Motor phase to phase short shows GF on VFD

[Saturn_Europa]

My facility uses the cheap automation direct Durapulse GS VFDs. Twice I’ve seen the drives show Ground Fault. Then the motor and the wire meg fine. In both instances I replaced the motor and the Ground Fault on the drive went away.

I’m assuming the motors had a phase to phase short and this is what caused the ground fault alarm on the VFD.

In a different instance. I’ve seen a terminal strip melt due to loose connections and the metal parts of the strip were shorting out 480v leg to leg. This would cause the ground fault meter on the switch board to switch from zero ohms to infinity constantly back and forth.

It’s my understanding that the ground fault meter on the switch board acts similarly to an Megger Insulation Tester. Where it sends a voltage and looks for a milli amp reading that it converts to M Ohms. So it makes sense that a leg to leg short would show up as a ground fault.

Do VFD ground fault indicators work in the same way ?

 

[jim dungar]

It’s my understanding that the ground fault meter on the switch board acts similarly to an Megger Insulation Tester.

Very very few Ground Fault devices work like this, except for those installed on ungrounded systems.
The majority of GF devices work by measuring the current 'going out and the current coming back' typically by passing all current carrying conductors through a single CT or by summing the outputs of several CTs.

 

[Jraef]

Very very few Ground Fault devices work like this, except for those installed on ungrounded systems.
The majority of GF devices work by measuring the current 'going out and the current coming back' typically by passing all current carrying conductors through a single CT or by summing the outputs of several CTs.

Correct. All VFDs I am aware of use a summing method on the 3 outputs (typically using Hall Effect Transducers on the DC side of the transistors) looking for a current imbalance that shows up as non-zero summing of the phase currents. It’s not measuring anything to ground, but it does assume that if the phase currents do not cancel out, that it is likely going to ground.

Most likely the motor had a turn-to-turn short in one of the windings, meaning that winding then had fewer turns, so less resistance and thereby a different amount of current outside of the tolerance band in the VFDs firmware. A Megger looking at phase-to-ground cannot detect that, because the short is not to ground. Very common on VFD driven motors when no attention was paid to whether the motor was capable of being run from a VFD, especially if it was 480V. It’s caused by the reflected wave phenomenon.

 

[herding_cats]

My facility uses the cheap automation direct Durapulse GS VFDs. Twice I’ve seen the drives show Ground Fault. Then the motor and the wire meg fine. In both instances I replaced the motor and the Ground Fault on the drive went away.

I’m assuming the motors had a phase to phase short and this is what caused the ground fault alarm on the VFD.

In a different instance. I’ve seen a terminal strip melt due to loose connections and the metal parts of the strip were shorting out 480v leg to leg. This would cause the ground fault meter on the switch board to switch from zero ohms to infinity constantly back and forth.

It’s my understanding that the ground fault meter on the switch board acts similarly to an Megger Insulation Tester. Where it sends a voltage and looks for a milli amp reading that it converts to M Ohms. So it makes sense that a leg to leg short would show up as a ground fault.

Do VFD ground fault indicators work in the same way ?

2 ohm phase to phase?

 

[ptonsparky]

2 ohm phase to phase?

I’m missing your point.

 

[herding_cats]

did the motor windings test 2 ohm phase to phase?

 

[ptonsparky]

did the motor windings test 2 ohm phase to phase?

Where did you come up with 2 ohms?

 

[herding_cats]

It's a standard test on all three phase motor windings. 2-3 ohm between phases.

 

[petersonra]

It's a standard test on all three phase motor windings. 2-3 ohm between phases.

Should be completely open between windings.

Each winding might be two to three ohms though.

 

[herding_cats]

each winding are 2-3 ohms. In a healthy motor.

 

[ptonsparky]

each winding are 2-3 ohms. In a healthy motor.

2-3 is not good enough. A turn to turn short in a winding might give 2.2 vs 2.8 in the others. Random #s.

 

[kwired]

My facility uses the cheap automation direct Durapulse GS VFDs. Twice I’ve seen the drives show Ground Fault. Then the motor and the wire meg fine. In both instances I replaced the motor and the Ground Fault on the drive went away.

I’m assuming the motors had a phase to phase short and this is what caused the ground fault alarm on the VFD.

I suspect it was the reflected wave issue like Jraef mentioned especially if 480 volts with either non inverter duty motor or a long circuit length between drive and motor.

Line reactor on output side of drive will help with this.

Motor will meg fine, drive will indicate a fault even at low volts/frequency during motor starting.

Connect motor across the line and it likely will show more eventful destruction that doesn't happen when connected to the drive because the drive is not only protecting the motor but also protecting itself.

 

[garbo]

It's a standard test on all three phase motor windings. 2-3 ohm between phases.

Less then one ohm on three phase motors over 100 HP.

 

[garbo]

I was told 20 years ago that VFD'S do not use the typical torrid coil to measure current flow difference between wires that bullet proof $15 residential GFCI receptacles use. Large place that I retired from had over 500 drives that we PM'ed 1 to 4:times a year. Over 85% of the time when a 5 to 250 HP drive tripped out on a ground fault they were false. Would megger the motor with a 1,000 volt megger and some had over 500 megohm to ground. Would look at the touch screen and read the last ten faults. Some would trip out on ground fault once a year. We had maybe 8 outdoor motors & drives from 40 to 75:HP for lab exhaust fans that ran 24/7. At least 15 times a year at least one of them would trip out on ground fault. Examined every pecker head connection and every motor meggered great. Finally asked our extremely talented Danfoss tech about this issue. Here Danfoss had a free warranty replacement for a ribbon cable. He replaced them and no more frequent false earth fault trips with those Danfoss drives. If motor meggered out good but a drive tripped frequently on ground failed I would bypass the drive and run the motor for a few minutes while taking ampere readings.We were fortunate that over 200 of the drives had built in bypass. ( Three contactors ).

 

[jim dungar]

We were fortunate that over 200 of the drives had built in bypass. ( Three contactors ).

+20 years ago I went on a trouble call for a VFD at a hospital.
The maintenance staff switched it to bypass mode and promptly blew dust out of some ducts and also collapsed some duct work. The BMS could not respond fast enough and had been never been designed to work at 100% speed.
That hospital system removed the manual switch controlling the bypass on all of its VFDs and changed their specs so the BMS was in control.

 

[kwired]

I was told 20 years ago that VFD'S do not use the typical torrid coil to measure current flow difference between wires that bullet proof $15 residential GFCI receptacles use. Large place that I retired from had over 500 drives that we PM'ed 1 to 4:times a year. Over 85% of the time when a 5 to 250 HP drive tripped out on a ground fault they were false. Would megger the motor with a 1,000 volt megger and some had over 500 megohm to ground. Would look at the touch screen and read the last ten faults. Some would trip out on ground fault once a year. We had maybe 8 outdoor motors & drives from 40 to 75:HP for lab exhaust fans that ran 24/7. At least 15 times a year at least one of them would trip out on ground fault. Examined every pecker head connection and every motor meggered great. Finally asked our extremely talented Danfoss tech about this issue. Here Danfoss had a free warranty replacement for a ribbon cable. He replaced them and no more frequent false earth fault trips with those Danfoss drives. If motor meggered out good but a drive tripped frequently on ground failed I would bypass the drive and run the motor for a few minutes while taking ampere readings.We were fortunate that over 200 of the drives had built in bypass. ( Three contactors ).

+20 years ago I went on a trouble call for a VFD at a hospital.
The maintenance staff switched it to bypass mode and promptly blew dust out of some ducts and also collapsed some duct work. The BMS could not respond fast enough and had been never been designed to work at 100% speed.
That hospital system removed the manual switch controlling the bypass on all of its VFDs and changed their specs so the BMS was in control.

Depending on what is being driven and other conditions you may need to at very least uncouple the driven load if starting a motor in bypass mode whether is a built in bypass or even if wiring around the drive?