Surge Testing Of DC Motor And Generator Armatures

of the armature
winding, as shown
in Figure 1. The
surge pulse is sent
through one section
of the winding, and
then an equal pulse is
sent through another.
The setup has three
brushes, or probes.
One is common and
is positioned in the
middle of the other
two. The surge pulse
By Cyndi Nyberg
is sent from the
EASA Technical Support Specialist
common to each of
the outside probes. Figure 1. Armature surge test.
In the April 2007 issue of The waveforms
CURRENTS, we covered surge testing resulting from the two surge pulses
anomalies, specifically for AC windings. are then superimposed on an oscil-
The surge test can be used for DC loscope screen and, if they are equal,
windings as well. It can be a useful only a single trace appears. If a second
tool for evaluating armatures and some trace can be seen, as shown in Figure
DC fields. 2, there is some inequality in the two
A note of caution: If a winding does sections being tested. This would
not have a minimum insulation resis- indicate a winding fault such as a
tance per ANSI/EASA AR100-2006, it ground, short or open; or the winding
is not safe to apply an overpotential has unequal turns and the bars spanned Figure 2. Bad surge test pattern.
test (surge or high potential). need to be changed to obtain an equal
Some surge tester manufacturers
Surge testing shunt fields may number of turns across the span; or an
offer a high-current booster attachment
not provide meaningful results if the equalized armature does not have the
to drive the surge pulse through the ar-
surge pulse decays too quickly — if same number of equalizers in the two
mature windings. One manufacturerʼs
it dissipates through only the first few sections being compared.
surge tester has a capacitor in the back
hundred turns. To obtain a test voltage As stated above, the surge test sends
of the unit with a separate lead position
high enough to test every turn would a pulse from the common to the two
for testing armatures; that supplies the
require too high a voltage. That high outside probes. However, the surge pulse
required current for the test.
voltage would overstress the ground- also travels from the common all the
wall insulation. way around the armature (CW) to the Voltage Level
outside of the “left” probe, and from the The bar-to-bar test voltage for the
Surge Testing Armatures
common around in the opposite direction surge test should be about 350-500 volts
DC motor or generator armatures
(CCW) to the “right” probe. Although per bar. However, to avoid over-stressing
can be tested with a surge tester. Al-
the strength of the surge pulse will dis- the groundwall insulation, the maximum
though the winding of an armature is
sipate through the armature winding, total surge test voltage should be 1500
different than that of an AC stator or
a short in the armature will still show volts for armatures rated less than 500
wound rotor, it will still be symmetrical
up on the surge trace. To prove this, try volts, and 2000 volts for armatures rated
if the winding is good.
shorting two bars outside the area under 500 volts and above. The relatively high
Surge comparison testing of an
test; the trace will show a separation. voltage peak of the surge makes it quite
armature, similar to the high-frequency
bar-to-bar test, compares two sections Continued On Page 4

EASA CURRENTS ■ July 2007 3

effective for probing the winding for only of the product of the distance
faults like cracked insulation or ther- between parallel plane electrodes. Further information on
mally degraded insulation or insufficient So how does this relate to surge testing of DC machines
voltage creepage clearances. testing? Stated a little differently, can be found in the
Paschenʼs Law means that an applied “Fundamentals of DC
Surge Comparison Test Example
voltage cannot bridge the gap of two Operation and Repair
Each set of bar-to-bar test probes
flat plates (i.e., shorted turns) with a Tips” manual, and in
must span an equal number of bars
certain gap between them until the several articles under
for a meaningful test. Otherwise the
voltage is raised above a minimum “Technical Articles” in
two compared sections will not be
level. This minimum is 350 volts per the “Members Only”
equal. Further, the total number of
turn. A surge test, at the proper voltage section of EASA’s Web
turns being compared must be equal.
level, is the testing method that can do site at www.easa.com.
If the armature has an unequal turn
this between the turns.
sequence, then the surge pattern in
the oscilloscope may show separa- Example:
tion. It may be necessary to adjust the For a 500-volt armature: is rotated slowly by hand, so that all
number of bars spanned to obtain a Surge test voltage = (2 x rated sections of the armature are compared
good pattern, as long as the minimum voltage) + 1000 = (2 x 500) + 1000 to the others.
voltage per turn is maintained. = 2000 volts
Series Fields And Interpoles
Also, when an armature is equal- This voltage applies to a new While the surge test may not be
ized, the sides under test must contain winding as well as one that has been a useful tool for evaluating shunt
an equal number of equalizers. In in service, since this is considered a fields, series fields and interpoles
either case (unequal turns or equaliz- non-destructive test. can be surge tested. The maximum
ers) the result is a pattern that changes Since we need to exceed Paschenʼs surge test voltage should not exceed
at regular intervals. Again, adjustment Law, determine the number of com- the AC high potential test voltage
of the number of spanned bars may be mutator bars spanned on each side of — twice the rated (armature) volt-
necessary. the test probes. The volts per bar will age plus 1000 volts for most DC
Paschen’s Law Requirements be the minimum 350 volts per turn, machines. Use the surge tester to
The test voltage must also meet divided by the number of turns in the compare each series field or inter-
the requirements of Paschenʼs Law winding. If the number of turns is not pole to the one on each side of it.
(i.e., at least 350 volts per turn) known, then assume that it has one That way, each one is tested twice.
without exceeding the test voltage for turn. Using our 2000 volts, and 350 If each pattern on the scope is iden-
the ground insulation. Paschenʼs Law volts per bar for a 1 turn coil: tical, then the fields or interpoles are
states that, at a constant temperature, 2000 volts 2000 volts not shorted.
= = 5.7
the breakdown voltage is a function Minimum volts per bar 350
Synchronous Rotors
Only 5 bars Depending on the number of
should be spanned turns in each rotor pole, the same
on each side of the issues arise with surge testing as
probes. Figure 3 would be present with shunt fields.
4 3 2 1 2 3 4 shows the setup for Further information on testing
5 5 this example. The of DC machines can be found in
surge pulse is sent the “Fundamentals of DC Opera-
from the center tion and Repair Tips” manual, and
through the 5 bars, in several articles under “Technical
and then through the Articles” in the “Members Only”
rest of the wind- section of EASAʼs Web site at
Figure 3. Armature surge comparison test. ing. The armature www.easa.com.

4 EASA CURRENTS ■ www.easa.com