Capacitors in Flux loop test of motors and generators ?

[edison123]

In flux loop test (aka ring test / core magnetization test / hot spot test) of motors and generators, is it advisable to use capacitors in parallel to balance the inductive current of the loop test ? This is to reduce the current requirement at the supply source.

I understand 100% capacitor balancing may create resonance. How about 80% ?



"Most people stop working when they find a job"

 

[Skogsgurra]

Resonance wouldn't be bad at all. It takes your current consumption down and there will not be any runaway currents or voltages at all.

I guess that your concern stems from the fact that 100 % compensation of an induction motor can/does create resonance when the motor is disconnected from the mains and still rotating with a decaying rotor flux. That is bad and should be avoided. Eighty percent compensation is a rule-of-thumb in such cases. But for the ring test - no problems using any compensation you want.

Gunnar Englund

http://www.gke.org

 

[electricpete]

If I understand correctly, you are thinking about hooking the caps in parallel with the test loop circuit (not the motor stator), correct?

If you are using a tester which is going to try to determine watts loss, in that case I think the capacitance in parallel could interfere with the measurement.

If you are just trying to excite the core for hotspot test (that's what your question sounds like), skogsgurra's response makes sense from a circuit analysis standpoint.

All the motor test loop cares about is the voltage supplied. It will draw current accordingly. Caps connected in parallel would serve to decrease the current which needs to be supplied by the test set to maintain that voltage.

I'm wondering if any extra precautions are required. I guess if you ramp down the voltage slowly to 0, (a normal precaution for an inductive circuit) there is no energy left in the circuit once the test is done. It still makes me a little nervous although I don't know exactly why. I think I would ground both legs of the loop before disconnecting anything just as an extra precaution. skogsgurra - have you ever done the test this way? I wonder why this is not a standard test method?

An extra note - Of course you already know the required capacitive compensation would not be based on the operating magnetizing current but would be based on estimated current during the test (similar calc as used for determining sizing for loop conducotrs)....IEEE56-97 recommends estimating current using amp-turns per inch of circumference (from manufacturer), times inches of circumference (based on mean diameter) divided by number of turns.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Skogsgurra]

Yes, I have done ring tests on several motors. (I hope that we are talking about how to find bad spots in stator iron. That is what I understand under ring test). But we (or the winder shop I use) has plenty of power available. So we never have had to compensate with capacitors.

But have no fear. A parallel circuit connected to a voltage source does what the voltage source says. No more - no less. And why do you think you need to ground both ends of the capacitor? The winding is low ohms and discharges the capacitor very effectively.

It is basic electricity - no room for superstition here.

Gunnar Englund

http://www.gke.org

 

[electricpete]

I agree when the voltage source is connected there is no concern.

Let's say the voltage source was abruptly disconnected while capacitor and loop still connected. Then the current and voltage would continue even without the voltage source. That is my concern that there may be an electrical hazard. To be double-safe: 1- ramp down voltage slowly to zero volts; 2 - ground both sides. (The winding discharges the capacitor for dc but not for oscillating ac at resonant frequency which can remain until dissipated by resistance).

Are both steps needed? Probably either one is good enough. I would do both.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Skogsgurra]

I can understand that, Pete.

But the resonance will be at 50 or 60 Hz. The Q of a resonance circuit with iron in it is probably well below 20 - I think that 10 is a safe assumption. So the oscillation will die out in something like a second. That is if you disconnect abrubtly. BTW, that is how some demagnetizing devices work.

Gunnar Englund

http://www.gke.org

 

[edison123]

Thx skogs and pete.

What about harmonics ? Someone indicated having the capacitors can create harmonics and I can't see how.

"Most people stop working when they find a job"

 

[Skogsgurra]

Tell that "someone" about resonance and that harmonics are outside the resonance frequency. Many people use words they don't understand - and hope that others don't either...

Gunnar Englund

http://www.gke.org

 

[edison123]

well, that someone was EASA. So, I had to ask here.

"Most people stop working when they find a job"

 

[Skogsgurra]

What does that say? Should I or EASA blush?

Still, I think that the problem is non-existant.

Gunnar Englund

http://www.gke.org

 

[electricpete]

I agree with skogsgurra - capacitors don't create harmonics.

But if there are harmonics present on the voltage, capacitors can draw high harmonic current in response, since Xc ~ 1/f
The higher the frequency, the more harmonic current drawn by the cap (for a given harmonic voltage amplitude).

If you expect heavy voltage harmonics on the power supply during the test it should be considered in rating the capacitor. I don't think it would be a concern in most situations. If for some strange reason you are driving the test from an electronic power supply rather than a transformer, it could be a problem for both the cap the supply.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Now I am thinking about something else. If the capacitors are far undersized, not less that full compensation, Xc>>XL at line frequency, then resonant frequency 1/sqrt(LC) is above Far line frequency. What happens if harmonic currens drawn by the core loop near saturation excite this frequency? It seems like you could get some circulating current at harmonic/resonant frequency circulating between the cap and the core loop.

Maybe that was the harmonic concern mentioned by EASA?

I would think that if you size the cap for full compensation this should not be an issue since resonant frequency is near line frequency and the power supply can effectively control voltage at this frequency which prevents any abnormal currents.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Typo alert..
"If the capacitors are far undersized, a lot less that full compensation, Xc>>XL at line frequency"

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[waross]

skogsgurra
My understanding is that the load will draw magnetising currents regardless, and the small capacitor will supply part of the magnetising current that otherwise would be drawn from the supply.
If there are harmonic currents they are not caused by the capacitors.
If I have this wrong, skog, please "Red flag" it out of here to avoid adding to the confusion.
yours

 

[Skogsgurra]

Then we are back to the question about compensation or not in harmonics-rich grids. And I have a feeling that it has been dealt with several times before in these fora and it is not special to the ring test as such. Is that what EASA is talking about? Compensation in general?

If you are doing the same ring test as we do, then you enclose the stator iron with a heavy gauge cable and feed it a low voltage from a variable transformer - or a big (1 MW)MG set as it happens to be in our case. We adjust excitation and sometimes frequency (speed of MG) to get the current we need. And that's it. So we don't have any direct connection to the grid. Only mechanical via motor shaft.

Gunnar Englund

http://www.gke.org

 

[oftenlost]

A ring test for shorted iron, as we accomplish it in my shop is to calculate the iron to determine ampere turns required to saturate. If we are using the little test board with limited current, we use more turns.

 

[electricpete]

I don't think grid harmonics would be a problem. Only if powered from electronic power supply. I don't know why there would be an electronic power supply for this purpose but I was trying to guess why harmonics might be a concern.

Going to the issue of motor excitation harmonics. Let's say resonant frequency =sqrt(1/LC)=5* line frequenc. By superposition, two different circuits, one at line frequency and one at 5th harmonic. The one at line frequency we know everything about it, since voltage is controlld by the source.

Now what about the 5th harmonic circuit.

If we model the 5th harmonic excitation current (due to saturation), as a current source, it circulates harmlessly between coil inductance and inductor. The 5th harmonic voltage is pretty darned close to 0 because the impedance of that circuit is 0.

If we model the 5th harmonic excitation (due to saturation), as a 5th harmonic voltage, it creates very high 5th harmonic current because the impedance to 5th harmonic is 0 in the resonant circuit.

I think the first scenario is closer to reality. In either event the chances of creating a resonance at an excitation harmonic are small and are 0 if you have full compensation (in which case resonance is at line frequency.). I don't see any big harmonic concern. It would help to ask the person who made the comment specifically what the concern was.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.