Version 1.
1
August, 2004
Table of Contents
Introduction. ……………………………..……………………………. 3
Theory of Operation. …………………………………………………. 4
Equipment Required. .……………………………..…………………. 6
Instrument Settings. …………………………..……………………… 7
Oscilloscope Settings. ………………….….………...……………... 8
Rotor Connections. ……………………………….………………….. 9
Final Calibration Procedure. ………………………………...……… 10
Instrument Diagnostics. …………………………..…………….….…11
Performing a Spinning Test. ………………………………………….15
Repair & Warranty. ……………………………………………………...16
2
Introduction
Shorted turns in generator cylindrical rotor field windings can contribute to vibration
problems due to rotor thermal bending from the uneven heating associated with non-
symmetrical dc current flow and watt losses in the windings. Shorted turns can also
cause unbalanced magnetic flux in the air gap that can also aggravate vibration
problems.
Since vibration signature analyses for rotor shorted turn problems is not always an
exact science, it is desirable to have confirming data from other testing before
proceeding with very costly disassemblies and repairs of large machines. Additional
tests for confirming the existence of shorted turns in generator rotor field windings
are commonly performed before committing to expensive repairs. At this time, the
following four test procedures are generally used in the industry to help verify
whether or not generator field windings have shorted turns:
1. Thermal Stability Testing - involves changing generator-operating
parameters (watts, vars, and cooling) and recording and analyzing the impact
on rotor vibration signatures.
2. Field Current Open Circuit Testing – If accurate pre-vibration electrical
measurements have been recorded, precise field current measurements can be
compared to the historical data; if there is a shorted turn, the field current will
be higher for an identical terminal voltage. If the machine is loaded, it is much
more difficult to compare, since var flow is dependent on system voltage
levels.
3. Flux Probe Analysis - utilizes an installed air gap probe to measure and
analyze the magnetic flux from each rotor slot as it passes by the location of
the sensor. Some generators are permanently equipped with flux probes and
many are not. Installing the probe normally requires a unit outage, especially
with hydrogen-cooled machines.
4. RSO (Repetitive Surge Oscilloscope) Testing – which is the applied principal
for the “Sumatron Generator Rotor Shorted Turn Analyzer” will be
discussed in more detail in the following text:
It should be noted that the foregoing testing (vibration analyses, thermal stability,
flux probe analyses, and RSO testing by themselves do not provide absolute certainty
that there is a shorted turn problem in the generator rotor. However, when
confirmed by other testing the probability of the field winding being the cause of the
vibration problem increases significantly. Shorted turn anomalies can be masked if
they are near the center of the winding or otherwise balanced, if there are multiple
shorted turns, if they are intermittent, if there are grounds, and if there are other
contributors to the overall machine vibration levels.
3
RSO testing has some advantages over other testing in that it can be used periodically
during rewinds to verify that windings are free of shorts and on both at rest and
spinning de-energized rotor windings.
Theory of Operation
The “Generator Rotor Shorted Turn Analyzer” or (“RSO” Repetitive Surge
Oscilloscope) produces a succession of step-shaped low voltage pulses. The pulses
are introduced simultaneously to the dc rotor winding (“field winding”) from both
ends. The resulting reflected signals can be viewed on a dual channel oscilloscope
screen as two separate waveforms, or after one of them is inverted, and both summed
as a single trace.
If no discontinuities are present in the winding (due to grounds or shorted-turns),
both traces will be nearly identical and if inverted and summed, a single trace will be
displayed as a horizontal straight line, with a minor blip at the origin and an almost
imperceptible ripple. Any significant discontinuity arising from a fault will be shown
as an irregularity on the summed trace. By estimating the location of the anomaly on
the screen, an inference can be made as to the approximate location of the fault. For
instance, large irregularities near the origin of the trace are attributed to faults close to
either end of the winding.
The injected pulses have amplitudes of between 8 to 12 volts. Thus, it is possible to
carry out tests on a rotor without undue safety concerns.
The best interpretations are obtained when the results are compared to other tests. It
is therefore suggested that for reference, results of tests performed with this
instrument be recorded and saved for future use. This is important both for
benchmarking a specific unit and for comparison with results obtained on other
rotors. In this manner a database can be compiled for future reference. Figure 1
below shows the typical connection arrangement:
Slip-rings
Ground wire
RSO 110 Vac
Oscilloscope 4
The foregoing figure shows the connection to a field winding with an external
excitation source. In the case of self-excited rotors or brush-less excitation systems,
the connections are made to the leads leading from the diode-wheel to the winding.
The leads between the diode wheel and rotor winding must be disconnected before
proceeding with the test. All connections “away” from the rotor must be open.
Otherwise, the RSO will be measuring not only the field winding, but also everything
that is connected in parallel with it.
Figure 2 below shows a typical inverted summed trace indicating that no shorted-
turns are present in the rotor under test:
Figure 3 below shows an actual winding trace indicating the presence of one or more
shorted-turns:
5
Equipment Required
1. Generator Rotor Shorted Turns Analyzer (RSO).
2. Oscilloscope (see “Note” below).
3. Camera to take pictures of the traces. Alternatively, a digital scope with the
capability to store traces that can be down loaded for display and analysis.
4. Interconnecting test leads (supplied with the RSO)
Note: There are numerous types of oscilloscopes in existence. Almost any analog
or digital oscilloscope will be acceptable for this application as long as it:
1. Can measure two independent traces.
2. Has an inversion function allowing one trace to be inverted.
3. Has a summation function allowing both traces to be added.
4. Has a bandwidth of at least 20 megahertz.
5. Has a voltage resolution of at least 0.5 volts per division.
Precaution: with digital scopes, the input channels may saturate at 0.5 volts
per division and result in erroneous measurements. Accordingly, only the
math or summed channel should be reduced to 0.5 volts per division and not
the input channels. When selecting a digital scope, ensure that increasing the
sensitivity or reducing the volts/div of the math channel does not impact the
input channels.
Testing Sequence
1. A rotor winding insulation test should be performed before RSO testing. RSO
testing is not valid for grounded rotor windings (most rotors will have at least
1.0 megohm to ground). The brushes should be lifted for the test. It is not
unusual that a thorough cleaning is required on the exposed insulation
between the collector rings and the shaft to remove the carbon/oil build-up
before getting an acceptable megohm reading.
2. The RSO should be verified as working properly using the RSO built-in rotor
simulator circuit per the instruction manual.
3. The RSO should then be carefully re-calibrated using the actual rotor with the
RSO ground wire attached to the rotor forging per the instruction manual.
6
Instrument Settings
1. Turn both the Inner and Outer potentiometers to their fully counter-
clockwise positions.
2. Connect the test leads to the oscilloscope as shown in figure 4 below:
GENERATOR ROTOR SHORTED TURN ANALYZER
FAULT
120 VAC INNER
.05 A
NORMAL OUTER
ON
TEST SIMULATOR
OFF
ROTOR SCOPE
GND I O 1 2 E GND
Channel 1 Channel 2 Optional Optional ground
External Sync if scope is not
equipped with a
ground reference
3. Power-up the instrument and the oscilloscope.
7
Typical Analog Oscilloscope Settings
• Trigger to EXT
• Trigger Mode to AUTO
• Trigger Coupling to AC
• INVERT button to OFF
• CH1/CH2 button to CH1
• MONO/DUAL button to DUAL
• AC/GND/DC knobs to the AC position
• Turn both channels to 2 VOLTS/DIV
• Set the TIME/DIV knob to 5 MICRO-SEC/DIV
Adjust the TRIGGER until two traces are displayed. Both traces should be nearly
identical (of about 8 to 12 volts amplitude).
For digital applications suggest the Pico Scope Laptop Converter because of the
clean waveforms and ease of use. Can be shipped with the RSO or purchased
locally.
Pico Scope Model 5242D RSO Settings
60 Megahertz – 1 GS
• 20 us time base
• 14 Bits
• A & B active – Red & Blue – DC - Scaling factors of 1.0
• A-B Library – Green Math Channel – Scaling factor of 3.0
• Trigger – Auto
• Advanced trigger – Logic – Rising 100 mv – External – 20% of screen
• 12V full scale for initial calibration
• 10V full scale for fine tuning
• 5V full scale for actual test
8
Rotor Connections
Connect the Generator Rotor Shorted Turn Analyzer instrument to the rotor
slip-rings (collector rings) or leads, if these are disconnected from the slip-rings
during repairs, testing or manufacturing as shown in figure 5 below. The BNC
cables to the rotor should be of equal length. The instrument banana jack
ground must be connected to the rotor body or forging since the tests are
made in relationship to ground. Make sure the inner “I” BNC jack is connected
to the inner ring, and the outer “O” BNC jack to the outer ring. The Inner ring
is the one closes to the forging, and the outer is the ring away from the forging.
In rotors without slip-rings, identify the leads connected to the inner “I” and
outer “O” instrument jacks. This will aid in identifying the location of the fault.
GENERATOR ROTOR SHORTED TURN ANALYZER
FAULT
120 VAC INNER
.05 A
NORMAL OUTER
ON
TEST SIMULATOR
OFF
ROTOR SCOPE
GND I O 1 2 E GND
Rotor Forging Inner Ring Outer Ring
9
Final Calibration Procedure
1. Disconnect the cable from the outer “O” BNC jack and adjust the
INNER potentiometer until the trace on Channel 1 is about one half the
amplitude of that on Channel 2.
2. Reconnect the cable to the outer “O” BNC jack and adjust the OUTER
potentiometer until the two traces are of equal magnitude, from ground
level to the positive peak. Change both channel settings to 1.0 volt per
division to make the adjustment more sensitive.
Note: for units equipped with counter knobs, our suggestion is to not
use the counter brake. If you decide on using the brake, please use
one hand to hold the counter to prevent turning or loosening the
potentiometer when setting or releasing the brake. Otherwise, you may
need to disassemble the instrument to retighten the potentiometers.
3. Invert one of the channels in the scope (some scopes allow only one of
the channels to be inverted; others allow any one or both). Set the
ALT/CHOP/ADD button to ADD, and the MONO/DUAL button to
MONO.
4. Position the resulting trace in the middle of the screen with the help of
the respective POSITION knob.
5. Set both channels to 0.5 volts per division and photograph the
resulting trace for determination of winding condition, or if a storage
function exists in the scope, store the traces to be downloaded at a
convenient time.
Please see the next section on “Instrument Diagnostics” for a sequence of
oscilloscope photographs that illustrate the calibration procedure.
10
Instrument Diagnostics
The Generator Rotor Shorted-Turns Analyzer has a built-in test simulator
circuit. This has been included to prove that the instrument is functioning
properly, for training, and to also prove that the test leads are intact.
To use the “Test Simulator”, follow this procedure:
1. Plug the outputs of the “Rotor” section (bottom left on instrument’s
panel) to the Test Simulator” binding post banana jacks located in the
“Test Simulator” Section (center of instrument panel). Use the short
banana jack cables and BNC adapter provided for this purpose. An
external ground connection is not necessary, since an internal ground is
provided.
2. Complete the “Final Calibration Procedure” explained previously, with
the exception that the connection is made to the “Test Simulator” circuit
instead of a generator rotor. During the calibration, the “Test Switch”
should be set to the “Normal” position. Two identical traces (or one
straight line) should be obtained on the scope’s screen.
3. After performing the calibration, set the “Test Switch” to the “Fault”
position. The two traces should now be different (or the otherwise
straight line will now show a “blip” at the front-end of the waveform),
mimicking a faulty winding.
The following photographs show the calibration procedure using the test
simulator circuit:
Figure 6 below shows both traces (2.0 volts per division) before the test leads
are connected to the “Test Simulator” circuit.
11
Figure 7 below shows both waveforms after connection to the “Test Simulator”
circuit. In this case, the inner (I) BNC is connected to the black binding post
and the outer (O) BNC is connected to the red binding post.
Figure 8 below shows both traces after the red binding post lead is
disconnected and the Inner potentiometer is rotated clockwise to reduce the
channel 1 amplitude by approximately one half.
12
Figure 9 below shows both waveforms superimposed after the red binding
post lead is reconnected and the Outer potentiometer is rotated clockwise to
reduce the amplitude of channel 2 until it matches channel 1.
Figure 10 below shows a single trace that was formed when channel 2 was
inverted and then added to channel 1. The test switch is in the normal position
and the straight line with a minor blip at the origin indicates that the winding
does not have shorted turns.
13
Figure 11 below shows the inverted summed single trace with the test switch closed to
the fault position. In this case, the relatively large negative going blip indicates a
shorted turn near the start of the simulator circuitry. If channel 1 had been inverted
instead of channel 2, the fault blip would have been positive going instead of negative.
The tests prove that the instrument is functioning properly. If the actual test leads
were used to connect from the instrument rotor BNC jacks to the test circuit binding
posts, the integrity of those leads would be proven as well.
The figure below shows the simulator fault test using the Pico Scope. As you can see
the two input wave forms (red and blue) are not symmetrical and are displaced
because of the fault. The math channel is displayed in green and illustrates the
difference or anomaly.
Repeat the same calibration sequence (figures 6 - 10), when testing an actual generator rotor
field winding.
14
PERFORMING A SPINNING TEST
The spinning test is mainly carried out for the purpose of ascertaining if the
rotor winding has shorted turns that tend to appear or disappear when the
rotor goes from full speed to rest (or vice-versa). Obviously this test can only
be performed on a rotor with an external source of excitation, given that the
connections to the winding are made via the slip-rings. The ground connection
to the rotor is made indirectly via the grounding-brush (sweeping against the
rotor’s shaft).
Figure 5 shows a typical arrangement for the test. In order to eliminate any
route for the injected waves other than the rotor winding, the excitation must
be disconnected from the slip-rings. This can be achieved either by opening
the leads/busses at a point close to the brush-rigging, or by lifting the brushes
and placing instead a set of brushes insulated from the rest of the brush-
rigging, but connected to the Shorted Turn Instrument.
Field
disconnects
Rotor
Excitation
Rotor winding
System
R
S
O
Scope
15
Repair
The unit is equipped with an internal 1.0 amp fuse. If scope traces are not
displayed and the “Power Switch” does not illuminate when closed to the “ON”
position even though the power cord is plugged into a 115 VAC outlet that is
verified to be energized, the instrument panel should be removed (after
unplugging the power cord) to access and prove the integrity of the fuse. The
panel wiring connections to the circuit board terminals should also be verified
tight when the panel is removed. For all other problems, the instrument should
be returned to the factory for repair.
Warranty
Sumatron warrants the instrument to be free of defects in material and
workmanship for a duration period of 18 months from the date of shipment.
Sumatron will either replace or repair the instrument free of charge during the
warranty period. Any other obligation or liabilities on the part of Sumatron is
expressly excluded. Sumatron shall in no event be liable for special or
consequential damages.
For repair, ship to:
Sumatron, Inc.
24601 Steffy Drive
Laguna Niguel, California 92677
Contact information:
(949) 360-0386
www.sumatron.com
info@sumatron.com
16
