View Full Version : Using a VFD can be a shocking situation !

07-06-2008, 01:28 AM
Wired up a Teco EV50 VFD to my new mill this last week. While making an adjustment on the side of the machine, one arm touched a basement support column and I got a good zing.

Double checked my wiring which is for 110v single phase input, 220v, 3 phase output. All looked fine except for the PE, protective earth grounds, appearing on each end of the VFD. I figured the PE is PE is PE, right? Wrong !

Here's where the wife's suggestion to RTFM has some merit.

A real shocker to me but the PEs were not common so I proceeded to common them. Took about 10 milliseconds to shut down power at the GFI outlet. Ok, what gives? First thought was some bad switching crap present on the ground but should it be there in a good design? Not sure.

Here's where the internet shines. A quick Google turned up rather definitive comments regarding the Teco VFD insisting on separate ground returns, not really a safe idea. Also discovered that residential GFIs don't like VFDs.

Wiring it up to a non-GFI circuit, all is well. The 145 volts from the machine to my basement support post (set in concrete / ground) was gone.

RTFM again, I noticed that they actually show a GFI in place in one of their installation scenarios but they also specify a high (flesh roasting 200mA) and long time to switch (100ms or more, can't recall now).

Just telling the story here for anyone else who may get zinged and wonder why.

A call to Teco Westinghouse is in order as PE should be PE should be PE ... or label the darned thing something else on each side of the drive :eek:


07-06-2008, 02:12 AM
When all else fails....


Read the F***ing manual.

Glad you survived the shock !

07-06-2008, 04:07 AM
I would call Teco and ask them why there are two "grounds". That really does not make sense.

The VFD GFCI issue is a common one. I bumped my head on it some time ago. I removed the GFCIs.

07-06-2008, 06:06 AM
My ABB VFD has two seperate grounds. IIRC, one side is specifically for grounding the shield on the wiring to the motor, and the other is for power ground. I am also fairly certain that my manual made a point of specifying that they should not be tied together.


J Tiers
07-06-2008, 10:33 AM
One of the grounds is often specifically for EMI filters. You have the option of how and where to ground it, although there are usually recommendations.

That may not be helpful... EMI filters and GFCIs are NOT compatible in many cases, since the trip point of the GFCI can be close to the filter "touch current". Any added leakage, such as from the VFD to motor leads (high frequencies cause capacitive leakage) may go OVER the GFCI limit.

ANY grounding of the filter may trip the GFCI, since the current comes from the mains, and does not return via them.

When the "safety" systems cause outages, they may be bypassed by annoyed users (as in this case), in which case they are no longer "safety systems". There is an inherent conflict.

07-06-2008, 04:56 PM
JT, Ventricular fibrillation occurs in the 60mA to 70mA range. Teco calls out a GFCI, if used, of "up to" 200mA and response time "up to" 100ms to prevent "high frequency failure".

Herein lies part of the problem ... the manual. They may be referring to a high frequency of tripping or failure in operation at high frequencies.

You bring up a good point and that is frequency. I'm operating at a high carrier frequency (to eliminate audible noise). The internal capacitors and the motor windings are going to pump more current into the ground leg because of that.

Even an "annoyed user" would be zapped at 200mA trip current. I hesitate to call it a safety system. Incidentally, the Teco manual shows both the inverter and the motor frame connected to earth ground.


J Tiers
07-06-2008, 05:01 PM
JT, Ventricular fibrillation occurs in the 60mA to 70mA range.

less than that if the pathway is right. it isn't a hard and fast rule. Plus the max current allowed by UL is 3.5mA.

Incidentally, the Teco manual shows both the inverter and the motor frame connected to earth ground.

Which is the correct way, and would eliminate the problem and the safety hazard.

The UL rules specifically ASSUME that the ground is non-functional. The test is for consumer equipment, and "consumers" are assumed to automatically cut off ground pins because their house has non-grounded outlets (mine does, for instance).

Alistair Hosie
07-06-2008, 06:25 PM
RTFM could also be return to the fabulous manufacturer:DAlistair