What's new
What's new

VFDs and the RCCB

gubboar

Plastic
Joined
Apr 10, 2019
Hey folks,

I searched the topic for 'RCCB' and found nothing particularly relevant and I figured others might be interested.

I have several 1 and 3 phase VFDs at home. This is 230/400-land, not that it should matter, I guess. All single-phase VFDs have been working fine so far (3 of them incl. the one built into the washing machine ;) ). Problem is, 6 out of the 7 three-phase units trip my 2 yr-old type A, 30 mA, 40 A ABL Sursum RCCB within 1 second after being powered up. The RCCB is practically right after the meter, protecting all the circuits on the property, so currently I can't really use most of my VFDs and this is becoming more and more of an issue.

My understanding is that type B RCCBs are supposed to be used when inverter-type loads are involved. As the cost of such a unit is quite steep, I'd exercise some caution before I buy one. My concern is that even though type B RCCBs are designed to protect against faults involving high frequency DC currents, I've seen no claim that they're immunized against false trips caused by the input stage of VFDs, even though it sounds pretty darn logical.

Actually I'm a bit wary of the possibility that I spend hundreds of *whatever* on a new 30 mA RCCB and it's still going to trip. I'm going to separate the circuit going to the living areas and the one heading to my shop in the upcoming weekend, as I think it's not that bad of an idea anyway, and I'll try and see if the old 300 mA unit still trips when I switch on the VFD.

However, I'd really appreciate your input on the matter.

Thanks,
Gabor
 
Although not official, this site is based in North America, so the term RCCB means nothing to most people here. We use the term “Ground Fault Protection” and we have two main “Classes” of it, Class A which we call a GFCI (GF Circuit Interruptor) which is required to trip at no more than 6mA to protect humans, and other classes (B, C, D) known as GF EP for “Equipment Protection”. which is roughly equivalent to your RCCB. We also sometimes use older terms like GFI, but that’s not official. If you search for GFCI, GFEP or GFI you might get more results.

As a general rule though, using anything like a VFD that is protected by any ground fault protective device is going to be problematic. The technologies are mostly incompatible because VFDs inherently create what’s called “Common Mode” (CM) noise, which most GF detection schemes will interpret as a ground fault. It isn’t really, but the difference is often lost on the GF circuit. But a 300mA RCCB should suffice to keep from nuisance tripping.

If not, or as an alternative, you can put an isolation transformer between the VFD and the RCCB. That usually cures this issue.
 
One problem with many VFDs (and computers) is that they have filters on the input to reduce the amount of interference they send back into the supply. Many/most of these filters pass some current from the live/neutral into the earth conductor. This causes priblems with RCCBs.

It might be sensible to use another fuse-box/consumer-unit/distribution-board with a 100mA RCCB for the circuits supplying the VFDs so you can have a higher leakage current with them, but keep the 30mA protection for the rest of the house. I have done this with my wife's pottery kiln in the garage (very leaky). The workshop has a separate supply with local RCCBs on the lights and general supplies.

I hope that this is understandable!
 
As a general rule though, using anything like a VFD that is protected by any ground fault protective device is going to be problematic. The technologies are mostly incompatible because VFDs inherently create what’s called “Common Mode” (CM) noise, which most GF detection schemes will interpret as a ground fault. It isn’t really, but the difference is often lost on the GF circuit. ...

Ah, that's me (jim raises hand). I have a new spare 20 amp GFI breaker available for reuse in my house when I fitted a Hitachi VFD to the lathe out in
my garage. Instant trip on 'run' command.
 
Thanks for the clarification about the situation in the US. I was aware that over there, fault current protection devices were commonly fitted to individual (or pairs of) sockets, but I figured there must be something very similar to those protecting equipment and machinery [edit: actually the people using them]. For what it's worth, since the statistics about mortality/injury related to electric shock are similar or even a bit more favorable on this side of the pond - not counting intentional uses (insert evil grin here) - any ground fault detection and shock prevention regime will work surprisingly well, let it be central or individual. Chance doesn't really play a role anymore; stupidity does, though. Every single shock I've heard of in the news was due to some person's meddling with things they didn't understand at all. The sad thing is, it's not always them who pay the price.

"The technologies are mostly incompatible because VFDs inherently create what’s called “Common Mode” (CM) noise, which most GF detection schemes will interpret as a ground fault." - I didn't elaborate the last time, because I though most people were not familiar with the subject at all. I'm quite happy I was wrong. On the other hand, I kinda hoped that the 25-fold price difference between the "regular" and "HF DC-aware" devices was justified by their ability to detect HF DC fault currents AND recognize a certain amount of CM noise as harmless.

I'll try with the 300 mA one and if it doesn't get tripped, I'll probably pick up a 100 mA unit for pennies (literally! I don't know what's going on, these things can be had new - name brands - from the German ebay for a dollar plus shipping, practically in quantity! If I needed them, I could score at least one every day...). If that doesn't get tripped either, I'll buy a 100 mA type B RCCB.

Errr... the idea of an isolation transformer crossed my mind only as long as I realized that sometimes I might use as much as 10 kW of power. Just the sheer amount of copper going into that transformer makes it unfeasible.

Thanks again!
 
"Many/most of these filters pass some current from the live/neutral into the earth conductor." - Exactly! The way I got one of the Danfoss drives to run was to bypass the filter unit. Some of them didn't even come with a filter, though, and still fail to even turn on for more than a second. With the filters on, the 7-seg display on the VLT would light up and bam! there went the RCCB.

I've actually disassembled all of those that didn't have an external EMI filter and found that some had built-in ones.
 








 
Back
Top