How to test an RCD module in a CU?

[Lobster]

I've recently fitted a split load CU (Screwfix 82204) to a new installation and am having problems with the RCD module. As soon as I turn on the CU's main switch, the RCD trips, even when all the circuits are switched off or disconnected.

Have checked wiring of the CU and RCD, and that looks OK. Is there anything that could cause this problem other than a faulty RCD module? I don't know how likely that is, but what I'd like to know is how I can test the RCD definitively and safely.

The CU is unused but have had it 'in stock' for some time so I'll probably have to buy a new RCD if it's duff - hence I'd like to be sure.

Thanks David

[OldBill]

Obviously you have wired the circuits so that those fed by the RCD only get their neutral from it as well?

[dom]

Have you done a 500V insulation test on all your wiring for all combinations of phase, neutral and earth?

The try disconnecting the output side phase and neutral from the rcd. If it still trips, then it would sound like it's faulty.

It's not something stupid like the test button stuck down?

[John Rumm]

The usual cause of this on a new install is having the neutral of a RCD protected circuit connected to the main neutral bus bar and not the RCD neutral bus bar (or the link between them is still in place.

If this is not the case, then a neutral/earth short would be another possibility.

For testing methods etc, you may find this helpful:

The CU is unused but have had it 'in stock' for some time so I'll

only £15 for TLC if it comes to that.

[OldBill]

Anyway, should a CU be changed by an amateur? isn't this covered by Part P of the Building Regs?

[Andy Hall]

It is, but this does not mean that it can't be changed by an amateur.

[Lobster]

Nope, that's all OK.

that would help in my case, that don't require specialist kit which I don't have.

If I was to remove the RCD from the CU and connect L and N to its input side, then if it trips then, doesn't that mean it must definitely be duff?

David

[OB]

Maybe but doesn't he have to get LA to inspect it?

[John Rumm]

Yup, there are a few tests to try. Disconnect both outlets of the RCD (i.e. the connection to the secondary neutral bus bar, and the live bus bar), and then apply power. If it trips, then it is indeed knackered.

If it does not, then use a DMM on a high resistance range to measure between earth and the RCD neutral bus bar on the CU (with main switch off, and main equipotential bonds disconnected). You should see an open circuit.

The only test you can't so easily do without test gear is insulation resistance.

[John Rumm]

Haven't you got an egg to go suck granny?

Part P does not prevent a baboon from changing a CU, so long as the relevant building notice has been submitted. (and I doubt many people care if it hasn't)

[Andy Hall]

Submitting a building notice is one way, yes.

[OldBill]

What many people care really isn't the point. If the OP wants to sell the property he may have a potential problem. e.g. suppose he doesn't make it official or mention it at sale, then a fault injures the new owner.

[OldBill]

Moronic Essex thug

[OB]

Do you actually get any work from that amateurish website in your sig? A 14 year old could do better than that. It would be ok for a plumber, but for an IT related outfit its poor.

[John Rumm]

Ah, got the highbrow team in tonight I see.

Don't worry, I am sure plenty of others had a similar problem first time they had a drink as well.

[Owain]

But if a fault injures the new owner it won't matter whether it's been Part Pee "approved" or not. In fact, having paperwork might increase his liability.

Owain

[Frank Erskine]

Possibly, but it does help if it works correctly first.

[Frank Erskine]

Hopefully a potential (no pun intended) new owner will have a survey done before he buys. If he doesn't, he's only got himself to blame. What's with this "official" bit? No "official" procedures (which imply government meddling) protect a purchaser.

[Lobster]

Ah - no open circuit on one if the ring circuits. Yes, of course - I'd been thinking that with the individual circuit MCBs switched off, then the problem couldn't be related to the circuit. Duh.

Today I tracked the fault down to a short between neutral and the CPC in one leg of a ring main, between sockets and buried under plaster. Grrr. The only possible reason I can come up with is a nail through a cable when capping was fitted before plastering (although I put the capping on, a pro later did the plastering so I'm going to blame him! I'm bloody sure I wouldn't have done it, anyway...). Either that, or it was a faulty length of cable. Hmm.

Anyway - I was cursing about the prospect at having to replace my chased-in cable (the cable was dropped down all the way down from the ceiling void), when I realised that I could just disconnect and isolate the damaged leg of cable, and then reconnect the socket to the ring as a spur, using the remaining undamaged cable drop. So, not too awful a result.

Thanks very much John and others for pointing me in the right direction! David

[John Rumm]

Thought that might be the case ;-)

Neutral CPC faults can be tricky to find... a diligent person would test the circuit after each alteration, but it takes a much greater degree of anal retentiveness to check all the circuits after ever nail you hammer in!

Easily done. I have seen someone nail a length of picture rail to a wall with the nail placed right in the middle of a nice new strip of plaster running down the wall from ceiling to light switch. (and the wall paper was off as well - so no excuse there).

That should do it...