Rotor and surge test: testing a motor or a stator

Does a motor need rewinding when the three surge curves don’t line up? We get that question a lot. To be honest, at that point you can’t tell yet.

In a surge test the tester applies a short voltage pulse to the winding. Winding and tester together form a resonant circuit. The circuit rings out, and that damped oscillation is the surge curve you see on the screen. On a three-phase winding you test the phases one after another and overlay the three surge curves. A symmetrical, healthy winding gives three nearly identical surge curves.

With the rotor in the stator, the rotor takes part. The pulse creates a magnetic field in the phase under test, just like AC voltage does, and that field couples into the rotor. How strongly depends on the position of the rotor.

Rotor and stator usually don’t have the same number of slots. So each phase faces a different part of the rotor: in one phase two rotor slots couple into the coil, in another only one. Each phase “sees” a different rotor, and that shows up in the surge curve.

The three surge curves can then differ quite a bit while the winding itself is fine. Reject such a motor on the phase-to-phase comparison alone and you may end up rewinding a good motor.

A stator without a rotor is the easy case. There is no rotor coupling differently into each phase, so the three phases should respond the same way. The phase-to-phase comparison works well here: the phases are each other’s reference. If one phase deviates, that is a real sign of a winding fault, such as a turn-to-turn short.

The MotorAnalyzer 3 compares the three surge curves with each other automatically, or with a stored reference object. In production the second option is common. You teach in the surge curves of a few good windings and compare every new stator with them.

For this the tester uses the correlation method, patented by SCHLEICH. It expresses the difference between two surge curves as a percentage. The SCHLEICH test method handbook gives 5 to 15 % as typical limits.

Rewinding a motor? Test the stator after winding, before the rotor goes back in. At that point you can still compare the phases cleanly. A surge curve of the loose stator is no good as a reference for the assembled motor later on, because the rotor changes the oscillation.

On an assembled motor you don’t compare the phases blindly. You look for the reference within the phase itself and test the same winding at a low and at a higher voltage. As long as the rotor stays put during the test, the coupling between rotor and stator stays the same. The rotor position then drops out of the comparison.

In a sound winding, raising the voltage only changes the amplitude. Damping and frequency stay the same. SCHLEICH compares it to a balloon with a face drawn on it: inflate it further and the face gets bigger, but it is still the same face. If the insulation breaks down at a certain voltage, the shape does change.

Peak-to-peak (P-P)

The tester starts at a low voltage and raises it automatically in fixed steps up to the final voltage. It compares each measurement with the previous one. From step to step the deviations should be about the same. If the deviation suddenly jumps, the winding has a voltage-dependent insulation fault and the test stops automatically. You do this for the three phases one after another, and you can see at which voltage things went wrong.

On the tester you specify whether you are testing a stator or an assembled motor. For a stator it compares the three phases at one voltage; for an assembled motor you use the peak-to-peak method. The maximum test voltage follows from the rated voltage you enter. If the motor is very dirty or damp, you set that maximum lower. How to determine the test voltage is covered in Maximum test voltage surge test.

As a rule of thumb, without a rotor you compare the phases and with a rotor you compare each phase with itself.

Peak-to-peak doesn’t catch everything, though. A fault that is already there at every voltage doesn’t change between two steps, so it stands out less. Read a P-P result together with the phase-to-phase traces and your other measurements, such as resistance and inductance. That takes some experience from whoever does the testing.

Measured three different surge curves on an assembled motor? Then, if you can, turn the shaft by hand and test again. Do this at a few positions over a full revolution.

If the deviation comes from the rotor, it moves along. In one position one phase deviates, a bit further on another one does. After one full turn the same picture is back. The animation below shows what that looks like.

A pattern that moves like this is no reason in itself to reject the motor. So make it a habit: when an assembled motor with a free-turning shaft is on the bench, always give the shaft a turn before you make a call.

If one phase deviates the same way in every position, the winding is the more likely cause. Investigate further, for example with the peak-to-peak method.

If the shaft can’t turn, for instance because the motor is still coupled to a drive, the per-phase peak-to-peak test is what’s left.

Turning the rotor in steps may sound familiar from the RIC test. There you measure the inductance at a fixed number of positions per revolution (for example every 15 or 30 degrees) to find broken rotor bars. That test looks at the rotor itself. More in Detecting broken rotor bars with the RIC test.

A deviating surge result on an assembled motor means looking further first. Rewinding comes later. The post Faulty surge measurement, replace or rewind the motor? already touched on this question. With the difference between motor and stator added, you can answer it better:

1. Loose stator: compare the phases with each other or with a reference.
2. Assembled motor: use the peak-to-peak method.
3. In doubt on an assembled motor: turn the shaft by hand and see whether the pattern moves.

Phase-to-phase comparison and peak-to-peak are both available on the MotorAnalyzer 3, the MTC2-R7 surge tester and the MTC3-R2 winding tester from SCHLEICH. Want to practise this with your own team? IONIO runs tailored training on off-line testing of electric motors.