Variable Frequency Drive tripping out the earth leakage in my house.

[AllTheGearNoIdea]

 I’m not sure if we have in UK electricians in the house but I need some advice.

I have just bought a new milling machine and it uses a variable frequency drive. Anyway when I switch it on it takes out the earth leakage breaker in the consumer unit. I have used VFDs in the past in industrial application and I’m aware they have some quite large filter capacitor and I’m guessing that’s my problem.

My test bench with a stack of Marconi analyser and computer sometimes trip them out a few times a month.  But this machine just trips instantly.

Desperate to try my new machine I’m running via an isolation transformer and it works fine but the transformer really underrated 750VA for 1.5VA machine.

Assuming the machines not faulty what are my options can a less sensitive RCBO be installed or should I just buy a bigger isolation transformer.

I have a PAT tester is this likely to damage anything like the VFD. What’s the easiest way to measure earth leakage or get to the bottom of what’s tripping.

Any advice welcome.

[Zero999]

If a larger transformer is not practical. The only advice I have is get it permanently wired in, preferably by an electrician, on a separate circuit to everything else, with a 100mA RCD.

[AllTheGearNoIdea]

Yes I don’t think it will be practical to wire in a new circuit. I am getting a quote for a large isolation transformer.

[jmsc_02]

When I was working in the elevator field the solution to that common problem in places where a VFD was installed is by replace the breaker by another called as "high immunity" (I don't remember the exact name).

[jmsc_02]

Note: some VFD has a little stick where you can decouple gnd and work isolated. We usually do this at the office (not industrial installation) when a test involving vfds is needed

[SeanB]

You might also look into installing the line reactor in the supply, which does a good job of increasing line inductance, and as well as taming the diode input current spikes it also has a big side effect of reducing leakage current through the RF filter capacitors to ground. As well install a big inrush limiter, either a capable NTC thermistor ( with 1.5kVA will be big) or just a  power wire wound resistor and a contactor to short it out instead after a 2 second time delay.

[AllTheGearNoIdea]

Thanks for comments the line reactor sounds interesting. I don’t think the time delay will work or the  NTC as the VFD is switched in out of circuit via a main contractor  interlock with the main guard. I’m thinking a large isolation transformer is the way to go as that how I have it work. It’s likely that the transformer itself is going to need some inrush protection, although I have specified a low inrush in my enquiry to a supplier.

I am guessing I need to determine if it is current inrush or the earth leakage.

[fourtytwo42]

I think you are following the right path (isolating trafo), the problem is it is an industrial drive that is also very noisy so it has large Y capacitors causing the leakage problem, of course they are designed not to trigger 100mA industrial rcd's. However residential rcd's are normally 30mA at most hence the tripping. You dont say if your consumer unit has individual circuit rcd's or there is just one on the incoming grid supply (as is often the case in rural installations).
The other problem you may encounter using industrial equipment in a residential area is complaints about tv and in particular broadband inteferance and you may have to fit additional line filtering to counter that!

[AllTheGearNoIdea]

I was just watching a big Clive YouTube Video with him using a current clamp meter to trace these RCB type proplems.  So I have order the same sensitive clamp meter to do some checks.  The machine also has a forced air cooling motor and a transformer. I think I will disconnect the VFD an double check if it still trips.  It could be a duff motor or mains transformer and I need to be sure I’m not masking a real problem using the isolation transformer.

I’m not exactly sure what’s installed in my consumer unit. I have one large consumer unit but it has two large breakers that look like main switches.  When the unit trips the lights in the house stay on but the sockets go off.  The machine is installed in my garage and the supply is a spur from the kitchen plug socket circuit I think.  The house installation is relatively modern about 20 years old.

Just had a look at the consumer unit and the device that is tripping is marked as an earth leakage device.

Many thanks for the comments and advice all are appreciated.

[fourtytwo42]

Just had a look at the consumer unit and the device that is tripping is marked as an earth leakage device.

So my guess is this is a combined 30AMCB and 30mA RCD, can you take a picture of it and post it ?

Cheaper than expensive test gear buy a 30AMCB without RCD for TEST ONLY and see if it still trips.......

[Edison]

When switching off the shield when using the inverter, it can (if otherwise work) mostly lead to an EMI filter.

[Kjelt]

I have the same issue with a small 3-phase machine.
Some suggestions:
-give the machine its own rcb, remember that the current unit already has many mA's leakage from your fridge, washer, tv etc etc, so it only has perhaps a margin of 20mA's left from the 30.

-use a ssr with zero phase switch on , you get less "spike" noise which also could cause the ecb to trip.

-check the mains filter. In industrial machines they dont give a hoot about leakage currents. I replaced my original mains filter for a 0,5mA low leakage Schurter, lowered the tripping margin significantly.

Then normally with modern vfds they recommend special classB rcb's but those cost over €500 and probably won't solve your problem. They are only less sensitive for the higher frequency distortions.

Good luck and keep us updated.

[Edison]

Just remember that - residual currents are summed
If the protector is in the common circuit with the other device, leakage may occur on another device, when the last leakage machine is connected, activation occurs.
If you do not sensitive enough ammeters Use with caution equivalent circuit.
Disconnect the GND in the device connector - connect the disconnected wire to the mAmeter, connect the second lead of the meter to the wire N the device will be grounded to N and will not switch off the protection. The replacement wiring will only work if it is not connected to an external GND. After the measurement, return everything to the correct terminals !!!

:idea: I do not know what the standards say about you, but in CZ - each circuit is a separate protector (due to the price someone gives a central = fault)
Circuit 230V combined fuseprotector or fuse + protector
Circuit 400V fuse + protector   ( for 400V combined protector is not )

[Edison]

Yes I don’t think it will be practical to wire in a new circuit. I am getting a quote for a large isolation transformer.

the value protector determines  of the environment, not a user  

[capt bullshot]

For lab use, we usually remove the Y filter caps from the VFDs (at the place where I do work, and for lab use only).
There's another not recommended way: isolate the VFD from earth and connect neutral to it's earth.

Note, either way breaks with some rules, so I clearly didn't recommend that ...

[AllTheGearNoIdea]

Thanks for all the comments and suggestions. I’m going to speak to the supplier Warco on Monday and see how helpful they are feeling.   I’m going to unplug all the other equipment in the house to reduce the other residual earth currents from any other equipment as my rcbo could be on the edge of tripping.

I guess I should also just try disconnecting the VFD as there quite a few other motors on the machine. The Vfd and forced air fan both come in together from the main contactor so either could be at fault.  But the filters on the VFD look most likely to me.

Thanks again I will keep you all posted with my progress.

Regards Chris

[Vtile]

Mathematical expression of the situation described above.
VFD + RCD = Problems

[Edison]

Mathematical expression of the situation described above.
VFD + RCD = Problems

YES 

[FrankE]

Is it wired into an isolator or plugged into a 13A socket?
If the latter have you tried it in a filtered socket eg K1816 WHI
The mitigation for circuits supplying equipment having high leakage current eg SMPSs in BS7671 is more about induced voltage, bigger cables 4mmsq for a ring main, dual earths and in more technical environments an earth grid with .

 As someone pointed out leakage currents are additive. At tower blocks the leakage currents I used to measure on the main earthing conductor were in the 10s of Amps.

Could you fit an inlet filter?

[Electro Detective]

...I have a PAT tester is this likely to damage anything like the VFD.

What’s the easiest way to measure earth leakage or get to the bottom of what’s tripping.

Whatever you do mate, DO NOT do any insulation tests on the VFD via your PAT tester or any megger, or you will be starting another thread here   


If it's easy to get into, I'd be having a quick look inside the VFD for any grounding issues and correct A/N polarity, and the stuff mentioned by members above

You could try disconnecting everything in the premises and see if the VFD causes earth leakage trips   

Also consider that the earth leakage device may be degraded if it's old and or weathered,
just changed out both of mine after 14?  years, both were suss,
the new ones pass every external leakage test I can throw at them, the old ones fail intermittently   

Perhaps try powering VFD through a 25 to 50 metre long rolled up extension cord to soften the turn on, might save cash on a transformer 
(Warning: urban myth backyarder jerry rig alert!) 

[AllTheGearNoIdea]

I spoke to the machine supplier  today and they were not unhelpful but clearly new a lot less about domestic installation and VFDs than I do. They just said that it worked fine when tested at the factory and said it should be run off a 16 amp supply and it’s design for industrial use. I said if that’s the case why is it supplied with a 3 pin 13 amp UK plug.

So yes is just plugged into a socket. The sales sheet says it can be run from a domestic supply. But apparently not one that RCD protected as all new domestic installations are.   I do need to pull the lid of my consumer unit at the weekend  and check the standing residual earth leakage.

When I said PAT test I wasn’t suggesting insulation resistant test. The pat tester  I have shows earth leakage by default. However switching the machine on takes out the breaker so disables the PAT tester. However thanks for the warning that’s a good thing.

So I ordered a 3KVA isolation transformer as the machine is working from a 750VA at the moment.   I am a little concerned that this could be masking a real earth fault and I need to investigate this a bit further.

My other concern is that if I do ever get a real live to earth fault in the future that the circuit breaker won’t trip as it will no longer be referenced to earth. I am wondering if should connct one of the secondary winding to earth and neutral conductor?

My house is PMA so I don’t have a real earth anyway as the  earth is bonded to the incoming neutral conductor.

I think the idea of the long extension lead to reduce inrush is a good idea. If I can actually power the machine up I can measure the residual earth current and disconnect VFD and cooling fan.

Thanks again for all your comments it’s been helpful. I’m going to investigate further at the weekend as I cannot keep put the wife in darkness every 5 minutes. 



Best regards Chris

[AllTheGearNoIdea]

[imghttp://][/img]

I was thinking about something similar to this. The reason im considering bonding the neutral to earth on the secondary of the isolation transformer is so that a real earth fault on the machine will clear the circuit breaker.

Does anyone have any thoughts on this.

Regards Chris

[AllTheGearNoIdea]

Just spoken with the distributor for the inverter  in the UK and they were very helpful.  There is a RFI link that can be cut taking the filter capacitors out of circuit.   This sound like the best plan rather than messing about  with isolation transformers.   Should have spoken with those guys first.

I will have a look at this next weekend and report back regarding if it works.

Regards Chris

[Kjelt]

Add your own low leakage filter or you will be jamming the neighbourhood 

[sibeen]

There is a RFI link that can be cut taking the filter capacitors out of circuit.   

Dear sweet WTF. Lets cut out the circuitry that lets the unit comply with IEC regulations for RFI and EMI

Or perhaps you could look into RCDs that are especially manufactured for just this problem. Look up type F & B RCDs. F generally for single phase VFDs and the B and B+ models for three phase industrial machines.