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

[AllTheGearNoIdea]

Do test the rcb with a resistor leaking 35mA to gnd see if it trips.

It will be 100mA now as I have installed a 100ma RCB on the output from the issolation transformer.  I think my fluke tester will display the trip current and the time delay.

Thanks Regards Chris

[Kjelt]

That means it is not safe.
35mA through the heart is critical, 100mA is instantenously death.

[Zero999]

That means it is not safe.
35mA through the heart is critical, 100mA is instantenously death.

It's unlikely though and a 100mA RCD is still better than no RCD at all.

The risk of electric shock on fixed machinery is fairly low, compared to portable devices with a flexible power cord, which as far as I'm aware, is why RCDs are not mandatory on machinery in workshops. 100mA RCDs are commonly used in industrial situations, as protection against fire, more than shock, although it will provide some protection against the latter.

[Kjelt]

He could have tried a 35mA first, as stated earlier the leakages in the rest of the house are now isolated so it could have worked.

[Zero999]

He could have tried a 35mA first, as stated earlier the leakages in the rest of the house are now isolated so it could have worked.

I agree, except I've never seen a 35mA RCD, 30mA is the usual value.

[Kjelt]

Ah yeah 30mA it is.

[AllTheGearNoIdea]

 Ok the clamp meter arrived today. I’m getting around 14mA at the mains lead and 10 mA at the VFD. There a few other fan motors and filter accounting for the other 4mA.  Also likely some measurement errors  as the clamps sensitive to position and adjacent metalwork.

Just googling the subject states that VFDs leak 20mA upwards.

Your correct I could have tried a 30mA RCD and perhaps I will. But let’s not get carried away, I grew up before RCDs were common and managed to survive. These mainly serve to protect the average very stupid user that decides to pick up the cord just cut by the hover mower.

Regards Chris

[Kjelt]

You know what I realized, if there is standard 20mA leaking that should be considered the zero state.
A RCB should have a nilling or delta knob to set this leakage as normal and ignore it, eg only trip if an additional 30mA leaks.
These things are too dumb to be really usefull in such cases.

[AllTheGearNoIdea]

Looks like I will be trying the 30mA RCD as that is what arrived in the post today. I did order the 100mA

[Electro Detective]

So far so good    

I reckon VFD may be fine with a 30ma RCD because the transformer should block all any recycling nasties going up and down the line/s causing issues

FWIW a good quality '30ma' labelled RCD should trip anywhere (depending on ambient conditions)  between 22ma and 27ma in zero and 180 degree testing, or a real world trip   

If 30ma is a fail, better to have a 100ma RCD than no RCD

I have one packed in a box somewhere    just in case, or as a future upstream RCD at the switchboard before the main breaker/isolator if ever needed


Doesn't hurt to own one, and not cheap or easy to get when you need one 
 

[AllTheGearNoIdea]

Well all up and running now with the 30mA RCD. Did my first cut with the new milling machine today, cutting a window for the RCD in a plastic box. The new VFD is much better having to constantly change those drive belts.

[Kjelt]

That is excellent news, so the 30mA does not trip? 

[Edison]

what I said, she wants to measure and not guess 

[AllTheGearNoIdea]

Here is a link to my first go on the new milling machine. I show some of the leakage current issues and the isolation transformer solution. Not completely happy with the earthing arrangement and would be happy for any comments etc. Not sure if I should have connected in the domestic earth connection.  The VFD information is about 8:51 in as many people won’t be interested in the machine part.

Thanks Regards Chris

https://youtu.be/F82FMSPw38o

[Jeroen3]

That a cool machine. I'm sure you can put it to good use.

Not completely happy with the earthing arrangement and would be happy for any comments etc.

I'd be unhappy with this way as well. But it runs for now. I have seen bad things happen when running VFD's from low impedance sources, they can't do load steps really well. But you won't be making those anyway. VFD's are really designed to run on the grid, those that are not are marine approved (and not affordable). And even then there are many problems.

If I would buy such machine, I would install (or have installed) a RCD-less socket in the workshop or a higher rated one that is VFD compatible*. Preferably with an other type of plug.
CEE 16A for example (the blue one).

*This is important, I'm actually in the process of developing an automated tester for RCD's, and VFDs create such high frequency and distorted leakage current, often also DC, that normal type A and AC RCD's refuse to trip if their coil is saturated with DC or this distorted current. Testing some units on the shelf shown that you could have seconds with up to 3A DC leakage before they would trip.
This problem can also occur with bad PFC power supplies... So something to think about.

[AllTheGearNoIdea]

Hello jeroen

Thanks for the comments. I did actually order a 100mA RCD for my power converter box but it was a 30mA version that turned up in the post.  I would have liked to have run directly off the mains but there was just no easy way to defeat the RCD in the consumer unit and still be compliant with the U.K. regulation.  The drive manufacturers say there is a link that can be cut to remove the filter capacitor and that was also an option. I guess I’m just going to have to try it and see how I go. These VFDs have really come down in price so not the end of the world if it pops.

Yes I’m aware of desensitisation Of breaker. I work in the rail industry and we still have some traction system that use DC and also tram system running from DC and it can be a real problem.  Network Rail specify the use of dc immune breakers in such applications and these are really expensive. Still not quite a big an issue of the cathodic erosion of steel structure and steel reinforcement bars.

Good look with the test gear be very interesting to see it when you have it done.

Chris

[Jeroen3]

there was just no easy way to defeat the RCD in the consumer unit and still be compliant with the U.K. regulation.

Over here, you can screw around all you like. No third party will ever validate what you do, until injury or you sell the place.

Good look with the test gear be very interesting to see it when you have it done.

Yeah, unfortunately some things stay proprietary.

[Kjelt]

Very nice machine Chris. 
Great you got it running.
I haven't used my vfd since for the drillpress but soon will use another chinese vfd for my cnc machine.
No one of the many tens that use these vfds on the cnc forum has complained about their ecb tripping, so I guess there are vfds that don't leak as much.
Searching for mains netfilters from Schurter I also found out that there are filters esp three stage ones that leak more than 20mA and there are ones with two stages that leak <0,5mA.

The ecbs fit for vfd cost over €500 so this might be two cheaper future ways to tackle it although Jeroen is right they might not trip while they should in case of vfds, difficult dilemma.

[AllTheGearNoIdea]

Yes having the VFD really makes a great improvement. Found that if you get resonant chatter of the tool you can slow down the cutting speed but also sometimes just increasing the speed helps. The ease of doing this on the fly as opposed to belt changing for speed control is a massive bonus.
Thanks

Regards Chris