VFD Tripping GFCI at stop (125VAC)

[theColonel26]

I just had a problem dumped on me.

We have been selling Machines with a Allen Bradley 22F-V2P5N103 4M 125VAC 1/2HP VFD.

A problem we had in the past was customer plugging them in to GFCI receptacles and the drives kicking the GFCI out right at the moment the motor comes to a stop.

So our Mechanical Engineer that usually deals with the products found a model of GFCI that seemed to fix the problem (LEVITON GFRBF-W). He also installed one of those GFCIs in the shop so the guys could test each machine before it went out. To solve the problem we just sold that GFCI to anyone that need to use a GFCI outlet.

I think the shop's testing was performed rather inconsistently... but that is a topic for another time.....



BUT NOW it seems that either Leviton or Allen Bradley has changed something and now the VFDs are even tripping this Leviton GFCI.

Things our MechE has tried

• A line reactor on the input side of the VFD, which made the GFCI trip sooner in the slow down step. Before it was tripping right when it came to a stop.
• Adjusting the carrier frequency all the way to the lowest and highest setting.
• A EMI Cap Filter, which I assumed before he even tried it would make the issue worse. I was correct. As soon as the machine was plug in the GFCI tripped. Didn't even need to run it. (again because it was draining to ground which is the problem)

They are using a Shielded cable on the VFD output. My first thought is, DON'T do that, because then you are inducing current in to the shield and it is draining through the ground and causing an imbalance. Anyone have thoughts on that?

Any one have any idea's?

 

[Robb B]

GFCI should trip when an imbalance in line current and neutral current occurs. As an initial guess, the line side of the VFD is discharging capacitors or something similar at full stop (is it ramp to stop, coast, plug?) causing the imbalance. I think most GFCI are designed to trip around 5mA imbalance.

RE: shielded output cable: I don't think that is going to cause a GFCI trip, as it's pretty far from the incoming supply, which is where the GFCI trip is occuring.

 

[theColonel26]

GFCI should trip when an imbalance in line current and neutral current occurs. As an initial guess, the line side of the VFD is discharging capacitors or something similar at full stop (is it ramp to stop, coast, plug?) causing the imbalance. I think most GFCI are designed to trip around 5mA imbalance.

He said he tried ramp to stop and coast.

RE: shielded output cable: I don't think that is going to cause a GFCI trip, as it's pretty far from the incoming supply, which is where the GFCI trip is occuring.

Why would it's location matter it is still down stream of the GFCI, so if you power leaks to the shield and back through the ground it would mean the Hot as more current coming in than the Neutral has flowing out, which would mean the Hot and Neutral are not balanced causing a trip.

 

[joseph_e2]

Could you try an isolation transformer above the drive? That should prevent a line/neutral imbalance.

 

[theColonel26]

Could you try an isolation transformer above the drive? That should prevent a line/neutral imbalance.

Could but that would add a lot of cost, and size, and I would have to LOL at UL but I figure that would be frowned upon.

 

[Robb B]

Semantics, I know, but I'd consider the VFD as downstream from the GFCI, and the output of the VFD as the furthest downstream power. Unless I have the concept completely wrong and you are powering a GFCI receptacle by means of a VFD?

The induced voltage into the shielded cable could certainly be enough to trip a GFCI, though I would expect that it would trip every single time, not intermittently as you are describing. It is possible that there's enough induced current on the bond/ground (bonded to the GFCI circuit I hope) to cause a trip, though typical GFCI only measure the difference in current between hot and neutral. Reminds me of a story one of my instructors told me one time about demonstrating a GFCI to a customer by shorting the hot and the ground. He instantly blew up the receptacle, the small jumper wire he was using, tripped the breaker, and scared the customer silly. The receptacle still worked after he cleaned everything up.

 

[theColonel26]

Semantics, I know, but I'd consider the VFD as downstream from the GFCI, and the output of the VFD as the furthest downstream power. Unless I have the concept completely wrong and you are powering a GFCI receptacle by means of a VFD?

Nah sorry I was being dumb I just edited the verbiage in that post to say down stream. you are right.

The induced voltage into the shielded cable could certainly be enough to trip a GFCI, though I would expect that it would trip every single time, not intermittently as you are describing.

No no, it happens every time.

And I just disconnected the EGC/Ground from the machine and tried it and guess what it didn't trip. Obviously can't ship them that way though LOL.

It is possible that there's enough induced current on the bond/ground (bonded to the GFCI circuit I hope) to cause a trip, though typical GFCI only measure the difference in current between hot and neutral.

Well yeah but if the Current is being induced to the EGC, then it is coming from the hot right? And then being split between the ground and the neutral which would kind be like an "auto" transformer, maybe.

but then the hot would have more current flowing in than is flowing out on the neutral because some of that is flowing out and around the GFCI sensor on the ground.

Reminds me of a story one of my instructors told me one time about demonstrating a GFCI to a customer by shorting the hot and the ground. He instantly blew up the receptacle, the small jumper wire he was using, tripped the breaker, and scared the customer silly. The receptacle still worked after he cleaned everything up.

LOL well that makes a lot of sense LOL He should have shorted it to the EGC/Ground LOL

 

[Robb B]

I think you've got the problem figured: Induction to the shield is draining through ground, and the GFCI is seeing more current on the hot than the neutral, thus tripping. Not sure about a solution though, short of not using a shielded cable.

Are you allowed to ship them without shielded VFD output cabling?

 

[theColonel26]

I think you've got the problem figured: Induction to the shield is draining through ground, and the GFCI is seeing more current on the hot than the neutral, thus tripping. Not sure about a solution though, short of not using a shielded cable.

Are you allowed to ship them without shielded VFD output cabling?

Yeah we can ship them without a shielded cable, but It seems the issue might actually be the motor case itself.


I disconnected the shield and ground from the VFD output cable and it made no difference. as the motor is bonded to the stainless frame and that is grounded. And we wouldn't be allowed to ship a motor out that has a metal case that isn't grounded anyways.

I am still playing around with it but it looks like this is no easy fix.

 

[Robb B]

Is it required to be connected to GFCI?

 

[jaden]

Look in the manual about disconnecting the MOV's that connect to ground. I have problems with the single phase input drives tripping the GFCI upon starting.