A customer wants to connect a piece of equipment with a VFD to a GFI circuit. It runs fine on a normal breaker but trips the GFI breaker. (It isn't the motor)

I would recommend replacing the drive if the new drive wouldn't have the same problem. A couple years ago, I wouldn't have thought that CSA/UL equipment would have a ground fault, but now I know better.

Have you ever put a VFD on a GFI?

It's not that they have ground faults...

It's that their wave forms are electronically perceived by GFCI circuitry as being faulted.

It's inherent in the design of a GFCI's 'balance' circuit that the hot is compared to the return.

VFDs create zany harmonics that rise above the threshold of instantaneous imbalance -- triggering a 'trip.'

Hence, one must install an Equipment Utillization Receptacle rated at a higher threshold.

"In the case of Heat Trace Cables, instead of using the Class A, 6ma Trip Threshold GFCI Devices for controlling excessive "Leakage", Ground Fault Protection for Equipment ("GFPE") Devices are used. The standard Trip Threshold for these GFPE Devices = 30ma (0.03A)."

Per Scott ^^^^ down at the end of the GFI Tripping thread.

I had the same problem, I installed ferrite beads (several)on the power lines between the VFD and the GFCI. It is running today trip free.

How are ferrite beads sized and from where are they purchased?

I don't know where you can buy them in retail quantities...

But every old, discarded TV or PC power supply has ferrite rings as part of the low-pass DC filter circuits.

I don't think I've ever seen them sold alongside resistors, caps and such at Fry's or Radio Shack.

Within these power supplies, ferrite cores are wrapped in copper. It acts as a magnetic dampener -- clipping the highest peaks off of non-harmonic wave forms.

They're required because PWM triggering circuits create sharp edged pulses that can degrade conductors via corona discharge through weakened insulation.

-----------

Ironically, they were once the 'core' of computer memories. All the way through the 1960s the DRAM of all computers was composed of ferrite core memories.

Hence, the term 'computer core'....

Early in 1970 silicon memory chips finally overtook ferrite core memory as the least expensive DRAM. Ferrite cores left the market entirely within twelve months.

Now I'm wondering if a line-side reactor would do the same thing. I know where to get those, but the ferrite bead is more interesting.

I could send you a few if you just want to experiment.

The ones I have are for line cords so they are about 3/8" id

you can get them from DigiKey
http://www.digikey.com/product-search/en/filters/ferrite-beads-and-chips/3408555

Just have to know what size you need

Thanks, but I'll have to pass on your offer. I would have loved to experiment, but I don't know the owner very well.

Why is the circuit GF protected? Is it protected by a class A device whick trips at 6ma or a GFI at 30 ma as Tesla outlined?
There are a lot of exceptions for specific equipment from GF protection because motors often trip GFCI devices. For example a domestic sump pump is located outdoors and the receptacle is in a location that would normally require GFCI protection. We have a rule in the Canadian code book that allows the pump to not be GFI protected as the pumps work is critical. Just need to use a single receptacle and in use cover with a lock to prevent casual use of the outlet. I would expect this is generally true for the NEC too?

I'm in Canada, Mike.

It's on (or would be on) a gfi because the customer wants it as added protection for his employees.

I didn't wire it. I'm involved because the customer tried to buy a new vfd because the existing one trips the breaker. They didn't know it was a ground fault issue. The local supply house sent me to make sure the new one would solve the problem.

I gave them a report and cost estimates. Now, I suppose they'll get their own electrician to do the profitable work.