Possible Consumer Unit Problem

[benb]

Hi,

I was hoping someone here might be able to give me some advice on consumer unit/RCD wiring.

I have a Volex consumer unit in my garage, which came with a Volex RCD, and is connected to the main consumer unit in the house.

The consumer unit has the RCD, and 4 MCBs, 1 for the sockets, one for the lighting, one for the freezer (this means if the sockets are turned off, the food doesn't defrost). I also have a spare for future expansion.

At the weekend, I was trying to add an additional socket to the sockets circuit, and flipped the MCB off, so I could work on it. However, I found when trying to wire the new socket in, that my finger tips were tingling, (luckily I don't think I was earthed), then when I was screwing the wire into the socket, the screwdriver touched the outer metal casing, there was a loud bang, bright flash, and the main RCD flipped (screwdriver also has a melted mark halfway up).

Obviously something has been wired in incorrectly, I'm thinking the neutral wire going to the N Bar, and the Bus bar are the wrong way around, I seem to remember when buying it that the RCD & bus bar were loose inside, although the neutral wire was screwed in to the N bar (and there were no instructions). The live is fairly obvious, as there is an 'L' sticker next to that screw.

I have created the following diagram showing the current layout:

'd be grateful if anyone can offer any suggestions!

Cheers

Ben

[philipuk]

layout:

> I'd be grateful if anyone can offer any suggestions!

You're a lucky lucky man. You shouldn't be here you should be a smoking corpse on the floor. You've mad a number of errors. Not switching off at the main isolator, not checking if the circuit you are working on is actually dead. The problem is pretty obvious but if you can't see it.... Also Part P.....

Philip

[Andy Burns]

and socket neutrals.

And neutral out of the RCD going to the live bar and the MCBs and socket lives

Are the colour in your diagram accurate?

[zikkimalambo]

layout:

> I'd be grateful if anyone can offer any suggestions!

It doesn't actually say, but I would guess that what goes into the left hand top (L) of the RCD comes out at the left hand bottom.

Therefore, "live" is connected to the neutral busbar, and neutral is being fed to the MCBs via the live busbar!

[<me9]

L and N are transposed at the RCD.

[John Rumm]

A bit more information would be useful...

Where is the sub main to the garage CU fed from? (spare way on house CU etc?)

What type of earthing does the house have? For the options, see:

where is the house (in the sense of in a town, or out in the sticks?)

Is the house earth exported to the garage?

How far from the house is it?

What type of cable is used for the submain?

You have a sub optimal arrangement there anyway since a RCD trip caused by leakage on any of your circuits will kill all the power (including the freezer) and lights.

The RCD, or the MCB as well?

Remember also that the neutral *is* a live conductor (as opposed what is colloquially called "live" which is really a "phase" conductor) and should be treated as such. How significant this issue will be partly partly depending on your answer to the earthing question above.

It is safest to disconnect the neutral from the CU bus bar when working on the circuit - it also saves causing the RCD to trip when you accidentally cause neutral to earth shorts.

Not necessarily... but certainly possible.

A quick check with a multimeter will tell you.

and black in reality? (or just an artefact of your diagram?)

*If* the incoming submain wire is brown (or red), and that *is* connected as shown, then you probably *do* have the phase connected to the neutral bus bar!

[benb]

Yeah I know, I realised after just how lucky I was.

The trouble is I like to try these things myself, and try to learn how to do it, rather than get someone else in to do it.

I'm fairly sure the problem is with the neutral and bus bar at the bottom, just needing swapping around. But if it's not, please let me know, as I'm liable to just go out and try that anyway!

[benb]

Yes, it was some armoured cable that I got from a friend. I believe the colour code conforms to the harmonised colours in page 1 of this:

[Andy Burns]

what about the blue/red goin gout to sockets, is that blue/brown, or black/red. Is it just the link between RCB and neutral bar that's blue?

If it was 2 core armoured in harmonised colours I'd expect it to be brown/blue, if 3 core I'd expect brown/black/grey

old colours would be red/black, or red/blue/yellow

you show brown/black, so is it 3 core and you're sleeved the grey with green/yellow? is so that's OK, what have you done with the armour?

If you moved the RCD to the

[Lurch]

On Tue, 27 Mar 2007 13:17:59 +0100, "benb" mused:

Leave it alone, get someone who is competent.

Did none of this show up when you tested the work after you fitted the board, or when you tested the circuit before working on it to ensure it was in fact dead?

[John Rumm]

It does highlight the importance of measuring both L & N to Earth as well as each other when confirming a circuit really is isolated!

[Cyberdog]

Could I ask. If I am working on a socket with the RCD switched OFF at the main consumer unit which feeds the ring main. and no power to the socket. If I was to short pos to earth the main RCD will trip, switching everything OFF. Is this correct?? Thanks

[Roger]

The message from Cyberdog contains these words:

I take it that the first RCD you refer to is actually not a RCD but a MCB (mini circuit breaker) and the 'pos' is the live side of the ring circuit.

AISI shorting the live to earth should have no effect as the MCB has disconnected that leg. However shorting the neutral to earth will as you are providing an alternative route for the circuits that are still live. That will happen whether or not you have a fully RCD controlled CU or a split load one. The RCD will be triggered taking out all the other circuits controlled by it that were still on but not any (such as the lighting on a split load unit) that are not.

[Lurch]

On Tue, 27 Mar 2007 20:38:55 +0100, Roger mused:

That's what I'm assuming as well for the purposes of this reply.

Possibly, but EMF can provide enough induced voltage to trip an RCD.

[Cyberdog]

Hi, Thanks for confirming that all is OK. You are exactly right as to the correct terminology. Thanks

[Andy Wade]

Not to mention the importance of the polarity test as part of the initial verification of an installation before first connecting the supply, as required by Reg. 719-03-01 in BS 7671. See also page 78 in the On-Site Guide or page 47 in Guidance Note 3.