THE COLLIERY GUARDIAN
AND
JOURNAL OF THE COAL AND IRON TRADES.

Vol. CXV.	FRIDAY, JANUARY 11, 1918.	No. 2976.

The Care of Commutators of Direct=Current Machinery.

By L. FOKES.

The weakness of all direct-current machinery lies in
the use of a commutator, and this is specially the case
where the conditions of service are severe, as in all
colliery work.

The simplicity of operation, speed control, and
efficiency of direct-current motors places them far
ahead of induction motors; and, in spite of the draw-
backs of commutators, a large amount of direct-
current machinery is in use throughout the coal fields,
and as the design of this class of apparatus has
reached such a high state of perfection, little or no
trouble should be experienced if reasonable attention
be paid to the details which make for the smooth and
efficient working of the commutator.

It cannot, however, be said that troubles are
unknown, as cases of oversight or neglect are unavoid-
able in everyday practice, and must always be
reckoned with.

The case of commutators not only implies proper
attention to repairs, etc., but more specially to points
which will obviate the necessity for repairs.

Simple Commutation.

Without going into the details of commutation
itself, it will be sufficient to explain that commuta-
tion must take place when the armature conductors
connected to the segments under the brushes are in
what is termed the neutral zone, or in such a position
that no electro-motive force is induced in them.
Obviously, that position will be between the pole pieces.
Owing, however, to armature reaction disturbing the
field, the neutral position is not equidistant from the
poles, but, in the case of a generator, tends to set
itself towards the direction of rotation. In the case
of a motor, the reverse occurs, i.e., the neutral posi-
tion shifts against the direction of rotation.

In any case, the brush position must be such as to
comply with the rule already mentioned, and it often
becomes necessary, under certain conditions of load,
to adjust the position of the brushes in order to get
them into the neutral position. When machines are
fitted with commutating poles, the position of the
brushes is set, and does not require to be altered to
meet varying loads.

Carbon Brushes.

At the outset, one may ask for what reason is carbon
used in practically all brushes for direct-current
machinery.

A correct understanding of this often enables one to
remedy troubles without difficulty, whereas otherwise
their origin would never be discovered. If we assume
for the moment that a machine is equipped with metal
brushes of negligible resistance, and that the path
through the armature from one commutator bar to
the next is of 0*05 ohm resistance, then, if the position
of the brushes is such that, when the segments are
short-circuited by a brush, a voltage of 5 volts exists
across adjacent segments, a short-circuit current of

C = l = W = 100amps-

will be induced in that armature coil.

It must be remembered that the brush has to carry
this current across its face, as well as deliver the load
current, and therefore it would not be surprising to
find it sparking badly.

In consequence of a number of factors into which it
is unnecessary to enter, the short-circuit voltage of
armature coils as they come under the brushes varies
from 4 to 5 volts in commutating pole machines, and
in non-commutating types it is almost impossible to
work always exactly in the neutral position, even if
such a position actually exists.

It is due to this short-circuit electro-motive force
that carbon brushes have been universally adopted, as
by their relatively high contact resistance the short-
circuit current is kept within safe limits. The com-
mutating conditions vary to such an extent between
different machines, that numerous grades of carbon
brushes are now available, and the proper selection of
brushes for a given machine often decides whether it
will work or not. The question of carbon brushes will
be again referred to when dealing with the various
causes of bad commutation.

Sparking at Brush Contacts.

This is direct evidence of bad commutation, and
many are the causes which contribute directly and
indirectly to its existence. A few of these will now be
briefly outlined, together with their remedy.

Wrong Position of Brushes.—This is probably the
first thing one looks to on a machine which is sparking,
and often it can be stopped by moving the brush rocker
one way or the other, so as to bring the brushes into a
more favourable position for commutation. This does
not, however, always cure the trouble.

Insuffidient Tension on Brushes. — This does not
appear at first sight to be serious, but it often leads
to bad sparking. The tension on all the brushes
should be as near as possible alike, as otherwise what

is known as selective commutation takes place, i.e.,
the brushes with most tension carry the most current,
while those with little or no tension carry scarcely any
current. Sparking wi}l probably occur with all.

Those under strong pressure will carry a heavy
current, which will cause them to overheat, and as
carbon has a negative temperature coefficient, the over-
heating lowers their resistance so that still more
current will flow in their direction.

Brushes with little or no tension would spark,
because the contact would probably be so bad that the
little current being carried would continually be inter-
rupted.

Particular attention should always be paid to
brushes which are working on the underside of the
commutator, as the springs need to be tighter than
those on the top, since they have to support thenveight
of the brushes, besides supplying the necessary tension.

Bad or Broken Brush Shunts.—The shunt of a brush
is the copper connection provided for the purpose of
conveying the current from the brush to the brush
arm. On no account should the holders of brushes be

Fig. 1.

Fig. 2.

allowed to perform the office for which the shunts were
designed. Broken or disconnected shunts often burn
the brush holders, and cause brushes to become fast,
besides causing unequal distribution of current between
the various brush holders on the same arm.

The shunts should receive special attention, and
preferably, on any brush arm or number of arms of the
same polarity, all the shunts should be of the same size
and length, otherwise the selective commutation
already referred to will result, and sparking is likely
to take place.

Brush Resistance Too Low. — Sometimes, after
checking the points already mentioned, sparking still
continues, and therefore the cause must be looked for
in other directions. If the load on the machine is

Fig. 3.

TOil TION

’OM M UTA TOW

known to be normal, and the brushes attain excessive
•temperature, regardless of position, then the trouble
may be looked for in the brushes themselves.

If the brushes be too soft, and their resistance too
low, then the heating is caused by the short-circuit
current reaching too high a value, and the remedy is
to obtain harder and higher-resistance brushes in
order to overcome the difficulty.

It is possible for the short-circuit current to attain
such high values that the brushes become red hot,
besides the armature coils attaining an excessive tem-
perature. The writer has heard of a case where the
short-circuit current attained such proportions that
the machine in question had to undergo structural
alterations, which, however, made no difference; but
the difficulty was eventually overcome by cutting slots,
in the brushes, parallel to the commutator segments.
The slots were cut deep by a hack-saw, and so intro-
duced a much higher resistance across the face of the
brush from bar to bar, but made little or no difference
to the resistance in the path of the load current passing
through the brushes.

High Mica and Under-Cutting.—The mica used in
the construction of commutators is usually of a very

hard nature, and, in fact, so hard that the copper bars
frequently wear down aijd leave the mica projecting.
This is a most fruitful source of sparking, but it has
become the general practice now to undercut the mica
between the barf? to avoid this difficulty. It may also
be overcome by using a harder and a more abrasive
type of brush, which will keep the mica down level
with the commutator.

The under-cutting of mica is often carried too far,
with the result that one source of trouble is removed,
but is replaced by another, the effects of which may do
more damage than high mica.

The depth to which the mica between the bars should
•be cut ought not to exceed V32in. If the slots are cut
too deep they readily become filled with foreign matter,
usually copper or carbon dust.

Under-cutting may be carried out by a special
machine designed for the purpose, but usually, on a
colliery, the point of a small file meets the case. It is
difficult to start the slots, but, once started, there is
no trouble in scraping the mica down.

If the mica be scraped too low, and the copper of
the commutator bars be soft, there is sometimes a
tendency to burr over, as shown in fig. 1. This occurs
on the trailing side of the bar, and may actually bridge
the slots where the distance is very small. To avoid
this as far as possible, the slots should be slightly
bevelled, as shown in fig. 2.

The most common cause of trouble with deep slots is
what has been referred to as ring-fire. A careful
examination of a running machine with this trouble
shows at once that the sparks observed do not come
from the brushes; in fact, they are brightest when equi-
distant from the brush arms. The sparking is caused
by short circuits between adjacent bars, due to the
difference of potential bridging across accumulations
of dirt or oil, which readily fill the slots when the
latter are very deep.

A reference to fig. 3 will make it clear as to why this
sparking occurs at points on the commutator farthest
removed from the brushes. Remembering that the
correct position for commutating is that in which
little or no difference of potential exists between seg-
ments, the suggestion at once arises that, across the
segments in front of and behind the brushes, the differ-
ence of potential across segments must be higher, as
they will be connected to armature conductors
approaching the poles of the machine. From this it
follows that, as we trace the segments round from a
given set of brushes, their corresponding armature
coils will be found distributed under the pole piece, the
centre segment between two brush positions, reckon-
ing round the commutator, coinciding with the centre
of the pole. Now, we know that the maximum flux
from a pole is about its centre, so that the segment
referred to will have a maximum electro-motive force
between itself and the one on either side, the electro-
motive force gradually shading off in each direction
down to the next brush position. This curve of electro-
motive force is shown dotted in fig. 3, and it will be
noticed that the maximum electro-motive force is
between brush arms, and across segments, the corre-
sponding coils of which are under the centre of a pole
piece.

It is assumed in the figure that there is no arma-
ture reaction, and therefore the field is not distorted.
The curve is not of the shape found on electrical
machines, but is suitable to illustrate the point.

Ring-fire appears more often on comparatively slow-
speed machines, whilst it is almost unknown on those
run at high speeds, the reason being that a spark has
not sufficient time to establish itself before it is out
of the sparking zone, or that region in which the differ-
ence of potential between segments is a maximum.
Ring-fire in itself is not dangerous, but it may in time
develop insulation troubles between the commutator
segments.

Nothing causes this arcing so quickly as oil getting
on the commutator and between the segments, whether
by accident or design. The oil attacks the binding of
the mica in the slots, besides collecting copper and
carbon dust, which sets up arcing, and burns the mica
away in small holes. These latter sometimes become
very deep, in which case serious trouble may result,
and perhaps the only cure would be to dismantle the
commutator and rebuild it. This may, however, be
avoided if ring-fire is dealt with promptly. Its source
should be traced, and, if small holes appear, they
should be thoroughly cleared of carbonised matter,
and filled in with dental cement, which has been found
to be very suitable for the purpose.

Spacing Brushes.

It is advisable, when setting.brushes, to stagger them,
/.e., the brushes on the several arms should not trail.
The idea is to cause even wear all over the commutator,
as far as possible. This is difficult to do in two-pole
machines, as there are only two sets of brushes, one
negative and one positive.

The positive and negative brushes do not affect the
commutator in the same way. In the case of a gener-
ator, the positive brush is negative to the commutator,
and electrolytic action takes place, depositing copper
upon the positive brush, so that it will usually be
found that the positive brushes wear the commutator