is desired to build a smooth-core machine the armature core
must be made 3^ ins. in diameter, and two grooves y$ in. wide and 5-16
in. deep must be cut in the face of the core at opposite points for the re-
ception of driving teeth. These are two pieces of fibre % in. thick, $4 in.
wide and 2 ins. long, set on edge in the grooves, and projecting 3-16 in.
above the surface of the core. The core must be thoroughly covered with
ONE-FOURTH HORSE-POWER MOTOR 23
two layers of micanite cloth. The number of coils is the same as before,
"but the coils will be 18 turns wide and 5 deep; and in this case they are
net superposed, the depth of a coil (5 layers) being- the total depth of the
winding. The guiding- diagram, therefore, must divide the periphery of
the armature into 36 spaces instead of 18, because each space now con-
tains one side of only one coil. The smoothest winding will be as
follows :
COIL NO. I 23 4 5 6 7 8 9 10 ii 12 13 14 15 16 17 18
STARTS IN SPACE NO. I ig J 2$ 3! 13 3 21 27 Q 33 15 23 5 II 2Q 35 17
ENDS IN SPACE NO. 18 36 24 6 12 50 2O 2 8 26 14 32 4 22 28 IO l6 34
Care must be taken in connecting- up either of the armature windings
to take the starting ends cf the coils in proper succession to the commu-
tator segments ; the outer end of each coil goes to the segment on the
ri jht of the one to which the starting end is led. The smooth-core arma-
ture is banded just as the slotted one is, except that soft brass wire
must be used instead of tinned iron.
CHAPTER IV.
ONE-FOURTH HORSE-POWER MOTOR WITH RING ARMATURE.
This machine has a field magnet of exactly the same design as the
one last described., the only difference being in the dimensions. The
instructions for fitting up the magnet shown by Figs. 14 and 15, therefore,
apply to this one. The size of the magnet core and yokes shown by Fig.
14 also apply to this magnet. Figs. 21 and 22 give all of the dimensions
for this magnet frame that differ from those of the previous one, excepting
the bore of the armature chamber, which is 5 3-16 ins. instead of 4 3-16
ins. The lugs that support the journal yokes are set one inch wider apart
than in the drum armature motor, and the seats for the ends of the journal
3'okes are bored or cut to 5^ ins. diameter. As in the former case, this
boring must be done before the frame is moved from the position it oc-
cupied during the boring of the armature chamber.
The journal yokes may be made of anything except iron and steel.
The bar is 3-16 in. thick and I in. wide., except near the ends, where it
flares to correspond with the width of the arms. At each end is a right-
angled lug, *X> in. thick after machining; these lugs fit the seats in the
ends of the iron arms, and the yokes should be fitted to. the magnet im-
mediately after finishing the machine work on the latter, and before it is
taken apart to put on the coil. The box portion is \y 2 ins. long over all,
3-16 ins. of its length being on the inside of the yoke and i J^ ins. on the
outside. As shown by the plan view of the yoke, Fig. 22, there are stiff-
ening webs starting flush 'near the ends of the yoke and attaining a width
of 4 in, at the box; these are % in thick. The box is j in. in outer di-
ameter, and bored to 17-32 in. inside; it is bushed to J6 in. diameter.
Most of the dimensions of the yoke and box may be varied to suit individ-
ual ideas, as may also the design of the box. The only essential measure-
ments are the length of the yoke-bar, the length of the box and the bore
ONE-FOURTH HORSE-POWER MOTOR
FIG. 22. PLAN OF JOURNAL YOKE.
American Electrician
FIG. 20. ELEVATION OF FIELD MAGNET WITH JOURNAL YOKE IN POSITION.
ELECTRICAL DESIGNS
J
s
Finishcd
Finished
FIG. 21. PLAN AND FACE VIEWS OF ONE FIELD-MAGNET POLE-riECE.
ONE-FOURTH HORSE-POWER MOTOR
27
cf the journal bulling. The journal yokes are held in place by *4 m -
cap-screws passing through the iron arms and tapping into the lugs of
the yokes. ) . .
The armature core, spider and shaft are shown, partly in cross-sec-
tion, by Figs. 23 and 24. The core is built up of charcoal iron (not steel)
discs 5 ins. outside diameter and 2^ ins. inside, not more t/.an 1-32 in.
thick ; these are assembled on a brass drum i^ ins. long (Fig. 25). which
FIG. 23. AP"VTURE CORE AND SHAFT.
FIG. 24. END OF ARMATURE CORE.
FIG. 25. ARMATURE CORE DRUM,
should be 2^4 m s. outside diameter before finishing, so that it may be
turned down to exactly fit the inner circle of the armature rings; the
wall of the drum is y% in. thick after finishing, and there are four equi-
distant projecting lugs, /,, ^3 in. wide and l / 2 in. long, on each end, by
which the drum is secured to the spider (see Figs. 22 and 23). The rings
forming the core, C (Fig. 22), are compressed and held on the drum, r,
by two brass washers, w, w, 3-16 in. thick and 3^ ins. outer diameter,
28 ELECTRICAL DESIGNS
which screw onto the lugs and ends of the drum. The core when com-
pressed is 1^2 ins. long, and has 20 slots 3-10 in. wide and */ 2 in. deep;
the washers, w w, must be set up as. tight as the threads will stand.
The spider, s (Figs. 23 and 24), is made of brass, and consists of a hub
(% in. in diameter, 2 r / 2 ins. long and $/$ in. bore) and four arms having
T-shaped ends, the wide part or heads of which project beyond the arms
at each end, the length of these heads being 2^8 ins. and their width Y*>
in. The heads of the spider arms are turned off to fit very closely in-
side the drum, r, which is mounted on the spider in such a position as to
bring the spider arms in alignment with the lugs, /, of the drum ; screws
through the spider arms into the lugs hold the drum and spider together.
The shaft, S, is 8}^ ins. long; the portion, a, is 2*/ 2 ins. long and fa
in. in diameter ; b is y 2 in. long and J/2 in. in diameter ; the part passing
through the core is 2J/2 ins. long and ^ in. in diameter ; d is i J4 ins. long
and 3/2 in. in diameter, and e is i*4 ms < l n g an d Y* in. in diameter. The
spider, s, should be secured to the shaft by a key, the key-seat being lo-
cated at the base of one of the arms. The front end of the commutator
must be located not less than 3-16 in. from the shoulder where d and c
join.
The armature is next prepared for winding by removing the drum
and core from the spider and insulating the ends and interior of the core
and the walls of the slots. Cut four rings of heavy drilling of a size to
cover the washers, zv 7C r , and the ends of the drum, r ; varnish two of them
on one side with shellac, and apply them to the ends of the armature body.
While these are hardening cut twenty strips of micanite cloth, 25-1000 in.
thick, \Y* ms - wide and 2 ins. long; in each end of each of these cut two
slits, }4 in. long, parallel with the sides and located 17-32 in. from each
side of the strip. Varnish these on one side, and when nearly dry fold
them into troughs to fit the slots ; fold them so that the varnish will be on
the inside of the trough.
When these are dry varnish the slots and the outsides of the troughs
and put the latter in the slots, bending the ends flat against the core and
securing them there with a little fresh varnish. Then varnish the ends
of the core (two cloth rings being on them), and one side of the two
remaining rings of drilling; put these rings on top of the first ones,
varnish them on the outside and put the core in an oven to bake. The
armature coils consist of No. 22 double cotton-covered wire, wound seven
turns wide and thirteen layers deep. Before winding them four strips of
wood 3 ins. long, y% in. wide and 3/2 in. thick should be screwed to the in-
ner wall of the brass drum, in line with the lugs, /, so as to preserve spaces
for the four arms of the spider. A double thickness of drilling should also
ONE-FOURTH HORSE-POWER MOTOR 29
be applied to the interior cf the drum to insulate the coils from it. The
connections are the simple Gramme ring arrangement. Before connecting
up to the commutator the band wires should be put on. Use No. 19 B.
W. G. soft tinned-iron wire, known by hardware dealers as "white stove-
pipe wire," for the bands., and put them on under as heavy pressure as
possible without endangering the armature shaft. Two bands of eight
turns each, J / 2 in. from each end of the core, will suffice. A strip of mica
between two strips of fullerboard must go under each band, and the bands
should be soldered at intervals, not all the way around. Four tin clips
located equidistantly, with a dab of solder at each, will give ample secur-
ity. ^
The commutator (not shown) must be bored to fit the l /2 in. portion,
d, of the shaft, and must not exceed ij4 in. along the shaft; it must have
a brush tread i in. wide. The lugs where the wires are attached to the
segments may project toward the armature l / in. or so. There must be
20 segments, and a diameter of 2 l / 2 ins. is recommended. The quadrant
carrying the brush-holders should be fitted to the inner end of the journal
box, and carbon brushes not smaller than J4 in. X H in. (one on each
side) on the contact surface should be used. If the machine' be used as a
dynamo (it will maintain five or six no-volt lamps) metal brushes of the
same surface should be used to reduce the resistance of the brush contact.
The field coil contains 37 layers of No. 28 double cotton-covered
wire. After the magnet is fitted as described in the beginning of the arti-
cle it is taken apart and two circular magnet heads of fibre l /% in. thick and
3^4 ins. outer diameter are put on with a driving fit, care being taken that
the distance along the core from outside to outside of the heads cor-
responds with the distance between the pole-pieces (5 ins.) when the
whole is assembled. A groove must be cut on the inner face of one head
from the center to the outer edge, in order to lead out the starting end of
the field wire, and this must be covered with two layers of oil paper toi
prevent short-circuiting the successive layers of the coil. The core must
be insulated with three layers of shellac muslin between the heads a and
the field wire put on evenly, care being taken not to "spread" the heads ;
if the winding is carefully done the coil will be 216 turns in length. The
number of turns in length is not a vital matter, but the depth must be 37
layers. The ampere turns are the same no matter what the length of the
coil, but it should be as long as practicable to reduce the heat loss.
After winding the coil and securing the ends one pole-piece is put on
solid and the other one slipped on until it begins to bind, when the journal
yokes must be inserted between their arms, and the bolts put in as far as
possible without jamming. Then by tightening up the journal-yoke bolts
30 ELECTRICAL DESIGNS
and the pole-piece bolt together, being particular never to draw the yoke
bolt hard against the arm, the frame will come together in its original
position. As an additional precaution it may be set on a true plar.c cur-
face, and if the base of the loose pole-piece gets out of alignment tup the
horn lightly until the frame is true on the bottom. The magnet frame
must be provided with a non-magnetic base ; hardwood is as good as any-
thing, the frame being secured by flat-head brass machine screws from
below, two in each casting, countersunk in the wood.
The technical data for the above machine are as follows :
TERMINAL E. M. F., IIO VOLTS.
Armature current, normal 1.9 amps.
" " maximum.... 2.3 "
" resistance, warm .... 4.15 ohms
Field current at no volts 25 arr.p.
" resistance, warm 440 ohms
C^loss in field y.-j^/z watts
C 2 J? loss in armature 15 "
Hysteresis loss in armature.... 48 "
Magnetic flux per square inch.
In field core 76,000 lines
In pole-pieces , 36,000 "
In air-gap 23,000 "
In armature teeth 65,000 "
In armature core 85,000 "
Co-efficient of leakage 1.4
Electrical efficiency 82 per cent.
Commercial efficiency (windage
and friction losses 10 p. c.,
estimated) 52 "
Revolutions per minute 2,000
Lamps
FIG. 26. DIAGRAM OF STARTING SWITCH.
It is advisable to provide a starting switch similar to the one shown
diagrammatically by Fig. 26, where b b are the brushes ; S the starting
switch lever ; m a magnet, and Szv a double-pole snap switch. The lamps
shown are 5o-volt, 32-candle-power lamps. The handle of the starting-
switch is provided with a spring tending to keep it in the position shown
by the sketch. This starting switch is also suitable for use with the motor
described in Chapter III.
If it is desired to build a smooth-core machine the armature core
must be made 4 l / 2 ins. in diameter, and two grooves y% in. wide and 5-16
in. deep must be cut in the face of the core at opposite points for the re-
ception of driving teeth. These are two pieces of fibre */6 in. thick, l / 2
in. wide and 2 ins. long, set on edge in the grooves and projecting 3-16 in.
ONE-FOURTH HORSE-POWER MOTOR 31
above the surface of the core. The core must be thoroughly covered
with two layers of micanite cloth. The number of coils is the same as
before, but the coils will be twenty turns wide and five deep on the out-
side; on the inside of the ring the coils must lap, making ten layers of
wire. The smooth core is banded just as the slotted one is, except that
soft brass wire must be used instead of tinned iron.
CHAPTER V.
HORSE-POWKR MOTOR, WITH DRUM ARMATURE.
For this size of motor three types of field magnet are described, the
single-coil Jenny, like those previously described, a bipolar one-piece
magnet of the so-called iron-clad type, and a similer form with four poles
FIG. 27. END ELEVATION OF FIELD MAGNET.
(Kapp type). The armature core and shaft are the same in each case,
excepting the number of slots in the four-pole machine. The machine is
a 5/2 horse-power motor to operate on a no-volt constant potential circuit
at a speed of 2,000 revolutions per minute. The single-coil magnet
ONE-HALF HORSE-POWER MOTOR
33
(Figs. 27 and 28) has a round core of commercial wrought iron 2 l / 2 ins. in
diameter and n ins. long over all. The ends are turned tapering, as in-
dicated by the dotted lines, to insure intimate contact with the yokes ; the
taper is from the full diameter to i^ ins., and begins with 2 l / 2 ins. from
each end. The pole-pieces are of cast-iron. The arms which support
the journal yokes are cast solid with the pole-pieces, and their horizontal
L_
FIG. 28. PLAN AND FACE VIEWS OF ONE FIELD-MAGNET POLE-PIECE.
thickness tapers from y 2 in. at the pole-piece to J4 i n - where the yoke is
bolted in place.
In fitting the magnet frame together the best procedure is to bore the
tapered holes in the lower part of each pole-piece and turn the ends at the
magnet core to the same. taper, but just a trifle larger; then dress each
34
ELECTRICAL DESIGNS
tapered end of the core down very gradually with a fine file (the core be-
ing run on a lathe) until the pole-piece can be pushed on by hand far
enough to bring the end of the core within 3-64 in. of the surface of the
cast-iron. The pole-pieces and ends of the core should be punch-marked
so as to insure finally mounting each pole-piece on the end which was
fitted to it. After dressing down the ends of the core as above described
drill and* tap in each end a hole for a 34 -in. machine screw, the purpose
of which will be made apparent by a glance at the right-hand end of
the complete magnet in Fig. 27, where C is a four-armed claw or spider,
with a hole through the center where the arms intersect. The arms are
3-16 in. thick, measured at right angles to the bolt, and taper from 3-16
to y% in. thick, measured parallel with it. One of these claws or spiders
is used at each end of the core, though the drawing shows it at one end
only.
After drawing one pole-piece home solid by means of the spider and
bolt, slip the other on the other end
of the core loosely and clamp the
pole-pieces lightly between two iron
plates with planed surfaces, applied
between the journal arms, so as to
keep the four horns of the pole-pieces
in alignment ; then force the second
pole-piece home and clamp the horns
hard between the iron plates. The
bottom surface of the pole-pieces are
then to be turned up on a shaper or
planer and the iron clamping plates
removed from the horns.
The next operation is boring the
armature chamber and the seats
for the ends of the journal yokes. The bore of the armature chamber is
4 3-16 ins.; the seats for the journal yokes are machined to a 4^-in. cir-
cle for 24 m. from the outer ends. These operations must be completed
before the original position of the frame on the lathe or boring machine
is altered. This completes the machine work on the magnet, with the
exception cf the holes through the ends of the supporting arms and holes
in the bottom surfaces cf the pole-pieces for bolting to the base.
The journal ycke must be made cf brass or some similar composition.
The bar is 3-16 in. thick and I in. wide, except near the ends, where it
flares to correspond with the width cf the supporting arms. At each end
is a right-angle lug, */g in. thick after machining ; these lugs fit the ma-
FIG. 29. JOURNAL YOKE.
ONE-HALF HORSE-POWER MOTOR 35
chined seats in the ends of the iron arms, and the yokes should be fitted
to these arms before the frame is taken apart to put on the magnet coil.
The journal box is ! J / 2 ins. long over all, 3-16 in. of its length projecting
on the inside of the yoke bar, and ij/g ins. on the outside. As shown by
the plan view of the yoke in Fig. 29, there are stiffening webs starting
flush near the ends of the yoke and attaining a height of ft in., where they
join the box ; these ribs are % in. thick. The box is I in. in outside diam-
eter and bored out 9-16 in.; the bore is bushed to ^ m - No particular
form of oiling device is specified, as any amateur of sufficient ability to
build such a motor will be fully competent to decide this detail for him-
self. The journal yokes are held in place by %-in. cap-screws passing
through the ends of the supporting arms and tapping into the lugs on the
yokes.
The field coil contains 35 layers of No. 26 double cotton-covered
magnet wire. After the magnet is fitted as above described it is taken
apart and two circular fibre heads 4^2 ins. in diameter and J^ in. thick
are put on the core with a driving fit, care being taken that the distance
from outside to outside of the heads corresponds with the space between
the perpendicular faces of the pole-pieces when the frame is assembled ;
this measurement should be taken prior to dismantling the frame. A
groove must be cut on the outer face of one head, from the center to the
outer edge, in order to form a channel for leading out the starting ends
of the coil when the frame is re-assembled, at which time two discs of oil
paper with one of mica between them must be threaded on the core out-
side cf this head to insulate the leading-out wire from the pole-piece. Be-
fore winding the coil insulate the core with a strip of muslin just wide
enough to go between the heads, and long enough to wrap around the
core three times ; this should be heavily shellacked before it goes on.
If the coil is carefully wound it will be 210 turns in length along the core;
the umber of turns in this direction is not particularly essential, but as
many should be put on as possible without jamming the insulation, in
order to reduce the heat loss. The depth of the winding must be 35
layers.
After winding the coil and securing the ends, put one pole-piece on
solid and slip the other on loosely. When it begins to bind bolt the jour-
nal yoke to the lugs on the pole-piece first put on, and insert the bolts
through the lugs of the one that is loose. Then tighten up the spider bolt
at the end of the core and force it into place, the bolts through the lugs
serving as guides to keep the pole-piece from twisting on the core.
These bolts should be set up little by little with the spider bolt, so as to
keep the bolt heads within 1-16 in. of the surface of the lugs. As an ad-
36 ELECTRICAL DESIGNS
ditional precaution the frame may be set on a true surface and tried at in-
tervals to see if it gets out of alignment ; if it does, tap the horn of the
loose pole-piece until the bottom surface agrees with the guide. The
magnet frame must be provided with a non-magnetic base, preferably
composition metal, but allowably of wood.
Figs. 30, 31 and 32 show an armature disc, the shaft and armature
core (the latter in cross-section), and the shell and head. The discs are
of charcoal iron, 4 ins. outside diameter, with a I in. hole in the center and
an y% in. key-seat, annealed after punching and key-seating; there are
1 8 slots 3/ in. wide and 3/g In. deep. The shell and one head are cast in
one piece (of brass), and consist of a barrel i in. outside diameter (when
finished), and 2 ins. long, with a head, j, at one end, 3^ ins. in diameter
and tapered in thickness from J4 in. near the center to 1-16 in. at the
periphery ; at the opposite end of the barrel is a cross-bar y& in. thick, cast
FIG. 31. ARMATURE SHAFT AND AXIAL SECTION OF ARMATURE CORE.
with the barrel and of the shape shown, being j in. wide where it joins
the barrel and $4 in. at the center. A ^2 -in. hole is drilled in the center of
this cross-bar and another in the center of the head, s, at the other end of
the barrel ; the shell is mounted on a mandrel, the barrel is turned down
to fit the hole in the armature discs, and both sides of the head are faced
off smooth. A J^ in. key-seat 3-16 in. deep is cut in the barrel, so as to
come in the center of one end of the cross-bar, as shown ; a l /& in. X 1 A
in. feather, or parallel key is laid in the key-seat, and the discs are thread-
ed on the barrel and compressed against the head by the collar, h, drawn
down by two bolts (not shown) passing through the collar and in-
side the barrel, and tapping into the head at the other end. This collar,
^, is of brass, 3^3 ins. in diameter and tapering from 3-16 to 1-16 in. in
thickness when finished. The opening in the center should fit the outline
of the cross-bar on the end of the barrel at least closely enough to pre-
ONE-HALF HORSE-POWER MOTOR
37
vent the collar from shifting under stress of centrifugal force ; the collar
must be finished up smooth on both sides. A disc of insulation should
be put on next to the brass head before the iron discs are put on, and
another insulating disc should go between the last iron disc and the
clamping collar, h.
If the slots are cut in the core with a milling machine the discs must
all come off the barrel to have the burrs removed, and also be re-
annealed ; the key-seat will insure their returning in the original angular
position, It is much better to have discs with the slots punched before
the first annealing. The shaft is 10^4 ins. long over all ; y 2 in. in diam-
eter in the largest part ; 7-16 in. where the commutator goes, and Y% in.
in the journals. A 1-16 in. X Yz in. collar, e, is shown back of the arma-
ture, the purpose of which is merely to "locate" the armature shell ; it is
not absolutely necessary, however, and may be left off if desired. The
FIG. 30. ARMATURE DISC.
FIG. 32. ARMATURE CORE DRUM AND HEAD.
easiest way to provide for it is to make the shaft of ^ in. stock, leaving
the original metal to form the collar when turning the shell to the proper
diameter. The armature shell may be keyed to the shaft or pinned
obliquely through the thick part of the head ; it must be positively secured
by some such means.
The commutator shell must be bored to fit the 7-16 in. portion of the
shaft, and must not exceed ij4 ins. along the shaft. The lugs where the
wires are attached to the segments may project toward the armature 3/
in. or so. There must be 18 segments, and a diameter of 2 ins. is rec-
ommended. The quadrant carrying the brushholders should be fitted
to the inner end of the journal box, and carbon brushes (one on each side)
not smaller than J4 in- X /^ in. on the contact surface should be used.
If the machine be used as a dynamo (it will maintain about ten no-volt
lamps) metal brushes of the same surface should be used to reduce the
.38 ELECTRICAL DESIGNS
resistance of the brush contact. The armature winding is divided into
18 coils, each having 32 turns of No. 20 double cotton-covered wire, eight
turns wide and four turns deep in the slot. The slots must be insulated
with troughs of muslin and mica, or preferably flexible micanite, 0.03 in.
thick. The troughs are easily made by cutting the material into strips
2^/2 ins. long by ij.^s ins. wide, and slitting the ends so as to permit the
projecting portion of the trough to be folded back flat against the core.
Before putting in the troughs a disc of heavy drilling 3^4 ins. in diam-
eter should be secured to each end of the core by means of varnish, and
the outer faces varnished and allowed to nearly dry. Then put in the
troughs and put on two more muslin discs, varnishing the whole, and
bake until thoroughly dry. Instead of winding each coil in diametrically
opposite slots, take slots lacking one of being precisely opposite.
A good plan is to make a sketch of an armature disc and number the
slots from left to right successively around the periphery. Then wind
the coils as follows, the coil numbers indicating the order in which the
coils are put on, not the order in which they are connected to the commu-
tator :
COIL NO. i 2 3 4 5 6 7 8 9 10 n 12 13 14 15 16 17 18
STARTS IN SLOT NO. I IO 13 4 7 l6 2 II 1$ 6 1 4 5 iS Q 3 12 8 17
ENDS IN SLOT NO. 9 18 3 12 15 6 io i 5 14 i 13 8 ij ii 2 16 7
Each pair of coils must be covered with muslin where they cross the
lieads before the next pair is put on, and before coil No. 8 is wound on top
of coil No. i in slot No. i the bottom coil must be insulated by a strip of
inicanite laid in the slot ; this is true of every bottom coil.
After the winding is on, and before connecting up to the commutator,
the band wires should be put on. Use No. 19 B. W. G. soft tinned-iron
wire, known by hardware dealers as "white stove-pipe wire," for the
Lands, and put them on under as heavy pressure as possible without en-
dangering the armature shaft. Two bands of eight turns each, y 2 in. from
each end of core will suffice. A strip of mica between two strips of fuller-
Loarcl must go under each band, and the bands should be soldered at in-
tervals, not all the way around. Four tin clips located equidistantly, with
a dab cf solder at each, will give ample security.
If cast steel be available, one of the iron-clad types of magnet, shown
l>y Figs. 33 and 34, is somewhat preferable because of the small amount of
machine work required. Of these two the four-polar type is considered
preferable by the writer, being much lighter in weight and having an
"open-head" armature winding. Each of the iron-clad magnets is a
single casting; the essential dimensions are shown in the sketches, with
ONE-HALF HORSE-POWER MOTOR
39
t"
n: mfi
|
u
a
B
S
o
it-
i
T
i -
i
40 ELECTRICAL DESIGNS
the exception of the bore of the armature chamber, which is, of course,
the same as for the single-coil magnet 4 3-16 ins. As the two mag-
nets require the same treatment, varying only in dimensions, the fol-
lowing remarks apply to both :
It will be noticed that the feet of the -machine project y% in. below
the body and that there is a transverse rib under the center of similar
depth. These are to give the machine a floor bearing which may be
trued up on a shaper or planer without finishing the whole bottom of the
machine. The first operation on the casting is chipping off the numerous
fins and lumps with which steel castings are invariably afflicted. An em-
ery wheel may be used for this purpose around the outside of the frame,
t
FIG. 34. SIDE ELEVATION OF FOUR-POLE IRON-CLAD FIELD MAGNET.
but in the corners of the coil spaces a cape chisel and lots of muscular
exertion will be required.
Next, the bearing surfaces are trued up, and J^-in. holes drilled
in the feet; then the magnet is mounted for boring out the armature
chamber and the seats for the journal yokes, all of which must be done
with one mounting. This finishes the magnet frame, unless it is de-
sired to put a terminal block on the machine instead of on the base
and do away with the latter. In this event four ^-in. holes are to be
drilled in the top surface of the frame and tapped for machine screws to
hold the block, which may be 2 by 6 ins. and ijX ins. thick. The journal
yoke and journal are the same as shown in Fig. 29, except that for the bi-
polar magnet the yoke is 7^ ins. long instead of 4^/8 ins.
The field coils for the bipolar machine consist of No. 25 double cot-
ton-covered wire wound 45 layers deep, 'and each coil is 80 turns lonj.
ONE-HALF HORSE-POWER MOTOR
The coils are to be wound in fibre bobbins, as shown by Fig. 35. The
heads of the bobbin must be 2% ins. apart, and the body must be
Y in. wider and longer than the magnet core, actual measurement. Be-
fore winding the coil the bobbin must be mounted on a wooden core of
proper size to fit the opening through the center, and having flanges or
heads at each end to "back up" the heads of the bobbins; one of these
heads is put on permanently and the other is secured by two screws
so as to be removable. A spindle of i-in. iron goes through the center of
the wooden core upon which to mount it in the lathe for winding.
When a coil is completed bend the wire back upon itself near the
end, tie a linen thread in the loop formed, and secure the end of the
coil by passing the thread several times around the coil and tying its
ends together. Then varnish the outside heavily and bake the coil at a
low temperature 100 to 125 deg. Fah. until the varnish is hard.
The coils for the four-polar machine are 35 layers deep, and no long,
FIG. 35. MAGNET COIL BOBBIN.
FIG. 36. ARMATURE DIAGRAM.
of No. 25 wire. The heads of the winding bobbin are 3 ins. apart. The
instructions for winding the coils for the bipolar iron-clad machine
apply to these also. In connecting the coils on the machine, however,
there is a difference. On the bipolar machine the final end of one coil
must be connected to the beginning of the other ; on the quadripolar the
reverse is true. Fig. 37 shows diagrammatically the manner of connect-
ing the field coils of the quadripolar machine. It will be noticed that
the exciting current passes around the cores in opposite directions. The
connection for the bipolar machine is exactly the reverse of that shown.
The armature core of the four-pole machine has 19 slots 3-10 in. wide
and y* in. deep, instead of 18 slots 3/sXH- There are 19 coils, each hav-
ing 28 turns of No. 20 wire, 7 turns wide and 4 turns deep. These may be
wound directly on the core, but it will probably be easier for an amateur
to wind them in a little frame, tie them at intervals with thread and put
42 ELECTRICAL DESIGNS
them on the core complete. The winding frame will be exactly like the
one for the field coils except in size. The "channel" formed between
FIG. 37. DIAGRAM OF FOUR-POLE FIELD-COIL CONNECTIONS.
the heads must be 9-32 in. wide and J4 m -deep. The body of the frame,
determines the length and width of the coil, is 2j4 ms - on e way
TABLE OF WINDING AND CONNECTIONS.
NUMBER OF
COIL
IN SLOTS
NOS.
BEGINNING END
GOES TO SEG-
MENT NUMBER
FINAL END
GOES TO SEG-
MENT NO.
I
I and 6
I
II
2
2 7
2
12
3
3 8
3
13
4
4 9
4
14
5
5 10
5
15
6
ii 16
ii
2
7
12 17
12
3
8
13 is
13
4
9
13 19
14
5
10
15 I
15
6
ii
16 2
16
7
12
17 3
17
8
13
18 4
18
9
14
19 5
19
10
15
6 ii
6
16
16
7 12
7
17
17
8 13
8
18
18
9 14
9
19
19
10 15
10
I
and 2}4 ins. the other. The coils are put on and connected up as indi-
cated by the accompanying table.
The numbers of the coils indicate the sequence in which they are put
ONE-HALF HORSE-POWER MOTOR 45
on the core, and this order should be observed in order to secure maxi-
mum symmetry of the wires across the heads of the core. The numbers-
of the slots and segments refer to the diagram shown by Fig. 36. Each
figure applies to the slot and segment between which it is located.
The brush quadrant for this machine is also different from that of
the other two ; instead of bearing upon the commutator at diametrically
opposite points, the brushes must be 90 deg. apart corresponding;
with the relative angular positions cf magnet poles of different signs. In
the bipolar iron-clad the "north" and "south" poles are, of course, oppo-
site each other ; in the four-pole machine the poles directly opposite are
of the same sign if one horizontal pole is "north" the other must also be
"north," and the other two, without coils, will be "south."
CHAPTER VI.
ONE HORSE-POWER BIPOLAR MOTOR, WITH DRUM ARMATURE.
The accompanying drawings and description will enable any one
\vith moderate machine-shop facilities to build a 1 horse-power motor
to work on a no-volt or a 22O-volt continuous-current circuit. Two
types of field magnet are given, the armature and shaft being the same
in both cases.
The armature is 4 ins. in diameter, outside, with twenty-four slots,
each 7-32 in. wide and y% in. deep. Fig. 38 shows the shaft and a cross-
sectional view of the armature core. The discs are compressed by two
cast-iron end plates, which are screwed on the shaft ; these plates are J/2
in. thick at the shaft, and taper to 3-16 in. thick at the outer edge, which
FIG. 38. ARMATURE SHAFT AND AXIAL SECTION OF ARMATURE CORE.
is rounded as shown, to avoid abrading the insulation between the core
and the windings. The full list of armature dimensions is as follows :
Core Shaft
Body. Heads. at 1. m. n. p. q.
Diameter 4 2^ # # i ^ y 2
Axial length 4 j 5 i# 5 3^ 2
The discs should have a shallow key-seat in the edge of the central
hole, and the shaft should be correspondingly key-seated, and a spline,
or perfectly straight key, J^-in. square, should be used to transmit the
movement of the discs to the shaft. If this is done, the slots in the per-
ONE HORSE-POWER BIPOLAR MOTOR
45
iphery of the discs may be milled out; the armature core must be dis-
mantled after the slots are cut, and the burr which is left by the milling
center smoothed off. If the key and key-seats are properly fitted the
discs will go back on the shaft in precisely the position which the slots
were cut, and the sides of the latter will be smooth. If the key is a loose
fit, however, it will be advisable to use a straight edge in one of the slots
to insure perfect accuracy in re-assembling the discs. It is scarcely
necessary to urge a very careful and close fit of the key and its seats. In
assembling the core, one of the cast-iron heads should, of course, be
screwed to place first; then^put on a disc of vulcanized fibre, 1-16 in.
thick, 4 ins. in diameter, and next thread on the iron discs. After the
last iron disc put on another fibre disc and follow with the end plate or
< J *< --K-- -*-
FIGS. 39 AND 40. PLAN VIEW AND END ELEVATION OF IRON-CLAD MAGNET.
head of cast-iron, which will have to be set up with a pin wrench. If the
discs are purchased with the slots already stamped out notches will have
to be cut in the fibre end discs to correspond with the armature slots ; if
the slots are to be milled the fibre discs will, of course, be cut along with
the iron ones. The latter must be not over 1-32 in. thick and preferably
thinner; care should be taken not to get steel discs, but the very best
possible grade of charcoal iron.
Of the two types of field magnets shown, the iron-clad is preferable
from a constructional standpoint, as the only operations are boring out
the armature chamber and the seats for the journal pedestals, and drill-
ing the bolt holes for the latter. Fig. 39 gives a plan view of the iron-
4 6
ELECTRICAL DESIGNS
clad magnet, Fig. 40 an end view and Fig. 41 a side elevation. The
thickness of the magnet core (the portion on which the coils are placed)
parallel with the shaft is 4% ins. except right at the pole face, where it
is rounded down to 4 ins. ; this is necessary in order to reduce the flow
of magnetism from the pole to the cast-iron end plates of the armature,
which produces waste of energy by heating. The complete measure-
ments of the field magnet are as follows :
INCHES.
A Thickness of yoke portion of magnet i%
B Inside length of horizontal part of yoke 8
C Vertical thickness of magnet core 4^
D Distance from core to yoke 2^
E Total outside width of magnet frame II
F Width of journal foot 3
G Radius to which journal scat is bored 4^
H Horizontal thickness of magnet core (see above) 4^"
J Length of journal foot, commutator side 4^
K Width of magnet yoke or frame, axially 8j
L Length of journal foot, pulley side 2^i
The bore of the pole pieces is 4 3-16 in. in diameter, and this figure
must be rigidly observed for best results, as all the calculations are based
upon this length of air-gaps. The above dimensions are intended to
apply to a magnet made of the best ^ -N.
grade of cast-iron; Scotch pig should
be; used if it is obtainable, and if not,
then the very best grade of soft/iron.
The casting should be allowed to re-
main in the mold until it is absolute-
ly cold care being taken not to remove
any of the sand from about the mag-
net proper. The sand can be scraped
away from the extreme end of the
longer of the two pedestal feet, so as
to enable the molder to ascertain
when the casting is cold. It is fre-
quently the case that a casting re-
quires as much as two days to thor-
oughly cool, but it should not be
... -, , r ... 11 FIG. 41. SIDE VIEW IRON-CLAD MAGNET.
disturbed before it is cold.
Fig. 42 gives outside and cross-sectional views of the journal pedes-
tal for this magnet; the two pedestals are alike in every particular, and
when in position on the projecting feet of the field magnet frame their
ONE HORSE-POWER BIPOLAR MOTOR
47
outer edges should be exactly flush with the ends of the feet. The
pedestals are of iron; the base is curved to conform to the arc of the
circle to which the upper surface of the foot is machined, and is j inch
thick. The standard consists of two ribs at right angles with each other,
each 3^ in. thick, with their edges curved as shown. The box is of the
ring-oiling type, with a single ring hung midway of the journal; the
bushing is easily made from thin brass tubing, J^ in. outside diameter,
and with a very thin wall (not over 1-32 in.), babbitted to fit the shaft
and having a slot y% in. wide cut half way through it, midway between its
ends. This bushing is shown in Fig. 43, which represents the bearing
for the other type of magnet, to be presently described. The bushing
is 134 ins. long; the oil ring is made of brass, one inch in diameter, in-
side, \}i ins. diameter outside, and }4 in. wide along the shaft. Refer-
ence to the side views of the journal pedestal will show a slot in the
FIG. 42. DETAILS OF JOURNAL BOX AND PEDESTAL.
FIG. 43. JOURNAL YOKE
FOR FIG. 44.
upper wall of the box portion, through which the oil ring is inserted
before putting in the bushing. A cover should be provided for this slot
to keep out dust, etc. The dimensions of the journal pedestals are as
follows :
INCHES.
g Length of base and journal box 2 }
h Width of base 3
j Outer diameter of reservoir 2
k Axial length of reservoir, outside I %
Internal diameter of reservoir 2^f
Internal length of reservoir I
The bore of the box portion of the pedestal must, of course, be
made to fit snugly the outer diameter of the tubing used for a bushing,
as the wall of the latter is too thin to admit of turning it down to fit a
predetermined bore in the pedestal. After boring the pedestal to fit the
bushing it should be mounted on a mandrel and its base turned to fit the
4 8
ELECTRICAL DESIGNS
circle of the foot on the magnet frame, namely, 9^ ins. in diameter.
Each pedestal should be fastened to the foot with two y\ -in. cap screws.
Fig. 44 gives a side elevation of a much lighter magnet, which may
be used in connection with the armature above described, if the builder
has sufficient skill and facilities to do the machine work accurately.
The magnet core is a round piece of
wrought iron, 3^ ins. in diameter,
with its ends turned down to 3^ ins.
diameter for a distance of 4 ins. from
each end; the total length of the
core is 12^ ins., so that the length
of the untouched portion will be 4^
ins. The pole pieces are of cast-
iron, only the very best possible
grade being suitable. Where the core
enters the cast-iron the latter is 4 ins.
square, with the corners rounded,
and having two ribs or flanges, /, /,
running along one edge; these con-
tinue clear up to the top of the pole
piece, and are i in. thick by 2 ins.
B *
FIG. 44. SINGLE-COIL MAGNET
FIGS. 45 AND 46. END ELEVATION AND PLAN VIEW OF SINGLE-COIL MAGNET.
wide. Fig. 45 shows an end view of the magnet frame, and Fig. 46 a
plan view. The hole occupied by the wrought iron core should be cored
out to 2% ins. diameter when the casting is made, and afterward bored
to a driving fit of the end of the core.
ONE HORSE-POWER BIPOLAR MOTOR 49
The first operation should be turning off the ends of the core ; next,
bore the holes in the pole pieces (or, more strictly speaking, the yokes).
Then drill a ^-in. hole through the yoke just below the lower edge of
the big. hole and at right angles with it, to accommodate the clamping
bolt shown in Fig. 45. Next drive one end of the core into one yoke
and set up the nut on the end of the clamping bolt ; then put on the other
yoke and twist it on the core until the four horns of the pole pieces are
exactly opposite each other, tighten up the second clamping bolt, and
plane off the bottom surfaces of both yokes. To bring the pole horns
into alignment, the simplest method is to cut out two heavy blocks of
hard wood, say 3 ins. thick and 3^ ins. square; bore a J^-in. hole
through the center of each block, run a y 2 -m. bolt, 12 ins. long, through
the two blocks, and apply them to each side of the pole pieces, the bolt
passing through the armature chamber in about the position to be occu-
pied by the shaft. Set up the nut on the bolt until the blocks are hard
against all four pole horns, and then tighten up the clamping bolt in the
foot of the loose yoke.
After planing off the feet of the frame, bore out the armature cham-
ber 4 3-16 ins, in diameter, and the seats for the journal yokes (in oppo-
site faces of the side flanges, /, /), 5 ins. in diameter, and then remove
one magnet yoke and put on the magnet coil. If the coil is separately
wound in a form (which is preferable) only one yoke need come off ; if it
is wound directly upon the core, both yokes must come off, of course,.
The base of the machine must be of wood or brass. Wood is better, as,
aside from its cheapness, it affords convenient space for the terminal
posts and fuse-block of the machine. The base should be 15 ins. X 18
ins., made of two pieces of hard wood each ij ins. thick, glued and
screwed together with the grains at right angles. The longer dimen-
sion of the base is to go parallel with the shaft, and the machine should
be so set as to allow the pulley to overhang the edge of the base-board.
The pulley should be 4 ins. in diameter and 2 ins. wide on the face ;
the latter should be crowned. The pulley should preferably be keyed
to the shaft, with a set-screw in the pulley hub on top of the key. If
only a set-screw be used to hold the pulley on the shaft, a "flat" must
be filed on one side of the shaft under the point of the set-screw.
The journal box and yoke for this magnet is shown by Fig. 43. It
must be made of brass or some other non-magnetic composition. The
design and dimensions of the oil reservoir, journal box and bushing are
exactly the same as those given for the journal box of the iron-clad mag-
net above. All the dimensions are given in the following list, along
with those of the magnet just described.
50 ELECTRICAL DESIGNS
INCHES.
A Distance between yokes 4 l /{
B Thickness of yoke 4
C Radius of outer curve of pole-piece 4^
D Length of pole horn I Y%
E Distance from pole horn to center of magnet core 3^
F Distance from floor line to center of magnet core 3j6
G Width of foot 2%
H Width of slot under core hole in yoke i Yz
] Width of yoke 4
K Width of flange 2
a Diameter of curve of journal yoke ends and seats 5
b Vertical width of journal yoke arms 2 }
c Length of machined portion of yoke arms i^
d Distance from end of yoke arm to inner end of journal box,
pulley end of shaft 2
Distance from end of yoke arm to inner end of journal box,
commutator end of shaft 4
e Length of bushing I ^
g Length of journal box. . ; 2\
h S !ot to let in the oil ring HXI ^
j Outer diameter of oil reservoir 2
Outer length of reservoir, axially l^
The armature 'core and field magnet frames may be wound for any
voltage desired, but the most efficient windings, as the cores now stand,
will be those specified below.
ARMATURE WINDING.
The armature core, after being finally assembled, is to be made
ready for windings by applying the insulation. Cut out four discs of
heavy canvas, 3 ins. in diameter, with a ^-in. hole in the center; varnish
two of them on one side with shellac varnish, and apply them to the end
plates of the armature core, varnished sides in. The edges will turn
over to cover the outer edges of the plates, and will have to be slitted
at intervals of y% in. all around to prevent bunching up. After putting
on these varnish their outer faces, and one face of each of the remaining
canvas discs ; when the varnish begins to thicken put on the two other
discs, one at each end, and apply considerable pressure to them until
they dry. This is best accomplished by boring a hole in a piece of plank,
large enough to pass the shaft, and setting the core on the plank, on end,
next putting a short piece of board (6 or 8 ins. square) with a hole in its
center on the upper end of the armature, and piling any convenient
pieces of heavy scrap on the top board.
Next insulate the slots with troughs of oil paper, 1-64 in. thick, such
ONE HORSE-POWER BIPOLAR MOTOR 51
as is used with the ordinary, office outfit for copying letters ; each trough
should consist of two thicknesses of the oil paper, and the floor of the
trough should be 4% ins. long, so as to project a little beyond the iron
of the core and rest upon the edges of the canvas discs, which were previ-
ously turned over to cover the edges of the end plates.
The coils may then be w r ound directly in the slots, each coil consist-
ing of twenty turns of No. 18 wire, four wide and five deep. Each slot
will contain, when the windings are complete, half of each of two sepa-
rate coils. It w r ill facilitate the winding and insure electrical balance (as
nearly as a core wound armature can be balanced) if the builder will
make a diagram of his armature disc, numbering the slots from I to 24
successively around the circumference, as shown by Fig. 47. Then the
winding will proceed as follows :
First coil starts
2d
3d
4th
5 th
5th
7th
8th
9 th
loth
nth
I2th
--- ' i, U
~J<-n
I4th
I5th
i6th
i7th " "
i 8th " . "
igth " ' "
2oth
2ISt " "
22d
23d
24th
After winding the first two coils, thin strips of varnished muslin
should be laid over them, across each armature head from slot to slot, so
that the next two coils will be insulated from the first pair; each suc-
cessive pair of coils should receive this treatment, and after the slots are
half filled (twelve coils being put on), a strip of oil paper 7-32 in. wide
and 4*4 ins. long must be laid in each slot on top of the coil already in
place before proceeding to put on the coil which next goes in that slot.
No.
No.
in slot i ^ ends
in slot 12
13^ "
" 24
" " J 7 ^ "
4
" " 5 V "
16
;; ;; * ;;
" " 20
8
_
14
< < jgl C
" 2
" 19 ^
6
< ,_ N
18
" " II* "
< " 22
23
10
22
" " II
10
23
IS
7
** " 6 "
19
14 "
3
. 2 t
15
20
9
8
<. 2I
" ./ " 16
5
4
17
24 "
13
it I2
" " i
52 ELECTRICAL DESIGNS
The starting end of each coil should l^e kept leading out straight
from its slot, and the finishing end should be brought across the head
and secured to the starting end by a turn around it. When the winding
is complete, untwist the finishing end of each coil from' its starting end,
and twist it and the starting end of the next coil to the right firmly to-
gether. This will leave twenty-four terminals to lead out to the commu-
tator lugs.
Before connecting the ends to the commutator, the binding wires
should be put on and the winding tested for grounds on the core. The
binding wires are put on in two bands, and consist of small tinned iron
wire ; they should be put on beginning i in. from each end of the core,
and making each band l / 2 in. wide. The binding wire should be wound
FIG. 47. WINDING DIAGRAM.
FIG. 48. CONNECTING DIAGRAM.
on strips of thin varnished muslin laid around the core two layers deep,
and the bands should be soldered at four equidistant points around the
armature surface, not all the way around. The wire used should be not
larger than No. 22 B. W. G. or No. 20 B. S. G.
Unless the machine is likely to be used in very dusty surroundings
it is better not to put any covering over the ends of the armature after the
winding is complete. If the instructions- for insulating each pair of coils
from the succeeding pair have been carefully followed out, any ordinary
collection of dust will not be liable to cause a breakdown in the heads.
The winding just described is intended for a no-volt machine. If it is
desirable to wind the armature for a 22O-volt circuit, use No. 21 double
cotton-covered wire, making each coil five turns wide and six layers
deep.
ONE HORSE-POWER BIPOLAR MOTOR 53
The commutator had better be purchased from any well-known
manufacturer of commutators, as its market price will be less than the
cost of material and labor necessary to make one properly. It must
have twenty-four segments and be not more than 2 ins. long along the
shaft ; the diameter does not matter particularly take one of a stock size
from the maker. In connecting up the coils to the commutator carry
the ends previously twisted together straight out to the commutator
segments. Fig. 48 shows the connections diagrammatically. The slots
are omitted and each coil is represented as having only one turn for the
sake of simplicity. The coils are lettered, to facilitate identification of
opposite ends. The ends leading straight to the commutator are the
starting ends ; those leading around being the final ends. The diagram
is not intended to show the relative radial positions of the coils, and care
must be observed to avoid becoming confused. For example, coil A
may or may not be under coil Z at its starting side ; they are both in the
same slot, but it does not matter which is on top. If coil A was the first
one put on, it will, of course, be in the bottom of both of its slots, and
coil Z will come on top of each side of it. The diagram only shows the
relative angular positions of the coils and the manner of connecting their
ends. The brushes should be of carbon, ^ in. thick, and of a width y$ in.
less than the length of the commutator face, which should be about itf
ins. The brush holders may be copied from any standard type to which
the builder of this motor has access.
The iron-clad magnet requires two magnet coils, one on each pole ;
for iio-volt circuits these coils consist of No. 22 wire wound in 6 1 layers,
each layer being 50 turns long. If both the coils are wound in the same
direction in other words, if they are precisely alike as to the manner of
winding, as they should be the beginning end of one must be connected
to the final end of the other, the two remaining ends being carried to the
terminals of the machine. The best arrangement is to connect the two
ends that are farthest apart, making this connection on the pulley side of
the machine. For 22O-volt circuits the wire must be No. 25 gauge, wound
to a depth of 75 layers, 65 turns to each layer. These coils should be
wound on a block, the cross-section of which is of exactly the same shape
as that of the magnet core at its largest part, but which measures ^ in.
more in each direction. After winding each coil, tie it at each corner
with coarse linen thread (cobbler's thread) and cover it with strips of
muslin wound at right angles to the direction of the wires, and so put on
as to have the edge of each convolution of muslin lay just alongside that
of its neighbor touching it but not lapping it. The muslin must be
one-fourth the width of the inner edge of one side of the coil, so that
54 ELECTRICAL DESIGNS
four turns will cover one side evenly. Put the muslin on in two layers,
the turns of the second layer covering the joint between the turns of the
first layer. Then wind strips over the corners cf the coils, two layers
deep. After the coil is covered with one layer, varnish the muslin cover-
ing heavily with shellac; when this is nearly dry, put on the next layer
and the corner strips, and after varnishing the whole, set the coil aside to
dry. Do not put any varnish on the wire itself. Next cover the iron
cores of the machine with a layer of muslin, this time lapping the edges
cf successive convolutions ; varnish the muslin, and when it and the coils
are thoroughly dry. put the latter on. Unless the pattern for the field
magnet has been very exactly made, and the casting is an unusually
perfect one, it may be necessary to file the corners of the pole pieces
slightly to get the coil between them in putting on the magnet core. In
filing these corners, be careful to round them, leaving no sharp corners
or edges whatever. It is advisable to do this, even if it is not mechan-
ically necessary for the introduction of the coils.
The Jenny type of magnet has only one magnet coil, which consists
of No. 25 wire, wound to a depth of 65 layers, with 148 turns to a layer,
for no-volt service. For 220 volts, use No. 28 wire, wound to a depth
of 81 layers, 186 turns to each layer. All the wire specified herein for
both armature and field winding should be double cotton covered. Cir-
cular magnet heads of vulcanized fibre should be used to protect the ends
of the coil, as the full voltage of the machine exists between these ends ;
these heads should be J^-in. thick and 63^ ins. in diameter, with a hole
to fit the magnet core snugly if the coil is wound directly on the core.
If not, a bobbin should be made, the center consisting of a tube of 1-32
in. fibre, 4^4 ins. long, and of an internal diameter to go easily over the
core ; the heads of the bobbin to be of 3/-in. fibre, as above. If the coil
is wound on the core, the latter must be covered with three layers of
muslin, each layer varnished with shellac. The whole must dry thor-
oughly before the wire is wound on.
The data of the machine are as follows :
IRON-CLAD TYPE.
110 VoltS, 220 VoltS.
Resistance of armature winding I ohm. 3.15 ohms,
Armature capacity, maximum 8.3 amp. 4.7 amp.
" normal....' 7 amp. 3.8 amp.
" loss, &R> normal 49 watts. 45 K watts.
" *' hysteresis, normal 11.7 " 13 "
" " eddy currents 1.3 " 1.5 '*
Total internal armature losses 62 " 60 "
Magnetic flux per sq. in. in armature core. 71,800 72,000
Revolutions per minute, loaded 1,800 2,000
ONE HORSE-POWER BIPOLAR MOTOR 55
Resistance of field winding 220 ohms. 682 ohms.
Current in field winding J^ amp. 0.322 amp.
Heat loss in field winding 55 watts. 70.85 watts.
Density per sq. in. in core 38,000 38,160
Density per sq. in. in gaps 25,500 25,600
Efficiency, approximately 70^ 75#
SINGLE COIL TYPE.
Armature data same as above. no volts. 220 volts.
Resistance of field coil 433 ohms. 1,400 ohms.
Current in field coil % &mp. 0.16 amp.
Heat loss in field coil 27^ watts. 35.2 watts.
Density per sq. in. in core 89,750 90,00
Efficiency, approximately 78$ 79$
An amateur motor builder will be wise not to attempt to make a
starting box for this size of machine ; one can be purchased for a mod-
erate sum from any of half a dozen reputable manufacturers, and, as
either of the motors here described is well worth the outlay necessary to
insure its protection in this particular, the writer advises buying the
starting rheostat.
If, however,, the reader particularly desires to make his own starting
rheostat, the arrangement shown by Fig. 49 will be found easier to con-
struct than anything in the shape of a wire rheostat. In the sketch, L is
the lever, pivoted on a J^-in. metal post, and normally forced down-
wardly by a coil spring of three or four turns (not shown), which is
located under the washer, W. A pin through the post secures the
washer, spring and lever. H is the handle ; a wooden handle, such as
coffee grinders are given, or a large porcelain knob will answer. B is
the contact brush of copper, slitted tangentially to the circles of the con-
tact strips, c, c, c, c, c, c, c\ these circles have 'their common center,
of course, in the center of the post on which L is pivoted. An end view
of the brush, B, is given by E, showing the convex shape given the
under face of the brush to enable it to pass smoothly over the contact
strips. The end of these should be beveled to avoid digging into the
brush.
The connections are shown diagrammatically. R is a bank of five
32 candle-power incandescent lamps, rated at no volts (100 will be
better, and they can probably be readily obtained) ; C is the motor com-
mutator; b, b, the brushes; F, the field winding; S, a double-pole com-
bined switch and fuse block, and M, the service mains. A glance at the
connections will show that the functions of the lever L are to first con-
nect in the field, next the armature in series with one lamp ; at each suc-
cessive step a lamp is added in parallel with the first one until all are in,
ELECTRICAL DESIGNS
and the last position of the lever cuts out the lamps, leaving the armature
in circuit direct.
The lamps should be mounted on the base with the lever and con-
tacts, and it is preferable, though not particularly urgent, that the switch
S be mounted on that base also. The sketch shows the lever in the "off"
position ; the switch, S, should never be closed, except when the lever L
is in this position. If the reader desires to make the arrangement auto-
matic he need only add a retractile spring to pull the lever L to the "off'*
x^\ /^N/"^ X-*
\O0OC
FIG. 49. MOTOR-STARTING RHEOSTAT AND DIAGRAM O? CONNECTIONS.
position ; a bar of iron j in. by J4 in. by 2 ins. on the right hand edge of
the lever, and a small magnet connected in series with the field, F, and
located on the base about Mg, in such a position that it will hold the
lever, by means of the bar of iron, when it is brought to the "on" posi-
tion.
CHAPTER VIT
ONE HORSE-POWER FOUR-POLAR MOTOR WITH DRUM ARMATURE.
For the four-polar one horse-power motor here described only one
type of field magnet is shown, namely, the familiar ring yoke with radial
magnet poles. This type combines more good points than any other,
hence the limitation to the one type. A choice is given, however, be-
tween cast-iron and cast-steel. The armature construction is the same
d-\"
FIG. 50.
DETAILS OF ARMATURE CONSTRUCTION.
for both types of field magnet, the only difference being in the length of
the core along the shaft, and, consequently, the length of the shaft.
Fig. 50 shows the shaft and a cross-sectional view of the armature
core. The discs are mounted on a cast-iron drum, d, which has a flange,
/, and a hub, h*, at one end, and a hub, h, at the other end. Fig. 5 1 gives
a transverse cross-sectional view of the drum, and Fig. 52 is a perspective
58 ELECTRICAL DESIGNS
view, from the flangeless end. The wall of the drum is thickened at two
places, diametrically opposite, as shown in Fig. 51. This is necessary
on one side in order to provide sufficient metal under the key-seat ; it is
necessary on the opposite side to obtain a mechanical balance.
The discs are held endwise by a clamping ring, r, which may be
either screwed onto the end of the drum, d, or held on by four flat-headed
screws with large heads. The discs are held from turning by a key. At
each end of the magnetic core a disc of fibre, indicated by heavy black
lines, should be placed. These discs must be exactly like the iron discs,
except that they are 1-16 inch thick.
The iron core discs are 5^ ins. in diameter and 1-40 in. thick, with
32 slots, each y m - wide and 9-16 in. deep. The slots have parallel
sides. The discs must be of the best charcoal iron ; the hole in the center
is 3 ins. in diameter, key-seated. The flange, /, and the clamping-ring,
r, must have their outer edges rounded off to avoid cutting the insulation
of the winding. The dimensions of the core drum are as below :
INCHES.
Length of drum, d 4^
Inner diameter of d 2j
Outer diameter of d 3
Diameter of flange, f, and ring, r 4^
Thickness of flange, f , and ring, r. : }
Thickness of d at thickest point l />.
Diameter of hubs, h and h 2 ij
Bore of hubs, h and h 3 i>
Length of hub, h ij^f
Length of hub, h 3 I
Length, a, of disc portion of core 4
The shaft measurements are as follows :
At v x y z
Diameter, inches %i "fa i/i Y%
Length, inches 3 2 l /2 7>^ 6
The shoulders where v and x meet and where y and z meet should
be slightly rounded off at the corner and filleted in the angle. A key
should be used to fasten each hub to the shaft, but the machine will doubt-
less give satisfaction with only one key, that one being in the hub, h, at
the pulley end.
Figs. 53 to 56 inclusive show end and side views and cross-sections
of a journal pedestal and box. The two bearings are alike in every par-
ticular, and are made of cast-iron. The base or foot is tooled to conform
to the circle to which the pedestal seat, on the magnet frame, is machined,
and is j/2 in. thick. The standard or pedestal consists of two ribs at
right angles to each other, y 2 in. thick and having curved edges, as
ONE HORSE-POWER FOUR-POLAR MOTOR
59
shown. The box is of the ring-oiling type, with a single ring hung about
midway of the journal ; the bushing is easily made from thin brass tubing,
J/s in. outside diameter, and with a very thin wall (not over i~32d in.),
babbitted to fit the shaft and having a slot y% in. wide cut half way
through it, nearly midway between its ends ; accurately, slot must be J4
in. nearer one end than the other. The bushing is 2^/4 ins. long; the oil
ring is made of brass, iJ/2 ins. in diameter inside, I 11-16 ins. diameter
outside, and $4 in. wide along the shaft. Reference to the side views
of the journal pedestal will show a slot in the upper wall of the box por-
tion, through which the oil ring is inserted before putting in the bush-
ing. A cover should be provided for this slot to keep out dust, etc.
The dimensions of the journal pedestals are as follows :
INCHES.
B Radius of arc, pedestal seat $%
g Length of circular oil reservoir 2
i Length of journal box 3
Bore of journal box ^
j Diameter of oil reservoir 2^
Internal diameter of oil reservoir 2
J Width of pedestal foot 3
k Length of pedestal foot 3
The bore of the box portion of the pedestal must, of course, be made
to fit snugly the outer diameter of the tubing used for a bushing, as the
r- J
FIGS. 53, 54, 55, 56. DETAILS OF JOURNAL BOX AND PEDESTAL.
wall of the latter is too thin to admit of turning it down to fit a prede-
termined bore in the pedestal. After boring the pedestal to fit the bush-
ing it should be mounted on a mandrel and its base turned to the radius
B y of 55/2 ins., which is the same as the radius of the circle of the foot on
6o
ELECTRICAL DESIGNS
WG, 50. PLAN VIEW OF FIELD-MAGNET; EITHER CAST-IRON OR CAST-STEEL.
FIG. 6l.
SHAPE OF MAGNET CORE.
FIG. 60. SIDE ELEVATION OF FIELD MAGNET.
LIBR
OFTH1
62 ELECTRICAL DESIGNS
the magnet frame. Each pedestal should be fastened to the foot with
four %-in. cap screws.
Of the two field mag-nets shown, the cast-iron one will be found
much easier to make because there is less tooling to be done and iron
castings are smoother than steel, requiring little or no finishing else-
where than the pedestal seats and pole faces. Fig 57 shows the cast-iron
magnet frame and Fig. 58 the cast-steel frame. Fig. 59 is a plan view
of either frame and Fig. 60 is an edge view.
The measurements for the cast-iron magnet are as follows:
INCHES.
A Bore of armature chamber 5 ^
B Radius to which pedestal seat is bored 5^2
C Outer diameter of yoke ring 13
D Distance between parallel inner faces of yoke ring 10
E Width of plane surface behind coil 5
F Width of magnet coil 2}/ z
F 2 Breadth of magnet core 4
G Distance from core to angle of yoke i^
H Width of frame foot 2
Hg Length of double foot 4^
J Width of pedestal lug and seat 3
K Length of pedestal lug commutator side 5^
L Length of pedestal lug pulley side 3^
k Length of pedestal seat 3
M Axial width of magnet yoke 7
Fig. 6 1 shows the cross-section of a magnet core, from which it will
be seen that the corners of the core are rounded off. The radius of the
curve here is % in. The only machining that should be required for this
frame is boring the armature chamber and pedestal seats and drilling 12
bolt-holes. The frame should be clamped to a lathe carriage with its
center true with the lathe centers, and the boring done at one setting by
means of a boring bar and tool. Both pedestal seats should be cut be-
fore the frame is moved from its original position.
The magnet must be made of the very best grade of iron obtainable ;
use Scotch pig if possible. It should be allowed to remain in the sand
until it is cold, care being taken not to remove any of the sand around
the magnet portion until the casting is ready to come out. The longer
of the two lugs might advantageously be placed uppermost in putting the
pattern in the sand, and after the casting has been cooling for 24 hours
the sand may be scraped away from the end of this lug so that its tem-
perature may be noted .
The steel field magnet is much preferable if the reader has the skill
and facilities to make it properly. The difference from the cast-iron
magnet consists in making the magnet cores round instead of oblong, anci
O.V HORSE-POWER POUR-POLAR MOTOR
putting on pole-shoes. The length of the machine is thereby reduced
one inch, but all the transverse measurements remain unchanged. The
magnet ends arc machined, exactly as in the case of the cast-iron frame,
but the bore, A* t is greater, namely, 6*/ ins.
The pole-pieces arc made in one piece, called a polar-bushing, like
Fig. 62, and this had better be done before the magnet is bored out. This
bushing is a simple cylinder of cast-iron with four openings in its wall,
equidistant from each other. Fig. 63 shows the exact shape of each of
these openings. The measurements of the bushing are these :
INCHES.
A Bore of bushing, finished 5^$
A 2 Diameter of bushing, finished 6y&
a 2 Length of bushing, finished 3 ^f
b Length of openings in wall 3
c Radius of curve, side of opening 4
e Maximum width of opening itf
The casting for this bushing should be about 3^ ins. long, 6^ ins.
in diameter and 5^ ins. bore in the
rough. After it has been turned
down to the finished diameter, mount
the magnet frame and bore out its
K Q
American Electrician
FIGS. 62 AND 63. MAGNET POLE BUSHING. _ , FIG. 64. FIELD COIL CONNECTIONS.
polar circle to such a size that the bushing is a snug fit not quite a driv-
ing fit, but tight enough to prevent turning by hand. Then insert the
bushing so that the openings in its sides come half way between the
magnet cores, and scribe the outlines of two opposite cores on its surface.
Remove the bushing and set a steel pin at each extremity of each
6 4
ELECTRICAL DESIGNS
ellipse scribed on the surface. Then put the bushing back and bore it
out for the armature chamber. The pins will take up against the edges
of the magnet cores and prevent the bushing from turning. After bor-
ing it out, turn on the ends of the bushing so as to leave the connecting
webs from pole piece to pole-piece % in. thick.
The objection to this magnet is the difficulty of fitting the bushing to
the magnet with sufficient accuracy to make good magnetic contact and
still leave it loose enough to permit removal without breaking the thin
connecting webs. This could be obviated by bolting the pole-pieces to
the ends of the magnet cores by means of long, slender machine screws,
put in from the outside of the yoke through holes in the centers of the
magnet cores. Then the connecting webs could be sawed out entirely,
leaving each pole-shoe independent of the others. This construction is
also magnetically preferable, and if the builder has means for drilling a
FIG. 65. CROSS SECTION OF ARMATURE SLOT, SHOWING ARRANGEMENT OF WIRES.
J^-in. hole from the outside of the ring to the end of the magnet core (a
distance of 3^ ins.), the pole-shoes should be held on this way.
With the steel magnet the following measurements must be substi-
tuted for those previously given :
INCHES.
F Diameter of magnet core 2^6
M Width of magnet yoke 6
a Length of disc part of core 3
Length of drum, d 3#
Outer diameter of drum, d 2^
Inner diameter of drum, d 2j^
The four field coils for the cast-iron magnet frame described in the
preceding chapter are of No. 21 single-cotton-covered magnet wire. The
depth of the winding must be ij4 ms -> as nearly as possible, and the
length along the core should be 2 ins. Careful and close winding should
give 40 layers of wire, with 58 turns to a layer. Whatever number of
turns the reader may obtain, that number must be precisely the same in
?ill four coils. In order to attain uniformity the coils should be wound
upon a frame and the turns religiously counted.
It will be found advantageous to tie a knot in the starting end of
ONE HORSE-POll'ER FOUR-POLAR MOTOR
each coil before taping it so that it may be identified afterward. The
coils must be connected up as shown by the diagram, Fig. 64, so that the
starting end of one connects to the finishing end of its neighbor. This
presupposes that all four are wound in the same direction, as they
should be.
The coils for the cast-steel magnet are of No. 24 single-cotton-cov-
ered wire, i l /i ins. deep and i% ins. long. Good winding will enable
the reader to put on 50 layers of wire and 75 turns to a layer. As in the
previous case, however, the depth in inches is the essential point, though
,it is advantageous to get as many layers in that depth as possible. The
coils are, of course, wound, insulat-
j ed and connected up exactly like the
I oblong coils of the cast-iron frame.
t
WIDH
A.
01, m
32
FIG. 66. DIMENSIONS OF ARMATURE COIL,
FIG. 67. WINDING DIAGRAM.
The armature core for either of the magnet frames will contain 32
coils ; each coil consists of No. 21 double-cotton-covered wire, wound five
turns wide by four layers deep. Each slot contains one side of each cf
two coils, so that the cross-section of the winding in a slot will be as in
Fig. 65, except that the wire will lie closer together than the sketch indi-
cates. All armature coils should be wound on a forming bobbin so that
they will all be exactly alike. Fig. 66 shows what the essential dimen-
sions should be. The width of the hollow of the coil is the same for
both armature cores. As the armature core to be used with the steel
magnet is an inch shorter than the other one, the coils for this core must
be an inch shorter ; hence the two dimensions for coil lengths.
Fig. 67 is a winding diagram and shows the first four coils in posi-
tion. The coils are indicated by a single line across the head and doit
66
ELECTRICAL DESIGNS
in the slots for simplicity. The builder should note that the left-hand side
of each coil is in the bottom of the slot and the right-hand side is on top ;
this should be true of every coil. The starting ends should be
knotted for identification, and all the knotted ends should occupy the
same relative position on the core. For smoothness of finished heads
the coils should be put on the core in the following order:
Coils i, 2, 3, 4 in Slots I, 9, 17, 25.
Coils 5, 6, 7, 8 in Slots 2, 10, 18, 26.
Coils 9, 10, ii, 12 in Slots 3, n, 19, 27.
L Coils 13, 14, 15, 16 in Slots 4, 12, 20, 28.
Coils 17, 18, 19, 20 in Slots 5, 13, 21, 29.
Coils 21, 22, 23, 24 in Slots 6, 14, 22, 30.
Coils 25, 26, 27, 28 in Slots 7, 15, 23, 31.
Coils 29, 30, 31, 32 in Slots 8, 16, 24, 32.
If put in properly, the coils will give a regular sequence of knotted
ends and straight ends, one each projecting from each slot. The con-
TERMINAL
FIG. 68. COMMUTATOR DIAGRAM.
FIG. 69. CONNECTIONS BETWEEN
BRUSHES.
nections to the commutator are then simple. Carry all the knotted ends
straight out to the commutator, and the straight ends one segment less
than a quarter circle backwards from their corresponding knotted ends.
Thus, the knotted end from slot No. i goes to commutator segment No.
i (see Fig 68), and so on, all the way around. Then the straight ends
go to the commutator as follows :
From Slot No. i to Segment No. 26.
From Slot No. 2 to Segment No. 27.
. From Slot No. 3 to Segment No. 28.
ONE HORSE-POWER FOUR-POLAR MOTOR 67
From Slot No. 4 to Segment No. 29.
From Slot No. 5 to Segment No. 30.
From Slot No. 6 to Segment No. 31.
From Slot No. 7 to Segment No. 32.
From Slot No. 8 to Segment No. I.
From Slot No. 9 to Segment No. 2.
From Slot No. 10 to Segment No. 3.
From Slot No. u to Segment No. 4.
From Slot No. 12 to Segment No. 5.
From Slot No. 13 to Segment No. 6.
From Slot No. 14 to Segment No. 7.
From Slot No. 15 to Segment No. 8.
From Slot No. 16 to Segment No. 9.
From Slot No. 17 to Segment No. 10.
From Slot No. 18 to Segment No. n.
From Slot No. 19 to Segment No. 12.
From Slot No. 20 to Segment No. 13.
From Slot No. 21 to Segment No. 14.
From Slot No. 22 to Segment No. 15.
From Slot No. 23 to Segment No. 16.
From Slot No. 24 to Segment No. 17.
From Slot No. 25 to Segment No. 18.
From Slot No. 26 to Segment No. 19.
Frcm Slot No. 27 to Segment No. 20.
From Slot No. 28 to Segment No. 21.
From Slot No. 29 to Segment No. 22.
From Slot No. 30 to Segment No. 23.
From Slot No. 31 to Segment No. 24.
From Slot No. 32 to Segment No. 25.
The commutator must have 32 segments, as indicated by Fig. 68,
and should be purchased already built for assured satisfaction. The
brush surface of the commutator must be i% ins. long", at least, so that
carbon brushes i in. wide and J4 in. thick can be used. The diameter
of the barrel of the commutator should be not less than 3, and preferably
4 ins. The brush holders and yoke may be copied advantageously from
any of the standard machines now on the market. Four brushes must
be used, and the two diametrically opposite are connected together, as
shown by Fig. 69.
The windings just described are for machines to work on a 110-115-
volt circuit. If windings for 220-230 volts are desired the armature coils
should be of No. 25 wire, each coil five layers deep and eight turns wide,
making ten layers of wire per slot. The field coils of the cast-iron mag-
net must be of No. 24 s.c.c. wire, wound to the dimensions specified
above, namely, \Y\ ins. deep and 2 ins. long. The coils for the cast-steel
magnet will be of No. 27 wire wound to a depth of i 1 /^ ins. and a length
18 ins.
68
ELECTRICAL DESIGNS
The principal magnetic and electrical data of the two machines are
below :
II5-VOLT MOTOR.
Cast-iron.
I ohm
9 amp.
81 watts
4. 1 A watts
Resistance armature winding
Normal armature currents
C 6 R loss armature
Hysteresis and eddy currents ,
Approximate speed 1600
Resistance field winding 183 ohms
Normal field current 628 amp.
C 8 R field loss 72.22 watts
Flux per pole
Density in field cores 48,000
Density in air gap 27,500
Efficiency, assuming 10 per )
cent friction and windage \
75 percent
Cast-steel.
0.9
9 amp.
73 watts
3 watts
1600
424 ohms
.27 amp.
31 watts
400,000 lines
93,000
37,500
80 per cent
As in the preceding case, it is by far preferable to buy a starting box
from one of the standard rheostat manufacturers. If the reader insists
upon having a home-made one, however, the arrangement shown by Fig.
49 and described on pages 55 and 56 will answer.
CHAPTER VIII.
TWO HORSE-POWER FOUR-POLAR MOTOR WITH TWO-PATH DRUM
ARMATURE.
For this motor, as in the preceding design, only one type of field
magnet is shown, namely, the familiar ring yoke with radial magnet
poles ; a choice is given between cast-iron and cast-steel field magnets.
The armature construction is identical for both types of magnet, there
being a difference only in the length of the armature and shaft.
Fig. 70 shows the shaft and a cross-sectional view of the armature
core. The discs are mounted on a cast-iron drum, d, which has a flange,
/, and a hub, Ii2 } at one end, and a hub, h, at the other end. Fig. 71 gives
FIG. 70. ARMATURE SHAFT AND CROSS-SECTION OF ARMATURE CORE.
a transverse cross-sectional view of the drum, and Pig. 72 is a perspective
view, from the flangeless end. The wall of the drum is thickened at two
places, diametrically opposite, as shown in Fig. 71. This is necessary on
one side in order to provide sufficient metal under the key-seat; it is
necessary on the opposite side to obtain a mechanical balance.
The discs are held endwise by a clamping ring, r, which may be
either screwed onto the end of the drum, d, or held on by four flat-headed
screws with large heads. The discs are held from turning by a key. At
each end of the magnetic core a disc of fibre, indicated by heavy black
lines, should be placed. These discs must be exactly like the iron discs,
L except that they are 1-16 in. thick.
ELECTRICAL DESIGNS
The iron core discs are 6]4 ins. outside diameter and 1-40 in. thick,
with 43 slots, each ]/[ in. wide and 9-16 in. deep. The slots have parallel
sides. The discs must be of the best charcoal iron ; the hole in the center
is 35^ ins. in diameter, key-seated. The flange, /, and the clamping-
ring, r, must have their outer edges rounded off to avoid cutting the in-
sulation of the winding. The dimensions of the core drum are as below :
Length of drum, d 5^
Inner diameter of d 3
Outer diameter of d ^H
Diameter of flange, f, and ring, r., 5^4
Thickness of flange, f, and ring, r. $
Thickness of d at thickest point. . . Y 2
Diameter of hubs, h and h 2 2
Bore of hubs, h and h 2 itf
Length of hub, h 1*4
Length of hub, h 8 l*^
Length of a, of disc portion of core 4^
FIGS. 71 AND 72.
CORE DRUM AND HEAR.
The shaft measurements are as follows :
At v x y z
Diameter, inches %{ I ij %(
Length, inches 3^$ 3 9 7^
The shoulders where v and x meet and where y and 2 meet should
be slightly rounded off at the corner and filleted in the angle. A key
it- . *
r d
FIGS. 73, 74, 75, 76. DETAILS F JOURNAL BOX. AND PEDESTAL.
should be used to fasten each hub to the shaft, but the machine will doubt-
less give satisfaction with only one key, that one being in the hub, A, at
the pulley end. The hub, h, must be exactly % ' m - fr m tne shoulder
on the shaft.
Figs. 73 to 76j inclusive, show end and side views and cross-sections
TWO HORSE-POWER FOUR-POLAR MOTOR JI
of a journal pedestal and box. The two bearings are alike in every par-
ticular, and are made of cast-iron. The base or foot is tooled to con-
form to the circle to which the pedestal seat, on the magnet frame, is
machined, and is % in. thick. The standard or pedestal consists of two
ribs at right angles to each other, % in. thick and having curved edges,
as shown. The box is of the ring-oiling type, with a single ring hung
cbout midway of the journal ; the bushing is easily made from thin brass
tubing, I in. outside diameter, and with a very thin wall (not over i~32d
in.), babbitted to fit the shaft and having a slot 7-16 in. wide cut half way
through it, nearly midway between its ends ; accurately, the slot must be
J4 in. nearer one end than the other. The bushing is 3^ ins. long; the
oil ring is made of brass, 2 ins. in diameter inside, 2*4 ins. diameter out-
side, and Y in. wide along the shaft. Reference to the side views of the
journal pedestal will show a slot in the upper wall of the box portion,
through which the oil ring is inserted before putting in the bushing. A
cover should be provided for this slot to keep out dust, etc. The dimen-
sions of the journal pedestals are as follows :
INCHES.
B Radius of arc, pedestal seat 6
g Length of circular oil reservoir 2
i Length of journal box 3f
Bore of journal box I
j Diameter of oil reservoir 2j
Internal diameter of oil reservoir z%
J Width of pedestal foot 3^
k Length of pedestal foot 3)6
The bore cf the box portion of the pedestal must, of course, be made
to fit snugly the outer diameter of the tubing used for a bushing, as the
wall of the latter is too thin to admit of turning it down to fit a prede-
termined bore in the pedestal. After boring the pedestal to fit the bush-
ing it should be mounted on a mandrel and its base turned to the radius
B, of 6 ins., which is the same as the radius of the circle of the foot on the
magnet frame. Each pedestal should be fastened to the foot with four
5-16 in. cap screws.
Of the two field magnets shown, the cast-iron machine *^vill be found
easier to make because there is less tooling to be done and iron castings
are smoother than steel, requiring little or no finishing elsewhere than
the pedestal seats and pole faces. Fig. 77 shows the cast-iron magnet
frame and Fig. 78 the cast-steel frame. Fig. 79 is a plan view of either
frame and Fig. 80 is an edge view.
ELECTRICAL DESIGNS
TWO HORSE-POWER FOUR-POLAR MOTOR
73
FIG. 79. PLAN VIEW OF FIELD MAGNET; EITHER CAST-IRON OR CAST-STEEL,
Fa
FIG. 8 I.
SHAPE OF MAGNET POLE.
d.
f ivr
- H;-
FIG. 8O. SIDE ELEVATION OF FIELD MAGNET.
74 ELECTRICAL DESIGNS
The measurements for the cast-iron magnet are as follows :
INCHES.
A Bore of armature chamber 7
B Radius to which pedestal seat is bored 6)
C Outer diameter of yoke ring 15^
D Distance between parallel inner faces of yoke ring 12
E Width of plane surface behind coil 6
T Width of magnet core 3
F 8 Breadth of magnet core 4>
G Distance from core to angle of yoke i^
H Width of frame foot 2)4
H 8 Length of double foot 5
J Width of pedestal lug and scat 3^
K Length of pedestal lug; commutator side 6j
L Length of pedestal lug; pulley side 3^
K Length of pedestal scat 3^
M Axial width of magnet yoke 8/75-
Fig. 8 1 shows the cross-section of a magnet core, from which it will
be seen that the corners of the core are rounded off. The radius of the
curve here is 0.3 in. The only machining that should be required for
this frame is boring the armature chamber and pedestal seats and drill-
ing 12 bolt-holes. The frame should be clamped to a lathe carriage with
its center true with the lathe centers, and the boring done at one setting
by means of a boring bar and tool. Both pedestal seats should be cut
before the frame is moved from its original position.
The magnet must be made of the very best grade of iron obtainable ;
use Scotch pig if possible. It should be allowed to remain in the sand
until it is cold, care being taken not to remove any of the sand around
the magnet portion until the casting is ready to come out. The longer
cf the two lugs might advantageously be placed uppermost in putting the
pattern in the sand, and after the casting has been cooling for 24 hours
the sand may be scraped away from the end of this lug so that its tem-
perature may be noted.
The cast-steel field magnet is much preferable if the reader has the
'".kill and facilities to make it properly. The difference from the cast-
_ron magnet consists in making the magnet cores round instead of ob-
long, and putting on pole-shoes. The length of the machine is thereby
reduced ij4 ms -> but all the transverse measurements remain un-
changed. The magnet ends are machined exactly as in the case of the
cast-iron frame, but the bore, A 2 , is greater, namely, /J/2 ins.
The pole-pieces are made in one piece, called a polar-bushing, like
Fig. 02, and this had better be done before the magnet is bored out.
!This bushing is a simple cylinder of cast-iron with four openings in its
TWO HORSE-POWER FOUR-POLAR MOTOR
75
wall, equidistant from each other. Fig. 83 shows the exact shape of each
of these openings. The measurements of the bushing are these :
INCHES.
A 8 Diameter of bushing, finished 7/^
A Bore of bushing, finished 7
a s Length of bushing, finished 3^
b Length of openings in walls , 3/2
c Radius of curve, side of opening 4j
e Maximum width of opening 2>
The casting for this bushing should be about 4 ins. long, 7^ ins. in
diameter and 6^J ins. bore, in ihe rough. After it has been turned
down to the finished diameter, mount the magnet frame and bore out its
polar circle to such a size that the bushing is a snug fit not quite a driv-
K e -w
FIGS. 82 AND 83. MAGNET POLE BUSHING. FIG. 84.FIELD COIL CONNECTIONS.
ing fit, but tight enough to prevent turning by hand. Then insert the
bushing so that the openings in its sides come half way between the mag-
net cores, and scribe the outlines of two opposite cores on its surface.
Remove the bushing and set a steel pin at each extremity of each
ellipse scribed on the surface. Then put the bushing back and bore it
out for the armature chamber. The pins will take up against the edges
of the magnet cores and prevent the bushing from turning. After bor-
ing it out, turn off the ends of the bushing so as to leave the connecting
webs from pole-piece to pole-piece ]/% in. thick, measured axially.
The objection to this magnet is the difficulty of fitting the bushing
to the magnet with sufficient accuracy to make good magnetic contact
76 ELECTRICAL DESIGNS
and still leave it loose enough to permit removal without breaking the
thin connecting webs. This could be obviated by bolting the pole-pieces
to the ends of the magnet cores by means of long, slender machine
screws, put in from the outside of the yoke through holes in the centers
of the magnet cores. Then the connecting webs could be sawed out
entirely, leaving each pole-shoe independent of the others. This con-
struction is also magnetically preferable, and if the builder has means
for drilling V^-m. holes through the magnet cores from the outside of the
yoke ring to the inside of the bushing (a distance of 4^ ins.), the pole-
shoes should be held on this way.
With the steel magnet the following measurements must be substi-
tuted for those previously given :
INCHES.
F Diameter of magnet core 2 7 /&
M Width of magnet yoke 7 T s s
a Length of disc part of core 3>
Length of drum, d 3j|
Length of y, on shaft 7^
This is to say, the machine must be exactly 1^4 ins. shorter, axially.
The four field coils for the cast-iron magnet frame are of No. 22
single-cotton-coverecl magnet wire. The depth of the winding must be
IjX ins., as nearly, as possible, and the length along the core should be
2*4 ins. Careful and close winding should give 50 layers of wire, with
70 turns to a layer. Whatever number of turns the reader may obtain,
that number must be precisely the same in all four coils. In order to
attain uniformity the coils should be wound upon a frame and the turns
religiously counted.
It will be found advantageous to tie a knot in the starting end of
each coil before taping it so that it may be identified afterward. The
coils must be connected up as shown by the diagram, Fig. 84, so that
the starting end of one connects to the finishing end of its neighbor.
This presuppose! that all four are wound in the same direction, as they
should be.
The coils for the cast-steel magnet are of No. 23 single-cotton-
covefed wire, 1^2 ins. deep and 2 ins. long. Good winding will enable
the reader to put on 56 layers of wire and 70 turns to a layer. As in the
previous case, however, the depth in inches is the essential point, though
it is advantageous to get as many layers in that depth as possible. The
coils are, of course, wound, insulated and connected up exactly like the
oblong coils of the cast-iron frame.
The armature core for either of the magnet frames will contain 43
r:ro HORSE-POWER FOUR-POLAR MOTOR
77'
coils; each coil consists of No. 16 double-cotton-covered wire, wound
three turns wide by three layers deep. Each slot contains one side of
each of two coils, so that the cross-section of the winding in a slot will be
as in Fig. 85.
All armature coils should be wound on a forming bobbin so that they
FIG. 85. CROSS-SECTION OF A SLOT.
FIG. 88. CONNECTING DIAGRAM.
will all be exactly alike. Fig. 86 shows what the essential dimensions
should be. The width of the hollow of the coil is the same for both
armature cores. As the armature core to be used with the steel magnet
is ij4 ins. shorter than the other one, the coils for this core must be cor-
respondingly' shorter ; hence the two dimensions for coil lengths*
Fig. 87 is a winding diagram and shows four coils in position. The
coils are indicated by single lines across the head and dots in the slots for
_ A NGO'LAR SPREAD
FIG. 86. DIMENSIONS OF ARMATURE COIL.
FIG. 87. WINDING DIAGRAM.
simplicity. The builder should note that the left-hand side of each coil
is in the bottom of the slot and the right-hand side is on top ; this should
be true of every coil, but it is not imperative. The machine will work
just as well if half of the coils are put on with both sides bottom and the
other half on top of them, but the job will not be so neat on the armature
78 ELECTRICAL DESIGNS
heads. The spacing or pitch of the coils must be exactly as indicated
IO slots in between the two sides of each coil.
The starting ends should be knotted for identification, and all the
knotted ends should occupy the same relative position on the core. If
put in properly, the coils will give a regular sequence of knotted ends and
straight ends, one each projecting from each slot. The connections to
the commutator are then simple. Carry all the knotted ends straight out
to the commutator, and each straight end to a segment 22 bars from the
one to which the knotted end is connected. Fig. 88 represents one coil
connected, and shows that there are 21 segments between the two to
which the coil ends go, reckoning around that side of the commutator
nearest the coil itself. This spacing must be observed throughout.
The commutator must have 43 segments and should be purchased
already built for assured satisfaction. The brush surface of the commu-
tator must be i^ ins. long, at least, so that carbon brushes i }/\ ins. wide
and 2/8 ins. thick can be used. The diameter of the barrel of the commu-
tator should not be less than 4, and preferably 5 ins. The brush holders
and yoke may be copied advantageously from any of the* standard ma-
chines now on the market. Only two brushes are to be used, and these
set precisely a quarter of a circle apart, reckoning around the barrel of
the commutator. _
The windings just described are for machines to work on a 110-115-
volt circuit. If windings for 220-230 volts are desired the armature coils
should be of No. 19 wire, each coil five layers deep and four turns wide,
making ten layers of wire per slot. The field coils of the cast-iron mag-
net must be of No. 26 s.c.c. wire, wound to the dimensions specified
above, namely, ij/2 ins. deep and 2j4 ins. long. The coils for the cast-
steel magnet will be of No. 27 wire wound to a depth of ij^> ins. and a
length of 2 ins. For 5oo-volt service use No. 23 double-cotton-covered
wire on the armature, six turns wide and seven layers deep, per coil ; 84
wires per slot. On the cast-iron magnet use No. 29 double-covered
wire and on the steel magnet No. 30.
The principal magnetic and electrical data of the two machines are
below :
II5-VOLT MOTOR.
Cast-iron. Cast-Steel.
Resistance armature winding I ohm 0.9
Normal armature current 18 amp. 18 amp.
Approximate speed 1325 r. p. m. 1325 r. p. m.
Flux per pole 340,000 lines
Density in field cores 33, 700 70,000
Density in air gap 26,000 38,000
CHAPTER IX.
THREE HORSE-POWER MOTOR.
The motor design which forms the subject of this chapter, although
somewhat similar to those described in Chapters VII and VIII, differs
considerably in the constructional details of the magnet. Here a cast-
iron ring and wrought iron cores are employed with a view to simplifying
the work as far as possible without sacrificing the efficiency of the
machine, and also without making it unduly heavy. The cast-iron ring
is preferably made in a single piece and the wrought-iron cores are
turned to a very slight taper and drawn into holes in the yoke ring by
means of a bolt and heavy washer from the outside. Unless the builder
has excellent machine-shop facilities, however, and is an expert machinist,,
this construction will be found rather difficult, as it is necessary to have a
perfect fit between the taper of the magnet core and that of the hole in
which it is seated.
As an alternative the magnet frame can be cast in two pieces, the
division being along the line, x, Fig. 91. If the motor is built in this
way, each half must be chucked and the joint faced off fairly smooth, al-
though it is not necessary to have a perfect joint, as no magnetic lines of
force cross the break. After truing up the abutting facer of each half of
the magnet ring the two halves should be clamped together with i-32-in.
of cardboard in between them and four straight holes bored for the re-
ception of the field-magnet cores.
Fig. 89 is a semi-sectional elevation of the field magnet complete
without the journal pedestals ; a field-magnet core is shown by Fig. 90.
The cast-iron pole-pieces must be accurately fitted to the ends of the
cores and pinned permanently in place with iron pins. The magnet ring
and pole-pieces should be of the best grade of pig iron obtainable. The
ELECTRICAL DESIGNS
FIG. 89. SEMI-SECTIONAL ELEVATION OF THE FIELD-MAGNET FRAME.
magnet cores should be made of Norway wrought iron. The corners of
the pole-pieces should be heavily rounded so that no sharp edges are left.
The length of the pole-piece parallel with the shaft is the same as its width
at right angles to this dimension. The outside diameter of the magnet ring
THREE HORSE-POU'ER MOTOR
81
FOOT
FIG 90. MAGNET CORE.
FIG. QI. SIDE ELEVATION OF MAGNET FRAME AND SECTION THROUGH BEARING.
is 2oys ins. The extreme breadth of the ring parallel with the shaft is 8
ins. The other dimensions are as below :
INCHES.
A Bore of armature chamber 8
B Axial length of pole-face 3&
C Width of pole-piece, tip to tip, in a straight line 3^
82
ELECTRICAL DESIGNS
D Radius to which pedestal seat is cut , 8}
E Thickness of yoke ring i/^
F Distance between ribs 7
G Width of plane surface back of magnet coil 4/4
H Height, base line to armature center 12
O Axial length of pedestal seat 4
P Straight-line width of pedestal (width of pedestal scat is the same). . 7
Q Length of pedestal foot, commutator side 7
The dimensions of the magnet core (Fig. 90) are as follows :
At y m p
Diameter, ins 2^4, 3^ lYz
Lengthens 2 3 I 1 X
Fig. 91 is an edge view of the field-magnet frame, including one jour-
nal pedestal shown in perspective and the other in cross-section. After
fitting the magnet cores into place in the ring, the pole-pieces should be
bored and the pedestal seats cut, at one setting of the frame. Fig. 92 is
a croSvS-section of the armature
core, showing the details of con-
struction. The discs are mounted
on a cast-iron drum, which is
provided with a flange head and
a hub, y, at one end, the other
end being open. The discs are
clamped in place by a cast-iron
ring which is provided with a
hub, /, similar to that at the other
end of the drum, and drawn to
place by means of six /^-in.
bolts passing through holes in
the clamping ring and tapping
into the end wall of the cast-iron drum. The center lines of two of these
bolts are indicated by b b. The cast-iron drum should have a key-seat cut
in it so that the discs may be positively driven.
Both the hub on the end of the drum and that on the clamping ring
should be keyed to the shaft so that there will be no opportunity for dis-
placement. At each end of the core structure a disc of fiber 1-16 in. thick
should be provided, as indicated by the heavy black lines in the engraving.
These discs must be toothed exactly like the core discs so that the ends of
the magnetic core will be entirely covered. The iron core discs are 7^4
ins. in diameter, with a central hole 4^/4 ins. in diameter, and 47 slots j4 m -
wide and ^4 m - deep ; the slots have parallel sides. The discs must be of
FIG. 92. SECTION THROUGH ARMATURE CORE.
THREE HORSE-POWER MOTOR 83
the best grade of charcoal iron, 1-40 in. thick. The dimensions of the
armature core structure are as follows :
INCHES.
I Length of hub clear through 2%
Bore of this hub iX
J Length of hub clear through 2
Dore of this hub I
K Diameter of hubs 2^
L Internal diameter of core drum 4
M Outer diameter of core drum 4^
N Diameter of flange and clamping ring 6^
Thickness of flange and clamping ring fi
The two journal pedestals are exactly alike and made of ordinary
cast-iron. The base must be turned accurately to conform to the circle
to which the pedestal seat on the magnet frame is machined. The
standard, or pedestal, is an open frame of j/2-in. metal; the box is of the
ring-oiling type, with a single ring hung exactly midway of the journal.
Fig. 93 is a transverse cross-section of the pedestal and box. The box is
bushed ; the bushing may consist of a brass casting turned to shape, or it
may be made by babbitting a piece of thin brass tubing, the outer di-
ameter of which is a snug fit in the box. The oil slot across the center of
the box must be provided with a suitable
covering to exclude dust. Each pedestal
should be bolted to the magnet frame with
two y% -in. cap screws. The pedestal and box
measurements are below :
INCHES.
P Widest part of standard 7
R Axial length of oil well 2^
S Inner length of oil well 2
T Inner width of oil well 3
t Radius line to indicate origin of circle ij
U Outside diameter of box 2%
V Inside diameter of box I j
W Outside length of box 4
Y Length of bushing 3^
Bore of bushing I
Z Projection of box beyond oil well wall. -Jf
fc I~?"-'l."p f 'l"-~^ >| Diameter of oil ring 2^
FIG. 93. CROSS-SECTION OF Bore f oil rin S ^
PEDESTAL AND BOX. Width f ' d rin S ^
A bearing must be turned on the outside of the inner end of the
pedestal on the commutator side of the machine,, as indicated in Fig. 91, to
84 ELECTRICAL DESIGNS
accommodate the brush-holder yoke, which may be copied from any of
the standard makes. Only two sets of brushes are required, each set
comprising two carbon brushes y$ in, thick and ij/s ins. wide; the two
sets must touch the commutator exactly 90 (n^4 segments) apart, center
to center. The commutator must have 47 segments, and must measure
3 ins. along the shaft, extreme length. The commutator core must be
bored to fit the portion, c, of the shaft, and key-seated to correspond.
The diameter of the barrel should be not less than 4 ins., and the diameter
measured at the connecting lugs must not exceed 63/2 ins. The brush
surface, measured parallel with the shaft, must be 2.^/2 ins. long. It will
be best to buy the commutator complete from one of the several makers
cf this class cf apparatus.
Fig. 94 is the armature shaft. The key-seats are all J-fJ wide and 3-16
deep. The dimensions of the shaft are below:
, At > Total
f a c j Length
Diameter, inches I i l X- 1% I
Length, inches 6-^ 6j\ 5-^ 3^ 21^
The field-magnet coils may be wound directly on the cores or on
bobbins made of thin vulcanized fibre. If they are wound directly on the
cores, the latter must be wrapped first with three layers of unbleached
cottons and painted with shellac varnish, two circular coil heads of hard
fibre being first fitted to the large part of each core. The coils consist of
No. 22 single cotton-covered wire, wound to a depth of ^4 in., exactly.
The exact number of turns is immaterial, except that all four coils must
contain the same number of turns, and as many turns should be put on
as can be got in the space available. With careful winding, the builder
should get 2,565 turns in each coil. For a 23<>volt motor use No. 25
double cotton-covered wire, and for 500 volts use No. 28 double cotton-
covered, wound to the depth specified. Should the reader prefer to wind
the coils in bobbins, the magnet core need not be wrapped, of course.
After each coil is completed, secure the outer end and cover the outside
layer with unbleached cottons two layers deep, heavily varnished. Fig.
95 indicates how the field coils should be connected up.
The armature coils of the ii5~volt machine are of No. 13 wire, each
coil containing eight turns. The winding must be two wires wide and
four layers deep, per coil, so that when the coils are in place there will be
1 6 wires in each slot two wide and eight deep, as shown in Fig. 96.
There are 47 coils, connected up wave-fashion. In winding the coils it
will be advisable to bend a hook in each starting end and leave the final
ends straight. The armature coils must be wound in a former so that the
THREE I!ORSE-PO::'ER MOTOR
FIG. 95. FIELD COIL CONNECTIONS.
outline of the opening
through each coil is a
square measuring 4^
ins. on each side. The
ends should lead out from
two corners, and each
coil must be wrapped
carefully and firmly with
two layers of German
linen tape, each layer
being painted with shel-
lac varnish. The slots in
the armature core should
be provided with insulat-
ing troughs of press
board 1-64 in. thick.
Put the coils on the armature all the same way bent
ends to the left and straight ends to the right, facing the
commutator. There must be twelve teeth between the two
slots in which any given coil is placed, and there must be 22
commutator segments between the two to which the terminals
of any given coil are connected, as indicated by Fig. 97. It
will be found best to first put on 1 2 coils in regular right-
handed rotation, pressing the ends down closely where they
lap, and slipping a bit of
thin oiled paper between
the crossings. This will
put one layer of coils in
24 of the slots. Then
put coils in the 23 vacant
slots in the same fashion;
there will then be 46 half
filled slots and one filled.
Continue the second
layer of coils right along
from the 24th coil,
following the same plan
as before. At the finish,
there will be a bent end
and a straight end pro-
PIG. 96. SLOT. FIG. 97. CONNECTING DIAGRAM. jecting from each slot.
86 ELECTRICAL DESIGNS
Carry the bent ends 1 1 or 1 2 segments to the left, around the commutator,
and put them all in the segment slots. Then take any one of the straight
ends, find the bent end which is the other terminal of its coil, and con-
nect the straight end, as shown in Fig. 97, with 22 segments between it
and its mate. The other straight ends may be put in in regular order
without tracing, if the coils have been put on the core properly and the
bent ends in the commutator lugs in strict sequence.
If it is desired to build the machine for 230 volts, wind the armature
with No. 1 6 double cotton-covered wire, putting 15 turns in each coil
three wide and five deep so that each slot will contain 30 wires, 3 wide
and 10 deep. For 500 volts, use No. 19 wire, putting 28 turns in each
coil 4 wide and 7 deep so that each slot will contain 56 wires, 4 wide
and 14 deep. The principal technical data for the Ii5-volt machine are
given below :
Revolutions per minute 1,320
Armature resistance, warm 0.4
Armature current, normal 22
Armature and brush drop, volts about 10
Per cent regulation about 9%
Flux density in armature core and teeth, per square inch.. . 75,000
Flux density in air gap 29,300
Flux density in magnet cores 93,000
Flux density in magnet yoke : 48,000
Leakage coefficient 1.25
Resistance of field winding, ohms 169
Exciting current, amperes 0.68
Copper loss in field, watts 78.2
Copper loss in armature, watts IQ4
Core loss in armature, watts 46
Approximate efficiency, allowing 5 per cent for friction and
windage, per cent 82
The starting box should be purchased from any of the standard
rheostat builders ; a satisfactory home-made one of this size is rarely
produced.
CHAPTER X.
ONK KILOWATT COMBINED ALTERNATING AND DIRECT-CURRENT
MACHINE.
There are presented in this chapter designs and working drawings
for a type of combined alternating and current machine which it is
thought will prove generally useful for experimental and laboratory work
in alternating and direct currents, and which is applicable on most of the
electric-lighting circuits found in practice.
The design contemplates working the machine in a number of dif-
ferent ways :
1. As a direct-current generator or motor.
2. As a single, two or three-phase generator or motor.
3. As a rotary converter, changing single, two or three-phase to
direct current.
4. As an inverted rotary converter, changing direct current to
single, two or three-phase alternating currents.
5. As a phase transformer, changing alternating- current of one
phase to that of any other phase.
Some of the foregoing functions may be in operation at the same
time ; for instance, Nos. i and 2 combined would give a "double-
current" generator. Also No. 3 or No. 4 may be in operation simultane-
ously with No. 5.
Three sizes of this type of machine will be described, of I, 2 and 4
kilowatts capacity, respectively, and in all of these the same scale of volt-
age has been adopted, namely, no volts for the direct current, 80 volts
for single or two-phase alternating, and 70 volts for the three-phase alter-
nating. These voltages admit of considerable adjustment, however, by
varying the field excitation or speed in case of a generator. The values
given represent about the maximum which can be developed continu-
ously.
In operating on single-phase alternating circuits it is necessary to
adopt some device which will make the machine self-starting, and this
88 ELECTRICAL DESIGNS
has been provided in the shape of a special switch located in the base
of the machine and which, at starting, temporarily changes the connec-
tions to those of a series motor which, as is well known, readily starts
when alternating current is turned on. The armature is allowed to reach
a speed slightly above synchronism, and the switch is then thrown over
to the running position, where the machine operates as an ordinary
synchronous motor.
In starting on two or three-phase circuits, the same switch is utilized
to break up the field winding into a number of short sections on open
circuit, thereby avoiding the high induced e.m.fs. which would otherwise
be produced on turning the alternating current into the armature wind-
ing. It will be understood that where two or three-phase currents are
employed the machine is self-starting without any special device, by
virtue of the rotary field principle. If the starting current, with this
arrangement, is found to be objectionably large, it can be avoided by
starting on a reduced pressure supplied from small auto-transformers.
The general features of the design are multipolar field having a
circular yoke of cast iron with laminated wrought-iron poles cast in.
This type is selected because it admits of high magnetic density and short
air gap. and consequently much greater output than does an all cast field,
while at the same .time it is only slightly more expensive or difficult to
construct. An all cast-iron field of the same general design will have
only a little more than half the output, and an all cast-steel field, while
good magnetically, is scarcely to be considered at present owing to the
difficulty in securing steel castings on short notice.
Field coils wound in two or more sections each, and provided with
terminals for connection to the starting switch. This is necessary in
order to obtain a sufficient reduction in the impedance by connecting the
various sections in multiple at the start.
A distributed armature winding, with collector rings tapped in at
appropriate intervals on the commutator for alternate-current working.
A toothed armature core with deep and narrow slots, and provided with
a formed-coil winding, as in direct-current practice.
The minimum number of slots and coils is determined by the number
of poles and by the consideration that taps must be made for both two*,
and three-phase working. The quotient obtained by dividing the number
of coils or commutator segments by the number of poles must be divisible
by two for two-phase working and by three for three-phase working, and
hence by two times three for both together. Thus 24 coils and segments
are appropriate for a four-pole machine, 36 for a six-pole, and so on.
Six collector rings will be required; ordinarily seven would be
ONE KILOWATT DOUBLE-CURRENT MACHINE 89
necessary, four for two-phase and three for the three-phase. By making
one of the two-phase rings the starting point for the three-phase, one ring
serves for two, and the total number may be reduced to six.
It would be possible, of course, to use but four rings, obtaining
three- phase current by means of two-phase three-phase transformers, but
it is preferable to add two rings and obtain all phases directly- from the
machine.
The hollow base plate, which is cast in one piece with the bearing
pedestals, serves as a housing for the starting switch already referred to.
This switch is operated by a lever on the outside, at the front or direct-
current end of the machine, and has two positions 120 degrees apart, the
starting and running positions respectively. In the starting position the
various sections of the field winding are in parallel with each other and
in series with the direct-current end of the armature.
In the running position the field sections are in series, giving the
maximum resistance, and are placed across the direct-current brushes,
at the same time alternating current from the single-phase mains is
turned into the collector rings.
A pulley having a heavy rim for the purpose of securing a consider-
able fly-wheel effect will be found advantageous in adding to the smooth
running of the machine, particularly when used as a rotary from the
alternating-current end.
A pulley of this kind will also be useful where the machine is to be
used as a generator direct belted to a gas or gasoline engine. The need
for a considerable amount of momentum in the running parts of a rotary
is real and genuine, for without it there is a disagreeable oscillation or
"pumping," which makes synchronism unstable and sometimes causes the
machine to break out of step even before full load is reached.
The bearings are of the ring-oiling type, and of a form which gives
good lubrication without the disadvantage of having oil thrown oil
outside the bearing.
The running qualities of these machines will doubtless prove quite
satisfactory. There is not likely to be trouble from sparking, in spite of
the fact that the armature is multiple wound, in which' ordinarily, a slight
lack of symmetry in field strength would cause heating and sparking.
The connections already made to the collector rings for another purpose
serve also as equalizers, which permit equalizing currents to flow and
thus counteract any slight inequality in the various field poles.
Armature reaction may be guarded against by clipping off the
corners of every third lamination in the field poles. This will have the
effect of increasing the density in the pole tips to practical saturation, thus
90 ELECTRICAL DESIGNS
avoiding further distortion by armature currents and giving practically a
fixed point of commutation for all loads.
Heating in the armature and field windings should not prove serious,
for the current densities employed are moderate, considering the size of
machine. In the pole pieces, heating would ordinarily be expected, due
to the short air gap and high density, but their laminated construction will
entirely obviate this difficulty.
While primarily intended for use on 125-cycle circuits, modifications
will be indicated enabling these machines to be used on 6o-cycle circuits
also. This involves either a reduction in speed of one-half, with a
correspondingly reduced output and voltage, or a reduction in the number
of poles to one-half, keeping the speed and output the same, but necessi-
tating a somewhat more difficult change in connections and winding.
Referring now to the one-kilowatt machine, Fig. 98 shows an end
yiew of the field magnet and base. There are four poles cast into the
yoke, which forms a separate casting and is bolted to the base plate by
four 7-i6-in. by i l / 2 -in. hexagon cap screws. The poles are built up of
plain rectangular strips of soft iron about No. 22 gauge, which are
clamped between two heavier plates by one or more long flat-head bolts.
The pattern for the field casting should be made just as though it
were for an all cast field, the laminated pole pieces being laid in the
mould after the pattern has been drawn, and the iron poured in around
them. The natural shrinkage of the metal on cooling will cause the poles
to be tight and secure. It would give additional security, however, to
notch the poles before casting in as indicated by the dotted lines. Still
another plan is to leave the end plates short, and to spread the laminations
apart where they enter the yoke. This will allow the iron to fill in the
interstices and so obtain a good hold on the pole. As the poles have
teen left with square ends, they must now be bored out 3 5-16 ins. and
the corners slightly rounded.
Fig. 99 is a longitudinal half section of the assembled machine, which
shows the construction of the armature, bearings, commutator, and col-
lector rings, and also the location of the starting switch in the base.
The armature core is built up of soft-iron discs about No. 27 gauge ;
two heavier discs of wrought iron, 3-16 in. thick, are provided at the ends
as a reinforcement for the teeth, and the whole is clamped between two
cast-iron flanges run up on threads cut in the shaft. These flanged pieces
serve also as a support for the "straight-out" winding.
Plain round discs may be used in building the core and the slots
milled out, being careful, however, to take the discs apart after milling and
insulate them with paper or japanning. The keyway in the dies insures
ONE KILOWATT DOUBLE-CURRENT MACHINE 91
their registering when reassembled, in spite of possible slight inaccuracy
in milling the slots. It is not necessary to insulate the discs from the shaft
if they are fairly well insulated at all other points.
The bearings have a central rib 3/ in. thick, which supports the
brass or bronze sleeve forming the journal proper. The oil pockets at
1
FIG. 98. END ELEVATION OF THE MACHINE WITHOUT THE ARMATURE.
either side of the web communicate by means of a slot cored out in the
web, so that the oil level may remain the same on each side. The oil
rings are ^ in. wide and ride on the shaft through grooves turned eccen-
trically in the sleeve.
ONE KILOWATT DOUBLE-CURRENT MACHINE
93
The commutator has a steel sleeve fitting the shaft, upon which are
two flanges, one solid with the sleeve and the other threaded on it and
tightened by means of a spanner wrench applied to holes drilled in its
face. Both flanges are undercut at an angle of about 60 degrees, to hold
the segments in place.
Probably the best way to construct the commutator is to turn up a
copper casting of the required section, and then slit the cylinder into 24
regments by means of a i-32-in. cutter, in a milling machine. The seg-
ments are then built up with i-32-in. mica between and insulated from
the sleeve by 1-16 in. of mica or other good insulation.
The collector rings are similar in construction. The two end rings
are counter-bored to let in the flanges of the sleeve, which, in this case,
need not be undercut. The other rings are plain round and are simply
clipped over the insulating sleeve, and separated from each other by
i-i6-in. fiber, or equivalent insulation, which is allowed to project some-
what above the surface of the rings.
Connections to the rings arc made by drilling in from the back side
and soldering in short wires, No. 12 or No. 14, which should be carefully
I-'IGS. 100 AND 101. BRUSH-HOLDER COLLARS.
FIG. 102. ARMATURE HEAD.
insulated where they pass through other rings by small fiber or rubber
tubes. These wire leads are made only just long enough to project a
short distance from the back ring and are there soldered to some thin
copper strips taped and laid in the bottom of the armature slots, six of
which have been cut 1-16 in. deeper than the rest to accommodate these
connections. It will be the more convenient to make all these connec-
tions permanently and test them before laying on the armature coils.
Fig. 100 shows the brush ring for the alternating current end, and
Fig. 101 the one for the direct-current end of the machine. They are
made in halves, held together by screws, which will facilitate in assemb-
ling the machine. The direct-current ring has four lugs for supporting
the brush holder and the alternating-current ring has six, one for each of
the six collector rings.
94
ELECTRICAL DESIGNS
Fig. 102 shows an end view of the armature core and Fig. 103 a
development of the armature winding. The core has 24 slots 3-16 in.
wide and 7-16 in. deep. Every fourth slot is made y 2 in. deep to allow
space for connections to the rings. The teeth are plain straight and the
armature must be banded after the coils are in place.
The armature winding is of the type known as "straight out" and is
composed of form-wound coils of No. 20 double cotton-covered wire, each
coil consisting of 16 turns arranged four wide and four deep. The coil
is wound as a simple straight loop, and after receiving a wrapping of tape
it is bent until it will span one-quarter of the armature circumference.
One side of a coil occupies the top of a slot and the other side of the
same coil occupies the bottom half of a slot 90 degrees, or six slots, in
advance of the first. Thus arranged, the coils interleave in a very
compact manner and the space required for cross connection is reduced to
a minimum.
The terminals are brought out at the apex of the coil and are con-
nected directly to the commutator segments, the beginning of one coil
T-
. -} - - s'-
FIG. 103. BARREL WINDING.
---* H r>-'
FIG. 104. BRUSH-HOLDER DETAILS.
and the ending of the adjacent coil connecting to the same segment
The advantage in bringing the terminals straight out to the commutator
in this way is that, in addition to being more convenient, it permits the
brushes to be placed opposite the poles, where they are more accessible
than when placed between the poles.
Fig. 104 shows details of the brush holders. The direct-current
holders are of simple construction, but neat in appearance, and are
intended for radial graphite or carbon brushes y% in. thick, \y ins. wide
and I in. long. The necessary tension on the brush is supplied by an
open-coil spring concealed in a hollow lug cast on the side of the holder,
and acting on a small pressure foot shown separately in the drawings.
ONE KILOWATT DOUBLE-CURRENT MACHINE 95
By lifting the pressure foot by means of the eye at its top and turning it
half around, a brush may be readily removed from or inserted into the
holder.
The alternating-current brush holders are carried upon studs sup-
ported from the brush ring, and have slots l in. by y in. for copper-leaf
brushes. There need not be any spring tension provided, as the natural
spring of the brush will be sufficient to insure good contact. Two thumb
screws are provided, one to hold the brush and the -other to clamp the
holder upon its stud in the desired position. The studs are of different
lengths, the dimension -marked X having the values 3J/ ins., 2?^ ins., 2J/&
ins., i y% ins., ij ins. and y% in. for the six studs. Quarter-inch brass rod
may be used to make these from, the collars being soldered or threaded
on and the ends threaded for a hexagon nut. All brush holders and parts
should be made in brass or bronze.
Figs. 105 and 106 are diagrams to be followed in making taps to the
collector rings. The 4-pole arrangement, Fig. 105, is intended for oper-
ating on i25-cycle circuits and the two-pole, Fig. 106, for 60 cycles.
These connections should be made at the back of the commutator before
it is placed in position on the shaft. In the four-pole arrangement, for
FTG. 105. FOUR-POLE CONNECTIONS. FIG. Io6. TWO-POLE CONNECTIONS.
instance, segments No. i and No. 13 are connected together and to a lead
marked No. i, which goes to collector ring No. I, and similarly for the
others. Thus connected, single-phase current may be obtained from
rings 1-2 or 3-4. Two-phase current from 1-2 and 3-4 and three-phase
current from 1-5-6. The output and voltage with these various connec-
tions are as follows : Direct current, 10 amperes at no volts ; single-phase
alternating, 10 amperes at 80 volts ; two-phase alternating, 7 amperes per
phase at 80 volts; three-phase alternating, 6 amperes per phase at 70
volts.
Fig. 107 shows a form of fly-wheel pulley which is recommended as
Conducing to smooth running, for reasons already referred to. This
ELECTRICAL DESIGNS
pulley is of cast-iron and should be turned perfectly true all over and
carefully balanced, as should also the armature. These rotating parts will
be required to run at 3750 r.p.m. on 125 cycles, and unless precautions are
taken the vibration will be excessive.
Fig. 1 08 is a detail of the armature shaft. This is designed to be
turned from a piece of 24-in. cold-rolled steel, and for this reason the
customary collar at one end has been omitted, and instead threads are
cut on both ends for receiving the end plates of the core. This does
away with expensive forgings and provides a shaft requ iV mg only a
FIG. I08. FLY-WHEEL PULLEY.
FIG. 109 A. DETAILS OF THE STARTER SWITCH.
minimum amount of turning. Small
grooves are provided at the journals
which prevent oil from creeping along
the shaft and being thrown off outside
the bearing. There are two keys oil the
shaft, one for the core punchings and
the other to hold on the pulley.
Fig. 109 shows the arrangement cf the switch cylinder and contacts
for the single-phase starting device. There are 30 contact fingers, each
5-16 in. wide, fastened to a strip of fiber % in. thick, which in turn is
screwed to the under side of the cast-iron base of the machine. Upon a
cylinder of hard wood or fiber 1^4 ins. in diameter are arranged two rows
of brass pieces, sunk in grooves cut on the cylinder and upon which the
stationary contact fingers press.
The cylinder may be rotated through an angle of 120 degrees by
means of a handle on the outside. The contacts on the cylinder are 120
degrees apart, which allows sufficient space for the first set to leave
98 ELECTRICAL DESIGNS
contact before the second comes into contact, this being essential to avoid
short circuit.
Fig. no shows a diagram of connections for the starting switch, by
means of which its action may be readily traced out. Numbers 1-12
represent the sectional field winding, there being four coils, each of which
is wound in three sections of approximately equal resistance. There are
then twelve pairs of ends which lead down into the base of the machine
and are connected to the stationary contact pieces, which are represented
by the upper row of small circles. The remaining three pairs of contacts
connect to the d.c. brush leads, the single-phase rings and the single-
phase mains respectively.
The lower rows of circles represent the contact pieces mounted on
the cylinder, and these are connected, as here indicated, by means of
wires laid in grooves upon the cylinder and occupying that portion of the
cylinder over which the contact fingers do not pass.
To operate the machine at no volts direct current or I25~cycle
alternating, no changes are necessary. For 6o-cycle alternating, how-
ever, the number of poles is reduced one-half by reversing the terminals
of any two successive field coils, and the armature winding must be
changed to a bipolar, one.
Another plan is to reduce the speed one-half, thus halving the volt-
age and output and connecting the field coils in series-multiple so that
they will still take the same current as at the higher voltage. In operat-
ing the machine as a converter, if it is desired that the direct-current out-
put be at no volts, the single or two-phase input must be at 80 volts. This
relation of voltage is fixed and can be expressed by d. c. volts X 707
= a.c. volts, and for three-phase by d. c. volts X -612 = a.c. volts. So
that if the alternating circuit is of 52 or 104 volts the machine should be
supplied at the proper voltage through a transformer. An old i5-light
transformer will serve for this purpose, and it should be arranged so that
its secondary voltage can be varied to some extent by changing the
number of secondary turns in circuit, thus giving a means of adjusting
the direct-current voltage.
The following is a brief summary of the data for winding and general
dimensions, and shows the method of calculating same :
Four-pole machine, 3750 r.p.m. ; armature, 3^ ins. diameter, 3 ins.
long; 24 slots, 3-16 in. wide, 7-16 in. deep; total number of conductors,
768 ; 24 coils, No. 20 wire, 4 wide, 4 deep ; No. 20 has 1021 circ. mils, di-
ameter d.c.c., .042 in. ; direct-current output at 400 c.m. per ampere, 10
amperes ; useful lines per pole - ^ = 240,000 ; total lines, 330,000.
700 /\ 02.5
ONE KILOWATT DOUBLE- CURRENT MACHINE
99
Part
Material
Total lines
Cross sect.
B.
H..
L.
Amp. turns
Armature
Wrought iron
120,000
2.25 sq. ins.
53.300
14
1.4 in.
20
2 air gaps
Air
240,000
4-5
53,300
16,800
.06 "
1,000
4 teeth
Wrought iron
240,000
2.
120,000
1 80
45 "
So
2 cores
Wrought iron
330,000
3-75 "
88,oco
20
1-5 "
30
I yoke
Cast iron
165,000
4-
41,300
74
4-25 "
315
Total 1445
The table above gives a total of 1445 ampere-turns or 725 ampere-
turns per coil ; mean length, I turn, II inches.
Circ. mils shunt wire IT X 11 X 7^5 _
25 X 12
Use No. 25 wire, 320 c.m., .028 inch d.c.c. 1155 turns (approximate)
per coil; 25 layers, 45 turns wide.
Wind in three sections. Bring out terminals from each section.
4 X !
Resistance of shunt field =
X ' 97 = 136 ohms.
Normal shunt current, .63 ampere. Use a rheostat of about 50 ohms
total resistance in shunt-field circuit.
Weight of wire in shunt coils= 4X III5 * TI X ' 97 4- 1 pounds.
1000 X 12
Length of wire, each armature coil = -- --- = 17.4 feet.
Total length of wire, armature, = 24 X 17.4 = 4*7 feet -
Total weight of armature wire =
= I>3 P oun< * So
'3 .26 ohm.
Resistance of armature v
1000 X io
Drop in armature at full load = 10.63 X .26 = 2.76 volts.
CHAPTER XL
TWO KILOWATT COMBINED ALTERNATING AND DIRECT- CURRENT
MACHINE.
The 2-kw. machine shown in the accompanying drawings is similar in
design, construction and operation to the four-pole machine described
in the preceding chapters. The present machine is somewhat larger,
runs at a slower speed, and has about double the output capacity of the
four-pole machine. Fig. in gives an end view of the field-magnet
frame. There are six poles of laminated iron cast into a circular yoke of
cast-iron, which, in turn, is bolted to the base plate by four hexagon cap
screws. After the poles are cast in and it is seen that all of them are tight
and firm in the yoke, they may be bored out to the proper diameter, 4.04
ins. The armature is to be finished 4 ins. in diameter, so that the air-gap
will be .02 in. across at each pole ; this will be ample for clearance if care is
taken in lining up the machine.
Fig. 112 is a section of the assembled machine which shows the
construction and relation of the various parts. The armature is of the
usual laminated construction, the core discs being held between two cast-
iron flanges screwed upon the shaft. If the armature slots are milled out,
the discs must be taken apart, cleaned up, and insulated before being fin-
ally assembled on the shaft. If this is not done the eddy current loss will
be excessive, causing heating and seriously reducing the available output.
Fig. 113 shows a detail of the armature shaft. This is intended to bo
made from i-in. cold-rolled steel. Threads are cut at both ends of the
core portion to receive the cast-iron flanges which clamp the core punch-
ings. Two keys are provided, as shown in the drawing.
The bearings are made with a brass sleeve fitting the shaft, supported
at its center by a projecting web cast in the bracket. Although it is pre-
ferable to bore the bracket for this sleeve, the machine work may be
avoided by coring the bracket somewhat larger and then babbiting the
sleeve into its support when the parts have been lined up in their proper
position. The oil rings are of brass l /& in. wide and iJ/ ins. inside diame-
TWO KILOWATT DOUBLE-CURRENT MACHINE
101
ter ; grooves are cut eccentrically in the sleeves to receive the rings, the
grooves being made about 5-32 in. wide in order to allow the ring a small
amount of play.
The commutator is built up on a machine steel sleeve, with the
flanges undercut at 60 degrees. The segments may be cast separately or
cast as a solid cylinder and afterward cut into segments on a milling
FIG. III. END ELEVATION OF THE FIELD-MAGNET FRAME AND ONE PEDESTAL.
machine. The segments should be of copper, with i-32-in. mica between
them, and i-i6-in. micanite, or equivalent insulation, between the sleeve
and segments. The number of segments is 36. The collector rings are
six in number and mounted upon a sleeve with 3-32 in. fiber discs between
the rings. Connections are made by drilling in from the back and solder-
ing in short leads of No. 8 or No. 10 wire, one to each ring. Each of the
TWO KILOWATT DOUBLE-CURRENT MACHINE
leads must be carefully insulated from all rings, except the particular one
to which it is electrically connected. Some thin copper strips are to be
provided with a wrapping of tape and laid in the bottom of the armature
slots, six of which must be made 1-16 in. deeper than the rest to accom-
modate the strips. These strips carry the current across the armature
and are connected to the commutator at the proper intervals.
At the alternating-current end of the machine the fly-wheel pulley is
shown in position on the shaft. This style of pulley will be found advan-
tageous in operating the machine as a rotary converter or in driving it by
means of a gas engine. If the machine be used as a motor an ordinary
pulley will answer. The pulley is for a 2^-in. belt, and is $y 2 ins. in
diameter.
Fig. 114 shows the brush-holder collar. This answers for both the
alternating-current and the continuous-current ends of the machine, as
FIG. 114. BRUSH-HOLDER COLLAR.
FIG. 115. DETAILS OF BRUSH-HOLDERS.
there are six collector rings and also six brush holders. Care should be
taken in drilling the holes for brush holders to have them equidistant, for
upon this the accuracy in spacing the brushes around the commutator
depends. At the alternating-current end this does not matter particu-
larly. The brush-holder collars are necessarily made in halves, as it
would be difficult to assemble the machine with a one-piece collar.
Fig. 115 shows details of the brush holders. These are of the same
type as those already described in connection with the four-pole machine.
The alternating-current brush holders have no spring tension and are
designed for leaf-copper brushes % in. thick and ^ in. wide. The studs
are of different lengths to suit the position of the rings ; the dimension,
X, is 32/s ins., 2% ins., 2*4 ins., ift ins., I 3-16 ins. and 11-16 inches for
104
ELECTRICAL DESIGNS
the six studs. They are made of 5-16-111. brass rod. The continuous-
current brush holders are designed for radial carbon brushes j in. thick,
iJ4 ms - wide and 1% ins. long.
Fig. 116 is an end view of the armature core. There are 36 slots,,
each 3-16 in. wide and 7-16 in. deep; every sixth slot is made y 2 in. deep
to allow space for the connection strips referred to above. After the
coils are in place the armature must be banded at three points, one band
to go around the center of the core, and one around each end of the
winding where it projects beyond the core. A groove must be turned in
the periphery of the core to accommodate the central band, so that the
thickness of the band will not be added to the length of the air-gap. This
groove may be turned on the core before the slots are milled out, or it
may be done afterward by filling in the slots temporarily with hard-wood
strips. It should be about 1-16 in. deep and ^g or 7-16 in. wide.
Fig. 117 shows a development of the armature winding. This is of
the "straight-out'' type, and is composed of 36 form-wound coils of No. 20
wire, 16 turns per coil. One side of a coil occupies the top half of slot
No. i, and the other side of the same coil occupies the bottom of slot No.
7; that is to say, each coil spans one-sixth of the circumference of the
core. The terminals are brought out at the apex of the coil, and each is
FIG. Il6. END OF ARMATURE CORE.
FIG. 117. DEVELOPMENT OF WINDING.
connected to the nearest commutator segment ; the inside terminal of one
coil and the outside terminal of the adjacent coil connect to the same
segment. The point of commutation will be found at or near the center
line of the pole pieces.
Fig. 1 1 8 is a diagram of the connections for the collector rings. This,
arrangement is for a six-pole field. The leads numbered I to 6 pass
across the armature and are connected to the six collector rings at the
alternating-current end of the machine. Connected in this way, single-
phase current may be obtained from rings I and 2 or 3 and 4, two-phase
TWO KILOWATT DOUBLE-CURRENT MACHINE
105
currents from rings I and 2 and 3 and 4, and three-phase currents from
rings i, $ and 6.
The output and voltage with each of these various methods of work-
ing are as follows: Direct current, 15 amperes at 115 volts; single-phase
alternating, 15 amperes at 80 volts; two-phase alternating, n amperes per
phase at 80 volts; three-phase alternating, 9 amperes per phase at 70
volts.
Fig. 119 shows the outline of one of the field coils. These are wound
on a form, and each coil is divided into two sections of approximately
FIG. 1 1 8. DIAGRAM OF TAP CONNECTIONS.
FIG. 119. FIELD COIL.
equal resistance, with separate terminals brought out from each section.
The size of wire is No. 23, B. & S. gauge.
The arrangement employed for starting the machine as a motor on
single-phase circuits is as shown in the description of the four-pole ma-
chine (see Fig. 109 and no, and the description on pages 96 and 97, with
the single exception that in the present machine there are six coils of two-
sections each instead of four coils of three sections each.
To operate the machine at no volts, continuous current, or 125
cycles alternating, the speed should be 2500 r.p.m. For 60 cycles the
only method available is to reduce the speed to 1200 r.p.m and to connect
the field winding in series multiple. This is most conveniently done at
the starting switch by changing the wiring of the last row of contacts on
the switch cylinder, so that when the switch is in the running position
the two sections of each field coil will be in multiple and the six multi-
plied pairs in series and connected across the continuous-current brushes.
This will reduce the voltage to about one-half of its value at the higher
speed, and the output will then be as follows : Continuous current, 15
amperes at 55 volts; single-phase, 15 amperes at 40 volts; two-phase, II
amperes at 40 volts ; three-phase, 9 amperes at 35 volts. It is probable
io6
ELECTRICAL DESIGNS
that by adjusting the field excitation, the voltage could be brought up to
45 or 47 volts, and thus admit of working directly on single-phase circuits
of 50 or 52 volts as a motor or rotary without the use of an individual
transformer. For other voltages a transformer will be necessary.
The following is a summary of the data for winding and general
dimensions : Speed, 2500 r.p.m. on 125 cycles, or 1200 r.p.m. on 60 cycles.
Cast-iron yoke, laminated-iron poles cast in. Armature, 4 ins. in diame-
ter, 3 ins. long; 36 slots 3-16 in. wide, 7-16 in. deep; every sixth slot yl
an. deep, 36 coils of No. 20 wire; 16 turns per coil, four wires wide and
Jour deep. Total, 1152 conductors. At 15 amperes continuous-current
output the cross-section of armature conductors is 400 circ. mils per
ampere.
Useful lines per pole, at 115 volts and 2500 r.p.m. :
1152 X 41-6
US X 10"
240,000.
TOTAL LINES PER POLE, 32O,OOO
Part
Tptal lines
Cross sect,
sq. in.
B.
H.
Length
Ampere
turns
120,000
a.
40,000
IO
1.5"
je
:2 air gaps
240 ooo
4.
60 ooo
1 8 800
O4
7SO
240,000
2.
1 2O OOO
1 80
44
7Q
160,000
a.c
46,000
1 02
3.
1 06
^Z cores . ..........
320,000
I -JC
85 ooo
18
2t
AS.
23) =
Total ampere turns in field winding, 995. Circ. mils field wire (No.
ii X 12 X 500
= 345. Mean length per turn, 12 inches.
i6X 12
Turns (approx.) per coil, 500; 16 layers of 32 turns each. Resis. of field
winding (coils in series), 63 ohms. Normal shunt current, I ampere
(nearly). Use rheostat of about 40 ohms total in field circuit.
Mean length of wire per armature coil, 16 feet. Total length of
armature wire, 36 X 16 = 610 feet. Total weight of armature wire, 2
pounds. Resistance of armature, 0.17 ohm. Drop in armature winding
at full load, 2^/2 volts.
CHAPTER XII.
FOUR KII ins. diameter, 4 ins. long, 48 slots 3-16 in. X /^ in. ;
3-16 in. = 188 in. width of slot, allowing three No. 17 wires and insulation
of 18 mils; 5 in. depth slot, taking eight No. 17 wires and insulation and
FOUR KILOWATT DOUBLE-CURRENT MACHINE
bands of 45 mils., giving 24 conductors per slot. Total, 1,152 conductors.
Direct-current output at 470 circ. mils per ampere = 36 amps.
There will be 455 ampere-turns in each field coil. Mean length of I
turn = 13.5 inches circ. mils shunt wire =
_ ii X 13-5 X 455 _
= 465-
12 X 12
Use No. 23, having 509 circ. mils, .034 in. diam., d.c.c. ; 620 turns
(approx.) per coil, in 14 layers, 44 turns wide. Wind in two sections and
bring out individual terminals from each section.
Resistance of shunt field = 8 X 620 X_i3 : 5_>Og:3 _ 1I4 ohms
I ,OOO /\ 1 2
Normal shunt current = - = 75 am P- Use rheostat of about 40
ohms in shunt field. Weight of wire in shunt coils :
8X620X 13-5 X i.54 = g " lb
1,000 X 12
Total length of wire on armature :
4 X .8 X
12
Total weight of wire on armature :
864 X 6.2
1,000
Resistance of armature :
1,000 X 64
= 5-37
= .068 ohms.
864 X 5-04
Drop in armature at full load :
36 X .068 = 2.45 volts.
Part
Material
Total
lines
Cross-
section
B.
H.
L.
Ampere
turns
I armature..
Wrought iron
170,000
4. in.
42.500
4
1.5
6
2 air gaps . .
Air
340,000
6.
56,600
17,880
.0313
566
5 teeth
Wrought iron
340,000
3.
113,000
1 08
.5
54
Wrought iron
420,000
5.
85.000
18
1.7
31
Cast iron
210,000
4.7]
42,000
68
2.7
252
Total ampere-turns 909
CHAPTER XIII.
SINGLE-PHASE RECTIFIER.
The accompanying drawings and description constitute a design for
a machine to "rectify" single-phase alternating current; that is, to change
it into a pulsating direct-current without changing its e.m.f.
Fig. 134 shows an end view of the field magnet with coils in place,
FIG. 134. END ELEVATION OF RECTIFIER FIELD MAGNET.
two of which are represented as cut away to show the shape of the poles
and cores. The circular yoke, poles and arms to support the bearings
are cast in one piece, thus avoiding joints in the magnetic circuit, and re-
ducing the machine work to a minimum. The pattern for the field casting
SINGLE-PHASE RECTIFIER
should be made in two pieces, the parting being made along the line AB
in Fig. 135.
The field casting is mounted on the face plate of the lathe one of 9
ins. swing will do and the poles are bored out to the required diameter,
2 l /2 ins., at the same time the arms for the bearings are finished to the arc
cf a circle 5 T 4 i ns - i n diameter. These may, however, be filed and fitted
by hand, if this is more convenient, but the method of boring is the more
accurate.
Fig. 137 shows a section through the armature, commutator and
bearings. The armature core is built up of laminated iron in the usual
way, and held together by heavy washers of wrought iron threaded on
the shaft at either end. A cast-iron core may be used, though it will, of
course, heat up more than the laminated one.
The commutator and collector rings are preferably made of copper,
though brass may be used. A piece of ingot copper can be obtained and
Ei.CC.
FIG. 135. SIDE ELEVATION OF RECTIFIER FIELD MAGNET.
forged into a circular blank suitable for turning up into rings and a com-
mutator. The commutator is turned as a solid cylinder of the required
section, and it is then cut into four equal segments with a milling ma-
chine, or with a sharp hacksaw and hand power, if a milling machine is
n8
ELECTRICAL DESIGNS
not available. The segments are then built up with mica insulation and
clamped in a brass sleeve fitting the shaft.
It is well to follow the design of arc machine commutators to some
extent and to allow about 1-16 in. air insulation between the segments at
the top. Thus the mica insulation will not be injured if sparking occurs.
Oil and copper dust should not be allowed to accumulate in the air spaces
thus formed,, as this would cause a severe short circuit.
Connections are made between the collector rings and commutator
bars with some strips of copper about 1-32 in. thick and y% in. wide, laid
in the armature slots, which have been made about 1-16 in. wider than
K^
/T; .'
H H a--T I
* :
I
u.
-r-
W^
V i i
0|1D
.1
10
'3
i S
iJt
3
FIG. 136. PLAN VIEW OF RECTIFIER FIELD MAGNET.
would otherwise be necessary in order to accommodate both the coils
and the connection strips. The back collector ring is drilled with % in.
holes at four equidistant points, two of these holes, the opposite ones, go
through the ring and part way into the front ring, the other two holes are
drilled only part way in. Some short pieces of copper wire, about No. 10,
are soldered one into each hole, the two wires from the front ring being
insulated from the other ring where they pass through it. The four wires
are soldered one to each strip, and these carry the current across the
armature and connect to the commutator segments.
SINGLE-PHASE RECTIFIER
119
120
ELECTRICAL DESIGNS
Diametrically opposite segments of the commutator are thus con-
nected to the same collector ring, and neighboring segments have be-
tween them the whole potential difference of the alternating circuit. It
is much better not to connect the copper strips permanently to the
commutator until the builder has decided where he wishes to place the
brushes ; the commutator may then be twisted around on the shaft to the
correct position and the connections made permanently. The brushes
may be placed wherever they will be most convenient, the only restriction
being that they must be 9.0 apart and must pass from one segment to the
next at the same instant that an armature tooth is exactly under a pole.
Fig 138 shows an armature. coil and the method of placing the coils
upon the core. The coils are wound on a form, and, after being taped,
I
w
_
_^
r )
\
\
1
i
' t
, i i i
H
PJC.BROSH HOU.PCR.
TWO F THtst .
ftC. BROSH HOUDE.R .
TWO Of THfcSfc.
Ru.
Of
-AMER. CLEC.
FIG. 141. DETAILS- OF JOURNAL YOKES, BRUSH COLLAR AND BRUSH-HOLDERS.
are slipped over the top of a tooth. The slack is then taken up by bend-
ing the ends down in a semi-circular shape and fastening them in this
position by screws which carry small fibre or hardwood bushings,
In addition to this the armature should be banded at one or two
points with No. 28 brass or German silver wire, small notches having
been turned in the core to receive the bands and allow them to come flush
with the surface of the core.
SINGLE-PHASE RECTIFIER 121
With an iron-clad armature like this, the clearance need not be
more than from 1-64 in. to 1-32 in. ; just how much it will be depends
somewhat on the builder's skill ; 1-64 in. clearance has been indicated on
the drawings, and with care taken in adjustment it should not be difficult
to obtain this figure. Fig. 139 is an end view of the rectifying commu-
tator. Fig. 140 is a flywheel pulley which it will be found advisable to
use in order to obtain smooth running.
Fig. 141 shows the bearing, brush yoke and brush holders. The
bearings, while not so simple in construction as some other designs,
have proven very satisfactory. The rings carry up a plentiful supply of
oil, and what runs out at the end returns to the well and does not fly off
outside the bearing and spatter the surroundings with grease spots.
In finishing the bearings the cap is first fitted and fastened by two
machine screws. The bearing is then placed in the chuck and bored out
to a diameter of y& in. clear through. At the same time the ends are
turned off 534 ins. in diameter to fit the arms on the field casting, and
the outside of the boss is turned off i>< in. diameter where it is to receive
the brush yoke. The sleeve which forms the bearing proper is turned a
tight fit for the central part of the box, then when the cap is screwed
down it will be held firmly in place. The grooves for oil rings can be
cut in the sleeve conveniently by mounting it eccentrically in the chuck
and using a thin cut-off tool.
The brush yoke is shown with two arms 90 deg. apart. Another
pair of arms and brushes might be added if it is desired to have more
current carrying capacity.
The direct current brushes had better be larger than the alternating'
current brushes, and the holders should have spring tension, unless a
very springy brush is used. Copper brushes are better for this purpose
than carbon, as they make better contact and cause less sparking.
Figs. 142 and 143 show the field and armature coils respectively, with
the forms upon which they are wound. The former is best made of hard
wood, and consists of a block and two flanges, all held together by two
wood screws and having a ]A in. hole through the center for placing on
a mandrel in the lathe. The block should be made a trifle larger than
the pole over which the finished coil is intended to go, and it should be
given a slight taper of about 1-16 in., so that it can be readily slipped out
of the finished coil.
Before beginning the winding a short piece of tape is laid in the long
sides of the former, with the ends left sticking out. When the form is
wound full these pieces of tape are tied tightly over the coil, and will
hold it in shape while the former is taken apart and the coil is receiving
122
ELECTRICAL DESIGNS
its wrapping of tape. After being shellacked and dried, the coil is placed
on the poles and hard wood wedges driven in between coil and pole, thus
holding it securely in place.
The same form may be made to serve for both field and armature
coils, if the field coils are wound first, and then the block reduced in
thickness from ^J in. to y% in., the armature coils having the same inside
dimensions as the field coils, but being only half as thick.
The fields are wound with No. 28 double cotton covered wire and
the armature with No. 23. If the coils are wound to the specified di-
mensions they will have nearly enough the required number of turns.
Fig. 144 shows a diagram of connections and Fig. 145 some e. m. f.
curves. For 100 volts the field and armature coils are connected four
in series, as shown. For 50 volts the coils may be connected two in
series and the twos in multiple. The armature terminals are tapped onto
f I tLP. BOBBIN.
Wm VIP TO sue.
c
WOOD Ftosnta.
FOR FIELD.
FIG. 142. A FIELD COIL AND THE
WINDING FRAME.
Woo 9
FORi-lfi-R.
AMER. C1.CC.
FIG. 143. AN ARMATURE COIL AND THE
WINDING FRAME.
COIL.
Mart: WiNP UP TO eiz&
WITH *2.3 P.CC
tHe collector rings, or what amounts to the same thing, placed across any
two successive commutator segments.
The connections on both armature and field should be such as to
produce alternate north and south polarity all the way round. If all the
coils have been wound in the same direction and placed on the poles the
same way, connect beginning to beginning and ending to ending, and
the polarity will be right. The machine will run at 1,800 r. p. m. on a 60
cycle circuit, and on a I25~cycle circuit it will have to make 3,750 r. p. m.
This it can easily do if the armature is well balanced as it should be.
Since the strength of the field has a considerable effect on the be-
SINGLE-PHASE RECTIFIER
123
havior of a synchronous motor, it is best to have an adjustable resistance
in the field circuit of this machine, so that the field can be adjusted until
the minimum armature current is obtained.
Referring now to the curves in Fig. 145, it is clear that if the rectifier
is running in synchronism and the angular position of the brushes is cor-
rect, the brushes will pass from segment to segment at the instant when
the e. m. f. curve reaches its zero value at the points, a, a, a, etc. As the
brushes in passing from segment to segment overlap two segments for a
brief interval, they form a dead short circuit on the alternating current
mains during the interval. This will not, however, result in any damage
if the e. m. f . becomes zero at the
same instance. If, however, the
brushes had been incorrectly placed
and commutation occurred at the
points b, b y b, etc. , an e. m. f . of value
equal to the ordinate at b would
be short circuited four times in a
revolution, and serious sparking
would result.
This state of affairs is easily
remedied by shifting the brushes,
which corresponds to changing the
angular position of the point of
commutation until a position such
as a a is reached, when all sparking
will disappear. If the armature
falls out of step, or if it is thrown
into circuit before complete syn-
chronism is reached, a short circuit
travels over every portion of the e.
m. f. wave, at a slow rate equal to
the difference between synchronous speed and the actual speed at that
instant, the result being a magnificent display of fireworks and probably
a fuse blown.
To obviate this latter difficulty a resistance or choking coil should be
placed in series with the alternating current end at the moment of start-
ing, and cut out when it is seen that the machine has settled down to
steady running. Such a resistance will not have any appreciable effect
on the small current drawn by the armature and field windings, but
should be of such a value as to limit the current to about 10 amperes,
should a short circuit occur.
FIG. 144. RECTIFIER CONNECTIONS.
FIG. 145. SOME E. M. F. CURVES.
I2 4
ELECTRICAL DESIGNS
The ordinary method of a synchronizing lamp is not easily applicable
here on account of the small size of the machine ; and, moreover, a little
practice will enable the operator to judge by ear the proper instant for
closing the circuit.
Thus by making slight changes in the connections, as already pointed
out, this machine may be used as. a rectifier on single-phase circuits of 50
or 100 volts and 60 or 125 cycles. The amount of rectified current which
may be drawn is not limited in any way by the horse-power capacity, but
will generally be limited only by the capacity of the transformer which is
supplying the current. Thus from 50 to 100 amperes may be drawn, de-
pending somewhat on the nature of the load into which the rectifier is
feeding current.
The machine may also be used as a self-exciting synchronous motor,
FIG. 146. END AND SIDE VIEWS OF THE COMPLETE MACHINE.
developing from i-io to % horse-power, according to the strength of
field ; and finally it may be driven by belt as a self-exciting alternator, sup-
plying either an alternating current or a rectified direct current, or both,
up to about 100 watts output.
Fig. 146 was reproduced from photographs illustrating a rectifier
built from these designs.
CHAPTER XIV.
UNIVERSAL ALTERNATOR FOR LABORATORY PURPOSES.
The design of the machine illustrated in the accompanying engrav-
ings was adopted for the following reasons:
i. Simplicity of construction by students in the engineering shops,
without special tools or dies; 2. Its similarity to a bi-polar dynamo so as
to illustrate one, two or three-phase-current generation,- but without 3
F
r
uJ-
-4K*
Am.Elfc.
FIG. 147. JOURNAL PEDESTALS AND BOXES.
low limit to the frequency; 3. To illustrate practically the effect of com-
bining e. m. fs. differing in phase, in a variety of ways.
To accomplish these objects, both the field and the armature were
made with poles, the latter having two more than the former. The field
126
ELECTRICAL DESIGNS
revolves and is of the C. E. L. Brown type. The armature is made of
sheet-iron rings held together by bolts between cast-iron plates. The
spaces between the poles of the armature were milled out after the rings
had been bolted together. The field is made of two identically similar
steel castings, each with five poles symmetrically spaced and pointing in
the same direction parallel to the axis. The field has thus ten poles,
alternating in sign, and the armature twelve. The following are some of
the dimensions :
Diameter of armature pole-faces, ioj4 ins. ; length of faces parallel
to shaft, 4 ins.; width of pole-faces, iy 2 ins.; pitch of poles on armature,
2.68 ins. ; depth of poles, JX in. ; net cross-section of pole in square
inches, 5.4. The armature poles have forty turns of No. 16 wire each.
Am.Eled.
FIG. 148. SECTION OF ONE
MAGNET POLE.
FIG. 151. COMPLETE MACHINE WITHOUT BASE.
The double air gap is % in. The field coil contains 1012 turns of No. 16
double-cotton-covered wire.
The armature coils were wound in reverse order from pole to pole
in the usual way. They are connected in pairs, and the terminals of each
UNIVERSAL ALTERNATOR FOR LABORATORY PURPOSES 127
pair are brought up to binding posts on a board bolted to the top of the
machine.
The machine will give i , 500 watts when connected in three-phase
zig-zag mesh fashion and driven at i , 650 r. p. m.
The completed machine is shown in Fig. 151, although the engraver
has left off the base-plate on which the armature and pedestals are
mounted.
An inspection of the diagram (Fig. 152) will show that diametrically
Opposite poles of the field are of opposite sign, while the corresponding
coils of the armature are similarly wound. Hence with a closed-coil
128
ELECTRICAL DESIGNS
armature the e. m. fs. balance exactly as with a bi-polar dynamo. If,
therefore, connection be made with the armature at two opposite points,
the current in the external connecting circuit will be alternating. Further,
two such circuits connected at right angles will convey currents in quad-
rature. By connecting at points 120 degs. apart, three-phase currents
will be obtained.
Again, the coils may be joined either in mesh or star fashion by
means of the binding posts at the top, and we may zig-zag across either
\vith two or three-phase connections, so as to connect opposite coils by
FIG. 152. SIDE ELEVATION OF THE MACHINE.
twos or threes with no-phase difference between the two opposite groups.
This connection, of course, gives the highest e. m. f.
It is evident that the phase difference from coil to coil is 30 degs. or
one-twelfth of a period. Hence the voltage for a given magnetic flux cut
per second, calculated in the usual way, must be first divided by V 2 to
reduce from maximum to virtual volts, and then the equal e. m. fs. gen-
UNIVERSAL ALTERNATOR FOR LABORATORY PURPOSES 129
crated by the several coils must be added geometrically with a phase dif-
ference of 30 degs. from coil to coil. Since the e. m. fs. of the several
coils are equal and differ in phase by one-twelfth of a period, the series
may be represented by a regular polygon of twelve sides (Fig. 150).
Hence, if E be the e. m. f. of one coil, the following will be the e. m. fs.
of the several groups of coils :
AC, E. M. F. of two coils, 2 E cos 15 = 1.93 E.
AD, " "three " E -f 2 Ecos 30 = 2.73 E.
AE, " " four " 2 E (cos 15 -f cos 45) == 3.346 E.
AF, " "five " E -f- 2 E (cos 63 + cos 30) = 3.73 E.
AC, " "six " 4 cos 15= 3.86^.
The phase difference between the coils reduces the e. m. f. of the
six in series on either side to 3.86 -f- 6 or 0.643 f what it would be were
there no such phase difference.
S ' g
Volts i- i-
g 3 g S S
G Am.Elec.
FIG. ISO. DIAGRAM OF E. M. FS.
FIG. 153. MAGNETIC CURVE AND EX-
TERNAL CHARACTERISTICS.
It will be seen from the subjoined table that the observed e. m. fs.
agree very closely with those computed from the foregoing equations.
One coil. . .
Two coils. .
Three coils.
Four coils .
Five coils. .
Six coils . .
Observed.
21
40.3
55-5
. 'Vi 69
75-5
77-5
Computed,
20.5
39-6
56
63.6
76.5
79.1
With about 2000 ampere-turns on the field coil, the following ob-
served voltages were obtained for the several connections described in
130 ELECTRICAL DESIGNS
the first column. The computed values are readily obtained from the
preceding expressions.
Connection. Observed. Computed.
Two-phase mesh 79 79
" " star 112 112
Three-phase mesh 69 68. (
" " star 121 nS.S
" " zig-zag 136 136.8
The alternator was driven by a motor on a power circuit and the
voltage varied a good deal. Some of the irregularities of voltage in the
generator are accounted for by the variation in speed of the motor.
Fig. 153 shows the curve of magnetization and the characteristic with
3 amperes in the field. The armature was connected as a closed coil, and
only a single alternating current was drawn from it. The total drop for
full load is ii volts; of these, about 4.5 volts are due to drop in the arma-
ture, and the rest must be set down to self-induction.
CHAPTER XV.
ONE-QUARTER HORSE-POWER SINGLE-PHASE INDUCTION MOTOR.
The induction motor described in this article was designed to be
built by amateurs, and the aim has been to make it simple and easy to
construct. It is designed for a single-phase alternating circuit of 104
volts and a frequency of 60 cycles per second. It has four poles, and,
therefore/ its synchronous speed would be 1,800 revolutions per minute.
The actual speed of the motor at load will be about 10 per cent, less than
the synchronous speed.
The primary or stator has a plain ring winding, and the secondary
or rotor a so-called squirrel-cage winding, consisting of bare copper
conductors, embedded without insulation in an iron core, all conductors
being connected at the ends. The bearing supports are provided
with an oil chamber, and either a ring or felt self-oiler may be used.
In making the calculations for the motor we will follow the method
much used in transformer calculation, which consists in assuming the
various losses, and from these losses determine the dimensions of the
parts in which they occur. We will set down- the efficiency of our ma-
chine at 60 per cent, and the power factor at 75, figures which obtain in
machines of this size on the market. The output being J4 horse-power or
1 86 watts, the intake will be 2- = 310 watts, and the total losses are,
therefore, 124 watts. We will make a preliminary division of this loss as
follows :
Primary C'R loss, 25 watts.
Secondary C 2 R loss, 15 watts.
Hysteresis, 40 watts.
Friction and eddy currents, 34 watts.
The primary current at full load will be ~- = 4 amperes.
From fne C'R loss and the current strength we may now find the resist-
ance of the primary circuit, R =-^|-= 1.56 ohms.
4
132 ELECTRICAL DESIGNS
Allowing a current density in the primary winding of 2,000 amperes
per sq. in., we find that the wire to be used is No. 16 B & S. The length
of this wire, which will have a resistance, when warm, of 1.56 ohms, is
354 ft.
The ohmic component of the e.m.f. in the primary circuit is
1.56x4 = 6.24 volts, and the induction or counter e. m. f., therefore
V 104' + 6.24--* X 104 X 6.24 X. 75 = 9942 volts '
In a single-phase induction motor the strength of the rotating field
is not constant, but fluctuates. We will base our calculations, however,
on an equivalent rotating field of constant strength.
Fig. 154 shows a diagram of the magnetic field in the motor. Let O
be the flux of magnetism that passes through the teeth of the stator, be-
tween A and B. Then, as the magnetic field rotates with a velocity of
FIG. 154. MAGNETIC CIRCUIT.
FIG. 155. DIAGRAM OF STATOR WINDING.
1,800 r.p.m., and the conductors are stationary, each conductor cuts this
number of lines 120 times per second. The mean value of the e. m. f .
^> \/ I 2O
induced in one conductor will be - , and the square root of the
mean square,
i 2O \f
io c
. But this value, multiplied by the number of
conductors, is the maximum e.m.f. induced in the primary winding 1 .
Representing by n the number of cycles per second (60), we may write :
(0
10'
We will allow a maximum magnetic density in the stator core of
25,000 lines per sq. in., and make the axial width of the core double
ONE-QUARTER HORSE-POWER INDUCTION MOTOR 133
its radial depth. The maximum magnetic flux at any cross section of
the core is -. The area of the cross section will, therefore, be *
2 50000
The periphery of this area is 6A and taking the length of turn 10
100000
per cent, greater than the periphery of the core, we have for it the value,
i ooooo
The number of conductors is equal to the total length of wire divid-
ed by the length of one turn :
L
6.6
I OOOOO
By substituting this value of C in equation (i), transforming and re-
ducing, we obtain.
i.8^ 2 .io 12
n 2 L z
We may no>w substitute as follows : E = 99.4, n = 60,
L = 4,248 (inches). This gives 3> = 275,000.
The magnetic cross section of the field core is = 5.5 sq. in.,
50000
/
, 6.6 \
and the length of one turn, 6.6 \ =. nins.
i ooooo
The number of conductors is - = 386. As we have four
poles, the number of conductors should be divisible by four, and we will,
therefore, take 384 conductors.
These conductors may be distributed in 32 slots, giving 12 con-
ductors per slot. We will wind these conductors 2 wide and 6 deep.
The diameter of No. 16 double cotton-covered magnet wire is 61 mils.
For insulation we allow 30 mils on each side of the slot and 30 mils at
the bottom ; also a clearance space of 30 mils below the surface of the
teeth. This gives for the dimensions of the slots a depth of 426 mils and
a width of 182 mils.
The cross sectional area of the core was 5.5 sq. ins., and the ratio of
axial length to radial depth, 100. We will make the core 3^ ins. long
by i 9-16 ins. deep, which approximately satisfies the above two condi-
tions.
The mean magnetic density in the teeth may be taken at 32,000 lines
per sq. in. There are in all 550,000 lines, each passing through the teeth
134 ELECTRICAL DESIGNS
twice, which is equivalent to 1,100,000 passing once. The magnetic
cross section of the teeth should, therefore, be = 344 sq. ins.
32,000
Dividing this area by 3.5 ins. the length of the teeth, we obtain 9.83 ins.,
the circumferential space taken up by the teeth. The circumferential
space taken up by the slots is 32 + .182 = 5.82 ins. The total inner cir-
cumference of the stator is, therefore, 15.65 ins., and the diameter practi-
cally 5 ins.
Some of the lines of force that pass through the primary circuit do
not enter the secondary circuit, but leak around it. The leakage co-effi-
cient is about 1.2. The total lines of force crossing the gap number,
therefore, - - = 916,600. By dividing this number by the surface
of the gap we obtain the mean magnetic density in it, = 16,600.
jj
The output of our motor is 186 watts, and the allowance for friction
30 watts. The total energy transformed from the electrical to the me-
chanical state is, therefore, 216 watts. The C 2 R loss in the secondary
is 15 watts. It is a well-known fact that the ratio of the motor speed to
the speed of synchronism is the same as the ratio between the energy
transformed in the secondary to the energy absorbed by it. The speed
of our motor at load will be, therefore, 1800 = 1680 r p. m
231
When a rotor of 5 ins. diameter, running at 1,680 r. p. m. develops
mechanical energy at the rate of 216 watts, the tangential force on its
circumference is 4. 34 Ibs. Now a conductor a inches in length, carrying
an alternating current whose I/ mean square value is I in a sinusoidal
field of mean intensity, B, has a mean force of Ibs. exerted
10,180,000
/"* 7" ft 7?
on it. If their be C conductors, the force is Ibs. We may
10,180,000
substitute the values of a and B and equate this expression to our cir-
. ... C/X 3.5 X 16,600
cumferential force, = 4.34,
10,180,000
Hence, + 34 X 10,180000 = -^
3.5 X 16,600
The number of secondary conductors, multiplied by the amperes
per conductor is, therefore, 760. We will put 15 conductors on the
secondary, which will give a current of a little over 50 amps, per con-
ductor. The loss in each conductor is one watt, and the resistance per
conductor, therefore, ohm. This includes, of course, the resistance
ONE-QUARTER HORSE-POWER INDUCTION MOTOR 135
of the soldered joints and the return on the ends, which cannot be ex-
actly calculated, but which may be taken at one-half of the total, so that
the actual resistance of the conductor is only - ohm. The size of
5000
wire of which a length of 33/2 ins. has a resistance of - ohm is No.
8 B. & S., which we will use for the cage winding of the motor. The
maximum density of our equivalent rotating field of constant strength
was 25,000 lines per square inch. We have to make a small allowance
for the insulation of the discs, and also take into account the fact that
our actual rotating field is fluctuating, which subjects parts of the core
to a considerably higher magnetic density. At 35,000 lines per square
inch the hysteresis loss per cubic foot at 60 cycles is 360 watts. There
are 129 cubic inches of iron in the stator core below the teeth, and the
hysteresis loss would, therefore , be 28 watts. In the teeth the
magnetic density will reach 60,000 lines per square inch. The hysteresis
loss at this density and frequency is 840 watts per cu. ft. There are 17
cu. ins. of iron in the teeth, and the hysteresis loss is, therefore, 8.5
watts. In the rotor the hysteresis loss is practically nil, as the frequency
of the reversal of magnetism is proportional to the slip of the rotor,
which at full load is only one-fifteenth the frequency of reversal of the
magnetism in the stator core. Our total hysteresis loss is, therefore,
36.5 watts, which is well within the limit of our allowance for it.
CONSTRUCTION OF MOTOR.
The construction of the motor will now be explained, reference be-
ing had to the accompanying drawings. As the motor is symmetrical in
both vertical planes, the drawings of the stator show it part in full view
and part in section.
The stator is built up of discs stamped from No. 27 transformer
iron. These discs have an internal diameter of 5 ins. and an
external diameter of 9 ins. They should be varnished on
one side with an insulating varnish, thinned down so as to
form a thin, uniform coating on the surface of the discs. The discs must
be dried before being assembled. Two disc-shaped brass castings of the
same internal and external diameter as the discs and a thickness of 3-16
ins. serve as end plates. These castings have four lugs on the outer
edge, strengthened by ribs. Through these lugs the clamping rods
pass.
A round piece of sheet iron, 10 ins. in diameter, should be procured,
from which to make a templet. Find the center of the sheet iron and
ELECTRICAL DESIGNS
lay out a circle of 4^3 ins. radius. Divide the circumference into four
equal parts and centermark the division points. Drill the center and the
division joints on the circle with a small drill. (About No. 40.) This
templet is clamped on the brass castings so that the holes on the circle
come about over the center of the lugs. With the same small drill used
before a hole should now be drilled through the brass casting. These
holes arc then enlarged with a %-in. drill. The clamping rods are ^4 -in.,
either Bessemer or cold-rolled steel rods, cut oil to 6^-ins. With a
34 -in. standard die a thread is cut on each end of the rod to a distance of
i% ins.
The disc may now be assembled. The clamping rods are not strong
enough to properly compress the discs, and this should be done under a
drill press or in a vice while the nuts are tightened up. Enough discs
must be put on to make the length 3^ ins. when tightened up. After
FIGS. 156 AND 157. END AND SIDE ELEVATION, HALF IN SECTION.
the core has been put together it should be chucked in a lathe and a
light cut taken out of it to make the inner diameter 5 1-64 ins. Care
must be taken not to make the bore too large, as this would much reduce
the efficiency and capacity of the motor.
The rotor is built up of discs of the same material as the stator, $ in.
internal diameter and 5 ins. external diameter. Two disc-shaped braes
castings serve as end plates. The hole in the center of these castings
should be finished to ^ ins.
The shaft is turned up from a piece of cold-rolled steel of xi-in.
ONE-QUARTER HORSE-POWER INDUCTION MOTOR
diameter. The middle part of the shaft is turned to such a diameter that
the discs fit over it and the ends so as to be a running fit in |a ^-in. hole.
A f6-in. thread is cut on each end of the middle part of the shaft.
The discs for the rotor need not be insulated but can be built right
up on the shaft and clamped by the two hexagon nuts shown in the
drawing". A hole is drilled into the shaft through the end plates, as
shown, and a steel rod or round spike is driven into the hole and saved
off. The rotor is now put into a milling machine, and with a J/&-in.
FIG. 158. FACE AND EDGE VIEWS OF AN END DISC.
milling cutter, 15 slots are cut 3-16 in. deep. The No. 8 copper wires
have to be driven into these slots. They are soldered at the ends to the
brass end plates., the solder being applied liberally and made to fill up all
around the wire. The rotor should now be put in a lathe and turned
down to a diameter, of 4 63-64 ins. Care must also be exercised here to
avoid taking too large a cut. While the rotor is in the lathe, the nuts
on the shaft are turned to y% in. in length.
We now take the bearing supports, and by means of our templet
drill the holes for the clamping rods. In fastening the templet to the
ELECTRICAL DESIGNS
castings a center should be made to coincide. In one of the castings
two y&-\n. holes for the binding screws are drilled through the
center of the bosses provided for this purpose. A piece of soft wood is
cut that will fit into the oil chamber. A 3^ -in. hole is drilled through
tills piece of wood, through which the shaft may pass.
The rotor is now wrapped with paper until it fits tightly in the
stator. It is put into place and the bearing supports are slipped on with
the wood in place in the oil chambers. The bearing supports are fast-
ened down by means of nuts on the clamping rods. Some babbitt metal
should now be melted in a ladle, the motor set on end and the outer parts
b b
FIG. 163, BOBBIN CORE,
FIG. l62. BOBBIN CORE SHEET.
FIG. 164. COLLAR.
over the outside six rectangular collars of vulcanized fibre, like Fig. 164;
these are 1-16 in. thick. Bend the flaps, c, Fig. 163, outwardly at right
angles with the wall of the "box," so that when the collars are finally ad-
justed into place along the outside, each end one will be held on by four
flaps, as indicated in Fig. 165. The surfaces of the flaps may be coated
with thick shellac varnish in order to keep them in place against the faces
of the end collars.
Next, mount the complete bobbin, Fig. 165, on its wooden core
block, Fig. 161, and prepare it for the primary winding. The partitions
I 4 2
ELECTRICAL DESIGNS
or collars must be adjusted at equal distances apart, as in Fig. 165, and
the spacing maintained temporarily by means of wooden blocks. Before
applying the winding, the seam where the edges of the fuller board meet
must be covered by a strip of the same material laid clear across the side
of the box in each of the compartments and secured in place by varnish.
Then the coils may be wound on, care being taken to observe rigidly the
prescription of Table II as to number of turns per section. The starting
and finishing ends of each coil or section must be of heavier wire than
that of the coil itself, and rubber-covered with an outer braid; No. 18
wire is a good size for the two smaller sizes of transformer, and No. 14
for the two larger ones.
After winding and securing the ends with heavy linen thread, tag the
ends, marking the inner or starting ends, "B," and the outer or final ends,
"F." Then tape each section thoroughly and lead the terminals of the
various sections lengthwise along the outside of the whole winding to one
FIG. 165. COMPLETE BOBBIN.
FIG. 166. CORE-CLAMPING DOGS.
end, securing these terminal wires to the surface of the structure by
means of a few extra turns of tape. All of the terminals should project
from one end of the complete winding, and they should be laid side by
side in regular order.
The secondary bobbin or box is made in exactly the same way as the
primary, but has only five collars instead of six, and is larger in size, as
Table I shows. After winding the secondary, tape it on the outside and
tag the ends, like the primary; varnish both heavily with either P. & B.
or shellac varnish, and set them aside to dry.
Next mount between two pairs of wooden dogs the requisite number
of core plates to make the proper thickness, as indicated in Figs. 166 and
167, drawing the dogs up snug, and wrap the core plates tightly with
three layers of plain linen tape along the portion between the dogs.
SIMPLE TRANSFORMER IN FOUR SIZES
'::'i:>""!l!l
FIG. 167. CORE-CLAMPING DOGS.
FIG. 168. JOINT CLAMP.
FIG. l6g. TABLET BOARD DIAGRAM.
144 ELECTRICAL DESIGNS
Varnish the outer layer of tape lightly; remove the pair of dogs, and slip
on the primary and secondary windings separately. The windings should
be so disposed that the sides out of which the terminals lead are on top.
When the windings are in place, put back the pair of dogs previously re-
moved and bend the core plates around the ends, lapping them as indi-
cated in Fig. 160; one-half of the plates should be carried around one
side and one-half around the other. The lapped joints must be tightly
clamped together by means of two wooden strips iy 2 ins. wide and 2 ins.
thick, drawn together with ^-in. iron bolts, as shown by Fig. 168.
To the upper ends of the four dogs which hold the core, screw a
tablet board to which the terminals of the windings are led and from
which the main transformer connections are made. The tablet board is
most easily made of "soapstone slate." which is merely a very soft grade
of light gray slate. Wood or fiber will not do and hard rubber would be
expensive. On it mount eight single-pole double-throw "baby" knife-
blade switches, two primary terminal blocks, eight heavy binding posts,
and two secondary terminal blocks, as indicated in Fig. 169. The sketch
also shows how the coils are connected to the switches and binding posts.
The transformers are all designed for 1,000 volts primary e.m.f., with
all of the primary coils in series, and 100 volts secondary e.m.f. with all of
the secondary coils in "series. The primary winding is divided into five
equal sections of 200 volts each, so that it can be grouped for 200, 400,
600, 800 or 1,000 volts. In grouping for 400 and 800 volts one section
must be left open, and in grouping for 600 volts two sections must be left
disconnected from the main primary terminals. The original object of
the writer in dividing the primary into 2OO-volt sections was to permit the
transformer to be supplied from either an ordinary i,ooo-volt primary
circuit or a 2OO-volt motor circuit. The reader will readily understand
that this arrangement is not compulsory ; the primary may be wound in
a single coil without any partitions, if desired, although it will be found
more reliable if two or three partitions be used to reduce the voltage per
section.
The secondary winding as designed is divided into two sections of
about 1 8 volts each and two sections of about 32 volts each.
No switches are used for making the various combinations
because too much space and complication would be required.
The connections are to be made between the various binding posts by
means of short lengths of heavy iron No. 6 or No. 8 gauge! In order to
avoid excessive ohmic loss the wire should fit the hole in the binding
post snugly. All of the binding posts should have two holes and binding
screws each.
SIMPLE TRANSFORMER IN FOUR SIZES
With primary switches all thrown inward, the primary sections are
in series for 1,000 volts between the terminal posts. With all of them
thrown outward the primaries are in multiple for 200 volts. Other com-
binations may be easily traced out. At the secondary end several com-
TABLE I. MECHANICAL DIMENSIONS.
Size of transformer watts . ...
200
20 ins.
I#
1%
rA
4 7 A
4 3 A
I*
tf
12^
4 7 /s
4/8
3 l /&
l /2
3J/2
I
xtf
i#
2^
500
22 ins.
2^
2^
10^
5*
4*
1
15^
4#
4ti
3 7 /*
%
4X
i
2
Itf
3
750
30^ ins.
2^
2^
5 T Jr
2^J
*6
1^/2
5*
5 T>
4 1
4 l /2
I
2
1%
3
IOOO
31% ins.
3
3
!2#
5/8
4^
^
l ^
4 7 /s
4*A
H
4%
I
2^
2
3K
Length of core plates
Width " "
Thickness of compressed core d
Primary Bobbin.
I ength of sheet C
Width " L
Length of finished bobbin
Width of one side S
Depth of flanges . . . .
Secondary Bobbin.
Length of sheet, C
Width " L
Length of finished bobbin
"Width of one side S
Depth of flanges
Core Dogs, D
Depth of core clamp e
Thickness parallel with the core
Thickness parallel with the bolts
Height foot to core, f
Length of foot g
TABLE II. ELECTRICAL AND MAGNETIC DATA.
At 1000 volts primary and 100 volts secondary; 133 cycles.
Output of transformer watts .
200
COO
IOOO
Primary Winding.
No. 26
No 22
No 20
No 10
Turns per section
640
2C.O
220
3,200
I, COO
I,25O
I,IOO
Depth of winding layers
20
12
II
128
3O
18
C 8 R loss, full load
10 1 A
.fix
Secondary Winding.
Size of wire B &S
No 16
No ii
No 10
No 9
Turns per small section
27
23
20
" large "
103
48
4O
35
320
ICQ
126
no
5
4
3
Resistance hot
2
O.2
o. 175
C 8 R loss, full load
8
oy
Hff
Losses, Full Load.
16
21.5
30^
Hysteresis.
_
8
IO.7
jT/
i
0.8
i
7
Total losses
28
25
33
40
QS53X
06 Y
i^O ELECTRICAL DESIGNS
binations are obtainable, as shown by Fig. 170; the most serviceable will
doubtless be found to be the one on the left of the sketch.
The tablet board must be kept enclosed by a box cover when the
transformer is in use. The fuse blocks and master switches should be lo-
cated at a little distance from the transformer, and the cover should never
be removed except when the primary switch is open.
CHAPTER XVII.
CONSTRUCTION OF A REACTIVE COIL.
It is well known, of course, that reactive or "choking" coils are used
in alternating-current work instead of ordinary resistance coils for the
purpose of reducing the e.m.f. in a portion of a circuit because they are
much less wasteful than resistance coils or rheostats. As the uses of re-
active coils are so diverse no one design can be given which will fit all
cases ; hence only one form will be described here in detail and rules will
be given by means of which anyone can modify the design to fit any case.
The reactive coil here described is designed for use in series with
one, two or three open arcs on a constant-potential circuit of 100 or no
FIG. 171. THE CORE.
FIG. 172. TEMPLATE.
volts, or with one lamp on a 50-55 volt circuit, or with three enclosed arc
lamps fed from a 200 volt circuit. Its dimensions and windings are based
upon a magnetic density of 30,000 lines of force per square inch in the
core and it may be adjusted to pass any current from y 2 ampere to 15
amperes. The apparatus is of the adjustable core type, and if desired, the
winding can be tapped at various points and the coil can be used as an
auto-transformer.
The core consists of rectangular plates of No. 27 transformer iron,
14 ins. long and 2 ins. wide, bent into the form shown by Fig. 171. The
thickness of the core must be 2 ins. The easiest way to assemble it is to
make an iron template like Fig. 172, of J/2-inch strap iron and bend the
successive plates of the core into shape over the template, one at a time,
148
ELECTRICAL DESIGNS
leaving them in position as they are bent. Clamp the strips to the tem-
plate as shown in Fig. 173 and bend the ends down without resorting to
any hammering whatever. The core must measure 2 ins. in thickness
when clamped tightly. As each strip is bent down into shape its ends
should be squared off with a pair of tinners' snips, so that when all are
FIG. 173. CORE STRIPS ON TEMPLATE.
bent the ends will all be flush, forming a laminated pole-face at each end
of the core, as indicated in Fig. 171.
When the last strip is in place, remove the template, replace the
clamps near the corners of the core and bind the core strips tightly to-
gether with heavy cord (at least 1-16 in. in diameter), winding a full layer
from bend to bend, and pulling each turn just as tight as the cord will
FIG. 174. WOODEN CLAMPING BLOCKS.
stand it. The best way is to take a couple of turns at one corner and tie
the cord ; then loosen the clamp and move it an inch away, setting it up
tight again ; wind the cord over this inch of space and move the clamp
another inch, continuing this procedure until the whole core is covered
between the two bends. After binding the core with cord in this manner,
cover it with two layers of insulating tape, carrying the tape around the
bend and out almost to the end of the right-angle poles. Then wind on
that portion of the core between the bends 200 turns of No. 8 double cot-
CONSTRUCTION OF A REACTIVE COIL
149
ton-covered magnet wire in four layers. When this is done secure the
ends of the core between three clamping blocks mounted on a base
board, as shown in Fig. 174.
The yoke which completes the magnetic circuit is somewhat similar
in form to the core just described, as Fig. 175 indicates, but the right-
angle projections at each end of the yoke are much shorter than those of
the main core. The exact length of these projections is immaterial ex-
cept that it should be not less than an inch and not more than two. The
yoke will preferably be built up in the same way as above described in
connection with the main core, and after it is bound together with twine
it should be mounted in the clamp shown in Fig. 176 in such a way that
the center of the spindle projecting from one side of the clamp will coin-
cide with the center of the yoke structure. The clamp jaw should be
made a snug fit for the yoke so that the bolts will not need to be drawn
up very tightly. The bolts must be insulated from the metal of the clamp
Nut
FIG. 175. THE YOKE.
FIG. 176. CLAMP FOR YOKE.
by bushings of either hard fiber or rubber in order to avoid forming a
closed circuit around the yoke, which would result in a heavy flow of
current through the clamp and bolts.
The spindle is seated in a bushed hole through the center of the
long clamping block shown in Fig. 174, so that the ends of the yoke can
be brought into alignment with the pole-faces of the main core. A
round-nose set screw through the upper edge of the clamping block will
serve to hold the spindle in any position to which it may be adjusted.
Two adjustments are available, one in a rotary direction about the cen-
ter of the spindle, and the other in a straight line toward and away from
the pole-faces of the main core. This latter adjustment is preferably made
by means of a thin nut fitted to a fine screw thread on the spindle, as in-
dicated in Fig. 176, the set-screw in the clamping block being used mere-
ly to secure the core in any position to which it may be adjusted. To use
the apparatus as a choking coil, connect the winding in series with the
lamp or lamps, adjust the regulating nut on the spindle so as to secure
the length of air gap specified in the accompanying table and secure finer
gradations by twisting the yoke into or out of alignment with the pole-
150
ELECTRICAL DESIGNS
faces of the main core until the exact choking effect is secured, when the
set-screw may be used to hold the yoke in that position. The length of
air-gap given in the body of the tables refers to each of the two gaps, not
the sum of the two.
If it should be desired to use the apparatus as an auto-transformer
the yoke should be brought into accurate alignment with the pole faces,
pushed up solidly against them and held in this position permanently by
means of the set-screw. Used in this manner, the winding will have to
LENGTH OF AIR GAP ; OPEN ARC LAMPS.
I Lamp;
Lamps on ic>4-volt circuit.
Ampere
52-volt
circuit.
I
2
3
4
&
A
A
%
6.6
X
A
TV
y*
7-5
*J
A
H 7 T
H
10
H
A
A
i#
12
A
A
A
*X
14
%
7
7T
A
i#
LENGTH OF AIR GAP J ENCLOSED ARCS.
i Lamp;
Lamps on 2Oo-volt circuit.
Ampere.
loo-volt
circuit.
i
2
3
4
5
I
A
X
6.6
7-5
1
A
. & --
10
A
A
TV
13
'X
TV
A
I
be tapped. Each turn of the winding will represent 1-200 of the e.m.f. of
the circuit, so that to operate a single 33-volt lamp on a 5O-volt circuit
one terminal of the lamp must be tapped into the winding of -the auto-
transformer 133 turns distant from the other terminal, or preferably at
2-3 the distance from one end, as indicated in Fig. 177. Fig. 178 indi-
cates the arrangement of the taps for three enclosed arc lamps on a 200-
volt circuit. Taps would be taken out at exactly the same points to sup-
ply three open arc lamps on a loo-volt circuit.
Should the reader desire to construct a reactive coil to suit any other
conditions, the following simple formulas will give the required dimen-
CONSTRUCTION OF A REACTIVE COIL
15*
sions and winding. For work on a 133-cycle circuit the choking effect in
volts for a coil built for the average conditions of practice is given by the
formula :
In this formula A is the area of the core cross-section and T is the num-
ber of turns of wire. In order to make this formula hold good the num-
ber of turns of wire must agree with the formula :
34 L - T (2)
C
= T
In this formula L is the length of the magnetic circuit within the
FIG. 177. CONNECTIONS AS AUTO-TRANS-
FORMER 1 ONE LAMP.
FIG. 178. THREE ENCLOSED ARCS ON 2OO-
VOLT CIRCUIT.
iron, including the yoke, 'and C is the current in the winding. For 60-
cycl,e circuits
o.n XA X T = E ...... ( 3 )
in which the number of turns must agree with the formula :
_, - = ^ :,- ..... (4)
The two joints between the yoke and the main core when they are in
actual contact are equivalent to about 36 ins. of iron under the conditions
assumed in formulas (i) and (2), and to about 52 inches under the con-
ditions upon which formulas (3) and ('4) are based ; therefore, in figuring
152 ELECTRICAL DESIGNS
the closed circuit either 36 or 52, as the case may be, must be added to
the actual length of the magnetic circuit in the iron.
The size of the wire to be used on a coil may be ascertained by al-
lowing 1,000 circular mils of cross section per ampere of current. In
designing a reactive coil the dimensions for maximum output should be
calculated, first assuming the yoke in contact with the main core, because
anything below the maximum effect can be obtained by adjusting the dis-
tance between the yoke and the main core.
As an example, suppose it were desired to build a coil to regulate the
impressed e.m.f. of a circuit of 6.6-ampere lamps fed from a 2OO-volt
transformer, on a 133-cycle circuit, the range of regulation or choking
effect being from 25 volts to 150. Transposing formula (i) into:
o.i8x r"
and assuming temporarily 50 as the number of turns, the area of the core
will be for the maximum reactive effect of 150 volts, 2p sq. ins. Making
the core 4x5 ins. will give this area.
Now transposing formula (2) to read :
34
and substituting for C and T the values of 6.6 and 50 respectively, we find
that the length of the magnetic circuit (if it were all iron) would need to
be 97 ins., which, of course, would be a ridiculous dimension. It must
be remembered, however, that an air-gap is equal to practically 1, 800
times its length of good sheet iron under the conditions assumed in
formulas (i) and (2) ; therefore it is only necessary to make the iron core
long enough to accommodate the winding and then insert an air-gap of
such a length as to bring the total reluctance equal to that of 97 ins. of
iron.
No. 12 wire will be large enough to carry the current, and the diam-
eter of this over the insulation is 0.092 in. ; 50 turns side by side, therefore,
will make a coil 4^ ins. long, which is not excessive. The total length of
the iron part of the magnetic circuit may be made about 20 ins., so that
the two air-gaps must be made equivalent to 97 20 = 77 ins. of iron.
As each inch of air-gap is equal to 1,800 ins. of iron, the total length of
air-gap required will be :
. _
so that each air-gap will be about 7-64 inch long in order to bring the
choking effect of the coil down to 150 volts with 6.6 amperes flowing
through the winding.
CONSTRUCTION OF A REACTIVE COIL 153
The foregoing- formulas and instructions are based on a magnetic
density of 30,000 lines per square inch in the core of a coil to operate on.
a i33-cycle circuit, and 52,000 lines per square inch in the core of a coil
to work on a 6o-cycle circuit. While the density can be carried somewhat
higher than this and the size of the core correspondingly decreased, more
satisfactory results will usually be obtained by employing the densities
here given, because with higher densities the hysteresis loss in the core
will cause it to overheat and jeopardize the coil. The two densities above
specified give the same core-loss, to wit : o .68 watt per cubic inch of
core iron>
CHAPTER XVIII.
THE CONSTRUCTION AND CALCULATION OF RHEOSTATS.
The resistance material of rheostats for the regulation of current or
potential in electrical circuits may be metallic wire, carbon or graphite,
or acidulated water. In the present article rheostats in which the first
named material is employed will only be considered.
The different conductor materials used in the construction of com-
mercial rheostats are iron, German silver and copper. Each of these
materials has advantages in particular cases. The advantages of iron are
cheapness and the ability to withstand high temperatures. German silver
lias a high resistivity or specific resistance and a low temperature co-effi-
cient. Copper is only used where large currents have to be carried, as,
for instance, in electro-plating work, where one dynamo supplies several
tanks requiring different voltages, and regulation is effected by inserting
resistance into the circuits requiring the lower pressure. In this case,
copper, by virtue of its higher conductivity, makes it possible to use
smaller conductors, thus facilitating the construction of the rheostat.
The table on page 155 gives the carrying capacity of tinned iron
wire under different conditions. The last column gives the length of
wire having a resistance of one ohm.
In designing motor- starting rheostats, the values given under the
heading "Safe current for one minute " should be used, while the carry-
ing capacities given in the other two columns apply to dynamo field rheo-
stats, motor regulators and such other rheostats as have to carry current
continuously. No definite resistivity and carrying capacity can be
assigned to German silver, as it is an alloy, and different makers use
different proportions of the elements. In the tables given by Matthiesen
the resistivity of German silver is given as 2.2 times that of iron. For
the same rise of temperature a German silver wire would, therefore, carry
about two-thirds the current of an iron wire of the same size. Most
commercial German silver has, however, a specific resistance higher than
that indicated by the above ratio.
CONSTRUCTION AND CALCULATION OF RHEOSTATS
155
The wires of rheostats are mounted in a number of different ways.
They may be embedded in enamel or some other refractory insulating
material; they may be wound on a plate or slate; they may be wound
on a framework of iron rods insulated with asbestos, or on
Size of Wire,
B. & S.
Safe Current in
Wood Frame.
Safe Current in
Iron Frame.
Safe Current for
One Minute.
Feet per
Ohm.
S
17.4
20.3
43-6
250
9
14.6
17.1
36.6
173
10
12.3
14-3
30.8
137
ii
10.3
12.0
25.8
108
12
8.7
10. 1
2L7
86.4
13
7-3
8.5
18.3
68.5
14
6.1
7-i
15-3
54.3
15
5-i
6.0
I2.Q
43-r
16
4-3
5-0
10.8
34- r
17
3-6
4.2
9-1
27.1
18
3-00
3-5
7-6
24.3
19
2.52
2.9
6.3
16.5
20
2.17
2-5
5-4
13-5
21
1.82
2.1
4.5
10.7
22
1-53
1.77
3-8
8-49
23
1.28
1.49
3 2
6.73
24
i. 08
1. 20
2-3
5134
insulated metallic spools with layers of asbestos between the
layers of wire. Finally, the wire may be wound into coils which are
stretched between insulators on an iron frame or in a frame of insulating
material. When the wires are embedded in enamel, they are placed on the
surface of and in close proximity to a cast-iron base plate which assists in,
radiating the heat. Slate is also quite a good conductor of heat, and
plates of slate are often used for smaller rheostats. Spool-wound coils
of wire sometimes present an advantage where the rheostat is only used
for a short period at a time, as, for instance, in motor-starting rheostats,
as this method of winding permits of getting a large amount of wire into
a small space, and the capacity of the rheostat under such conditions de-
pends more on its capacity for taking up heat than on the radiation. When,
spiral coils are employed, they are generally placed in a case with open-
ings to facilitate the circulation of air.
The diameter to which spiral coils of wire are wound varies with the
size of the wire. If for a given size of wire the diameter Is taken
too large, the coils must be stretched considerably to obtain the necessary-
stiffness. No. 24 (B. & S.) iron wire may be wound into coils of y 2 in.
diameter, while No. 16 may be wound into coils of from 4 m - to % in.
diameter, and other sizes proportionally. The wires arc wound close on.
156 ELECTRICAL DESIGNS
a mandrel in a lathe and are stretched as they are put in position. Fot the
larger sizes of wire a stretching of 20 per cent, is sufficient, while coils of
No. 24 of 6 ins. or more in length must be stretched to about double their
length. Some manufacturers place asbestos tubes inside coils of smaller
wire, which, as they stiffen the coils, permit coils of larger diameter and
reduce the stretching required.
Dynamo Field Rheostats. Shunt and compound-wound generators
are generally regulated by means of a rheostat in the shunt field circuit.
In Figs. 179 and 180 are shown two dynamo field rheostats, both of which
are of fire-proof construction. The form shown in Fig 180 is intended for
small machines, while that at Fig. 179 is adaptable to any size.
FIG. 179. DYNAMO FIELD RHEOSTAT. FIF. l8o. DYNAMO FIELD RHEOSTAT.
In the factory it is generally easy to experimentally determine the
resistance required to cut down the voltage of a machine to the desired
lower limit. Cases may, however, arise, where this is not handy, and the
resistance can then be calculated, provided the excitation-voltage curve
of the dynamo and the resistance of its field are known. The calculation
may be illustrated by a practical example. The main curve in Fig. 181
is the excitation-voltage curve of a i.5~kw 55-volt generator.
The machine is run at such a speed that without any load and without
I any extra resistance in the field circuit, it generates 65 volts. A rheostat
is required which will cut down the voltage to 40. The field resistance is
36.4 ohms.
From the curve we see that at 65 volts the exciting ampere-turns are
6,200, while at 40 volts they are only 2,300. The ampere-turns are pro-
portional to the voltage applied to the shunt. When 65 volts are being
generated, the voltage at the terminals of the shunt is 65. At 40 volts it
/^ to which are fastened, by
shellac, a plane glass mirror from y 2 in. to 5-8 in. in diameter, the upper
set of steel needles, and a thin glass fibre about 1-50 in. in diameter.
The object of the mica vane is to dampen the vibrations and help bring
the needle system to rest and, for this purpose, it should be as large as
the space in which it swings will allow. The space which contains the
needle system and the silk suspending fibre should be made as air-
tight as possible to prevent external gusts of wind or currents of air
from entering and causing the needle system to vibrate and shake. When
finished and set up ready for use, the mirror should come opposite the
glass covered hole in the front of the instrument. It is difficult to make
a good mirror and it is more satisfactory to purchase one from an electrical
instrument maker. However, if the reader desires to make one he will
find the necessary directions on page 178.
The glass fibre used to connect the lower set of needles to the mica
vane on which the upper set of needles is fastened, must be perfectly
straight and very light. By a little practice a good one can be made
from a small glass tube by heating it over a fish-tailed gas burner and
THOMSON ASTATIC GALVANOMETER
191
drawing it out. Make a number of these and select the best one. To
the top of this secure, by shellac, a minute hook made of No. 36 or No.
38 bare copper wire, and to the lower end glue a very thin circular piece
of mica upon which is glued one set of steel needles. The two pieces
of mica must be in the same plane. This needle system should be as light
as possible (not over ten grains), to make its moment of inertia small.
Other things being equal, the smaller this moment of inertia the quicker
it will come to rest.
The steel needles may be made from steel piano wire about 1-40 to
1-50 in. in diameter, or No. 8 guitar string. The temper should first be
drawn and the wire straightened and cut into convenient lengths about
Battery
FIG. 219. METHOD OF MAGNETIZING NEEDLES.
6 ins. long. These pieces must then be tempered glass hard. This
should be carefully done, as the entire piece should be of the same hard-
ness. From this wire cut or break off twelve pieces of the proper length
(as indicated in Fig. 214) for the needles and secure them in place with
shellac, six above and six below, being careful to have a thin air space,
about 1-50 in. between each needle, so that they will not touch each other.
The best way to magnetize the needles and obtain the astatic system
that is one set magnetized equally but oppositely to the other is to mag-
netize the set by an apparatus shown in Fig. 219. A and B are soft-
iron rods, 7-16 in. in diameter. C, D, E and F are soft-iron pole pieces
between which the needles to be magnetized are placed, as shown. G
and H are two coils, each consisting of 350 turns of No. 18 B. & S.
copper wire. The coils must be connected in series so that the magnetic
192 ELECTRICAL DESIGNS
potential of each will produce a magnetic flux and poles as indi-
cated in the figure. The trough is merely a variable liquid resistance
containing some liquid, such as salt water or a solution of washing soda,
and two metal plates or electrodes, one of which can be moved from one
end to the other, so that the current may be varied from o to about 5 am-
peres. Any other convenient variable resistance which will do this may,
of course, be used.
After making connections and adjusting the strength of the solution
so that the proper current can be obtained, place the needles between the
pole pieces, as in Fig. 219, and put on the full current of 5 amperes and
then gradually and slowly reduce the current to zero. Repeating this
process several times should magnetize the needle sufficiently, and if the
steel were properly hardened, it may never need remagnetizing.
FIG. 22O. TELESCOPE AND SCALE.
For suspending the needle system, get a single fibre of unspun silk at
least 10 ins. long. To remove all initial torsion, hang this up for a day or
so, having fastened to its lower end a small weight of non-magnetizable
material, such as a brass screw. Pass this silk fibre through the 1-32 in.
hole in B, Fig. 212, and secure one end to C, by tying it through the 1-32
in. hole in that piece. One or both of the coil cups being removed, work
the free end of the silk fibre down through the glass tube. Fasten this
free end to the small hook on the upper end of the needle system with
shellac. When, the shellac is dry,, wind up the fibre on C, until the lower
needles hang in the center of the coils. Then level the galvanometer
until the lower needles hang midway between the two coils and turn
perfectly free. With both coil cups in place and the galvanometer prop-
erly leveled, the needle system should still vibrate free and smooth. It
will depend upon the resultant polarity of the astatic system whether the
THOMSON- ASTATIC GALVANOMETER 193
instrument should be set facing- East cr West. This can best be deter-
mined by trial.
If a plane mirror is used, a telescope and scale for observing the deflec-
tion of the needle system is more convenient than a lamp, scale and lens.
A telescope, suitable for this purpose, can be purchased for $1.80 or less,
and a one-half meter scale on cardboard can be obtained for twenty-five
cr fifty cents. If a terrestrial telescope is used, it may be greatly im-
proved, for this purpose, by removing the set of rectifying lenses, in
which case the figures on the scale must be reversed and inverted. Fig.
220 shows a suitable stand for this purpose, which may be easily made.
It is customary to place the telescope and scale just one meter from the
mirror. The rear support for the telescope and the board upon which
the paper scale is fastened are arranged to slide up and down for adjust-
ing. The scale should be as much below the level of the mirror as the
object glass of the telescope is above it.
Quite a number of galvanometers of the above type with the telescope
and scale have been in regular use in the electrical laboratory at Lehi^h
University for the past four or five years. There are also several four-
coil galvanometers of a similar design, whose resistances run up as
high as 5,000 ohms, but for general use the two-coil instruments are sen-
sitive enough. Considerable cf the work on these galvanometers was
done by student's taking the electrical course during their third year.
CHAPTER XXIII.
A CHEAP TESTING SET.
The Wheatstone bridge is one of the most valuable instruments that
can be devised, not only for its primary office of measuring resistances,
but for testing faults of any kind. Most beginners are not aware that a
convenient substitute can be made that will answer in many cases, and the
expense of which need not exceed two or three dollars. Following a de-
scription of such an easily made apparatus is given,, and it may be of in-
terest to those anxious to possess a testing set and who do not feel that
they can afford the large sum usually asked for such an outfit.
Select a board of some dry wood and shape it nicely to the dimensions
of 42 ins. long by 8 ins. wide. Procure some flat copper rod which is at.
least y 2 by y\ of- an inch in sectional area. Of this there should be one
continuous bar 36 ins. long, and two shorter pieces about 3 ins. long.
Then secure a good straight piece of German silver wire, about No. 14
B. & S. gauge. Go over this with a micrometer, testing it at every inch
in its length, and be sure that the piece selected is of uniform diameter
and has no nicks or marks of any kind.
With a suitable size of drill bore holes 34 of an inch from the ends ci
the short pieces of copper rod, and J/ of an inch from the flat sides, as
shown in Fig. 221. The wire should now be soldered into the copper
block, care being taken that it comes exactly flush with the block, no
drops or beads of solder appearing around the joint. To do this, the
'better way is to heat the copper block very hot and thoroughly tin the
inside of the hole and fill it with solder, which, if the block is hot enough,
will remain melted. Having previously tinned the outside of the wire,
insert it in the hole thus prepared, pushing it straight in until the proper
distance has been reached. Be very careful not to pull it outward unless
the -solder does not fill well around the hole where it enters, because any
outward pull of the wire would draw a meniscus of solder around it which
would be undesirable. The length of wire between the two connecting
blocks, as shown in Fig. 222, should be exactly one meter long. This
should be most carefully adjusted for accuracy.
CHEAP TESTING SET
195
-"
M
FIG. 221. COPPER BAR.
FIG. 223. METHOD OF FASTENING TERMINAL BLOCKS.
FIG. 224. WEIGHTED INDEX.
-<- 65" ja
FIG. 226. IMPROVED BRIDGE.
ELECTRICAL DESIGNS
The copper blocks are provided with binding screws at their terminals
for the convenient insertion of known and unknown resistances, and are
secured to the board as shown in Fig. 223 by screws passing up from bc-
iieath. The blocks carrying the wires are arranged to hold it so that it
will be free from all kinks and twists, but not to put any mechanical strain
upon it, which might change its resistance ; in ether words, the wire
should be straight, not stretched. A little platform of thin wood or paste-
board should now be built up underneath the wire, so that it will support
it throughout its length, and on the top of this strip of pasteboard should
be pasted a meter scale ; if the wire has been accurately adjusted, it will
exactly fit between the terminal blocks. If these directions and the draw-
ings are carefully followed, the result will be a bridge of considerable
range.
To use this instrument, a source of e. m. f. is connected to the end
blocks and a galvanometer, or, as shown, a telephone receiver, is con-
nected to the long middle block, and its other terminal to a weighted
index, which can be moved along the board between the wire and the
long copper rod and make contact with the former at any point desired.
This index is preferably made cf a brass rod, one end of which has been
filed up to a V-shape and drawn out to a sharp point, as shown in Fig.
224. The other "end cf the brass rod is conveniently shaped up into a
binding post, to which the detector terminal may be atttached. The
weight may be a piece of lead cast about it and suitably shaped. The
sharp edge of the brass should be very soft in order not to mar the wire
at the point where it rests upon it, and for that purpose should be an-
nealed by heating it in a flame and allowing it to cool slowly. A few
standard resistances should be provided ; one, ten and one hundred ohms
will be sufficient. These may be conveniently made by measuring off the
proper amount of insulated wire with a testing set, coiling it about a small
wooden block and inserting it in a short piece of large brass tubing, whicli
is then filled with parafnne. These coils may be made of moderately
fine wire and stout, projecting terminals of as nearly no resistance as may
be, led out through the paraffine for purposes of connection. Fig. 225
shows a section of such a coil prepared for use.
The known resistance is selected as nearly as possible equal to the tin-
known resistance to be measured. For instance, if we are measuring a
series of field coil we know its resistance to be at most but a fraction of an
ohm, consequently we should use the lowest resistance coil that we have,
which in this case would be one ohm. Similarly, in measuring a shunt
field coil whose resistance may be 16 or 20 ohms, we should use the ten-
CHEAP TESTING SET i;7
ohm coil, and so on. The known resistance is connected across the bind-
ing 1 posts at one end of the bridge, and the unknown resistance is similarly
connected at the other. A source of e. m. f., two or three dry cells will
answer, should be connected from A to B. The weight index is adjusted
along the wire for a position where no deflection of the galvanometer oc-
curs, and the point is read on the meter scale.
If a telephone is used instead of a galvanometer, it is best to interpose
a key in the circuit, as shown in Fig. 222. The procedure is to then
seek a point on the scale where the opening and closing of the key pro-
duces no click in the receiver.
Multiply the known resistance by the length of the wire on the un-
known side and divide it by the length of wire on the known side, and the
result will be the value of the unknown resistance.
The bridge that has just been described is rather a clumsy construction,
being nearly 4 ft. long, but it can be modified as follows, and the result
will be more convenient :
A board 12 ft. x 14 ins. is provided with blocks of copper, as shown in
Fig. 226. Four wires, each 25 centimeters long (about 10 ins.), should
be soldered into them, as shown in the diagram, the blocks at the end be-
ing shaped as shown so as to be brought up conveniently near a third
one. Assuming the blocks to be of negligible resistance, the wire must
then be divided into four lengths of 25 scale divisions (one meter) each,
and thus a more compact and satisfactory instrument will be obtained.
CHAPTER XXIV.
CONSTRUCTION AND USE OF A PHOTOMETER.
In the photometer here described, and for the construction and use of
which directions are given, an attempt has been made to gather into one
instrument the good points of a number, and at the same time to avoid
their defects. In this an especial indebtedness is owed to a portable pho-
tometer made by the Electric Motor and Equipment Company, of New-
ark, N. J., which must be here acknowledged.
For the construction shown no materials are required which are not
readily procurable at small cost, nor are there any processes of manu-
C*\
FIG. 227. THE PHOTOMETER COMPLETE.
facture involved which are beyond the capabilities of a man with very
simple tools and ordinary mechanical ability. In many things some sim-
ple change might make a more finished instrument, but more machine
work would be involved. Such modifications will readily suggest them-
selves to anyone undertaking the construction.
A general drawing of the instrument is given in F,ig. 227. The frame,
which supports everything, is of J^-'m. white pine 3 ins. wide, set together
in an open rectangle 5 ft. io l / 2 in. long by 6^4 ins. wide over all. This
frame may be supported at a convenient height by brackets on any side
wall. At the left is an argand gas burner, in front of which is a board
with an adjustable slot cut in it, the slot being directly in front of the ar-
flame. At the other end, on a suitable stand, is the incandescent
CONSTRUCTION AND USE OF A PHOTOMETER
199
lamp whose candle-power is to be measured. Between them is a frame
on which travels a car containing a piece of paper set in a plane at right
angles to a line joining the two lights. This paper has a grease spot on
it, and in use the car is moved backward and forward until a place is
found where the grease spot disappears. The candle-power of the lamp
is then read off on a scale immediately below the car. For adjusting
the e. m. f. on the lamp terminals a rheostat is provided, shown between
the incandescent lamp and the track for the car at X in the drawing.
The instrument is in principle a Bunsen photometer.
In Fig. 228 the car is shown in detail. The car proper is of wood,
open in front, closed at the back, with ends cut away in openings 2 l / 2 ins.
in diameter. Half way between these two ends is the screen seen in
x
-\
s
L
/'i-i
^,
V
/B,
a \
frv
fl/,
\ 0,1
fp
}| ! *
" '' l(
i
\
\\
n *?
c
^*-
~r
V-
y'fll
H
J~7^
T
*f i: . :; E
;
^vji*$^ > '
IT
/ ^
j
|
D;
G
1
st/fi/f ..
* i
10.5
29.055
29/T
25
30.000
30
10
30.384
3off '
9-5
30.790
soli
9
31-219
3i ft
8.5
31-672
341
8 r"
The candle-power of 8-c.p. lamps can readily be obtained with this
photometer also. Put a standard 8-c.p. lamp, at some marked voltage,
in the lamp socket ; put the car at 16 candle-power and adjust the slot at
the argand burner until the slot disappears. Then proceed to measure
the 8-c.p. lamps as though they were i6's, halving the value of candle-
2 o8 ELECTRICAL DESIGNS
power on the scale in each case for their real candle-power. A similar
method may be used also in determining the value of the candle-power
procured from 32-c.p. lamps. Use a standard 32-c.p. lamp as in case of
the standard 8-c.p. above, and when readings are made on the candle-
power scale double the value obtained in each case. Unless the gas used
is intrinsically of very good candle-power and the argand burner is put
very close to the slot board, this last cannot be done since the gas burner
will not give enough light. If it is a possible plan carefully avoid any
flickering edges of flame showing through the slot when it is viewed from
the car position. The following alternative method of measuring 32-c.p.
lamps will always be satisfactory.
Set up a second lamp socket (number 2) fifteen (15) inches to the
right of the one already provided in the regular construction, or seventy-
five (75) inches from the slot. Construct a second scale (number 2) under
the one already made (number i), to be used only when the lamp under
test is in the more distant socket.
This scale will be constructed by measurements given in the table
on page 207, all measurements being from the slot before the argand
lamp.
Use i6-c.p. standard in socket number I, put the car at 16 candle-
power on scale number I, and set the argand lamp by it so the spot,
disappears. Then put the 32-c.p. lamp, whose candle-power is desired
in number 2 socket, and measure its candle-power in the usual way,
but reading its candle-power on number 2 scale. If the two sockets are
connected in parallel they will be ready for operation alternately at any
time, the same instruments and rheostat being used without any change
in connections.
CHAPTER XXV.
CONSTRUCTION OF A SIMPI^ STORAGE BATTERY.
The accompanying engravings show the construction of a plate and
a single cell of storage battery of the Faure type, which may be built with
no tools beyond a pair of heavy tinners' shears, a small punch and a slit-
ting saw of the sort used for cutting thin metals.
Each plate is made up of twelve strips of lead cut to the shape shown
in Fig. 236. This strip is 0.075 m - thick, 8 ins. long, over all, and ]4> in.
wide. Midway on each edge is cut a square slot, s, 3-16 in. wide and
deep ; the ends of the strip are cut down to 3/2 in. in width for a distance
of % in. from each extreme end, and two l /% in. holes are punched in the
narrow part of the strip, as shown. The edges are folded along the dot-
ted lines until the end view of the strip looks like B, Fig. 236.
Next three rubber forks, like D, Fig. 237, are provided for each plate
(not each strip, but each group of twelve strips). Each fork is made
A B.
FIG. 2^6. SHAPE OF STRIP.
from a strip of hard rubber, 3-16 in. thick, ^ in. wide, and 75/2 ins. long;
the slit down the center must be just wide enough to admit the thickness
of the lead strips forming the plate (A, Fig. 236) with no "lost motion,"
and must stop exactly i in. from the lower end of the rubber. When
these rubber forks are ready, a plate is built up in three of them as fol-
lows : Cut one of the lead strips in two longitudinally, exactly down its
center, and assemble the strips on edge in the rubber fingers so that the
end view of the lead part of the structure looks like E, Fig. 237. Fig.
238 shows the face view of the complete plate. The halves of the strip
that was cut longitudinally go at the top and bottom of the plate, and are
indicated by a and b, Figs. 237 and 238.
210
ELECTRICAL DESIGNS
After the strips are assembled in the retaining forks, D, D, D, tie the
upper ends of the ringers tightly with lead wire so that they clamp the
plate ; this tie-wire should go just above the top strip, at s, and it is im-
perative that no material other than lead be used, unless very strong,
short rubber bands are obtainable, in which case they may be used. Then
cut two strips of lead, C, c, each y in. wide, and of the same thickness as
the lead strips composing the plate ; the longer one, C, is 8^ ins. long,
and the other one, c, 6j^ ins. long. Rivet one of these to all the ends of
the plate strips at one extremity of the plate and the other one to the ends
at the opposite extremity, using lead rivets. The engraving shows holes
punched in the connecting strips, C, c, to correspond with those in thfe
>
^*
.1
e o-
c
7~]
c
o o
!
Tg
!
o e
TT]
O
13
s TI
FIG. 237. RUBBER FORK
AND PLATE EDGE.
FIG. 238. COMPLETED
ends of the plate strips, A. Bend the upper ends of the long connectors,
C, about half an inch from the end, at right angles to the main body of the
strip ;, the bending point is indicated by. x y Fig. 238. The edges of the
plate strips which were bent at an acute angle to the body strips (B, Fig.
236) are designed to serve as shelves to hold the paste or active material
which is to be applied to each plate. The process of pasting will be de-
scribed later on.
After the paste has been put on the plates and has hardened, nine
plates (four positive and five negative) are assembled in a glass jar, as
CONSTRUCTION OF A SIMPLE STORAGE BATTERY.
211
shown by the plan view, Fig. 239, where A is the lead portion of the out-
side plates; P, the paste, or active material; D, the rubber retaining fin-
gers ; /, the glass jar, and C, c, the vertical connecting strips. The strips,
C, are shown coming straight upward instead of bent over, to avoid ob-
scuring the view of the plate ends. The rubber fingers, D, in addition
to holding the twelve strips composing each plate in their proper posi-
tions, serve to prevent the plates themselves from buckling and shifting in
the jar. The edges of these rubber ringers are shown slightly separated,
but as a matter of fact, they must abut each other so as to keep the plates
firmly in position. The jar, J, is rectangular, and should be 8*4 ins. long,
5)4 ins. wide, and ^/ 2 ins. deep, inside measurement. None of its di-
mensions can be smaller than specified ; if the width or length be greater,
a strip of wood may be used to bring it down to the figure desired. A
c -.
(!*--
ft--.
ct-.
c
c t..
c*.
ct..
c - . .
P k p
/ / K I
|
r \
^'J'^-^ji <^^,
..-.>.- ^ft&Z*M: ;>%;'?,
ffiMfy'jM'%?&$ffi$S?;>$)Z'j
teS:;.:^^St
HT"
--C '
-c-
t -..' v-' -""-^'f.^Si-^^i-^
$ffiffitj^'^"w%*yy/fi'$
'w?#$&i$<3y*'ffifrp*$ s * fit
ffii.fay!,yfr '/,);:'/ = * ^JXTt'
tn
--c.
-c-
--C
~ r-rr^TT^. ;..-^,4^
Iffa^j^^fr^-v-l
-::v;.-^^-^^.-\v;i;^/----.-.-v.^:vi.--^
..v.vmir.'i
.-c
i-C-'
t&
\
V
* * >
- AMR
FIG. 239. PLAN VIEW OF BATTERY CELL.
sheet of soft rubber must be laid against each end and side wall of the jar,
and the plates must fit snugly against the rubber sheets so that they will
not shift in the jar ; soft rubber must be used for the wall sheets in order
that the plates may expand without tending to buckle.
As above stated, each cell contains four positive plates and five nega-
tives; the two outside plates are negative, and have paste on only one side
each, as the sides next to the wall of the jar are not active. The terminal
strips, C , on the right hand end of the cell are all negative; those
marked C + on the left hand end are the positives, as their signs indicate.
2-2 ELECTRICAL DESIGNS
The short strips, c, serve simply to connect the parts of the plates to each
other, and should not project above the upper edges of the plates.
All the positive ends, C +, at the left are riveted to a lead strip, T,
which is laid along the top of the bent-over ends of the strips, C +. The
negative connectors are similarly riveted to the other strip, T 2, and these
two strips form the terminals of the complete cell. The arrows at the
corners of the jar indicate the direction in which the ends of the terminals
are led from the cells to connect to an adjoining cell or to leading wires,
as the case may be.
Each cell of battery of the above dimensions, when properly pasted
and "formed/' will give an electro-motive force of about 2 volts during
the greater part of its discharge, and it may be discharged at the rate of
18 to 20 amperes. To operate any of the small motors described by the
writer on pages i to 14, inclusive, four of these cells will be required
(The battery winding given in each case must, of course, be used on the
motor.) After the cells have been "formed" as described below, they
may be kept charged sufficiently for light, intermittent service by con-
necting up 10 cells of gravity battery in series with the four storage cells,
the copper terminal of the blue-stone battery being connected to the pos-
itive terminal of the storage battery. This connection may be left per-
manently on, during ihe use of the storage cells as well as when they are
idle, the only attention necessary being the replenishment of the gravity
cells at comparatively long intervals.
The plates of the storage cells are pasted, the positives with a thick
paste made of red lead and dilute sulphuric acid, and the negatives with a
similar paste made of litharge and dilute acid. The acid should be one-
tenth concentrated sulphuric acid and nine-tenths water, and the water
should be distilled ; the proportions of one and nine parts are by weight,
not volume. In mixing, always pour the acid into the water, never the
reverse. The pastes must be mixed with wooden spatulas in glass or
earthenware vessels, and should be so thick (containing so little dilute
acid) as to appear almos't powdery. The pastes are applied to the sides
of the plates and pressed firmly in with the spatulas until the surface of
the paste is flush with the edges of the little shelves ; the entire surfaces
of the lead strips, except the edges of the shelves, must be covered evenly.
The best procedure will be to take all the positive plates first ; lay them
flat on a board, and apply red lead paste to one side. Set them aside
and mix the litharge paste (in a separate vessel and with a separate spat-
ula), and then treat one side of all the negative plates. When the plates
are all dry, turn them over and treat the other sides, being careful not to
jar out the paste already on the under sides. It should be remembered,
CONSTRUCTION OF A SIMPLE STORAGE BATTERY. 213
too, that two of the negative plates in each cell are to be treated on one
side only the side which comes next to the neighboring positive plate.
When the plates are all pasted, assemble them in their cells, as de-
scribed above, and then rivet the ends of the connectors, C -f- ,and C ,
to the horizontal terminal strips, T and T2. Connect the positive termi-
nal of one cell to the negative terminal of its neighbor, and fill all the
cells with a solution consisting of one part concentrated sulphuric acid
and four parts distilled water, measuring by weight. Connect the series
of cells in an arc light circuit, just as though they were arc lamps, and let
ihe current pass through them from the positive to the negative terminal
of the series until the paste on the negative plates has all turned color.
The cells will then be "formed" and ready for service. They should not
be allowed to remain charged long before being put into service, and it
will be advisable, therefore, to have the apparatus for which they are
to furnish current all ready to start up before putting the cells in circuit
for formation. The arc light circuit on which the cells are "formed" may
have any current value from 4 to 20, but as most of the circuits in this
country carry either 6.8 amperes or 9.6 amperes, one of these values will
doubtless be found in the charging circuit.
If the circuit is an intermittent one (does not run constantly, 24 hours
a day), care must be observed to take the battery out of circuit as soon as
the current is of: at each shut-down, so that it cannot discharge in case
the line is closed before current is restored.
CHAPTER XXVI.
CONSTRUCTION OF A CONSTANT-POTENTIAL ARC LAMP.
With a small screw-cutting lathe, a drill chuck and a few drills and
other small tools, any mechanic of average ability, having a fair knowl-'
edge of electrical apparatus, can, by using the accompanying sketches
as working drawings, make a reliable and efficient arc lamp for use on a
no-volt continuous-current circuit, in series with a resistance coil of 8
ohms, or a duplicate lamp and a resistance coil of 1^2 ohms, preferably
the latter. Fig. 240 shows the frame of the lamp, one-half in cross-sec-
tion. A is the top-plate; JB is the floor-plate; <7, C, are short side-rods;
J) y Z>, are long side-rods; , the yoke; F y the bottom carbon holder; /,/
and k, are insulating washers of hard fibre. The under side of the top-
plate, A, is shown by Fig. 241. It is of cast-iron or brass, and is provided
with two lugs, d } d, $4 m - m diameter and i in. long, drilled and tapped
J in. deep to take % in. gas pipe ; a lug, e, ^ i n - in diameter and i in.
long, drilled and tapped to take a 5-32-in. machine-screw, and a flange
around the outer edge, 1-16 in. thick and y 2 in. deep. At diametrically
opposite points, two pins, x 9 x, are set in the flange; these are of i-i6-in.
steel wire, y 2 in. long. The centers of the lugs, d, d, are 2^ ins. from
the center of the plate, and the center of the lug, e, is 2.y 2 ins. from the
center of the plate. On the upper side of the plate is a neck (see Fig.
140) i% ins. in diameter outside, and standing i in. above the upper sur-
face of the plate. This neck is bored J^-in. deep and tapped to fit a J^-in.
gas pipe; below this bore, a hole, 11-16 in. in diameter is drilled clear
through the plate. A fibre washer, i, shown enlarged in Fig. 247, is fit-
ted to this hole ; the larger diameter of the washer, i, must be such as to
allow it to slip down freely in the threaded neck, and the smaller diam-
eter must fit snugly the n-i6-in. hole at the base of the neck; the bore
at the top of the washer is 2/g in., and the recess in the under side is y 2 in.
in diameter. The washer is }/\ in. thick, and the flange is ]/ in. thick.
The short side-rods, CC, are pieces of J^-in. gas pipe, 5^ ins. long,
threaded on the outside at the upper ends and on the inside at the lower
CONSTRUCTION OF A CONSTANT-POTENTIAL ARC LAMP 21
ends; the long side-rods, DD, are similar pieces of gas pipe, 19^2 ins.
long, and the outside thread is an inch long. The short and long side-
reds are held together by a steel plug, /, J^ ' ms ' l n g> threaded the whole
length to correspond with the thread in the side-rods; the washers, jj t
shown enlarged in Fig. 247, are interposed to insulate the floor-plate, B y
from the side-rods and top-plate. These washers are ^4 m - m outer di-
ameter and 3/2 in. diameter at the neck; the bore is such as to allow the
threaded plug, Z, to slip through without having to screw it; the flange is
}i in. thick, and the total thickness is ^ in.
The floor-plate, D (Figs. 240 and 242), is of brass or iron, 5 15-16 ins.
diameter, J /s in thick, and has a round lug, c, ^4 in- in diameter, andi
FIG. 240. SEMI-SECTIONAL SKETCH
OF FRAME.
FIG. 241. UNDKKSIDK OF TOP- PLATE.
Kite.
FIG. 242. UPPER STDK OF FLOOR-PLATE.
~]/2 in. high. There are two */2-in. holes, b, b, a ^-in. hole, a, in the center,
and a hole through the center of the lug, c, threaded for a 3-i6-in. screw.
The distances, center to center, are 23/s ins. from a to each b and i % ins.
from a to c; the lug, c, is 90 degs. from the holes, b b. If it is more con-
venient, the floor-plate may be cut from sheet brass and the lug, c,
-soldered on, or screwed in and riveted.
The yoke, E (Figs. 240 and 247), may be cut from a strip of brass i%
Ins. wide and 3-16 in. thick ; the center hole is ^ in. and the others l / 2 in.
in diameter; the distances are 2^ ins. each way, from center to center of
2l6
ELECTRICAL DESIGNS
the large hole and each of the smaller ones, corresponding to the location
of the lugs on A, and the holes through B. The carbon holder, F, is
made of two pieces of brass tubing, the long one y 2 in. inside diameter
and the short one of a size to fit snugly over the long tube ; the two are
sweated together and riveted as a precaution against the loosening of
the solder by the heat from the arc when the carbons are almost burned
out. The long piece of tubing is threaded from the end of the short
piece }i in. down, and turned down to % in. outside diameter the balance
of its length to allow the nut, g, to slip up to the beginning of the thread,
and to form a mandrel for the globe holder. A y%-'m. pin, i in. long,
must be driven horizontally through the shank of the carbon holder.
Yi in. below the thread, after the nut is on ; the pin holes should be drilled
exactly across the center of the tube, so that when the pin is inserted the
ends will project from diametrically opposite sides of the shank. This
pin is to support the globe-holder, as will be explained further along.
The carbon-holder is 3 ins. long over all, and the short piece of tubing,
FIG. 243. ELEVATION AND FIG. 244. ELEVATION AND FIG. 245. PLAN VIEW OF
PLAN OF ARMATURE. PLAN OF BRASS FRAME. MECHANISM, MINUS CLUTCH.
which forms a sleeve over the holder-tube, is I in. long. A set-screw, h,
serves to clamp the carbon in the holder. The yoke is insulated from
the side-rods by washers, / and &; the former was described above, and*
the latter is a plain, flat fibre washer % in. thick, ^ m - diameter, with a
J/2-in. hole ; screws, m, m, hold the yoke to the side-rods.
The lamp is of the clutch type and the moving parts consist of a mag-
net-armature, a clutch and a carbon-carrying rod. The magnet is a.
straight, round bar, with a single coil ; the core (K in Fig. 245) is ^4 i n - m
diameter and 3 ins. long, with a shoulder 54 m - long at each end, the di-
CONSTRUCTION OF A CONSTANT-POTENTIAL ARC LAMP 217
ameter there being ^ in. The core is provided with two insulating heads
of fibre % in thick, the hole in which is a tight fit on the reduced ends
of the core, and after it is wound it is mounted between two brass frames,
N, N (Figs. 244 and 245). Each frame, N, has a y 2 in. hole drilled in its
base, the exact location of which may be found by reference to Fig. 244.
These frames have standards to which is pivoted the armature, P (Figs.
243, 245 and 246). The thickness of the metal is } in. throughout. The
back ends of the frames, A 7 ", N f are held down by a cross-bar, q (Fig. 245),
which has a J/-in hole in the center to allow the lug, r, to come through.
The lug is threaded on the outside to take a nut to hold down the cross-
bar, q. The bar is y in. thick, ^4 m - wide and a trifle over 2% ins. long
so that the ends may be filed to fit exactly between the pivot standards of
the frames, N t N.
The armature is a piece of flat Norway iron, 3-16 in. thick, ^ in.
wide and Iij4 ins. long, bent into a U, as shown in Fig. 243, and provided
with two ears, r, r, y& in. thick, which carry a round rod, s, of steel,
:G n^Q)
FIG. 246. ELEVATION OF MECH-
AXISM, COMPLETE.
Am.Ehc.
FIG. 248. CROSS-SECTION OF GLOBE-HOLDER.
3-16 in. diameter. At the back end a clip, t, of brass, is riveted on ; this
piece has a hole through it, tapped for -a 3~i6-in. machine screw. Pivot -
holes, whose centers coincide with the dotted line, p, p, are drilled in the
sides of the armature. The dimensions specified in the drawing must be
carefully observed.
The clutch H (Fig. 247) is a flat piece of brass y& in. thick, % in.
wide, and 2^4 ins. long, with a 7~i6-in. hole drilled J^ in. from one end
and a J^-in. slot, y 2 in. deep, sawed in the other end. The edges of the
hole must be very slightly rounded to prevent the clutch from cutting
2i8 ELECTRICAL DESIGNS
into the carbon rod ; v is a regulating screw to trip the clutch ; it is
Ins. long over all, 3-16 in. diameter at the threaded part, and j/s in. diam-
eter beyond the shoulder. The shoulder is one inch from the end. The
armature, frames, clutch and carbon rod are shown assembled in Fig.
246. The spring, y, which pulls the armature upward, is adjusted by
means of the screw, w, which screws into the lug, c, on the top-plate.
The screw is 5-32 in. diameter and the threaded end is an inch long.
The play of the armature is limited by the back-screw, s, by means of
which the length of arc first struck is adjusted; the screw, v t and the
spring, y, regulate the length of arc while burning. The spring is at-
tached to a stout brass wire strung from limb to limb of the armature.
The magnet is wound with No. 30 double cotton-covered magnet
wire, 32 layers deep, and 125 turns long, the starting end being connect-
ed with the core and the outer end with the bottom carbon holder. Bind-
ing posts may be put on if desired, but the writer prefers to carry a piece
of No. 12 rubber-covered and braided wire from the yoke up alongside
one of the side-rods, making this the negative lamp terminal ; the outer
end of the magnet coil may be connected to this terminal by means of a
piece of stout wire brought through the floor-plate, B, of the lamp, the
hole being bushed with insulation, and the positive terminal may be a
binding post screwed in the floor-plate (which is in electrical contact with
the carbon rod).
The case of the lamp is a piece of thin sheet brass, 6 ins. X 19 ins.,
"bent into a 6-in. tube and riveted at the lap ; at one end of the tube thus
formed and diametrically opposite each other, are bayonet slots which en-
gage with two pins projecting inwardly from the flange of the top plate of
the lamp. The carbon rod is a piece of brass tubing j-in. diameter out-
side, 24 ins. long, with a i-i6-in. wall. The upper carbon holder can be
purchased for a small sum and is not worth the trouble of making. The
"ball and shank must be made to fit the carbon-holder and the bore of the
carbon rod ; the shank should be an inch long, very slightlty tapered. Drill
a 1-16 in. hole clear through the rod, ^ in. from the end, before inserting
Ihe shank ; then tin the shank and the inside of the end of the carbon rod,
drive the shank in, and solder through the holes.
The globe-holder (Fig. 248) consists of a disc of brass (or iron)
3-16 in. thick and 5 ins. in diameter, having a piece of brass tubing
screwed into its center and three lips riveted at equidistant points around
the edge. The tube in the center must fit snugly over the shank of the
carbon-holder (Fig. 240) and it has two bayonet slots at the upper end
which fit over the ends of the pin driven transversely through the carbon-
CONSTRUCTION. OF A CONSTANT-POTENTIAL ARC LAMP 219
holder shank. This tube must measure ij^ ins. long above the disc.
The ears are simple brass strips each J^ in. thick, y 2 in. wide and 2 ins.
long, with y 2 in. of its length bent up almost at right angles to the bal-
ance ; 3-16 in. thumb-screws in the up-turned lips serve to hold the globe
by its rim.
Two arc lamps such as the one above described will work together,
in series with a resistance coil of ij^ ohms, on any no-volt direct-
current circuit.
CHAPTER XXVII.
AN EXPERIMENTAL NERNST LAMP.
This lamp, invented by the physicist, Nernst, of Gottingen, consists
of a rod of dense magnesia with platinum terminals. This rod is con-
nected in series with a dead resistance, and an e. m. f. (preferably alter-
nating) of from 200 to 600 volts is applied to the arrangement. Upon
heating the magnesia rod, by a blow pipe, for example, it becomes a con-
ductor and passes sufficient current to raise its temperature to that of in-
tense incandescence. In the more recent types Nernst uses a large pro-
portion of thoria in the rod.
An increase of current in the lamp causes a rise in its temperature
and a drop in its resistance and, at the temperature at which the lamp is
used, this drop in resistance is so great that considerably less e. m. f. is
required to push the increased current through the rod, so that the lamp
is unstable, and without the dead resistance the lamp would be destroyed
by the excessive current that would flow through it. The efficiency y of
the lamp, according to tests made abroad, is about 1.5 watts per candle-
power, including the watts lost in the dead resistance. The lamp gives
a beautiful and pleasant white light and its life is claimed to be very great.
A number of these lamps have been constructed at the physical
laboratory in Bethlehem, Pa., by Prof. W. S. Franklin and Mr. R. B.
"Williamson. After many trials the following procedure was found to
give good results : A mixture of calcined magnesium oxide (composition
of mixture given below) is tamped as compactly as possible into a smooth
bore brass tube lined with two or three thicknesses of stiff writing paper.
This paper should be fixed in place with a little glue and baked dry. The
tube full of magnesia is then slowly baked on a metal plate over a Bunsen
burner until the paper is completely charred, when the magnesia rod may
be pushed out. The rod is then calcined before a blow pipe, heating it
slowly and uniformly to avoid cracking by unequal shrinkage. The rod
is then broken to a length of about 2^/2 ins. and laid upon a bed of mag-
EXPERIMENTAL NERNST LAMP
221
nesia. Two ordinary arc carbons are brought up to the ends of the rod,
one carbon being fixed by weights, the other being preferably held in the
hand. Several hundred volts e. m. f. are applied to the carbons with
dead resistance in circuit and the magnesia rod is heated by the blow pipe
until the current starts. As the magnesia rod rises in temperature it
shrinks greatly, and it must be subjected to very slight end pressure to
prevent the formation of cross cracks ; too much pressure will cause lon-
gitudinal cracks. The current is then increased until the magnesia rod
becomes slightly soft, when it may be straightened if, as is likely, it has
curled up in shrinking. The rod is then allowed to cool and ground on
an emery wheel to the required shape, as described below.
The most convenient source of current for the purpose of this pre-
liminary heating and for operating the finished lamp is a step-up trans-
former with a rheostat in the primary circuit ; a secondary e. m. f. of 1,000
volts is satisfactory. This e. m. f., of course, falls off greatly when the
current starts, because of the action of the primary rheostat.
The magnesia mixture may be pure
calcined magnesia with a slight amount of
magnesium chloride ground up with it to serve
as a bond. A slight amount of soluble silicate
of soda is also a good bond. The mixture
should be only moist enough to pack like
flour ; it is better to have it perfectly dry than
too moist. A lamp made of pure magnesia or
of magnesia with i per cent or less of powder-
ed silica, has a very high resistance and can
scarcely be started with less than 1,000 volts,
and then with difficulty. After it is once
started, however, the resistance falls so that
even a pure magnesia rod will operate with,
say, 300 volts per inch of length. A lamp
which is very much easier to start is made by
mixing from 2 to 6 per cent of pounded glass
with the powdered magnesia. Perhaps a lime glass would be best for
this purpose.
The magnesia rod should be about i in. or ij^ ins. in length, and
about J/s-in. in diameter, with slightly enlarged grooved ends : platinum
wire is wound two or three times around these ends and covered with a
paste of magnesia, pounded glass, and water glass (or simply water). The
lamp is conveniently mounted by binding the platinum wires to the side
of a small glass tube. Fig. 249 shows the finished lamp full size.
FIG. 249. A NERNST LAMP.
222 ELECTRICAL DESIGNS
A lamp made as above described, with about I per cent, of pounded
glass and I per cent, of powdered silica, the rod being about 1)4 ins. long-
and J^-in. in diameter, operated on 250 volts (between platinum termi-
nals), takes 0.8 ampere, and gives fully 175 candle-power, although the
candle-power has not been measured at Bethlehem. It has been found
that the silicates of sodium and potassium (or perhaps simply the sodium
and potassium) are slowly expelled by the heat while the lamp is in use,
causing the resistance to become slowly greater.
Commercial magnesia (calcined Grecian magnesite) makes good
lamps without any admixture of silica, although its resistance is rather
high unless it is mixed with powdered glass.
An attempt was made to fuse magnesia into a compact mass in an
electric furnace (100 amperes at about 90 volts), but it was found that the
boiling point of magnesia (at atmospheric pressure) is about the same
as its melting point, so that the material vaporized about as rapidly as it
was melted. The operation would, no doubt, succeed under pressure.
During this work with the electric furnace it was necessary to keep a
close watch of the action, and a small piece of heavily smoked glass was
used to screen the eyes, leaving the forehead exposed, and a sever case of
sunburn was produced, although the heat on the face was not excessive.
A most striking experiment is to mount a glass tube as a Nernst
lamp. A large, thin walled tube gives the best effect. Wind copper wire
terminals about 4 ins. apart on a thin walled glass tube 3^-in. or J^-in.
in diameter. Connect to the secondary of a step-up transformer with a
rheostat in the primary. Heat the tube along one side. The current
starts along a narrow strip of the glass, heats it to bright redness, and
this heated strip gradually widens until the whole tube is melted dowti
This experiment was tried in Bethlehem with a i,ooo-volt secondary, but
it would certainly be possible to perform the experiment successfully
with as low an e. m. f. as 100 volts, and direct current would answer as
well as alternating. With low e. m. f. the distance between the copper
terminals should be much less than 4 ins., and, of course, a rheostat
should be included in the circuit.
CHAPTER XXVIII.
CONSTRUCTION OF AN INDUCTION COII,.
Since the advent of the Rontgen discovery the induction coil ha$
risen to a much more prominent place as a scientific and practical instru-
ment. It has very naturally been greatly improved in construction with-
in the past year, but inasmuch as these improvements are not generally
known and used, the writer has presumed to believe that a description
of them may be interesting.
The basis of the discussion will be the construction of a 6-inch spark
coil, but it may be profitably remembered that the average induction coil
built in sections may be, thus rebuilt, and oftentimes the length of spark
it is capable of giving thereby trebled, even though thirty or forty per
cent, of the secondary is removed in order to accomplish the construction.
Many modern coils are built on lines that make extensive internal
leakage a great possibility. Some coils are made with as much as twen-
ty-five pounds of wire in the secondary, and yet under the most favorable
conditions the spark obtained is but six inches in length. The makers of
such coils broadly claim that it is impossible to break down the insulation
of their apparatus, but in view of the fact that a 6-inch coil can be made
with a 5-pound secondary, it is easy to see that the coils just referred to
are broken down already, and that it is a case of spoiling a bad eg^j a
manifest impossibility.
The principal leak in an induction coil is from the secondary to the
primary, as is shown in Fig. 250. Between the points of leakage indi-
cated the full difference of potential of the coil exists. The %-in. of hard
rubber and the almost negligible air gap usually provided can scarcely be
expected to withstand the e. m. f. that will urge a discharge across a six-
inch air gap.
A second source of leakage is shown in Fig. 2 50- A, and exists at the
separator pieces between sections. The insulation between the primary
and secondary is broken in its continuity by these pieces, and as it is im-
possible to make an electrically tight joint, such insulation as is provided
22 4
ELECTRICAL DESIGNS
is no more effective than an equivalent gap of air. The insulation between
primary and secondary must be a continuous homogeneous mass, and
sufficiently thick to withstand the maximum e. m. f. of the coil. Fig. 251
illustrates the method of insulating a secondary section. The spaces, 5 S,
are to be filled with paraffine or some equivalent continuous insulator.
Covered wire for an induction coil is not necessary, and the use of
silk wire is a most expensive construction, from which absolutely no ad-
FIG. 250. SHOWING LEAKAGE FROM
SECONDARY.
MMMMMM&MM
Secondary
Insulation
\
Am.Elee.
FIG. 250A. LEAKAGE THROUGH JOINT.
FIG. 252.
Secondary
J Primary and Core
S
Am.EUc*
S
S
Secondary
S
Am.Ekc.
FIG. 253. DIAGRAM OF COIL CONNECTIONS.
FIG. 251. IDEAL INSULATION FOR
SECONDARY.
vantage can be gained. One way is to use bare wire, winding a thread
between adjacent turns, as shown in Fig. 252. Colored thread should be
avoided. The space between the layers should be at least four or five
times the thickness of the insulation between the turns. The insulation
between the turns of an induction coil is about 5 mils (.005 in.) thick, and
CONSTRUCTION OF AN INDUCTION COIL 225
experience has shown that this is none too much. A space of i-64-in.
can be used between the layers to advantage. This space should be
filled with absorbent paper that will readily soak up paraffine wax.
rig. 254 shows a regular sectioned dimension drawing of the 6-in.
spark coil already referred to. It would be idle, to enter into a long dis-
sertation on the various features of this coil that are common to every in-
strument of a similar nature, and only the novel ones will be discussed
and the quantitative measurements given. The secondary coils are con-
structed of bare wire, absorbent paper and cotton thread, substantially as
indicated heretofore. Care must be taken in the winding to keep away
at least 54~ m - with the wire from the edge of the paper layer, partly for
the added insulation between the layers and partly to prevent the annoy-
ance of the end turn slipping out when handling the section. If an old
coil is being rebuilt, it will not pay to thus rewind it. Sufficient wire from
the inside of the secondary sections should be removed to admit of reas-
sembling it as per drawing, a comparatively easy thing to do, and the
results will be nearly as good as with the coil here described.
The great feature of the coil is the method of supporting and insulat-
ing its primary and secondary. A long box is constructed as per draw-
ing, and from the geometrical center of the ends is supported the tube
that forms the enclosing envelope for the primary coil and its core. The
secondary coil is divided into six sections, each supported on a piece of
hard rubber tube with end collars of glass or hard rubber. This hard
rubber tube allows j/2-in. in the clear between its interior surface and the
primary envelope. The glass collars are square, and are of such a shape
that they just fit the inside of the box, and in their lateral dimensions are
a perfect measure of its interior section. The space between adjacent
sections is %-m., and between the last coils and the end pieces, 5/2-in. is
provided. The coil is wound to a diameter of 6 ins., the internal dimen-
sions of the box surrounding it being 8 ins. square.
Before assembling, the coils are boiled for a long time in paraffine,
and are removed therefrom only when the wax has cooled sufficiently to
attain a mushy consistency. They are preferably assembled while in this
state, for large soft clots of wax adhere to the coils and close in on the
bobbin on which the coil is put, thus filling up objectionable air spaces.
The assembling of the coil being complete, each secondary will be mount-
ed on a tube in the box and will rest in a partition made on two sides of
glass or hard rubber. Nowhere will any secondary section have any con-
nection with any primary section except through paraffine wax in a con-
tinuous mass that cannot be broken down unless penetrated. The great
CONSTRUCTION OF AN INDUCTION COIL
227
merit is the continuity of the insulation and the entire absence of joints.
To attain this result, the box must be filled with boiling paraffme at all
partitions, thus filling up all the air spaces, of course, first making the
proper connections. The top of the box is then put on and the paraffine
is allowied to set. In setting is will shrink a certain amount, and this
space must be filled with more paraffine.
The coil is to be mounted on a box containing the condenser in the
usual way. It will be well to divide the condenser into sections, as shown
in the diagrammatic connections of Fig. 253. If the coil is to excite a
FIG. 255. PERSPECTIVE OF VIBRATOR,
Crookes' tube, this in an important matter. Some tubes that are capable
of giving admirable results often signally fail to do so on a coil of great
capacity, but will operate perfectly on a smaller one. The reason of this
is found in the fact that the large coil may not be in as close resonance
with the tube as the smaller one. By the use of the variable condenser,
the resonance of the coil can be varied in pitch and its range of excitation
of tubes widened materially. The principal dimensions of the coil just
described are as follows :
Primary coil. Two layers of No. 12 B. & S. wire, single cotton cov-
228 ELECTRICAL DESIGNS
ered, wound on a fibre tube and surrounded with a hard rubber envel-
oping tube, as per drawing 1 .
Secondary coil. Five pounds of No. 36 B. & S. bare wire, wound in
six sections, as shown and described.
Support. A mahogany box supporting primary envelope, and glass
partitions, as described and shown.
Condenser. Seventy-five sheets of tin-foil 7 ins. X 9 i ns - alternated
with sheets of paraffined paper 8 ins. X 10 ins.
A word about the secondary connections may not be out of place
because of the confusion that has arisen among amateurs. It is custom-
ary to wind the secondary coils exactly alike with the outer lead on one
flat face and the inner lead on the other. If such similar coils are slipped
on the core in the same way, it will be necessary in order to connect them
in series to join the inner end of one to the outer one of its next neighbor.
This will require that the connecting wire must be brought up between
sections and in this position it will be very difficult to insulate. There-
fore the coils are slipped on in alternate reverse order. By this is meant
that if the first coil is put on in one direction, the next is put on so that
similar ends face each other. To connect the coils so placed in series, the
like ends must be connected. A moment's inspection of this connection
will show that the" -current travels about the core in the same direction
through all bobbins, and that the arrangement does not connect the bob-
bins in opposition, as has been popularly supposed.
The circuit breaker or interrupter is one of the most important parts
of the coil and little has been done to improve it. The ordinary vibrator
is perhaps the most convenient automatic circuit breaker, but it is very
defective in many respects. One of its chief faults is that it keeps the
circuit open too long and closed for so short a period that the core does
not have time to fully charge or the current to attain its full value. An
interesting modification that tends to achieve this result is shown in Fig.
255 and is drawn in suitable form to apply to the coil just discussed. Its
principle is as follows : The spring, C, presses tightly against its contact,
Kt at all times except when it is struck by the hammer of the vibrator,
when contact is broken for an instant. Thus the break is instantaneous
and the circuit is closed for a definite period cf time. The other screws
are to limit the motion and frequency of vibration cf the hammer.
As indicated in the illustration, a double-pole switch and a means of
varying the condenser are to be placed on this induction coil. A double-
pole double-throw baby knife switch is the most suitable for the reversing
device, and for the condenser a pair of plugs and plates will be found
CONSTRUCTION OF AN INDUCTION COIL 229
convenient. These are not shown on the drawing because they would
tend to confuse the more important details of the vibrator. It is obvious
that they should be placed in a convenient and symmetrical position and
that further mention of them would be more perfunctory than interesting.
It will be noted that this coil is designed on lines that seem to direct-
ly defy all laws of magnetic efficiency with regard to the distance between
primary and secondary. Many might hesitate before spending their
time and money on such a construction. The reader is assured that the
dimensions herein given are the result of a series of progressive experi-
ments, and each coil in the series was constructed with the idea of im-
proving the last. Not until the liberal insulation shown was adopted
were maximum results obtained and even now the advisability of carry-
ing the principle further is being considered. The smaller amount of
wire and its inferior magnetic position are more than compensated by the
absence of leakage and, moreover, the extremely low internal resistance
of such a coil enables it to produce a much more highly calorific spark
or as it is commonly termed, a fatter one, than if the older and more
conventional construction were followed.
CHAPTER XXIX.
CONSTRUCTION OF A TKSLA-THOMSON HIGH FREQUENCY COII,.
The following is a description of the construction of a Tesla-Thom-
son high frequency coil, large enough to give a five-inch spark and ex-
cite Rontgen ray tubes.
To excite the Tesla-Thomson coil, a high potential transformer of
from 10,000 to 15,000 is necessary. The construction of this transformer
will be first given. Fig. 256 gives a partial cross-section of the trans-
former, which is made as follows : A two-inch iron pipe, sixteen inches
long, is slotted the whole length, either in a milling machine, planer or
shaper. This slot need not be more than i-i6-in. in width. The pipe
PIPE SLOTTED AND FILLED WITH SMALL SOFT IRON WIRE
PAPER INSULATI
PRIMARY WINDINQ*13 B. 4 3.
FIBRED INSULATION
FIG. 256. HIGH TENSION TRANSFORMER.
is then insulated with ordinary wrapping paper to an outside diameter
of 2,y$ ins., shellac being freely used, and is then wound with No. 13 3.
&.S. double cotton-covered wire for its whole length (one layer). It is
then covered with paper and shellacked until the outside diameter is 2^
ins.
The next step is to fill the pipe with soft iron wires, No. 16 B. & S.,^
each wire being cut eighteen inches long. This completes the primary
winding of the high tension transformer.
The secondary winding of this transformer consists of ten roils wound
TESLA-THOMSON HIGH FREQUENCY COIL
in a form and thoroughly taped and insulated. This form is shown in
Fig. 257 and can be easily made cf wood. The wire is wound in Ihis
form, shellacked, removed, taped and baked. These ceils are then slipped
over the primary winding, between each coil being placed a disc of card-
board J*8ECONDARY C0.ll. 31 B, & 3.
.i-LoUCOOQUOOOOOOOOOOOOQOOQOOOOOQOoooOQO - ; ;
FIG. 260. PRIMARY AND SECONDARY COILS OF HIGH FREQUENCY TRANSFORMER.
The high frequency coil is made as follows : Wind an 8-inch paper
cylinder eighteen inches long with No. 31 B. & S. double cotton-covered
wire (or larger), leaving a margin at each end of about one inch. This
is the secondary winding. The primary winding is placed on a 1 2-inch
paper cylinder eighteen inches long and consists of fourteen turns of four
No. 8 B. & S. double cotton-covered wires in parallel. Each of these No.
8 wires is wound on separately, then the four ends at the beginning and
ending are soldered together. Between wires of different polarity, as
an extra precaution, two turns of cord are wound. The primary and sec-
TESLA-THOMSON HIGH FREQUENCY COIL 233
ondary coils are then shellacked and baked. After being, baked, the sec-
ondary coil is placed concentrically (see Fig. 260) inside the .primary and
the connections as shown in Fig. 261 then made.
The primary of the high tension transformer must be excited with an
alternating current. With a frequency of 60 cycles per second, 50 volts
will suffice, and for 125 cycles per second 100 volts. The length of the*
spark from the secondary of the high frequency coil will depend on the
width of the ' 'spark gap, ' ' consequently, in exciting a tube it is best to start
with the ' * spark gap ' ' very short, then gradually increase until the tube is
properly excited. When the terminals of the secondary high frequency
coil are separated farther than five inches, a spark will pass from the sec-
ondary to the primary of the high frequency coil. By the use of a good
insulating oil a much longer spark can be obtained from the high fre-
quency coil, but for exciting Rontgen ray tubes a five-inch spark will be
sufficient.
CHAPTER XXX.
CONDENSER FOR EXTREMELY HIGH POTENTIALS.
A condenser for high potentials that is commercial has been a prob-
lem that has long defied complete solution, and the demand for one that
will withstand the enormous potentials of so-called Tesla currents has
been only partially met by the clumsy and ineffective Leyden jar. The
writer's practical experience with the condenser herein described bears
him out in offering it as a complete solution for Tesla currents as usually
employed, and a partial solution for the problem cf how to get a con-
denser for high voltage commercial currents.
This condenser has a capacity of about .02 microfarad according to
the specific inductive capacity and thickness of the glass that is used. It
will replace a battery of fifty or sixty quart Leyden jars and will only oc-
cupy the space of a couple of them. If it is stacked in banks of fifty or
sixty, a capacity of one microfarad could be obtained, which is sufficient
for experimentation on commercial circuits. This condenser can be
made by the veriest amateur, at an expense not exceeding $2.50.
Procure of some good-natured photographer a supply of old nega-
tives five by seven inches in lateral dimensions. About one hundred will
be needed. Soak them in hot water till the gelatine film has dissolved,
rinse them off, and when dry and clean they are ready for use.
A dealer in photographic supplies will sell ferrotype plates 14 ins. X
10 ins, for not more than four cents each. As each plate of this size will
make four condenser plates, the total cost of the latter will not exceed
$i. The plates should be laid out and cut as shown in Fig. 262. Through
the center of each lug should be drilled a J/6-in. hole. Procure an Edi-
son-Lalande jar 5 ins. X 8 ins. in horizontal sectional dimension, this be-
ing a standard size> Select the jar with some care, being sure that the
bottom and sides are perfectly flat, for otherwise the condenser plates
will not pack in place nicely. Two pieces of ^J-in. brass rod should now
be obtained, together with a box of ^-in. copper rivet burrs, and some
'j6-in. standard tap brass nuts. The brass rods should be the length of
CONDENSER FOR EXTREMELY HIGH POTENTIALS
235
the Edison-Lalanclc cell, and should be threaded for some distance on
each end. Having obtained about one-half gallon of parafnne oil, the con-
denser is ready to be put together.
The first thing to do is to pack the jar full of glass and ferrotype
plates, so adjusting their number that there will be one less ferrotype
than glass plate. If the glass is not too thick, the jar will hold between
ninety and one hundred plates, and it should have just enough that the
FIG. 262. GLASS PLATES.
FIG* 263. ARRANGEMENT OF PLATES.
FIG. 264. TERMINALS OF PLATES.
FIG. 265. CONDENSER COMPLETED.
walls of the jar shall be effective in holding the plates together in a solid
homogeneous mass.
The plates of glass and sheet iron should now be arranged alternately,
as shown in Fig. 263. The lugs of each set of plates are to be threaded
with the brass rods before mentioned, and rivet burrs interspersed so that
when the nuts are set up as shown in the sketch of the complete con-
denser (Fig. 265) the tin plates will not bind the glass plates between
them. The terminals may be simple wires, but preferably a ball and
236 ELECTRICAL DESIGNS
knob arrangement as shown in detail in Fig. 264 and in position on the
condenser in Fig. 265. After the condenser is thus arranged, it remains
to fill it up over the tops of the lugs with paraffine oil and it is complete.
As described, the condenser would be suitable for potentials of io>-
ooo volts or less. For higher potentials the plates between the conduc-
tors may be made thicker. This will reduce the capacity of the con-
denser both by increasing the thickness of the dielectric and reducing the
number of plates that can be placed inside a jar of given dimensions.
The ball and knob arrangement is very simple. Some i-i6-in. brass
rod is bent into a J^-in. eyelet at one end, while on the other is cast al
round leaden bullet. These rods are bolted each to its system of plates
on the rod holding the plates together. They will serve to separate the
ten plates at the points where they are bolted in, instead of washers, and
will bind a sufficient amount to hold them in any position that they may
be placed ; as they are placed opposite each other the discharge gap may
be varied at pleasure.
For use with the higher potentials the jar had better be of hard rub-
ber, for it is liable to be punctured, and if this happens the jar may crack
and release the oil, to the great discomfiture of the experimenter.
CHAPTER XXXI,
CONSTRUCTION OF A WIMSHURST INFUJICNCE MACHINE.
This machine is the easiest of all static machines to make, and one
of the most satisfactory in its results. It is practically independent of the
weather conditions. If made as described herein, the machine will be
capable of giving a continuous stream of two-inch sparks, and will have
sufficient power to excite a small Crookes tube, provided that the termi-
nals of the tube are very near together.
The first and most difficult part of the work is to shape the glass discs.
There are two of these and they are made exactly alike. They are to be
twelve inches in diameter, and have a "^4 -in. hole in the center. Inas-
much as many are not familiar with the cutting of glass into such a shape
a few hints will be useful.
Select a piece of window
glass of the cheap green variety.
Better grade glass contains lead
and is less suitable. The hole in
the center should be bored first.
Prepare a solution of camphor
in turpentine and use it to keep
the boring tool moist. The bor-
ing tool may be made of a rat
tail file. The end should be snap-
ped off and the boring performed
with a twisting motion of the hand, care being taken to keep the file
moist. Patience is necessary, and when the hole gets so deep that it is
nearly ready to break through, it is necessary to proceed with extreme
caution. Once safely through, the hard part of the work is done. The
hole must now be cautiously filed to size, still using the camphor and tur-
pentine as a moistener. A mark to work by may be made by gluing a
piece of cardboard carrying a hole of proper size onto the side of the
glass.
FIG. 266. METHOD OF CUTTING GLASS DISCS.
ELECTRICAL DESIGNS
:fc 1^4-
FIG. 267. HUBS OF MACHINE.
FIG. 268. STDE SUPPORTS.
FIG. 269. SHAFT.
SIDE ELEVATION
ELEVATION
VIEW AT A B
Glass
FIG. 270. ELEVATION OF MACHINE.
W1MSHURST INFLUENCE MACHINE
239
Having completed the hole, it remains to trim the edge of the glass
into circular form. This is a comparatively easy matter. Erect on a flat
surface a little pillar of wood 24~ m - m diameter. Place the glass over this
so that the pillar protrudes through the hole. Prepare a loop of string of
such length that when it is looped around the pillar as in Fig. 266, the
glazier's diamond will swing in a twelve-inch circle. Be sure to use a
glazier's diamond, as the use of a cheap wheel glass cutter would be likely
to spoil all the work in boring the holes. It may be better to have a!
glazier snap off the glass if the operator is not experienced in such work.
Prepare the wooden hubs as shown in the drawing (Fig. 267). Bush
them with a brass tube 2^-in. in internal diameter. These hubs are se-
cured to the glass discs with cement. Major's cement or marine glue
FIG. 271.
is excellent, and bicycle tire cement answers very well. After this is done
the discs should be thoroughly shellacked with filtered shellac, and al-
lowed to dry. In the meantime other parts may be prepared.
The side supports are of wood, and hard maple is preferable. They
are finished to the size shown in the drawing (Fig. 268) and the holes in
the uppper part are of such size as to tightly fit the ^-in. shaft they sup-
port. This shaft does not revolve. The hubs with their glass discs re-
volve upon it.
The shaft carrrying the two wheels is the only part that requires the
240 ELECTRICAL DESIGNS
services of a metal lathe ; should this not be available, the metal parts can
be made for a small sum by a machinist from the figured drawings in this
article. In its largest diameter this shaft is 5^ -in., and all of this part is
threaded. The ends are turned down to y 2 -in. journals, as shown in the
drawing. One of these journals is sufficiently long to pass completeb
through its bearing and carry a small crank. The shaft is shown in Fig.
269, and the bearing in Fig. 270. This latter may be cast in brass from
a wooden pattern.
The remainder of the wooden parts of the machine may be built and
assembled as per drawing. They should be of hard, well-seasoned ma-
ple, and thoroughly varnished. The parts should be put together with
glue. Nails and screws are to be avoided They will be necessary to
hold the main supports of the machine in place and in some other places
where the strain is great, but they should be used sparingly. The whole
should be given a coat of shellac varnish.
-*
IP ^ ft )
\J_L_ l \ L/
us
FIG. 272. YOKE FOR CONNECTING OPPOSITE SECTIONS.
FIG. 273. COMB.
When the discs are thoroughly dry they are ready to receive the tin-
foil sectors (Fig. 271). There are twelve of these to each disc, and they
are secured in place at equal angular intervals thereon. Follow the draw-
ings closely and no mistakes can be made. Shellac is to be used as an ad-
hesive, and the edges of the sectors are to be covered with varnish, over-
lapping at least i-i6-in., to prevent dissipation of charge. This com-
pletes the discs.
Mounted on the 'disc shaft with a tight driving fit are two pieces of
hard rubber (Fig. 272). These carry stiff wires, on the ends of which are
WIMSHURST INFLUENCE MACHINE
241
light brushes made of tinsel. Each rubber piece carries two brushes, one
at each end, and the two brushes are electrically connected. They are
adjusted so as to just touch the sectors, as the discs rotate and thereby
put opposite sectors in contact. Their angular position can be easily
adjusted to the position where the working of the machine is bound
to be best.
Two U-shaped combs collect the output from the discs. They are
conveniently made by drilling a X' m - brass rod with holes at suitable in-
tervals and soldering pin points into the holes. The combs may then
be bent to shape. In Fig. 273
is illustrated themethodo? form-
ing the comb. The sides of the
enclosure are of hard rubber and
serve to support the combs. A
small binding post may be
threaded into a hole at the curv-
ature of tlie U of the comb, and
with the aid of a few washers the
comb is neatly and securely held.
See general view, Fig. 274.
The other sides of the en-
closure are of glass, both on ac-
count of its insulating quality
and transparency. The plates
are held in place by pieces of
rabbeted moulding mitered on to
the sides of the upright pillars.
If the construction is followed
out as shown in the cuts, the
glass and rubber plates will lift
like a window sash and render
the machine completely accessi-
ble. The discs are driven in op-
posite directions by means of a
straight and a crossed belt from the shaft below. In making the metal
parts of the machine, all sharp corners are to be avoided with great care,
for at every corner the charge disappears and leaks away.
The person building this machine must not be disappointed if at first
trial it does not work at once. If the shellac is the least particle damp the
machine will refuse to generate, but once dry it will generate without fail-
ure thereafter. The tinsel brushes must make positive contact with the
FIG. 274. GENERAL VIEW OF MACHINE.
242 ELECTRICAL DESIGNS
sectors or the machine will not start. They must be so adjusted as to
touch opposite sectors simultaneously. The best working angle for the
tinsel brushes is 45 with the horizontal. The discs should rotate from
the comb towards the nearest tinsel brush.
The entire cost of the machine,, assuming- that all of the metal work-
ing that requires the use of machine tools is hired out, should not ex-
ceed $5.
CHAPTER XXXII.
TELEPHONE TRANSMITTER AND RECEIVER.
The only thing that prevented Philipp Reis being honored the world
Over (as he is to-day in Germany) as the inventor of the telephone, was
the fact that he could not or those who have since tried cannot make
his first instruments talk. It is said that the difficulty now is to find a mi-
crophonic instrument of any kind his kind included that will not talk.
And all the reason in the world is that we know how to adjust a single
screw! The whole secret lies in keeping the electrodes together con-
stantly. This is the only real difference between the Reis telephone and
the Blake transmitter, which is in use all over the world and lias proved
the best all-around instrument on the market.
For talking, a Blake transmitter and a form of the standard Bell re-
ceiver will be found the best. The patents on both of these instruments
have expired, and they can, therefore, be made and used by anyone at
present.
The Blake transmitter is illustrated in Figs. 275 to 281, and the re-
ceiver in Fig. 282. The receiver is the easier to construct and will be
described first. Procure a straight bar magnet of the best tool steel,
hardened glass-hard and strongly magnetized (Tungsten steel is prefer-
able). It should be long in proportion to its thickness, in order to main-
tain its magnetism well say, J4~ m - thick by 7 ins. long.
Take three-sheet bristol board and cut out two discs, about i l / 2 ins.
in diameter. Thin hard wood discs will also answer. Cut at the center
of each a hole of the exact size of your magnet and slip the discs on one
end of the magnet, ^4 -in. apart. One-sixteenth of an inch of the end of
the magnet should protrude. Wind a single winding of thin paper around
the magnet, between the discs, for insulation. Then wind No. 36 silk-
covered copper wire carefully between the discs until the space is nearly
full. In winding, the best results can only be attained by the greatest
care. There must be absolutely no kinks or twists in the wire and no
breaks in the insulation. A little melted paraffine or a little shellac can
244
ELECTRICAL DESIGNS
be placed over the outside layer to hold the wires, and you should leave
several inches of ends for connecting up.
Now take a sound piece of hard wood and turn up a case. In the il-
lustration (Fig. 282) this is lettered R. It should be I in. in external
diameter, except at the large end, which should spread out in a bulbous
form, as shown, with a diameter of 2^2 ins. outside at the edge. Turn up
a cap of the same wood, in the shape shown. Its outside diameter is 3
ins., and it should fit neatly over the end of the case. The cap is lettered
A. A depression should be turned in the outer face, as at b, and a sim-
ilar, but shallower, depression, as shown in the inner face. In the center
should be a hole, j/2-in. in diameter. Drill small screw holes in the side
flanges for the screws, d.
Through the center of your case bore out a straight smooth hole just
FIGS. 275 AND 276.
large enongh to receive the magnet, and at the large end of the case
turn the hole out to 2 ins. in width and ^-in. deep, as shown.
Drill a screw hole for the set-screw, S, near the small end of the
case. Now slip the long end of the magnet into place. If it proves too
loose, wrap thin paper about it.
Take a pair of compasses and lay off on cardboard a circle with i }
in. radius to fit exactly between the cap and case. With this as a tem-
plate scratch a similar circle on a piece of photographer's ferrotype plate
"tintype" plate which you can buy for a nickel. Cut out the disc
with sharp scissors. Take care not to bend it, as this is the reason the
compasses are not used directly on it. Any bend or buckle spoils it.
This is the diaphragm. Now take two pieces of No. 16 wire, scrape the
TELEPHONE TRANSMITTER AND RECEIVER
245
ends and make a kink in each. Solder the ends of your fine wire coil to
them and pass them through the side of the case. They are shown at w
in Fig. 282. The object of the kink or knot becomes apparent when you
insert it. The knot comes against the side of the case, and prevents
any pull coming on the thin wire to break it. Of course, if you want to
take the trouble you can cut a channel each side of the magnet hole all
the way back to the rear cap, c, and put binding posts there. Now put
the diaphragm half over the large end of the case, and adjust the magnet
rntil it nearly touches; i~32-in. is the proper clearance. Then screw
down the set-screw, S. Put on the diaphragm (shown at c, place the
cap, a, over it, screw in the holding screws, d, and glue on a covering
disc, e, for the rear end, and your instrument is complete.
It is advisable, in making the cap, a, to have the inner face over the
Battery.
Line
FIG. 277-
diaphragm with a clearance of not more than i-i6-in. If there is too
much space between the cap and diaphragm the sounds received will be
muffled. The cap should be firmly adjusted against the diaphragm at the
edges so as to clamp it against rattling.
In making the transmitter the first thing is the case, A. This is
shown in Figs. 275 and 276. It should preferably be of hard wood, nice-
ly smoothed, filled, rubbed and polished. The door of the case carries
all the operative parts of the transmitter and is provided with brass hinges,
which form part of the circuit, so that by simply opening the door yot:
can get at the apparatus without breaking any connection. The mouth-
piece, a, is simply a depression turned in the door face, as shown in dot-
ted lines in Fig. 276.
246
ELECTRICAL DESIGNS
The transmitter proper is all carried by the iron ring, B. This is
best shown, with its dimensions, in Figs. 278 and 279. As seen, it is a flat
ring having a circular rabbet or depression to receive the diaphragm,
and an upper and a lower lug, b, and c. It is screwed directly on the
inside of the door by screws, as shown in Fig. 277. It should be turned
tip smooth and true, and the extreme thickness of metal may be about
3/3 in. on the edge. The upper lug, b } is tapped in its face for two screws,
, and the lower lug, c, is vertically tapped for one, the adjusting screw, d.
Carried on the upper lug, by the spring, c, as shown in Figs. 277 and 281,
is the bar D, to which the electrodes are attached. This bar has an
tipper or head lug, /, and an inclined foot, g. The head lug carries the
springs, h and /, carrying the electrodes, E and p. These springs are
simply clamped to the head lug, /, the first one, h, resting directly
FIGS. 278, 279, 280 AND 28l.
against the metal, and held on by the insulating block, i, against which
rests the spring, /, clamped in turn by another little insulating block.
The screws pass through both insulating blocks and are tapped into
metal head lug, /. To the spring, /, above the insulation, is soldered a
wire, forming one terminal of the transmitter.
In operation, the point, p, carried by spring, /, rests against the dia-
phragm, the spring, e, however, constantly tending to carry it away ; so
that by screwing up or down the screw, d, the pressure may be accurately
adjusted.
The spring, h, is of fine spring steel, i-ioo in. thick and 9-64 in.
wide. Secured to the end, either by clamping under a screw head, or by
cutting a channel across and upsetting the edges over the spring, is a
TELEPHONE TRANSMITTER AND RECEIVER
247
brass button, E, carrying the carbon electrode, k. The way to make this
is as follows :
From a solid piece of brass turn up the button of the size indicated,
leaving the edges very thin and sharp. Cut the carbon button, k, accu-
rately and channel around it a shallow groove just where the edge comes.
Then put the whole in the lathe and spin the edge around into the
channel so that it tightly embraces the carbon.
The carbon button, k, must be pure homogeneous carbon free from
grit, and highly polished on the contact surface. The way to get the
[FIGS. 282, 283 AND 284.
best polish is to rub the button for a while on a smooth sheet of the
same kind of carbon. As this is not usually available., however, you will
probably find most convenient the old reliable emery paper or cloth.
Take a piece of fine emery cloth about 6 ins. square and rub your button
(which you must leave about y% in. too high when mounting) on the
emery, in a 3 in. circle. Keep it moving always in the same direction
and after a while the carbon deposited on the emery will form a fine
polishing surface, and give you a glass polish. Be sure, however, to use
none but the finest emery.
Attached to the spring, / (which is of German silver, .005 in. thick,
y% in. wide), is the platinum point, p. This can best be secured by solder,
and should be 5-64 in. across. A tiny end of platinum wire put through
248 ELECTRICAL DESIGNS
a corresponding hole in the end of spring, /, and soldered, is all *hat is
required.
In the bar, D, is an opening, opposite the electrode to permit adjust-
ing. The diaphragm, C, rests in the depression in the ring, B. Around
its periphery is stretched a rubber band, o, to deaden or dampen the
vibrations to some extent. It is held in place by two spring arms, in and
, made of flat spring steel, and shown best in Fig. 281. The arm, n,
extends over beyond the rubber sleeve, covering the end that rests on the
diaphragm. This produces a dampening effect that is very necessary
because of the delicacy of the contacts in this form of instrument. The
other arm, m, simply extends on to the rubber, and serves merely as a
clamp. Both spring arms should press lightly on the diaphragm.
The diaphragm itself is to be made of sheet iron. Ferrotype iron,
much heavier than that used for the receiver, is required.
One side of the circuit through the transmitter leads from the iron
ring, B (to which the wire is soldered), to a spring on the lower hinge, H.
This spring (one on each hinge) makes a scraping contact with the other
leaf of the hinge when the door is shut, and so ensures a good contact
there. The other side of the circuit leads from the spring, /, to the upper
hinge. The current from the battery enters at binding post i, Fig. 277,
flows to the upper hinge, through the wire to the spring, /, platinum tip,
p, carbon, k, brass button, E, steel spring, h, iron bar, D, screw, d, lug, c,
ring, B, and wire, to the lower hinge ; thence to the primary of the induc-
tion coil, 7, to the second binding post 2, and so back to battery.
The secondary of the induction coil is connected to the binding posts
3 and 4. The induction coil itself may be made as follows :
Take a bundle of very soft and fine iron wires, 2j4 ins. long, and
enough to measure *4 m - or H m - through. Wrap a turn or two o( thin
tough paper about them, and fit on either end a square block of wood,
Y% in. thick and iy& ins. on a side. Wind between these blocks and on
the paper, about 35 ft. of No. 24 silk-covered wire. The ends should
be carried out through fine holes drilled in the wooden end pieces.
Over this primary winding lay on carefully about 600 ft. of No. 3?,
fine silk-covered wire, and carry the ends out at one end in a similar
manner. Cover the coil with a wrapping of binder's paper gummed
fast on the edges, and fasten the coil in the position shown in Fig. 277,
by two long screws through and through the ends into the back of the
case. Carry the ends of the secondary winding to binding post screws, 3
and 4, and solder them. Connect one end of the primary to binding
post i and the other to the lower hinge, as shown.
TELEPHONE TRANSMITTER AND RECEIVER 249
In winding it is important to wind in regular layers from end to end,
and to avoid the slightest kink or twist in the wire.
The connections of the instrument are clearly shown in Fig. 284.
The switches cut off the bell and put on battery, when, moved to the
right, for calling, and cut in the telephone and close the local battery for
talking when moved to the left.
If an outdoor line is used it is advisable to use some form of light-
ning arrester, which may be obtained from a dealer at small cost, outside
the instrument.
In fitting up a telephone line for communication there are four ele-
ments necessary at each end a transmitter, a receiver, a call-sending
device, and a call-receiving device. For the purposes of this article I will
presume that the telephone line is a short one, say less than a mile in
length perhaps 1000 yds., with a single wire, of iron, No. 12, galvanized.
At each station you bring the line indoors to the instrument by connect-
ing office wire at the window and leading it around the woodwork of the
room. The joint outside the window must be soldered, the joint taped,
and the wire bent down U-shaped before it comes in, to allow moisture
to drip off. At your instrument another piece of office wire should be
started and led off to the nearest water pipe. The end of the wire should
be stripped for 12 ins., cleaned bright, the pipe likewise scraped bright,
and the wire wound tightly around the pipe and soldered. If this is done
carefully at both ends of the line, you have a good circuit completed
over the iron wire from one station to the other and back by way of the
pipes and the earth. It only remains to connect your instruments to the
wire ends and you should be able to talk perfectly.
For such a line a push button and vibrator bell, with a battery, at
each end, will furnish as good a call as may be. The arrangement of
these is indicated in Fig. 284. They can be purchased of any supply
dealer more cheaply than you can make them.
CHAPTER XXXIII.
CONSTRUCTION OF A DRY' 'BATTERY
Dry batteries, so called, are only dry in the sense that there is no
fiuid spilled 1 or slopped over when they are shaken or overturned. In
every voltaic cell the current is derived from the chemical action which
goes on within its substance, and no chemical action can take place be-
tween solids alone, but in all cases there must be present a liquid or a gas.
Some few cells employ gaseous electrolytes, and some, fused salts, but
the vast majority use aqueous solutions of their respective chemicals, and
it is in this class that the ordinary dry cells are to be found. Even the
old dry piles of Zamboni and others, which consisted of discs of paper
coated with metals (gold and silver paper) laid up "dry," in reality con-
tained a very small amount of moisture in the paper, and if the paper is
really perfectly dry, the piles will not work. If the ordinary dry cell then
requires moisture to make it work, and is in fact only a non-spilling wet
cell, it is a natural inference that the wetter the cell, consistent with it not
spilling or slopping over, the better. This inference is absolutely cor-
rect. The more fluid a dry cell contains the better, for many wet cells
would be improved for having a larger amount of electrolyte than they
do have.
It might seem in view of what has been said, that any wet cell, if well
sealed up, would do for a dry cell, but such is not the case, for several
reasons. Many cells will not stand sealing up tight, because they give
off gases, and these must have free vent, and again it is not always prac
ticable to seal up a cell so that it will not leak at all when inverted. Again,
it is not worth while to seal up a cell, except one of the kind that will
last for considerable time before it gives out or even needs replenishing.
Another desideratum of a dry cell is that it should not be easily broken,
as there are many places where cells are liable to fracture as well as up-
setting, etc. For these reasons, it is customary to dispense with the
glass jar, and to make the zinc serve the double purpose of containing
jar and electrode, and further, to use an absorbent substance that wil?
CONSTRUCTION OF A DRY BATTERY CELL 251
take up and hold the fluid electrolyte like a sponge, so that the seal is
rather to prevent evaporation and creeping- cf salts, than spilling or slop-
ping. The types of cells giving the best results on open circuit work,
as wet cells, naturally do the best when put up in the dry form ; conse-
quently, as might be expected, the vast majority of dry cells on the
market are some form or variety of the sal-ammoniac type. Several of
the manufacturers of dry cells claim to have valuable secrets relating to
their manufacture, but however true this may be in regard to the details,
the main requirements are well understood.
In making a dry cell, the first thing requiring attention is the jar.
This, as before remarked, is usually made of zinc. The cell is usually
cylindrical, although sometimes square or oblong in section, but in any
case a piece of moderately heavy sheet zinc is bent into the required
form, the edges soldered together and a bottom soldered in. Any one
who tries to solder zinc for the first time may be very much surprised
and disgusted to find that it does not take kindly to soldering like tin
plate, but balks and makes lots of trouble. However, by observing the
proper precautions, zinc may be soldered with comparative facility.
Thoroughly clean all the parts where it is intended that the solder
should stick, by scraping ; use clean chloride of zinc for a flux, and apply
the solder as near the point where it is needed as possible, not trying to
make it flow over the surface of the zinc as can be so readily done with
tin plate, for the more the solder alloys with the zinc, the more intract-
able it becomes. Learn to make a joint quickly on the first application
of the soldering bit, as the more you fuss and tinker with it the rougher,
more unsightly and more uncertain it becomes. Another pleasant little
habit of zinc is to strip the tinning off the soldering bit. You may have
tinned your bit with the utmost care, but after using it a short time find
it completely stripped. Some persons prefer an iron bit to the usual
copper one, claiming that it holds the tinning better, although somewhat
more difficult to tin in the first place.
Having made the jar, the next thing to attend to is the contents.-,
One of the most important constituents of the contents is the absorbent.
Several materials have been used for this purpose, among which are
plaster cf paris, gelatinous silica, gelatine, so called, which is really the
starchy mass obtained from boiling Irish moss ; gelatinous magnesium
oxychloride and a material made from the granular portion of the rind
of the cocoanut, called cofferdam. There are other materials, but these
r.re the more important ones. As the process of filling the cell differs
somewhat for each kind of filling, it is better to describe each one sep-
252 ELECTRICAL DESIGNS
arately. The zinc usually has a brass binding post soldered to its rim
on one side (there being no objection to this structure in a dry cell,
because the electrolyte cannot possibly come in contact with the junc-
tion), and the other electrode, which is usually of carbon, has a binding-
post of the same kind, fastened in one of several different ways.
The seal is made of pitch or some similar material, which will form
an air and water tight stopper, and also resist the tendency of the salts
to creep, as does the paraffin coating on the upper part of the jar of an
ordinary sal-ammoniac cell. It is simply melted, poured in on top of the
charge and allowed to cool in most dry cells, but some manufacturers
make a sort of safety valve or pressure regulator, by inserting a small
piece of rattan so that it passes completely through the seal, its ends pro-
jecting slightly above and below the pitch. The natural porosity of the
rattan is sufficient to relieve any pressure generated by the escape of
gases, but it will not of course provide for the swelling of the more
solid portion of the contents which sometimes takes place when the ceil
is subjected to too high a temperature, and cells are frequently destroyed
by bursting when placed in boiler rooms and other situations where the
temperature rises to an inordinate degree.
One of the important qualities of a dry cell is long life on open cir-
cuit, which means, that the local action should be negligible, and it is in
the prevention of local action that some manufacturers claim to have
valuable secrets. Bi-sulphate of mercury is sometimes used 10 keep the
zinc amalgamated, as in wet cells. This, of course, does some good, but
it is not all. Anything that will tend to keep the chemical composition
of the electrolyte uniform in all parts of the cell will help to prevent
local action.
We will now consider some of the particular forms of dry cells. The
Cox cell is formed by boiling Irish moss in sal-ammoniac solution until
it is thoroughly gelatinized, and pouring it into the zinc jar, where the
carbon electrode has been already placed. Bin-oxide of manganese in
conjunction with the carbon as a depolarizer is, of course, used as in the
wet form. The inventor also mixes a little bi-sulphate of mercury with
the electrolyte. Some of the other sal-ammoniac cells as described do
not use it, but there is no reason why they should not, and the reader
should understand that he may use it or not in the other cells described.
When the moss solution is cold it sets to a firm jelly, and is then ready
to be sealed. Obach's cell is made by mixing plaster of paris with tlie
sal-ammoniac solution and pouring it into the jar to harden. Mehner's
cell is made by mixing the sal-ammoniac solution with chloride of c^?-
CONSTRUCTION OF A DRY BATTERY CELL 53
/
cium and calcined magnesia, forming a paste of about the consistency
of cream, which is poured into the jar, and in two or three days forms a
stiff jelly, owing to the formation of oxy-chloride of magnesium. In
Gassner's cell the following composition is used : Oxide of zinc, I part ;
sal-ammoniac, I part ; plaster of paris, 3 parts ; chloride of zinc, i part,
and water, 2 parts, all by weight. The oxide of zinc is intended to make
the plaster more porous, giving this cell an advantage over the simple
plaster cell before described.
Gelatinous silica is precipitated when any strong acid is added to a
solution of. silicate of soda, and several inventors have used silicate of
soda to gelatinize the electrolyte in storage cells. This, of course, intro-
duces sulphate of soda into the electrolyte, but this does no harm, and is
even regarded as beneficial by some. The charge and discharge rate of
the cell is much reduced, and it will not do to allow much gas to be
generated into the jelly, neither can the cell be sealed up perfectly tight,
but a vent must be left for the escape of gas. The spattering or spraying
during charge, however, is cured, even if no cover is used.
Gelatinous silica may be used in any electrolyte in which the sodium
salt, formed at the same time with the silica, is not detrimental. It is very
difficult to wash the silica, and it does not pay to do it for such a pur-
pose as this, so if it is intended to use the silica with sal-ammoniac it is
better to precipitate it with hydrochloric acid, instead of sulphuric, as in
that case the precipitated silica will contain chloride of sodium, instead
of sulphate.
The Germain cell, which at one time attracted considerable atten-
tion, used the material known as cofferdam, previously mentioned in this
article. The containing vessel of this cell is made of wood boiled in
paraffine, the carbon plate is imbedded in lumps of peroxide of manga-
nese and carbon, and the rest of the space is filled with cofferdam satu-
rated with sal-ammoniac, the zinc, which is well amalgamated, is laid on
top, and the cover (or rather the side) is screwed on, slightly compressing
the contents.
It will be observed that some of the directions for the manufacture
of dry cells are very particular about the amalgamation of the zinc,
even when the electrolyte is sal-ammoniac, but, as a matter of fact, most
of the dry cells on the market are made with unamalgamated zincs, which
is a pretty good proof that the amalgamation is superfluous. There have
been many other dry cells proposed, some quite elaborate in composi-
tion and others less so, but the different forms of the sal-ammoniac cell
are at the present time used to the exclusion of everything else.
254 ELECTRICAL DESIGNS
One practical difference between wet and dry cells is the different
manner of using- them. In the wet cell, we endeavor to use some of the
parts, such as the containing jar, indefinitely, and the porous jar, if there
is one, at least a very long time. We also endeavor to use up the zinc as
completely as possible, excepting to renew the electrolyte several times
before the zinc is all gone. With dry cells, however, it is tbe common
practice to throw them away as soon as they fail from any cause to do
their work. The zinc being also the containing jar, it manifestly cannot
be all used up, but on the other hand must be discarded as soon as it is
perforated, if indeed something else does not give out before this event,
as is intended.
Dry cells are intended more particularly for a class of consumers
who do not care to be bothered with the manipulation of wet cells. When
they come from the factory they are ready for use without any prepara-
tion whatever, and when they are exhausted they may be thrown away
Vv r ithout compunction, for they are made for such a low price that it does
not pay to spend much time or trouble on them, even when they happen
to be in such a condition that they may be restored, which is not usually
the case.
The zinc electrode-jar must be insulated in some way, especially
when several of the cells are used in series,, as in this case any external
conductor touching two of the cells would short circuit at least one of
them. For this reason they are usually varnished, and often in addition
placed in strawboard boxes.
Judging from the variations in the different cells on the market, and
which work satisfactorily, it would appear that the exact proportions of
the ingredients are not very important. In the nature of things the elec-
trolyte is the first thing to give out, and there is no danger of getting
too much of it, and the solution should be saturated. The carbon and
manganese should not take up too much space, and as in the wet sal-
ammoniac cells the manganese is usually thrown away before it is ex-
hausted, from the fact that the inner part of each lump is unavailable, it
is evident that a small amount of manganese Vv'ill answer the purpose, if
arranged so as to be available, for which purpose it is better to have it
rather finely broken than in large lumps, and in as intimate contact as
possible with the carbon. There being no danger in a dry cell that the
manganese will become displaced after having been once fixed, there is
no objection to its being i:i a state of powder.
CHAPTER XXXIV.
SOME HANDY COMMUTATOR TOOI.S.
Direct current dynamos and motors have now come into very gen-
eral use, exceeding in number, perhaps, steam, gas and oil engines com-
bined. As is well known these electrical machines are subject to mechan-
ical wear t only the bearings and the commutator, which have to be
replaced from time to time. The very general demand for dynamo and
motor repairs has been met to some extent by electric repair shops that
have come into existence at many points through the country, but aside
from these there is hardly a regular machine shop of any size that is not
constantly called on for more or less work on the worn parts of dynamos
and motors.
As bearings in electrical machines are usually fitted with bronze
bushings, with standard reamed holes, well equipped machine shops are
usually in position to make these parts, but in the matter of commuta-
tors there is not one regular shop in fifty with the simple tools necessary
for their renewal, and the average machinist has but slight conception of
how this work must be done to insure satisfactory results.
In electrical repair shops there are usually some tools for handling"
commutator work, but they are frequently of the crudest kind and such
as to require too much labor, and even then lack certainty in results.
Some even among dynamo and motor manufacturers lack the few and
simple tools necessary to insure first-class commutator construction with
a minimum of labor. As usually constructed, a commutator contains
a number of copper "bars" or "segments," separated from each other
and the clamping parts by mica strips and rings. These bars are parts
of true circular sectors, though not of the exact circle of the commutator
surface, and are held, with the intervening mica strips, by a sleeve and
clamps.
Some of the main requirements of commutator construction are that
each segment be insulated or free from metallic contact with any other
or the sleeve and clamps, that each segment be so firmly held that the
ELECTRICAL DESIGNS
forces of expansion, due to heat when the machine is in use, shall not alter
its relation to the other segments, and further that the relative position of
segments and mica strips shall remain the same after the commutator has
cooled.
Were all parts of commutators metal, the above requirements would
make careful work necessary, but as each segment must be held entirely
by contact with mica, the problem is much more difficult, in fact, it has
FIGS. 285 AND 286.
required more study and experiment than any other mechanical question
that electrical manufacturers have had to meet.
Mica lias come into very general use for commutator insulation be-
cause of its high insulating properties, non-injury by heat and power to
sustain great pressures with but small compression. On dynamos and
motors of moderate capacity in common use the number of commutator
segments varies from six to eight to about two hundred according to the
purpose and capacity of the machine, the most common numbers of seg-
ments being from twenty-four to one hundred.
SOME HANDY COMMUTATOR TOOLS 257
As for every segment there must be a strip of mica, in addition to
the mica rings, the number of separate parts that must be held in their
exact position in a commutator will vary from fifty to about four hun-
dred in machines of moderate capacity and common use. In order to
hold so many pieces of materials, to a large extent contrary as to their
qualities, rigidly together, it has been found necessary to assemble them
with great pressure, and then set the permanent clamps as tightly as pos-
sible before the external pressure is removed.
The two most common and successful methods of clamping com-
mutators are shown in Figs. 285 and 286, the solid black lines in each
case representing the mica strips and rings. The better class of segments
are forged or of drawn stock, so that no labor is required oil the sides
before assembly in the way shown. As soon as assembled it is necessary
to compress and hold the segments securely so that the surfaces which
come in contact with the mica rings may be machined.
A method to compress and hold the segments, common in many
electric repair shops and with some manufacturers, employs a solid
forged ring turned on the inside to the diameter which the segments are
estimated to have when compressed, and tapered slightly at one edge
so as to start easily over the segments. In the correct use of this
ring it should be forced over the segments with a pressure of
some tons, as this source of pressure is the only one to bring
the segments and mica strips solidly together. To do a good job
with this solid ring a hydraulic or large screw press is necessary, and
in many cases machine and repair shops are without either of these
presses, so that the ring can only be forced on and off the segments with
a hammer, a very unsatisfactory method. A serious objection to this
solid ring method is that it is very hard to estimate the exact diameter
to which the ring should be turned in order to properly compress the
segments, and the trials to see how hard the ring crowds on all take
time. Again a solid forged ring must be had for every size of commuta-
tor, even though they vary by only a small fraction of an inch in diam-
eter, and there is great temptation when a ring goes on too easy to let it
go as "good enough." To do away with the necessity for presses, also
forged rings for every commutator, save time and insure means for the
desired compression in every case, the tool shown in Fig. 287 is now
much used.
This tool consists of two rings, the outer a solid forging and the
inner an iron casting, split along one side so that its diameter may be
slightly changed.
258 ELECTRICAL DESIGNS
The outer forged ring is fitted with six, eight or more radial set
screws, which bear upon the inner split ring. The commutator segments
and mica strips having been assembled in circular form, the split cast-
iron ring, having been turned as near as possible to the correct -diameter,
is pushed over them and pressure applied by means of the large set
screws in the forged ring. Any desired pressure can be obtained by the
combined action of the heavy set screws, the split ring readily conforming
to the slightly reduced diameter of the segments. The work can also be
done much more quickly than when a solid ring and press are used. I:
is not necessary to have one of the above tools for each size of commuta-
FIG. 287.
tor, as only the inner cast ring need be changed until the commutator
diameters vary by as much as three inches, so that the forged ring be-
comes either too small or so large as to be unhandy. In this way two or
three forged rings will cover a large line of commutators, while the cast
ring for each commutator is cheaply and quickly made.
When the segments are securely clamped in the ring the next step
is to turn up the surfaces to which the mica rings transmit the pressure ;
the lathe chucks are the only means for holding the clamped commutator
segments while the surfaces at each end are machined, and here comes
a waste of time and inaccurate results due to the effort, after one end
of the segments has been turned up to reverse them in the chuck, so as to
machine the other end in line with the first. The tendency with work
done in this way is to force the permanent clamps slightly out of line
with each other, and this may result ia loose segments at some point in
SOME HANDY COMMUTATOR TOOLS 259
tne commutator. Proper expansion mandrels, as shown in Figs. 288 and
289, not only enable the finished surfaces at the two ends of the seg-
ments to be brought practically into line, but also save much time on the
work.
Fig. 288 shows a mandrel adapted for use with segments of the
type in Fig. 285, where there is no undercut work to be done, while the
mandrel of Fig. 289 is more convenient for undercut segments. The
mandrel of Fig. 288 mounts in the usual way on lathe centers, has a taper
of one in twenty-four, is fitted with cast-iron expansion sleeve and a screw
collar at each end, to force the sleeve on and off. The cast-iron sleeve
should be cut entirely through once along its entire length and nearly
through, say, to within one-fourth or three-eighths inch at three or more
other points, that it may expand as evenly as possibly when forced onto
the mandrel.
A sleeve for this mandrel should be turned outside to correspond
with the inside diameter of the commutator segments it is intended to
mount, and if very accurate work is desired the inside of segments
should be turned out before mounting them on the' expansion sleeve,
though some makers think this unnecessary. When the segments are
mounted on the sleeve this latter is expanded by forcing it on the man-
drel with the screw collar, and the ends of segments in Fig. 285 can be
turned up without changing their position.
The segments shown in Fig. 286 can also be turned on the above
mandrel, but the work on the undercut is done at a disadvantage and
much time can be saved by the use of the mandrel of Fig. 289. This
mandrel, like the other, is on a taper and fitted with expansion sleeve,
but one end is forged into a flange, adapted to bolt to a face plate and
allow free access to one end of the commutator segments, so that the
undercut can be made quickly. One end of the segments being fin-
ished, they are forced off the mandrel with the sleeve by the screw collar,
and then put on again reversed and the other end finished. There is,
of course, no reason to take the mandrel from the face plate until both
ends of the segments are finished, so that time is saved and the finished
surfaces brought very nearly in line. Quite a number of expansion
sleeves can be used on the same mandrel for different commutators, so
that two or three mandrels will be enough for a factory turning out a
fair line of machines.
Having turned up the ends of the segments, it is next necessary to
mount them on the insulating rings and permanent clamps. As seg-
ments are held only at the finished surfaces, it is necessary that each
r
\N
4--
9/
<>*,
-n
b-:'-L
262 ELECTRICAL DESIGNS
make solid contact with all of the mica rings, as the side pressure of the
other segments is not sufficient to prevent motion either up or down.
The double ring clamps are of special value when the segments are to
be brought to a firm bearing on the mica rings, since, when necessary,
the set screws can be let up a little in order to allow the segments to
slide over the horizontal mica rings, and then set up until each segment
beds firmly on the mica. The permanent clamps once in place, the out-
side rings are removed and the surface of the segments finished up, first
by turning with a diamond-pointed tool and later with a single cut file
or the finest grade of sand paper.
It may be well to add the oft repeated warning that emery cloth
should never be used on a commutator, as the emery sticks in the copper.
The above simple and inexpensive tools will multiply several times
the amount of commutator work a man can turn out daily with the de-
vices now common in some shops devoted to the repair and even the
manufacture of electrical machinery.
ALGEBRA MADE EASY.
BY
EDWIN J. HOUSTON, PH.D. and A. E. KENNELLY, Sc.D.
CONTENTS.
CHAPTEU I. Introduction. II. The Symbols Commonly employed in Algebra with Their Meanings.
III. Powers and Roots. IV. Radicals. V. Logarithms. VI. Trigonometry. VII. Differen-
tial Calculus. VIII. Integral Calculus.
Cloth. 1O1 pages ivitli Diagrams, 75 cents.
This work is of great value to all students, but particularly to the students and
laymen who are deterred by mathematical formulae from reading otherwise intelligible
scientific works.
The book is specially designed for the beginner, and for all who have not been
able to avail themselves of a college education.
THE INTERPRETATION
OF
MATHEMATICAL FORMULAE
BY
EDWIN J. HOUSTON, PH.D. and A. E. KENNELLY, Sc.D.
CONTENTS.
CHAPTER I. Addition. II. Substraction . III. Multiplication. IV. Division. V. Involution.
Powers. VI. Evolution. Roots. VII. Equations. VIII. Logarithms. IX. Trigonometry.
X. Hyperbolic Trigonometrical Functions. XI. Differential Calculus. XII. Integral Calcu-
lus. XIII. Determinants. XIV. Synopsis of Symbols.
Cloth. 225 Pages, ages. Price, Cloth, $2. 5O ; Leather, $3.OO.
This Dictionary contains. 11,000 words and about 15,000 definitions used in the
science. There are several thousand technical words now used that have never ap-
peared in any other single book.
It is published in a convenient form for those who only seek brief definitions. The
pocket edition contains all the definitions of the larger work, but without the illustra-
tions and encyclopedic matter.
ELECTRICITY
ONE HUNDRED YEARS AGO AND TO=DAY,
BY EDWIN J. HOUSTON, PH.D. (Princeton).
Cloth. 199 pages, illustrated. Price, $1.OO.
In tracing the history of electrical science from practically its birth to the present
day, the author has, wherever possible, consulted original sources of information. As
a result of these researches several revisions as to the date of discovery of some im-
portant principles in electrical science are made necessary. While the compass of the
book does not permit of any other than a general treatment of the subject, yet numerous
references are given in foot notes, which also in many cases quote the words in which
a discovery was first announced to the world, or give more specific information in
regard to the subjects mentioned in the main portion of the book. This feature is one
of interest and value, for often a clearer idea may be obtained from the words of a
discoverer of a phenomenon or principle than is possible through other sources. The
work is not a mere catalogue of subjects and dates, nor is it couched in technical
language that only appeals to a few. On the contrary,- one of its most admirable
features is the agreeable style in which the work is written, its philosophical discussion
as to the cause and effect of various discoveries, and its personal references to great
names in electrical science. Much information as to electrical phenomena may also be
obtained from the book, as the author is not satisfied to merely give the history of a
discovery, but also adds a concise and clear explanation of it.
AflERICAN ELECTRICIAN COMPANY,
BEARD BUILDING, NEW YORK.
Practical Features of Telephone Work
BY A. H. DOBBS.
CONTENTS.
Pitfalls in Starting. Poor versus Good "Work. Starting a New Exchange. Wire. Aluminum Conduc-
tors. Weatherproof Wire. Country and Toll Lines. Exchange Lines and Circuits. Size of the
Return Wire. Locating Lines and Poles. Poles. Insulators, Guys, Bases, Etc. Cross Connec-
tion. Terminal Poles. Tree Trimming. Cables. Underground Conduits. Manholes. Elec-
trolysis. Fuses and Lightning Arresters. Selection of Instruments. Transmitter. Induction
Coil. Receiver. Wiring. Instrument and Line Troubles. Switchboards. Batteries. Cross
Connecting Boards. Exchange Management. Wire Tables and Formulas. Supporting Capacity
of Galvanized Strands.
134: Pages, 61 Illustrations. Price, 75 Cents.
The matter contained in this book is entirely practical in its bearing, and the result
of the author's experience covering fourteen years of active work. All branches of
practical telephonic construction are treated. Much attention is given to the pole line,
including its location, pole-setting, stringing conductors, transpositions, cross connec-
tions, corner and junction poles, etc. Underground construction and cable work form
the subject of several chapters. In the part devoted to the exchange, the switchboard
is treated, and a detailed description given of the central battery system. The book is
particularly adapted for those actually engaged in every day telephonic work, who,
from its pages, will be enabled to derive much information to assist them as new
problems arise.
EXPERIflENTS WITH
ALTERNATING CURRENTS
OF HIGH POTENTIAL AND HIGH FREQUENCY.
BY NIKOLA TESLA.
ClotJi. 14=6 pages, with Portrait and 35 Illus. Price, $1.OO.
Since the discovery of the telephone few researches in electricity have created as
widespread interest as those of Nikola Tcsla into alternate currents of high potential
and high frequency. The currents of enormously high frequency and voltage gener-
ated by "Mr. Tccla developed properties previously entirely unsuspected, and which
produced phenomena of startling character. The subject is popularly treated, and as
the author is the master of a simple and agreeable style the book is fascinating reading.
Copies of this or any other electrical book published will be sent by mail, postage prc*
id) to any address in the ivorld> on receipt of price.
AMERICAN ELECTRICIAN COflPANY,
BEARD BUILDING, NEW YORK.
ELECTRICITY MADE EASY
BY SIMPLE LANGUAGE AND COPIOUS ILLUSTRATION
BY
EDWIN J. HOUSTON, PH. D., and A. B. KKNNELL,Y, Sc. D.
CONTENTS.
CHAPTER T. The Turning of an Electric Lamp in the House. IT. How the Electric Wires n re Distributed
Through the House. HI. How the Electric Street Mains Supply the House. IV. How the Street
Mains are Supplied with Electricity. V. The Electric Lighting Station. VI. How the Incandescent
Lamp Operates. VII. How the Incandescent Lamp is Made. VIII. How the Electric Current Sup-
plied to the House is Measured. IX. How the Arc Lamp Operates. X. How the Light of Electric
Lamps is Best Distributed. XI.^The Voltaic Cell and How it Operates. XII. The Electric Bell
and How it Operates. XIII. The Electric Telegraph and How it Operates. XIV. How the Dynamo
Operates. XV. How the Electric Motor Operates. XVI. The Telephone aud How it Operates.
XVII. Some Other Applications of Electricity.
Cloth. 348 2>ages, 207 illustrations. Price, $1.25.
The authors have taken great pains to tell the story of electricity in a clear, com-
prehensive style so that beginners and laymen rannot fail to follow understandingly the
text. Many analogies are given which simplify what is usually a difficult technical
subject, and the book is entirely devoid of mathematics. Everyday operations in con-
nection with electrical apparatus, usually performed in a mechanical and wholly
unknowing spirit, are fully and clearly explained.
Recent Types
of
Dynamo Electric flachinery
BY
EDWIN J. HOUSTON, Ph. D. and A. E. KKNNKLLY, Sc. D.
Profusely Illustrated with over 600 Magnificent Engravings Iv the best known
Process, shoivn in Color t 'including Tables of exceptional value.
CONTENTS.
JHAPTER I. Introduction. II. Direct-Driven Continuous Current Generators for Isolated Plants. III.
Belt-Driven Continuous Current Generators for Isolated Plants. IV. Continuous Current Central
Station Generators. V. Central Station Arc Lamp Generators. VI. Some Miscellaneous Types of
Continuous Current Generators. VII. Alternating Current Generators. VII f. Multiphase Alterna-
tors. IX. Alternating Current Transformers. X. Continuous Current Motors. XI. Locomotors.
XII. Alternating Current Motors. XIII. Regulators for Alternating Currents Circuits. XIV.
Secondary Generators.
Cloth. 612 pages, 4=35 illustrations. Price, $4.00.
Although many books have been written on the subject of dynamo-electric machin-
ery, yet, so far as the authors are aware, none have yet appeared that have been
devoted entirely to American tvpes of machines. The book is not a treatise concerning
the principles of dynamo-electric machinery, or the theory of its operation, but a
description treatise of the various types of machines made by different manufacturers,
with their sizes, data, functions and capabilities.
AHERICAN ELECTRICIAN COMPANY,
BEARD BUILDING, NEW YORK.
Electrical Engineering Leaflets
BY
BY EDWIN J. HOUSTON, Pn.D. and A. E. KENNELLY, ScD.
In Three Grades.
Elementary Grade. 296 Pages, 121 Illus. Price, $1.5O.
I intermediate Grade. 3OO Pages, 14O Illus. Price, $1.5O.
Advanced Grade. 296 Pages, 121 Illus. Price, $1.5O.
This series has been prepared for the purpose of presenting, con-
cisely and accurately, the fundamental principles of electrical science as
applied in practical work. Each of the three grades is complete in itself,
though one may be used as a stepping stone to the next higher grade.
The Elementary Grade is intended for those electrical artisans, linemen,
motormen, central station operators or electrical mechanics generally,
i^vho have had no previous instruction in electrical science. Here the
i mathematical treatment is limited to arithmetic, and the principles are
illustrated by examples taken from actual practice. The Intermediate
I Grade is intended for those who have mastered the first volume of the
series, and for students of electricity in high schools and colleges. This
volume, moreover, contains such information concerning the science o
Electricity as should be acquired by those desiring general mental culture.
The Advanced Grade is designed for readers with some mathematical
preparation, and for students taking an electrical engineering course in
Colleges or universities.
Copies of this or any other electrical book published will be sent by mai? t
postage prepaid, to any address in the world, on receipt of price.
AMERICAN ELECTRICIAN COHPANY,
BEARD BUILDING, NEW YORK.
Elementary Electro=Technical Series.
BY
EDWIN]. HOUSTON, PH.D., and A. E. KENNELLY, Sc.D.
Alternating Electric Currents. Electric Incandescent Lighting.
Electric Heating. Electric Motor.
Electromagnetism. Electric Street Railways.
Electricity in Electro-Therapeutics. Electric Telephony.
Electric Arc Lighting. ELctric Telegraphy,
Cloth. Price per Volume, $1.0O.
The publication of this series of elementary electro-technical treatises
on applied electricity has been undertaken to meet a demand which is
believed to exist on the part of the public and others for reliable informa-
- tion regarding such matters in electricity as cannot be readily understood
by those not specially trained in electro-technics. The general public,
students of elementary electricity and the many interested in the subject
from a financial or other indirect connection, as well as electricians desir-
ing information in other branches than their own, will find in these works
precise and authoritative statements concerning the several branches of
applied electrical science of which the separate volumes treat. The repu-
tation of the authors and their recognized abilities as writers, are a
sufficient guarantee for the accuracy and reliability of the statements con-
tained. The entire issue, though published in a series of ten volumes, is
nevertheless so prepared that each book is complete in itself and can be
understood independently of the others. The volumes are profusely illus-
trated, printed on a superior quality of paper, and handsomely bound in
covers of a special design.
Copies of this or any other electrical book published will be sent by
postage prepaid, to any address in the world, on receipt of price. ,
AMERICAN ELECTRICIAN COflPANY,
BEARD BUILDING, NEW YORK.
AMERICAN
ELECTRICIAN
A Journal of Practical Electrical and Mechanical Engineering.
THE LARGEST PAID CIRCULATION OF ANY
ELECTRICAL JOURNAL IN THE WORLD. . .
A PRACTICAL PAPER FOR, PRACTICAL MEN.
ITS POLICY consists in printing only matter of in-
trinsic value, prepared by thoroughly competent
writers, and presented without a burden of theoretical
discussion or mathematical analysis, and yet without
sacrifice in accuracy or thoroughness. It has solved the
problem of a practical journal appealing alike to the pro-
fessional graduate and to those who have not had the
advantage of a technical education.
Among its features are descriptions of Central
Station, Electric Railway and Transmission Plants,
articles on Steam and Mechanical Engineering, Interior
Wiring, Telephone Practice, Construction of Apparatus,
Electric Measurements and numerous other subjects of
direct practical interest.
SUBSCRIPTION PRICE, $1.00 PER YEAR.
AMERICAN ELECTRICIAN COMPANY,
Beard Building. NEW YORK.
THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW
AN INITIAL FINE OF 25 CENTS
WILL BE ASSESSED FOR FAILURE TO RETURN
THIS BOOK ON THE DATE DUE. THE PENALTY
WILL INCREASE TO SO CENTS ON THE FOURTH
DAY AND TO $1.OO ON THE SEVENTH DAY
OVERDUE.
sJA,i 25 1939
APR 5 1943
--,
.d5
n*
8Aor'59BP
REG'D LD
MAR 2 5 1959
LD 21-95m-7,'37
0409
YC