208/230V/460V 3 phase motor windings different resistances

[Jesse Livingston]

Hi there. New to the forum. I recently completed a water park project and the filter system included a 5hp 3Ph 460V motor that was controlled by an Allan Bradley Powerflex 525 VFD rated at 7.5HP.

After about 6 weeks the VFD started tripping out on overcurrent. It was all pre-programmed and came as a complete control cabinet so I don't know what parameters it was set at.

The VFD will ramp up to about 10hz then the screen on the VFD goes blurry and you can no longer read the frequency and then it trips out.

I megged all the cables and windings and everything checked out fine, no fault to ground. When I measured the resistance on the windings they were all different A/B 2 Ohms, A/C 2.9 Ohms and B/C 3.6 Ohms. I know they should all be identical as all the windings should be the same length I would assume.

So now we are trying to determine what caused this? The VFD operates as expected without the motor connected. The motor dead shorted when I tried to start it across the line, but did no damage to the windings as I checked all over again after. The motor suppliers have suggested power fluctuation but one would again assume that if the VFD was set up right that motor should never be allowed to draw more than the S.F.A correct.

Any insight or other testing suggestions would be much appreciated. Thank you in advance.

 

[Jraef]

So now we are trying to determine what caused this? The VFD operates as expected without the motor connected.

Doesn't necessarily mean anything. It needs to be connected to a motor to be meaningful.

The motor dead shorted when I tried to start it across the line, ...

Bad sign, very bad sign.

... but did no damage to the windings as I checked all over again after.

I find that statement to be in direct conflict with the previous one.

The motor suppliers have suggested power fluctuation but one would again assume that if the VFD was set up right that motor should never be allowed to draw more than the S.F.A correct.

Unless the motor is bad, see above.

Any insight or other testing suggestions would be much appreciated. Thank you in advance.

Record all of the programming of the drive, quick while it is still alive.

Off hand, I'd venture to guess that there is a turn-to-turn short in the motor winding, a breakdown of the insulation WITHIN an individual coil, but it is not going to ground (yet) so your megger referenced to ground would not pick it up. The drive however will. The default settings of the PF525 are to be in SVC operating mode, meaning Sensorless Vector Control. If left that way, the installer would have (hopefully) performed an autotune procedure on the drive when connected to that motor. Once the windings begin to fail, the drive is seeing the response not being as it expects it to be in SVC, so it is artificially limiting the speed to maintain the motor current within spec, the result being a 10Hz speed, and likely falling as more of the insulation burns.

The insulation failure between conductors within the first turn is a well known consequence of someone being cheap on the design of a system with VFDs on motors. Most likely it is not an "inverter Duty" motor, or is a cheap version of one where they basically lie about it, and they took no mitigation strategies into consideration such as filters, reactors or motor terminators. It also might be a factor of the distance from the VFD to the motor. On that drive, if you have a standard / cheap motor, meaning one designed withOUT phase paper (as many motors 10HP and under are), the maximum distance is 25ft, total circuit length (up-over-down etc.) if the carrier frequency is left at the default setting of 4kHz. So a number of things could make that even worse; unshielded or improperly shielded motor lead cables, someone increasing the carrier frequency to make the motor quieter, certain types of twisted cables etc.

Download and thoroughly read this document. The principles involved by the way are applicable to any and all VFD installations. In the back are the tables that tell you the tested maximum drive-to-motor distances based on the motor design and different circuit elements.
http://literature.rockwellautomation.com/idc/groups/literature/documents/in/drives-in001_-en-p.pdf

 

[GoldDigger]

You might also try using an impedance meter that uses an AC signal instead of DC to make the measurement of the coils.

 

[kwired]

I'm leaning toward the motor windings were damaged by the high voltage peaks that the drive puts out, whether it be because of a motor not designed for use with a VFD, long circuit run, a little of both. When I have seen motors damaged this way, the drive trips before any significant arcing occurs so you don't see any obvious damage. Connect across the line though and boom! Damage after that still may be where you can't easily see it, but the insulation is compromised.

Here is an article I found that can explain what is going on here better then I can.

This problem is worse with 480 volts then 208/240 volts because the peak voltage of the spikes is well above 600 volts in a 480 volt application.