earth fault tripping rcd

[drenched]

Hi all. I am new to the site so please bear with me. I am not a registered electrician however my background is armed forces electrical technician and then medium voltage supply and terminations bu tnot a jointer so I am electrically savvy in most aspects. I am now retired. Got up the other morning and found the rcd protecting two circuits had tripped. All the other house circuits are protected by another rcd which was fine. One circuit was for all the house lighting and the other was for a couple of sockets in a cupboard and one of these sockets was used for the microwave. Reset the RCD which made but as soon as I reset the light mcb rcd tripped. Left the two mcb's tripped and reset the rcd which held for about 4 mins and then tripped. As the rcd tripped with mcb's off I assumed a dicky rcd or a neutral to earth fault. Rcd would not reset straight away with mcb,s tripped but did eventually after a period of time. Remade the sockets mcb and rcd held but as soon as I tried to make the lights mcb the rcd tripped immediately. So deduced a neutral - earth fault on the light circuit. No work has been carried out to suggest a possible area of fault however I have had some problems with one of the led lights in the bathroom. I decide therefore to try and chase down the problem by going to the neutral connections for the lights which were in the loft and disconnected the neutral leads for the lights. RCD still tripping so decided to remove the neutrals for the upstairs lights and the bathroom fan extractor. Still tripping so decided the go into the CU and disconnected the neutral for the light circuit from the neutral bar and fully expected the rcd to hold. No such luck so disconnected the earth from the earth bar and as expected the rcd held even with the other circuit mcb made. Connected the neutral and left the earth off and tried every appliance on the circuit and nothing tripped the rcd however one of the security lights didn't work so i thought the fault was internal to this so disconnected the supply to this, reconnected the earth on the cu and tried the rcd again but still tripping. I am now a bit stumped as to where to look and why the fault is apparently just on the earth. Any suggestions would be welcome.

 

[JohnW2]

... As the rcd tripped with mcb's off I assumed a dicky rcd or a neutral to earth fault.

That would seem to be a logical conclusion. The rest of what you go on to say is, as you imply, all a bit confusing ...

.... Still tripping so decided the go into the CU and disconnected the neutral for the light circuit from the neutral bar and fully expected the rcd to hold. No such luck so disconnected the earth from the earth bar and as expected the rcd held even with the other circuit mcb made. ... Connected the neutral and left the earth off and tried every appliance on the circuit and nothing tripped the rcd

That does seem weird. It seems to suggest that there is, indeed, a fault to that circuit's earth, but a fault between that earth to the neutral (or possibly L) of some other circuit - which doesn't really make a lot of sense (at least, not to me).

Maybe others have some better ideas that I do at present!

Kind Regards, John

 

[drenched]

Hi John

Thought about a line to earth fault exported from other ccts but the rcd for other circuits held and that also discounts a borrowed neutral from the other ccts ---well I think it does.

best regards

 

[opps]

Sorry. I am at pint number 5... I recently had to sort out a customer's RCD in his summer house tripping. I eventually traced it back to one on the many live/switched lives that run back to his extension.

Earlier this week I popped in to my local pub, the MCB was tripping (not RCD protected), I traced the fault back to a dodgy fan. I removed the live and the MCB reset.

Sorry, I tried to read your post but the lack of paragraphs meant that I didn't read the whole post- my bad, not yours, but if you want people to follow your post, I respectfully suggest that you make your post easier to follow/read.

Regards

 

[JohnW2]

.... Thought about a line to earth fault exported from other ccts but the rcd for other circuits held and that also discounts a borrowed neutral from the other ccts ---well I think it does. ...

I really should think a bit more before writing, but ....

... if there were a N-E fault on the circuit in question, then there would be no trip if there were no loads (on any circuit). However, if there were a load on some other circuit (protected by the other RCD), some neutral current from that load could flow from the supply side of the 'otherRCD' (hence no L-N imbalance, hence tripping, in that 'other RCD'), and thence backwards through the 'affected' RCD, finally going through the fault to earth. That 'affected' RCD could therefore trip because of the L-N imbalance due to that N current (from the other circuit) going 'backwards' through it.

I now need to do that thinking, to decide whether what I've just written makes any sense!

Kind Regards, John

 

[drenched]

Hi Opps

You are quite right to highlight my poor grammar and I will endeavour to remedy that in future posts.

John

Before working in the cu I turned off all circuits and then the main switch. Did the work on the neutral and earth of the affected side and then only energised those ccts with the tripping rcd so there was no load on the other rcd.

best regards

 

[JohnW2]

John .... Before working in the cu I turned off all circuits and then the main switch. Did the work on the neutral and earth of the affected side and then only energised those ccts with the tripping rcd so there was no load on the other rcd.

I think that I must have missed, or misunderstood, something, since this doesn't seem to add up.

Are you saying (as I thought you were) that with all the MCBs (those protected by both RCDs) switched off, one of the RCDs still trips even if the lighting circuit's neutral is disconnected, but not if its earth is disconnected? If so, then, since all L's are disconnected by MCBs, there can't be any 'conventional' current flowing anywhere within your installation, so no opportunity for any L-N imbalance within an RCD that would cause it to trip.

Even if you had a TN-C-S installation whose neutral was at a potential appreciably greater than true earth potential, the 'earth' within the installation would be at that same potential, hence no chance of current flowing between the two.

If I've understood correctly (as above) then it's hard (for me!) to see what's left other than an iffy RCD - but, as above, rather I suspect that I have misunderstood!

Kind Regards, John

 

[ericmark]

I can see @JohnW2 point, with a TN-C-S supply earth and neutral are the same voltage, so with no load, even if direct connection earth to neutral, no current will flow, so RCD will not trip.

However we don't at this point know if the supply is a TN-C-S, with a TT supply then very different.

I have in a hurry used a multi-meter to try and find faults, and some times it has worked, but a meter like this

costs £32 yes not as good as a mega, as no low ohm scale, but it should find the fault, I bought one when I mislaid my mega, and when I found mega the results matched.

You know you have an earth fault, and it has not been found with trial and error hunting, so now either you buy a meter, or you employ some one who has one.

 

[drenched]

Morning John/Eric

Thanks Eric I will get one of those.
System is TN-C-S with about 3-4 ohms between earth and neutral at the cu so just a question of getting for the meter.

 

[JohnW2]

I can see @JohnW2 point, with a TN-C-S supply earth and neutral are the same voltage, so with no load, even if direct connection earth to neutral, no current will flow, so RCD will not trip.

Quite so.

However we don't at this point know if the supply is a TN-C-S, with a TT supply then very different.

Well, we now do know that it's TN-C-S, but I don't think that I could think of a credible explanation (other than an iffy RCD) for what the OP is experiencing even if it were TT - could you??

Kind Regards, John

 

[JohnW2]

System is TN-C-S with about 3-4 ohms between earth and neutral at the cu so just a question of getting for the meter.

If that measurement is correct, it's far too high for the N-E resistance at the CU, given that they will be joined together, via pretty hefty cables, a metre or three from the CU.

Kind Regards, John

 

[EFLImpudence]

System is TN-C-S with about 3-4 ohms between earth and neutral at the cu so just a question of getting for the meter.

Are you sure it is not 0.3? Even that would be on the high side.

Ah, too late. That'll teach me not to do something else before posting.

 

[drenched]

Hi all

Before the meter came I did another check with multimeter in the loft and was getting approx 1.3. I split the circuit in the loft into upstairs downstairs and isolated fault to upstairs which had nearly all wiring for in the loft.
Went round all connections and wiring and lo and behold the fault cleared. So not very positive as it could be in a hager downlight terminal box which was installed when we had the bathroom re designed and the sparky who did it reconfigured everything in the loft and the box is jammed out with cables. Leaving it for now but at least I have an idea where to look next time ( hopefully never ) and I have got myself an insulation tester.

Thanks for all your assistance