Miffed & puzzled.

[ptonsparky]

CH VFD for a customer failed to start yesterday. Showed a fault that indicated either the motor load was wrong or an IGBT failure. The guys megged the motor and wires. No faults P-P or P-G with resistance of windings entirely within expectations. New AB drive went on this morning and immediately showed fault code of U-V. They called for help and we cussed & discussed the tests again. I had them repeat everything while I drove out. Naturally the first disconnect I open I find a phase wire that obviously, because I looked closely, has seen an intermittent fault to ground. We installed it. The miffed part. They removed all the wiring and pulled new without finding an additional fault. They are bringing the wire back to the shop so we can examine it closely. Old drive went back in and everything is fine. Puzzled why the drive showed a load side fault between conductors and not a ground fault. (Granted I may not be puzzled if I find other damage.) Basically a no charge day with a new AB VFD on hand that I don't need. Another miffed part.

 

[GoldDigger]

Puzzled why the drive showed a load side fault between conductors and not a ground fault. (Granted I may not be puzzled if I find other damage.)

A starting guess: As the VFD experts have mentioned in the past, reflections from the end of a long feeder wire between the VFD and the motor can cause high voltage spikes to appear on the load wires. Because of inductive coupling, current from the fault, during the spikes only, could have induced voltages in the other load-side wires that caused the VFD to give an incorrect (or at least misleading) fault indication.

Your mention of "first disconnect I opened" makes me suspicious that the installation does involve long leads.
If so, a combination of load-side reactors and/or shielded VFD cable would remove the problem. If the intermittent fault resulted from too close clearances, it may not have happened without the spikes, but the with proper wiring the spikes themselves might or might not have been a problem.

 

[Jraef]

A starting guess: As the VFD experts have mentioned in the past, reflections from the end of a long feeder wire between the VFD and the motor can cause high voltage spikes to appear on the load wires. Because of inductive coupling, current from the fault, during the spikes only, could have induced voltages in the other load-side wires that caused the VFD to give an incorrect (or at least misleading) fault indication.

Your mention of "first disconnect I opened" makes me suspicious that the installation does involve long leads.
If so, a combination of load-side reactors and/or shielded VFD cable would remove the problem. If the intermittent fault resulted from too close clearances, it may not have happened without the spikes, but the with proper wiring the spikes themselves might or might not have been a problem.

More likely:

The A-B drive is more tolerant of the loss of a phase (single phase feed) than the C-H drive is. The CH drive would have seen the exceptionally higher DC bus ripple caused by the single phase input as a possible load side fault as the current through the transistors varied in each pulse. The A-B drive has a built-in DC link choke, so that helps smooth the excess DC bus ripple from the single phase input. So then what happens is, the loss of the phase just results in a lower DC bus voltage, which is what it tripped on.

 

[ptonsparky]

This was a load side fault. Distance is less then 60'. The arc evidence I found was very small amount of smoke that almost looked like blurred inking on the wire until you looked closely. Details are coming in. My help says he may have misread the code and the AB, probably, did indicate that phase to ground. There was no other damaged wire. Another butt chewing was averted when discussion revealed that the customer had his hands in there in the not to distant past. Makes some of the time billable, but I still have a spare 480v VFD for 1 hp.