Nightmare of a day. RCD tripiing under any load

[brizospark]

Have had one of them days. I had a board change to do and a few other bits and bobs so everything was going great until I tried to put power back on.

Basically I have replaced a 3036 board with a split load Wylex board. One half of the board is working fine but the other half's RCD is tripping as soon as anything is plugged into it.

If I remove all plugs from sockets then the RCD holds but as soon as I plug anything in, kettle, sky box, TV, it trips instantly. On this side of the board is the cooker, 1 x lighting circuit and 2 x ring mains, funny enough the RCD isn't tripping when the lighting circuit or cooker is on.

Now I have checked that I have neutrals from correct circuits going to the correct neutral bar, several times and these are OK. I tried replacing the RCD itself with the one that was OK and it still tripped. What else is strange is that the kitchen ring is on the side of the board which is fine but if I plug anything into a kitchen socket the RCD which isnt even protecting this is tripping!! I opened up sockets and everything looked OK but it was getting late on so I replaced the offending RCD with a main switch I had in van just so the couple have power on tonight before I go back again tomorrow. Switched on with RCD removed and everything fine.

Anyone had experience with this problem before? Any advice much appreciated

 

[sima24]

Sounds like borrowed neutral to me. Have u tried putting the 2 lighting circuits on the same rcd?

 

[sima24]

Has it been tested to rule anything else out?

 

[Brightspark2]

Everything test out ok?...


Borrowed neutral?....or shared neutral whatever you want to call it.

 

[Marvo]

With a borrowed neutral, shared neutral or split neutral or whatever you call it the RCD won't trip until there's some current flowing. If there's no load current then there's no fault current so it will hold if you unplug everything and turn all the lights off.

Rather than using the RCD to do your fault finding use your normal tester and your mega tester if necessary.

 

[trev]

It could also be a neutral to earth fault but your IR testing ruled that out. Right?

 

[Dizzy_Maskell]

Yes sounds like to me like a borrowed neutral. Have they had extensions done you will need to break each circuit up and test in stages some where some one had tapped into a neutral or you could try try putting all socket circuits on one side of the RCD

 

[brizospark]

Thanks for replies guys yes everything tested out ok. So we are looking at a borrowed neutral most likely then?

 

[sledgehammer]

Neutral to earth fault, megger out.

 

[i=p/u]

Test before touch

 

[sledgehammer]

Sorry should've said, shows its self as you try to draw power. As you said its as soon as you plug in and use power.

 

[wersolaruk]

Sounds like a existing fault somewhere on wiring I.e nipped cable in box etc creating a existing fault on system once all the appliances are plugged in could just be cresting enough leakage to trip rcd. Had that one before, customer had used a 16mm back box whilst adding a new socket in kiddies bedroom!!! and cables were pushed against it! I came along changed d.b and found the fault 5 hours later! Pulling my hair out!

 

[trev]

It's either N-E fault in which case a simple global test will show it up and you can narrow it down to a circuit from there or it's a borrowed neutral, in which case a simple test will show it up.
Borrowed neutrals are usually on the landing light in my experience, should be an easy one to sort mate.

 

[brizospark]

What's confusing me is why when I plug anything into kitchen socket covered by RCD A is it tripping RCD B? Surely this means it can't be a N to E fault as it would trip RCD B if this was the case

 

[brizospark]

Can anyone explain that?

 

[SJD]

I'd also suggest neutral to earth fault somewhere, and not necessarily on the circuits on the problematic RCD.

 

[sledgehammer]

If its pme n-earth fault would common out both neutrals. I've had had exactly what you're talking about many times and its n-e.

 

[wersolaruk]

Could there be a borrowed neutral at that kitchen socket itself. Turn the rcd off covering that socket and turn other rcd on. Test between n-e at the problematic socket and if its showing voltage to earth it's a borrowed neutral! Sounds like a fault I would love to find!

 

[Brightspark2]

If its pme n-earth fault would common out both neutrals. I've had had exactly what you're talking about many times and its n-e.

Nice avatar :smartass2:....

 

[brizospark]

Could there be a borrowed neutral at that kitchen socket itself. Turn the rcd off covering that socket and turn other rcd on. Test between n-e at the problematic socket and if its showing voltage to earth it's a borrowed neutral! Sounds like a fault I would love to find!

It's not just this kitchen socket it is any socket in house covered by either RCD

 

[Richard Burns]

What's confusing me is why when I plug anything into kitchen socket covered by RCD A is it tripping RCD B? Surely this means it can't be a N to E fault as it would trip RCD B if this was the case

If you have an NE fault on a circuit on one RCD then loading a circuit on the other RCD can cause the first RCD to trip.
Did a diagram for you:

 

[wersolaruk]

You have a n-e fault, megger everything buddy! All fronts off etc, good luck with it mate!

 

[shockboy]

try lighting common fault landing light 2way neutral on upstairs circuit feed took from downstairs ciruit

 

[markythesparky]

Thanks for replies guys yes everything tested out ok. So we are looking at a borrowed neutral most likely then?

If its a borrowed neutral then how did it test out ok?

 

[bigspark17]

Borrowed neutral on light circuits.! Put all lights on same rcd and try.!

 

[trev]

Borrowed neutral on light circuits.! Put all lights on same rcd and try.!

Or check to see if there really is one, locate it and eliminate it.

 

[Engineer54]

Every time we get a thread about RCD's tripping after a CU change, the common consensus is either a N-E fault or a shared/borrowed neutral. And every single time the OP has failed to test the installation prior to replacing the old rewirable CU with one containing at least two RCD's.

How many more times do these guy's NEED to be told to test these installations before they start changing them out for god's sake??

What's more, why is this fault causing such a problem to find and rectify. I would have thought that in around an hour, an experienced electrician would have sorted this problem out, ...or at the very least well on his way to finishing off any remedial work...

 

[markc123]

Test before touch

:shame: that's where i've been going wrong.

Wondered where my afro had come from.

 

[brizospark]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

 

[trev]

A test for borrowed neutrals plus an IR plus an EFLI wouldn't take very long at all would it?
You've already done enough there to identify any major problems and you're doing the rest of the tests anyway on completion of the change so what's the problem?

 

[marcusone]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

Its not an additional cost. Its part of the cost of doing the job right from the start. Not every customer wants it cheap, most want it done once and done right.

 

[Guest111]

The problem you know have is a potentially peed off client,did you leave them without power last night? And possibly a lot more work to do if it's not a quick fix,if you test thoroughly before swapping the board any issues will be found and corrective action and any further costs can be discussed with client.

 

[Engineer54]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

What do you mean extra cost?? The Minimal IR/ELI/Shared Neutral tests are for ''YOUR'' benefit, so that you know, or are aware of any problems, (such as you have now) before you start!! Armed with that information, you can advise your customer of any extra costs that will be associated with the board change!!

 

[GLENNSPARK]

Sounds like borrowed neutral to me. Have u tried putting the 2 lighting circuits on the same rcd?

eh...?
its on the socket....

 

[GLENNSPARK]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

what a bunch of old twaddle this is...

having read all this its clear to me you are not competant to carry out board changes....

leave such stuff alone until you have gained the necessary experience

 

[Sharpend]

Brizo, no disrespect but you may need to shadow a more experienced spark, if not to improve your knowledge, them to learn the correct proceedure and why it's correct.
Are you fresh out of training?

 

[SJD]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

When you have completed the CU install, you will have a completed test sheet with the test data for each circuit (I hope), so you can just as well do some of that testing before doing anything to the old board, just in case it throws up things like N-E faults, or broken rings etc. There is minimal extra cost, but potentially saving a lot of embarrassment.

 

[Guitarist]

Eng54 do you honestly think in the real world customers would give the go ahead to carry out testing before changing a board? This would mean an additional cost on top of a board change The customer would simply get someone else to do the job!

I have to agree with Eng here. Basic tests at the time of quoting take 20 mins tops, and save a world of pain down the line. Also, the added bonus of carrying out a few basic tests gives you time to chat to the client and show them your shiney MFT. They get to know you, see that you are competent, and will usually give you the go-ahead there and then. Much better than someone who pops in for 2 minutes, gives a rough quote, then adds to the bill later on when the issues raise their ugly head.

 

[Brightspark2]

I know it's like an echo now but I agree with the above comments, testing should be done before you change the DB. I wouldn't change one without testing otherwise it just comes back to bite you like this. Then the customer rolls the old "well it was fine before".

 

[Chris sparky]

I know it's like an echo now but I agree with the above comments, testing should be done before you change the DB. I wouldn't change one without testing otherwise it just comes back to bite you like this. Then the customer rolls the old "well it was fine before".

Same as this. If they have had a quote off another "sparks" who's cheaper because he is going to lash it on and hope for the best, then let em do it. You may well get the call when it goes ---- up :smug: