Finding a Neutral Earth fault

Joined
6 Jun 2004
Messages
4,021
Reaction score
340
Location
London
Country
United Kingdom
One of those Ignorance is Bliss / Wish I'd Never Started This stories.

The boss wanted an Induction Hob and no spare slots on 8-way Consumer Unit. Also EXTREMELY old (although probably as safe as it ever was - I thought), so change it for a CU with more slots. And, given that MCBs are now the most basic option and not wanting the risk of a single RCD with no clue where the trip(s) might be coming from, I decided to use RCBOs. Ok - expensive. But you can always immediately tie problems to just one circuit.

So yesterday I stuck the thing in. And everything is OK including the new hob (with its dedicated radial using 6mm cable to meet its max load of 32A!!) but EXCLUDING a lighting circuit that immediately tripped the RCBO. Investigating revealed the cause straight away: a Neutral to Earth fault that had probably existed for a VERY long time but without any RCDs in the system, remained undetected.

Worse, several of the connections that might be wrong are lost in ceiling voids: the roses on several of the pendants have only one TC&E cable in them, indicating that some cowboy in the distant past moved pendants but (presumably) substituted junction boxes in the void instead of extending all the cables to the new positions. Understandable but irritating, since I've no idea of the original locations and the JBs are almost certainly under fitted carpets, bathroom floor, etc.

In the computer networking world, there's a gadget called a TDR for locating (Ethernet) cable damage but as far as I know, the only tool available for mains wiring is an ohm meter. At two locations where I've found a full complement of cores in the ceiling rose, the resistance Neutral to Earth is about 2 ohms. This is about the same value as on the end of the cable at the consumer unit.

Needless to say, being lighting and quite old, it's a radial circuit, not a ring. But I can't at present work out what the route is.

I guess the only good thing to come out of this is that 8 out of 9 circuits are working OK with RCBOs! If I'd chosen a split CU with 2 RCDs, I'd have spent a lot longer fiddling around to isolate the faulty circuit.

Anyone recommend an effective process for localising the fault to a smaller area than the whole circuit?
 
Sponsored Links
gadget called a TDR for locating (Ethernet) cable damage

You can get and use them on mains wiring, we use them all the time on the DNO networks up to 33kV for fault finding.
There is one sat in my car outside as I type this reply
 
It would be helpful if you could tell us what your test results for this circuit were before you changed the consumer unit and the test results after you changed the consumer unit.
 
Only disconnecting all the wires at what you think may be half way along the radial (record which wire is which for refitting) and then check each half to find which has the fault.

Then split the faulty half - and so on,

Have you separate circuits for upstairs and downstairs?
Often the landing light will have live from downstairs and neutral from upstairs circuits. This will cause tripping.
 
Sponsored Links
The whole of the affected circuit is on the First floor. The stairs/landing lighting is on the downstairs circuit and is OK on a different RCBO.

Other test results to this point are irrelevant: there's definitely (very) low resistance Neutral to Earth on the (single) cable leading to the start of the circuit. And the measured resistance N to E further down the circuit is similar.

Slightly comforted by the thought that a TDR might help. However, since I don't know the exact route, being told that there's a glitch 25 metres from instrument would not be much help!
 
Slightly comforted by the thought that a TDR might help

Also finding the speed of propagation might be an issue, though a test on a known length of similar cable will get you into the ballpark
 
The first thing I would check, if you haven't already, is EVERY ceiling rose/fitting.
I did a CU change not so long back where testing revealed a neutral earth fault on lighting - fortunately for me it was in the very first rose that I checked - rose had been screwed back on pinching the neutral and CPC together.

The problem with finding the 'route' of a radial circuit is that, a lot of the time they can be wired in a 'spider' - in other words, the main feed into a JB and then the feeds to individual lights fed from here.

If it's first floor lighting, can you not access most of the cable in the loft?
 
a lot of the time they can be wired in a 'spider' - in other words, the main feed into a JB and then the feeds to individual lights fed from here.

Yup - that's the sort of possibility that worries me! Actually the house is 3 floors total but half the first and all the second is on a secondary CU - I think there was a separate flat on the upper floors at one time. Then, when it was rewired later, they followed the same arrangement.

So the faulty circuit is between the fitted carpets, top bathroom floor, etc. on the Second floor and the finished ceilings on the First, with the added complication of 4 out of 5 ceiling fittings that were moved, leaving a hidden JB, before I got the house.

This could get messy!
 
You said this was an old install, how old? What kA where the previous breakers, 3 or 6?

Are you happy with the ELI & IRT readings taken for the install when you did the board change?

Have you done an IRT test on the faulty cct ?
If bad IRT then it could be trapped,damaged or deteriation of cable not a bad connection. Split down what you can get to end to end and eliminate the bad section and replace, forget looking for bad connections in a sealed ceiling or floor.

Can you renew the section or circuit? I am sure your boss will understand that its got to be sorted and if you cant get to hidden wiring it needs renewing. Not your fault you are doing it right to BS7671.

I have seen with lighting circuits spare neutrals used for earths and vice versa so any altered or added area may be the culprit of a crossed connection. Break it down, prove each conductor & IRT test it for damage etc.
It could be as simple as a neutral at a switch (T&E method) not in a connector touching back on the earthed backbox

And if it was a borrowed neutral from another cct it would trip both circuit RCBO's not 1.

Good luck, no pressure mate its only for the boss!
 
Other test results to this point are irrelevant
Nuff sed
I guess that it is 'nuff', really. Even if the OP had tested the circuit, and found the N-E fault, prior to changing the CU (which I presume is what the 'digs' are about), he would still presumably have been looking for advice in relation to the same question he has asked - how to locate the source of the known N-E fault. I can't see that any other standard test results would help you, or anyone else, to answer the question, would they?

Kind Regards, John.
 
Although pre-testing may have given prior warning it would still need correcting so really where one is not giving quotes and you know the job needs doing come what may then to pre-test only tells you if fault is yours or existed before so little point.

As to finding there are two methods.
One is remove neutral rose at a time or replace rose at a time.

At such a low ohm then easy to attach a battery bell or light and just remove one at a time. Once fault cleared replace in same order as removed in the hope of reducing the area under investigation.
 
Other test results to this point are irrelevant
Nuff sed
I guess that it is 'nuff', really. Even if the OP had tested the circuit, and found the N-E fault, prior to changing the CU (which I presume is what the 'digs' are about), he would still presumably have been looking for advice in relation to the same question he has asked - how to locate the source of the known N-E fault. I can't see that any other standard test results would help you, or anyone else, to answer the question, would they?

Kind Regards, John.

Thing is, John, if he was doing this for anyone else, he'd be in a bit of bother now - because, regardless of the cost to himself, he'd have to find the fault.

Had he done the absolute minimum testing before a cu swap (i.e. Whole Board IR Test), he'd have known there was a fault 2 minutes after getting his screwdriver and tester out.

Then he could have arranged to find the fault BEFORE fitting the new board.

The other thing, of course, is that he obviously hasn't done the tests required for 'Initial Verification' of the CU - i.e. IR testing..........otherwise he wouldn't have got to the point where the lighting circuit was tripping the RCBO - it shouldn't have been energised in the first place.Hence, the 'Digs'........shoddy work to say the least.
 
Although pre-testing may have given prior warning it would still need correcting so really where one is not giving quotes and you know the job needs doing come what may then to pre-test only tells you if fault is yours or existed before so little point.

Maybe, but as I've just posted - there's NO excuse for not doing the relevant dead tests before energising!!

The good old 'bang' test........
 
Thing is, John, if he was doing this for anyone else, he'd be in a bit of bother now - because, regardless of the cost to himself, he'd have to find the fault. Had he done the absolute minimum testing before a cu swap (i.e. Whole Board IR Test), he'd have known there was a fault 2 minutes after getting his screwdriver and tester out. Then he could have arranged to find the fault BEFORE fitting the new board.
No argument about any of that. However, my point there is that horse has already bolted, so nothing much is to be gained, at least in relation to this OP, by trying to close the gate - except in a headmasterly sense!

Kind Regards, John.
 

DIYnot Local

Staff member

If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.

Select the supplier or trade you require, enter your location to begin your search.


Are you a trade or supplier? You can create your listing free at DIYnot Local

 
Sponsored Links
Back
Top