VFD question

Greetings everyone,

I use ABB ACS-550 VFD's quite a bit on projects. Every once in a while I'll have one that trips for "earth fault". The manual states this is a sensitive fault, and sometimes produces false positive faults. In most cases I would just disable this earth fault protection.

My question is, what is the drive actually measuring or looking at to determine there is an earth fault?

TIA,

Greg

Can you give more specifications ?

- Rating of drive
- Type of filter in drive
- Cable : type, section and length
- Motor specifications
- Motor control mode : vector or scalar
- Motor load : pump, fan, ... ?

Are you always using the drives in this setup?

Regardless of size etc. an "earth fault" or ground fault usually indicates that the motor windings are failing and allowing some of the current to leak to ground through the motor frame. Dick DV can give you the actual mechanics of the measurement, I'm sure.

Ok so to me Earth seems to mean a Ground. AKA Ground Fault??

I would be curious as to how the drive would detect a ground fault without a reference. For example a GFI plug has a Neutral to reference the Ground with. If the fault in question works along the same lines, maybe you need a neutral on the panel that is somehow connected to the drive.

DickDV has answered questions similar to this before, search using his name.

As for earth faults, harmonics, especially those that are multiples of the 3rd can generate ground loops which drives can sense.

DickDV, especially since he is an ABB drive guy, can offer more details on what you need to do.

Earth Fault and Ground Fault is intended to mean the same thing. I think Ground Fault is North American ad Earth Fault is probably everywhere else, certainly Europe. Drives built as "worldwide" drives will usually say Earth Fault.

How is it measured? The simplest description is that the current leaving the drive off of the positive rail of the DC bus is compared to the current returning on the negative rail of the DC bus. Within very small tolerances, if there is a difference, the conclusion is drawn that some of the current that went out there to the motor didn't come back! As in leakage to ground! This difference monitoring creates the Earth Fault.

Unfortunately, there are numerous complexities that can make nuisance faulting a big headache. That's why the fault can be turned off in Group 30 Parameter 17.

I am not a drive repair person or a drive designer so the internals are not my strong suit. Having said that, externally, the drive Earth Fault detection works best when the incoming AC power is voltage balanced to ground and referenced to ground. That would be a grounded wye or star power source. Imbalance leads to false faulting due, as I am told, to capacitance between the two DC bus rails and the drive frame (which is grounded) causing imbalanced currents in the DC bus rails. Further complicating matters are EMI and RFI filter networks that are mandated by stringent European laws. These filter networks must contact the drive chassis ground and, again as I am told, try to pull the DC bus into balance with ground. If the supply is not balanced or is floating, false faulting results and, in some cases, the filter networks burn up. ABB states in their installation manuals that these networks must be disconnected from ground on floating and imbalanced supplies to avoid drive damage.

Here is a note from ABB about VFD earthing.

Thx for that ABB note, JohnW, but note that it is for the ACS600 product line which is a generation old--about 1998 or so.

The CE mandated EFI/RFI filtering has been made more demanding since then which makes the problem worse for everybody and floating networks have become more common in North America since then also.

I have seen grounded delta power cause ground/earth faulting, too. I have seen ground/earth faults caused by large DC drives on the same power grid. These commutation notches will fry those european filters on VFD's. My company had to deal with this a bit and we found that removing the european filter and using a 3% input reactor eliminated false tripping. This is FLOATING power with a high impedance grund fault detection system.

If you have a long cable run and the carrier freq is too high, this sometimes manifests as ground faults. Anytime I have more than 50 feet of cable I make a practice of setting the carrier to 2K or the lowest the drive will program to when carrier freq selection is fixed.

I have encoundered cases where a meggar test said the cable and motor insulation was fine but the drive tripped sporatically on ground fault. This was a new PWM drive replacing an old VVI that had no grounding issues. A hi-pot of the motor and cable showed excessive leakage current at 1500 volts. The electrician said a portion of the cable had been pulled during a turn around. He found two of the cables chaffed. They were replaced and things worked fine.
These are a few of the problems I have encountered over the years.