Mitsubishi Tripping RCD

[gwilliamson]

I have installed three Mitsi HI Inverter units in a domestic premises and all was fine. The customer has phoned and told me his RCD has blown on occasions but it cannot be pinned down to any one machine. Mitsubishi have told me that Inverter Air Conditioners naturally omit electrical noise down the earth wire and this is why his RCD keeps blowing. The solution is to get a sparky to create a split load board and take the air con off the RCD.
Does this sound correct or am i been fobbed off a bit?

[Andy W]

It makes sense but I have never heard it before, even commercial premises have RCD 's and I have installed loads of them. What I did have recently was an RCD which was "worn" and first thing in the morning when the lights of the shop , the fridges and the air con was switched on it would trip the RCD, we thought it was the air con at fault, but providing we started the air con first, then the lights then the fridges the RCD did not trip, start the air con last and it tripped, the solution was to get an electrician to fit a new RCD.

[mick2me]

I have installed three Mitsi HI Inverter units in a domestic premises and all was fine. The customer has phoned and told me his RCD has blown on occasions but it cannot be pinned down to any one machine. Mitsubishi have told me that Inverter Air Conditioners naturally omit electrical noise down the earth wire and this is why his RCD keeps blowing. The solution is to get a sparky to create a split load board and take the air con off the RCD.
Does this sound correct or am i been fobbed off a bit?

Thats going to be an expensive job.
It means replacing the whole board.

In a split load consumer unit there are two seperate MCB mounting bars for Lighting circuit MCB's, and and Power circuit MCB's

Assuming that the 15a (or 20a) MCB will fit on
the lighting circuit bar, as long as it is not fed through the RCD, the RCD should not trip.

The RCD could of course be triipping due to a fault
not related to Mitsubishi.

(If I dissapear from this thread, Sorry, I am flying (perhaps?) to Las Vegas tomorrow.)

[mick2me]

It makes sense but I have never heard it before, even commercial premises have RCD 's and I have installed loads of them. What I did have recently was an RCD which was "worn" and first thing in the morning when the lights of the shop , the fridges and the air con was switched on it would trip the RCD, we thought it was the air con at fault, but providing we started the air con first, then the lights then the fridges the RCD did not trip, start the air con last and it tripped, the solution was to get an electrician to fit a new RCD.

Yes good point Andy. The RCD could be faulty (replace?) But I would have thought it would have tripped before the mitsi was fitted. Could be a Current Leaking mitsi? Meggar Test it...

[rbartlett]

We covered this some time ago

http://www.refrigeration-engineer.co...ghlight=daikin

cheers

richard

[mick2me]

We covered this some time ago

http://www.refrigeration-engineer.co...ghlight=daikin

cheers

richard

Some useful stuff there Richard.

I am glad this subject was revived.
and will bear this in mind when fitting my split units.
I will try them on the RCD side of the board first.
(Glad I fitted the Split board now)

[Andy W]

Some good information on that previous thread, are we I wonder going to start to experience problems with the new generation invertors, are there precautions or recommendations that we should be taking because if we recommend split circuit boards everywhere just to cover our selves, we will start to lose the jobs to others.

I think more information or guide lines should be sourced.

[mick2me]

Some good information on that previous thread, are we I wonder going to start to experience problems with the new generation invertors, are there precautions or recommendations that we should be taking because if we recommend split circuit boards everywhere just to cover our selves, we will start to lose the jobs to others.

I think more information or guide lines should be sourced.

Good point Andy.

But if there are likely to be problems with Split air Inverters and RCD's the installer could have problems being called back all the time.

Perhaps we will see details of others experiences with the RCD problem?

[Andy W]

I spoke to a mate of mine this morning and he as already experienced the problems, I have come to the conclusion that maybe a dedicated supply from the incoming live side is maybe the way to go, in future I shall mention that on quotes.

[Peter_1]

Some additional information on the previous post and the posts Richard referred to of 2005.
I will try to explain this as good as possible.

Richard gave us a link how an RCD works. We all know that what goes in the RCD must go out. If not, there’s an unbalance and the RCD will then trip.
The RCD accomplish this by generating a magnetic flux in some build-in small coils and the sum of all the generated fluxes must be ‘0’
So, all the fluxes annul each other.

But a flux can only be generated by a AC current.

Now, due to the large capacitors inside a VFD, some of the AC current at the inlet of the VFD is send back as a DC current, superimposed on the AC signal.
Why this happens,... I don't know. This is beyond my knowledge. I only have seen it on a scope.
So, the total energy going in the RCD is the same as the energy coming back, but a small part of the energy flowing back is converted to a DC current, carried on the AC signal.
So what goes in the RCD comes also out.

But this DC signal can’t generate a flux in the coils of the RCD and the RCD will trip, although there’s no earth leak but only a ‘leak’ which is in fact only a partial conversion of current.
If your megger the circuit, you will find no leak.

What can be done by this: Siemens and Merlin Gerin and perhaps others have a special RCD for this (cost +/- 5 to 6 times the price of a normal RCD)
This RCD has a electronic circuit build in it which filters the DC signal out of the total sinewave and this DC signal is then via a build-in inverter converted back to an AC signal. This new AC signal is fed to an additional coil in the RCD, generating the ‘missing’ flux so that the total fluw becomes zero again.

Installing noise filters or RC filters doesn’t help anything because these filters ony filter out high-voltage spikes and ‘noise’.

Another way to solve the problem: installing a safety transformer with twice the same voltage (400 V/400V for example) so that the RCD doesn’t ‘see’ the leak because both are then completely separated from each other. The DC current of the first coil in the transfomer can’t flow through the second coil.

How can you measure this all? I noticed many times that few techs know that you can measure leakage with a special AC leakage current clamp (cost +/- 300 € for the cheapest type) I bought one of Meterman.

How can you visualise this phenomena? With a scoop able to perform a Fourrier analysis. This will mostly be an electronic version, connected to a PC.

How I learned this? Due to massive problems last year with two VFD’s constantly tripping the RCD of 30mA. We thought that there was something wrong with the VFD’s, then with the RCD’s, meggered all the components, … Finally someone did a analysis for us with a scope and we installed two safety transformers.
The leak we had was 80 mA/VFD which wasn't in fact a leak but a current conversion.

[Abe]

Peter

As usual you have produced a worthy account which will go a long way to helping us with future , current installations where we experiencing this problem.

Thank you from Me.

[Peter_1]

Did dome search via Google and found this link http://www.consumer.gov.uk/renewable.../dgcg00002.pdf

It's rather large and I haven't read through it but it seems that this article handles abouth the problems I described above.

Abeb, I'm allways happy if I can help here and share my experience with others (like most posters here)

[expat]

Hello everyone this is my first post on your marvelous site and I hope it will be helpful. I must say I do feel rather humbled in the presence of some very experienced people, anyway here goes.

I have recently experienced the same problems at least twice. Both on new installs and both in shops occurring in the morning when the shop was opened, the lights were switched-on followed by the aircon followed by a call to the tech that did the fitting to tell him his unit had tripped everything out again.
I went with the chap to one of the sites and we took amp readings but really couldn't find anything wrong.
My friend finished up telling the shop owner that the fault lay with her as the filters weren’t clean (I personally hate telling people that kind of b***$***t). From then on it was the usual Marseille behavior when confronted with the possibility of a bill: arm waving lots of shouting, so we left.

Here I am a couple of months later digging through your site for pearls of wisdom and after reading Peter's last post the proverbial penny drops.

Peter, your explanation has prompted me to recall that when building my house a few years back the new French normes for electrical installations required no fewer than 3 (type AC) r.c.d.s for a house >100m² and in all cases 1 (type A) r.c.d exclusively for the cooking and washing machine circuits.
I didn't understand this at the time and sales people gave me all sorts of incomprehensible spiel which gave me the impression that they didn't understand it either!

It seems that while a type AC rcd will detect leaks of a sinusoidal nature the type A will also detect DC leaks.

see legrand website

[Peter_1]

Bonjour Expat,

Je vous souhaite une (ou est-ce que c’est ‘un’?) bienvenue sur cette forum et j'espère que vous trouver ici tous les réponses sur votre questions.
N’hésiter pas de poser toutes sortes des questions et d’aider aussi les autres avec votre expérience.

Did you moved from England to France?
Your nickname Expat' and your good knowledge of the English is very rare for French peoples.

A schematic of what I tried to describe in my previous post can be found on page 7, left bottom

http://library.abb.com/GLOBAL/SCOT/SCOT209.nsf/VerityDisplay/12C65BC5E082F51DC1256DE400377F36/$File/3051-96%20S0002%20RCCB%20F%20804.pdf


If you look to this schematic, you will see that all the lines are measured and conencted together and going into some block (with the triangle) This is the DC/AC inverter that filters out first the DC and generates then again via new generated AC current a flux in the 5th coil.
Some pages further, there's a general recommendation when using VFD's.

I was once in Marseille while visiting Profroid in Aubagne.
I also remember we passed with the TGV just by the plant of Aerospatiale
We then also ate also in a fisher village Cassis.

[expat]

Thanks for the link Peter.
Yes I did move to France from England, from suffolk to be exact.
I have a friend who works in the commercial departement for Profroid his name is Christophe Luna.
I live in La Ciotat which is just over the hill from Cassis where I'm sure you had a fine albeit expensive meal (let me guess did you have loup de mer or dorade royale? ).

[Peter_1]

It was indeed dorade but if it was expensive, i shouldn't know it, Profroid paid the bill for us.
We had there contact with Mr. Bajolle.
What a small world we're living in.

There was some months ago a show on the television where they said that you only have to go 5 persons far to reach everybody in the world.

So we take the boss of Profroid, I know you, you know your friend, and he know his boss (so 3 persons) so the chain is complete.