[Footnote: From a paper read before the recent meeting of the American
Institute of Electrical Engineers, New York, and reported in the
_Electrical World_.]
By Prof. ELIHU THOMSON.
The actions produced and producible by the agency of alternating
currents of considerable energy are assuming greater importance in the
electric arts. I mean, of course, by the term alternating currents,
currents of electricity reversed at frequent intervals, so that a
positive flow is succeeded by a negative flow, and that again by a
positive flow, such reversals occurring many times in a second, so that
the curve of current of electromotive force will, if plotted, be a wave
line, the amplitude of which is the arithmetical sum of the positive and
negative maxima of current or electromotive force, as the case may be,
while a horizontal middle line joins the zero points of current or
electromotive force.
[Illustration: FIG. 1]
It is well known that such a current passing in a coil or conductor laid
parallel with or in inductive relation to a second coil or conductor,
will induce in the second conductor, if on open circuit, alternating
electromotive forces, and that if its terminals be closed or joined,
alternating currents of the same rhythm, period, or pitch, will
circulate in the second conductor. This is the action occurring in any
induction coil whose primary wire is traversed by alternating currents,
and whose secondary wire is closed either upon itself directly or
through a resistance. What I desire to draw attention to in the present
paper are the mechanical actions of attraction and repulsion which will
be exhibited between the two conductors, and the novel results which may
be obtained by modifications in the relative dispositions of the two
conductors.
[Illustration: FIG. 2.]
In 1884, while preparing for the International Electrical Exhibition at
Philadelphia, we had occasion to construct a large electro-magnet, the
cores of which were about six inches in diameter and about twenty inches
long. They were made of bundles of iron rod of about 5/16 inch diameter.
When complete, the magnet was energized by the current of a dynamo
giving continuous currents, and it exhibited the usual powerful magnetic
effects. It was found also that a disk of sheet copper, of about 1/16
inch thickness and 10 inches in diameter, if dropped flat against a pole
of the magnet, would settle down softly upon it, being retarded by the
development of currents in the disk due to its movement in a strong
magnetic field, and which currents were of opposite direction to those
in the coils of the magnet. In fact, it was impossible to strike the
magnet pole a sharp blow with the disk, even when the attempt was made
by holding one edge of the disk in the hand and bringing it down
forcibly toward the magnet. In attempting to raise the disk quickly off
the pole, a similar but opposite action of resistance to movement took
place, showing the development of currents in the same direction to
those in the coils of the magnet, and which currents, of course, would
cause attraction as a result.
[Illustration: Fig. 3]
The experiment was, however, varied, as in Fig. 1. The disk, D, was held
over the magnet pole, as shown, and the current in the magnet coils cut
off by shunting them. There was felt an attraction of the disk or a dip
toward the pole. The current was then put on by opening the shunting
switch, and a repulsive action or lift of the disk was felt. The actions
just described are what would be expected in such a case, for when
attraction took place, currents had been induced in the disk, D, in the
same direction as those in the magnet coils beneath it, and when
repulsion took place the induced current in the disk was of opposite
character or direction to that in the coils.
[Illustration: Fig. 4]
Now let us imagine the current in the magnet coils to be not only cut
off, but reversed back and forth.
For the reasons just given, we will find that the disk, D, is attracted
and repelled alternately; for, whenever the currents induced in it are
of the same direction with those in the inducing or magnet coil,
attraction will ensue, and when they are opposite in direction,
repulsion will be produced. Moreover, the repulsion will be produced
when the current in the magnet coil is rising to a maximum in either
direction, and attraction will be the result when the current of either
direction is falling to zero, since in the former case opposite currents
are induced in the disk, D, in accordance with well known laws, and in
the latter case currents of the same direction will exist in the disk,
D, and the magnet coil. The disk might, of course, be replaced by a ring
of copper or other good conductor, or by a closed coil of bare or
insulated wire, or by a series of disks, rings or coils superposed, and
the results would be the same. Thus far, indeed, we have nothing of a
particularly novel character, and, doubtless, other experimenters have
made very similar experiments and noted similar results to those
described.
[Illustration: FIG. 5]
The account just given of the effects produced by alternating currents,
while true, is not the whole truth, and just here we may supplement it
by the following statements:
_An alternating current circuit or coil repels and attracts a closed
circuit or coil placed in direct or magnetic inductive relation
therewith; but the repulsive effect is in excess of the attractive
effect.
When the closed circuit or coil is so placed, and is of such low
resistance metal that a comparatively large current can circulate as an
induced current, so as to be subject to a large self-induction, the
repulsive far exceeds the attractive effort_.
For want of a better name, I shall call this excess of repulsive effect
the "electro-inductive repulsion" of the coils or circuits.
[Illustration: FIG. 6.]
This preponderating repulsive effect may be utilized or may show its
presence by producing movement or pressure in a given direction, by
producing angular deflection as of a pivoted body, or by producing
continuous rotation with a properly organized structure. Some of the
simple devices realizing the conditions I will now describe.
[Illustration: FIG. 7.]
In Fig. 2, C is a coil traversed by alternating currents. B is a copper
case or tube surrounding it, but not exactly over its center. The copper
tube, B, is fairly massive and is the seat of heavy induced currents.
There is a preponderance of repulsive action, tending to force the two
conductors apart in an axial line. The part, B, may be replaced by
concentric tubes slid one in the other, or by a pile of flat rings, or
by a closed coil of coarse or fine wire insulated, or not. If the coil,
C, or primary coil, is provided with an iron core such as a bundle of
fine iron wires, the effects are greatly increased in intensity, and the
repulsion with a strong primary current may become quite vigorous, many
pounds of thrust being producible by apparatus of quite moderate size.
The forms and relations of the two parts, C and B, may be greatly
modified, with the general result of a preponderance of repulsive action
when the alternating currents circulate.
Fig. 3 shows the part, B, of an internally tapered or coned form, and C
of an externally coned form, wound on an iron wire bundle, I. The action
in Fig. 2 may be said to be analogous to that of a plain solenoid with
its core, except that repulsion, and not attraction, is produced, while
that of Fig. 3 is more like the action of tapered or conically wound
solenoids and taper cores. Of course, it is unnecessary that both be
tapered. The effect of such shaping is simply to modify the range of
action and the amount of repulsive effort existing at different parts of
the range.
[Illustration: FIG. 8.]
In Fig. 4 the arrangement is modified so that the coil, C, is outside,
and the closed band or circuit, B, inside and around the core, I.
Electro-inductive repulsion is produced as before.
It will be evident that the repulsive actions will not be mechanically
manifested by axial movement or effort when the electrical middles of
the coils or circuits are coincident. In cylindrical coils in which the
current is uniformly distributed through all the parts of the conductor
section, what I here term the electrical middle, or the center of
gravity of the ampere turns of the coils, will be the plane at right
angles to its axis at its middle, that of B and C, in Fig. 4, being
indicated by a dotted line. To repeat, then, when the centers or center
planes of the conductors, Fig. 4, coincide, no indication of
electro-inductive repulsion is given, because it is mutually balanced in
all directions; but when the coils are displaced, a repulsion is
manifested, which reaches a maximum at a position depending on the
peculiarities of proportion and distribution of current at any time in
the two circuits or conductors.
[Illustration: FIG. 9.]
It is not my purpose now to discuss the ways of determining the
distribution of currents and mechanical effects, as that would extend
the present paper much beyond its intended limit. The forms and relative
arrangement of the two conductors may be greatly varied. In Fig. 5 the
parts are of equal diameter, one, B, being a closed ring, and the other,
C, being an annular coil placed parallel thereto; and an iron core or
wire bundle placed in the common axis of the two coils increases the
repulsive action. B may be simply a disk or plate of any form, without
greatly affecting the nature of the action produced. It may also be
composed of a pile of copper washers or a coil of wire, as before
indicated.
[Illustration: FIG. 10.]
An arrangement of parts somewhat analogous to that of a horseshoe
electro magnet and armature is shown in Fig. 6. The alternating current
coils, C C', are wound upon an iron wire bundle bent into U form, and
opposite its poles is placed a pair of thick copper disks, B B', which
are attracted and repelled, but with an excess of repulsion depending on
their form, thickness, etc.
[Illustration: FIG. 11.]
If the iron core takes the form of that shown by I I, Fig. 7, such as a
cut ring with the coil, C, wound thereon, the insertion of a heavy
copper plate, B, into the slot or divided portion of the ring will be
opposed by a repulsive effort when alternating currents pass in C. This
was the first form of device in which I noticed the phenomenon of
repulsive preponderance in question. The tendency is to thrust the
plate, B, out of the slot in the ring excepting only when its center is
coincident with the magnetic axis joining the poles of the ring between
which B is placed.
If the axes of the conductors, Fig. 5, are not coincident, but
displaced, as in Fig. 8, then, besides a simple repulsion apart, there
is a lateral component or tendency, as indicated by the arrows. Akin to
this is the experiment illustrated in Fig. 9. Here the closed conductor,
B, is placed with its plane at right angles to that of C, wound on a
wire bundle. The part, B, tends to move toward the center of the coil,
C, so that its axis will be in the middle plane of C, transverse to the
core, as indicated by the dotted line. This leads us at once to another
class of actions, i.e., deflective actions.
[Illustration: FIG. 12.]
When one of the conductors, as B, Fig. 10, composed of a disk, or,
better, of a pile of thin copper disks, or of a closed coil of wire, is
mounted on an axis, X, transverse to the axis of coil, C, through which
coil the alternating current passes, a deflection of B to the position
indicated by dotted lines will take place, unless the plane of B is at
the start exactly coincident with that of C. If slightly inclined at the
start, deflection will be caused as stated. It matters not whether the
coil, C, incloses the part, B, or be inclosed by it, or whether the
coil, C, be pivoted and B fixed, or both be pivoted. In Fig. 11 the
coil, C, surrounds an iron wire core, and B is pivoted above it, as
shown. It is deflected, as before, to the position indicated in dotted
lines.
[Illustration: FIG. 13]
It is important to remark here that in cases where deflection is to be
obtained, as in Figs. 10 and 11, B had best be made of a pile of thin
washers or a closed coil of insulated wire instead of a solid ring. This
avoids the lessening of effect which would come from the induction of
currents in the ring, B, in other directions than parallel to its
circumference.
[Illustration: FIG. 14.]
We will now turn our attention to the explanation of the actions
exhibited, and afterward refer to their possible applications. It may be
stated as certainly true that were the induced currents in the closed
conductor unaffected by any self-induction, the only phenomena exhibited
would be alternate equal attractions and repulsions, because currents
would be induced in opposite directions to that of the primary current
when the latter current was changing from zero to maximum positive or
negative current, so producing repulsion; and would be induced in the
same direction when changing from maximum positive or negative value to
zero, so producing attraction.
This condition can be illustrated by a diagram, Fig. 12. Here the lines
of zero current are the horizontal straight lines. The wavy lines
represent the variations of current strength in each conductor, the
current in one direction being indicated by that portion of the curve
above the zero line, and in the other direction by that portion below
it. The vertical dotted lines simply mark off corresponding portions of
phase or succession of times.
[Illustration: FIG. 15]
Here it will be seen that in the positive primary current descending
from m, its maximum, to the zero line, the secondary current has risen
from its zero to m¹, its maximum. Attraction will therefore ensue, for
the currents are in the same direction in the two conductors. When the
primary current increases from zero to its negative maximum, n, the
positive current in the secondary closed circuit will be decreasing from
m¹, its positive maximum, to zero; but, as the currents are in opposite
directions, repulsion will occur. These actions of attraction and
repulsion will be reproduced continually, there being a repulsion, then
an attraction, then a repulsion, and again an attraction, during one
complete wave of the primary current. The letters, r, a, at the foot of
the diagram, Fig. 12, indicate this succession.
In reality, however, the effects of self-induction in causing a lag,
shift, or retardation of phase in the secondary current will
considerably modify the results, and especially so when the secondary
conductor is constructed so as to give to such self-induction a large
value. In other words, the maxima of the primary or inducing current
will no longer be found coincident with the zero points of the secondary
currents. The effect will be the same as if the line representing the
wave of the secondary current in Fig. 12 had been shifted forward to a
greater or less extent. This is indicated in diagram, Fig. 13. It gives
doubtless an exaggerated view of the action, though from the effects of
repulsion which I have produced, I should say it is by no means an
unrealizable condition.
[Illustration: Fig. 16.]
It will be noticed that the period during which the currents are
opposite, and during which repulsion can take place, is lengthened at
the expense of the period during which the currents are in the same
direction for attractive action. These differing periods are marked r,
a, etc., or the period during which _repulsion_ exists is from the zero
of the primary or inducing current to the succeeding zero of the
secondary or induced current; and the period during which _attraction_
exists is from the zero of the induced current to the zero of inducing
current.
But far more important still in giving prominence to the repulsive
effect than this difference of effective period is the fact that during
the period of repulsion both the inducing and induced currents have
their greatest values, while during the period of attraction the
currents are of small amounts comparatively. This condition may be
otherwise expressed by saying that the period during which repulsion
occurs includes all the maxima of current, while the period of
attraction includes no maxima. There is then a _repulsion due to the
summative effects of strong opposite currents_ for a _lengthened
period_, against an _attraction_ due to the summative effects of _weak
currents_ of the _same direction_ during a _shortened period_, the
resultant effect being a greatly _preponderating_ repulsion.
It is now not difficult to understand all the actions before described
as obtained with the varied relations of coils, magnetic fields, and
closed circuits. It will be easily understood, also, that an alternating
magnetic field is in all respects the same as an alternating current
coil in producing repulsion on the closed conductor, because the
repulsions between the two conductors are the result of magnetic
repulsions arising from opposing fields produced by the coils when the
currents are of opposite directions in them.
Thus far I have applied the repulsive action described in the
construction of alternating current indicators, alternating current arc
lamps, regulating devices for alternating currents, and to rotary motors
for such currents. For current indicators, a pivoted or suspended copper
band or ring composed of thin washers piled together and insulated from
one another, and made to carry a pointer or index has been placed in the
axis of a coil conveying alternating currents whose amount or potential
is to be indicated. Gravity or a spring is used to bring the index to
the zero of a divided scale, at which time the plane of the copper ring
or band makes an angle of, say, 15 degrees to 20 degrees with the plane
of the coil. This angle is increased by deflection more or less great,
according to the current traversing the coil. The instrument can be
calibrated for set conditions of use. Time would not permit of a full
description of these arrangements as made up to the present.
In arc lamps the magnet for forming the arc can be composed of a closed
conductor, a coil for the passage of current, and an iron wire core. The
repulsive action upon the closed conductor lifts and regulates the
carbons in much the same manner that electro magnets do when continuous
currents are used. The electro-inductive repulsive action has also been
applied to regulating devices for alternating currents, with the details
of which I cannot now deal.
For the construction of an alternating current motor which can be
started from a state of rest the principle has also been applied, and it
may here be remarked that a number of designs of such motors is
practicable.
One of the simplest is as follows: The coils, C, Fig. 14, are traversed
by an alternating current and are placed over a coil, B, mounted upon a
horizontal axis, transverse to the axis of the coil, C. The terminals of
the coil, B, which is wound with insulated wire, are carried to a
commutator, the brushes being connected by a wire, as indicated. The
commutator is so constructed as to keep the coil, B, on short circuit
from the position of coincidence with the plane of C to the position
where the plane of B is at right angles to that of C; and to keep the
coil, B, open-circuited from the right-angled position, or thereabouts,
to the position of parallel or coincident planes. The deflective
repulsion exhibited by B will, when its circuit is completed by the
commutator and brushes, as described, act to place its plane at right
angles to that of C; but being then open-circuited, its momentum carries
it to the position just past parallelism, at which moment it is again
short-circuited, and so on. It is capable of very rapid rotation, but
its energy is small. I have, however, extended the principle to the
construction of more complete apparatus. One form has its revolving
portion or armature composed of a number of sheet iron disks wound as
usual with three coils crossing near the shaft. The commutator is
arranged to short-circuit each of these coils in succession, and twice
in a revolution, and for a period of 90-degrees of rotation each. The
field coils surround the armature, and there is a laminated iron field
structure completing the magnetic circuit. I may say here that
surrounding the armature of a dynamo by the field coils, though very
recently put forth as a new departure, was described in various
Thomson-Houston patents, and to a certain extent all Thomson-Houston
machines embody this feature.
Figs. 15 and 16 will give an idea of the construction of the motor
referred to. CC' are the field coils or inducing coils, which alone are
put into the alternating current circuit. II is a mass of laminated
iron, in the interior of which the armature revolves, with its three
coils, B, B², B³, wound on a core of sheet iron disks. The commutator
short-circuits the armature coils in succession in the proper positions
to utilize the repulsive effect set up by the currents which are induced
in them by the alternations in the field coils. The motor has no dead
point, and will start from a state of rest and give out considerable
power, but with what economy is not yet known.
A curious property of the machine is that at a certain speed, depending
on the rapidity of the alternations in the coil, C, a continuous current
passes from one commutator brush to the other, and it will energize
electro magnets and perform other actions of direct currents. Here we
have, then, a means of inducing direct currents from alternating
currents. To control the speed and keep it at that required for the
purpose, we have only to properly gear the motor to another of the
ordinary type for alternating currents, namely, an alternating-current
dynamo used as a motor. The charging of storage batteries would not be
difficult with such a machine, even from an alternating-current line,
though the losses might be considerable.
* * * * *