rcd tripping

[GMC]

Hi there.

I needed to replace two socket plates in my parents kitchen, i bought the relevant 2 gang sockets from Screwfix, a brushed stainless Volex socket which will replace the white plastic ones currently installed. As soon as I connect either of the new metal sockets the RCD in the consumer unit trips when the power is turned back on. I have triple checked that I am connecting it up correctly. The fault goes away as soon as the plastic socket is reinstated. I initially thought that the socket may be faulty (!) so used the second one which gave exactly the same behaviour. It seems that the metal sockets are directly related to the RCD tripping but I dont understand why, can anyone explain this to me or offer a solution please? (my solution may be to replace the sockets with plastic ones!)

Thanks

Gerry.

[Edward]

I assume that you've earthed them, right?

[John Rumm]

If you connect the metal socket, but leave it hanging on the wires and don't screw it to the backbox, does it still trip?

I take it there have been no other changes other than the sockets? (no re-wiring etc or new circuits)

Is this a split load CU, or one with a single RCD protecting all the circuits?

Some metal sockets are quite low profile and hence leave less room in the box. This can compress the wires more, and also allow the fixing screws to penetrate deeper. That makes it more likely that any small nicks in the insulation could cause problems. Check the wires carefully

- especially where they pass into the box (it ought to have a rubber grommet if its a metal back box to protect the wires - but these are often omitted)

[GMC]

The socket is earthed exactly the same as it was with the plastic sockets.

The socket does still trip if i leave it hanging on the wires.

The CU is a split load in a fairly new build.

There have been no other changes at all, just the socket swap.

I did notice that there was a rubber grommet where the cable enters the back box but did not scrutinise the cables for nicks etc.

I dont understand how it works ok with the plastic units but trips with the metal ones! Google did not have much to say about it either!

Thanks for the help so far.

[Frank Erskine]

On Mon, 06 Jul 2009 09:33:45 +0100, GMC had this to say:

This might sound silly, but have you checked that you've connected the wires to the right terminals? The layout of the terminals can vary from one make to another.

[GMC]

Not silly at all. I did check this and it all looks correct.

[Andy Dingley]

What if you remove it (carefully!) and test it with no socket connected at all?

What if you do this without the socket faceplate, but with the earth bonded to the back box?

If you've damaged insulation and you've physically moved the cable ends, then there's a possibility you've now got a short onto the back box where the cable enters.

[GMC]

I will try what you suggest.

But it trips with the faceplate hanging loose and no earth connection back to the back box.

I will scrutinise the wires for nicks etc.

Thanks

[GMC]

Ok, further to a few suggestions I have checked the insulation on the cables for nicks etc and cannot find any, indeed it all looks perfectly normal.

As a test i decided to fit a new faceplate to a *different* socket. Same result, RCD trips. This is with the socket hanging on its own not screwed to the backbox and not earthed to the backbox.

Im confused.

[Dave Osborne]

Sounds like the faceplate may be faulty. Can you take the faceplate off and measure neutral-earth resistance at the connection terminals. This should be infinity. If it is low or zero, then there is a connection between neutral and earth on the faceplate that shouldn't be there.

[Graham.]

It is odd, I can only suggest they are both faulty or have been exposed to moisture or other contaminant. Disconnect them completely and use a multi-meter set to its highest resistance range between L&E and N&E. If there is any measurable leakage at all, take them back.

[Frank Erskine]

On Mon, 06 Jul 2009 11:15:18 +0100, GMC had this to say:

Is it actually the RCD that trips, or the MCB? An MCB trips to over-current, and an RCD trips to a L/N current imbalance.

[NT]

maybe your nw metal sockets are a faulty batch. I'd test them. A multimeter will pick up basic faults.

NT

[Peter Parry]

I'm assuming it is the RCD which is tripping. There is another possibility which is that somewhere in the house is a neutral/earth fault. These are commonly introduced when a new split load consumer unit is fitted to old wiring or when a floorboard nail is put through a cable.

The new metal faced socket may introduce a very small inductive current, more than the plastic one will. I had a similar fault once where simply plugging an extension lead (any extension lead, nothing connected to it) in to a particular socket tripped the RCD. The fault was eventually traced to a nail linking neutral and earth in an entirely different circuit in another part of the house. The trip was caused by the small transformer effect the extension was producing.

It was found by isolating all the neutrals in the CU and measuring each individually for insulation to earth.

[Andy Dingley]

Use a high voltage resistance meter to check this, not just an average multimeter. Don't complete electrical installation work without one. If you don't have one, eBay's your friend.

For this sort of fault, a cheap DVM wil probably find the fault, However I wouldn't "test" the system and claim it didn't have other faults unless I'd tested the insulation at high voltage. You don't need a few hundred quids' worth of 17th-ticketed Part-P-tricorder to do this, just an old hand-cranker or a tenner's worth of cheap Robin from eBay.

[Pete Cross]

Are the sockets illuminated ? maybe they are badly constructed and use a neon between switched live and earth ?

[george (dicegeorge)]

I want one... like this one?

number: 360165628074 Digital Insulation Meter Volt Resistance Tester 1000V £50.90

[g]

[Dave Osborne]

Yes, like that, but if you're going to buy one, you should consider getting one that is compliant with the wiring regulations (introduced in the 16th Edition). This requires (among other things) that the short circuit current is a nominal/minimum of 1mA. You will note that all major branded product will either state "16th/17th edition compliant" in the general features or state this explicitly in the specification:

e.g. from Megger website:

"Test Current on Load = 1 mA at min. pass values of insulation (as specified in BS7671, HD 384 and IEC 364)."

The unbranded Chinese model you have seen does not explicitly state this, which *may* mean that the 500V generator circuit is not capable of delivering a nominal 1mA into a short circuit, which may in turn mean that under certain circumstances it's readings may not be trustworthy.

[Dave Osborne]

/\ Sorry for the aberrant apostrophe.

[John Rumm]

Check the live to earth resistance. If it were a neutral to earth bridge then it would probably cause a trip even with the circuit MCB off. (assuming the RCD is shared between a few circuits)