RCD Randomly Tripping

[JohnW2]

At the moment, as you have wired the lights in a ring, disconnect one side and see what happens.

If the ring is currently unbroken, then doing that would obviously not make any difference. However, as both the OP and myself have said, it affords the opportunity to split the ring and then see what happens with each of the sides, separately.

Kind Regards, John

 

[EFLImpudence]

I'm having trouble following you this last couple of days - could be me.

I thought we were trying to find a fault causing the RCD to trip.

 

[JohnW2]

I thought we were trying to find a fault causing the RCD to trip.

We are - and the OP thinks he has narrowed down the problem to one (ring) lighting circuit.

However, as I said, if the OP were to do as you say and disconnect one end of that ring at the CU, that would make absolutely no difference if the ring were intact (which he could confirm with his multimeter) - since any L-E or N-E fault would get back to the CU (hence RCD) down one leg of the ring even if the other end was not connected to anything.

Of course, if you are postulating a broken ring and a fault to earth, that would be a bit different, but I would personally be inclined to look for the 'far more probable' (i.e. a single fault) first!

Kind Regards, John

 

[EFLImpudence]

I didn't say disconnect one end of that ring at the CU.

I said disconnect one side and see what happens.

In response to

My only other thought was to un-wire a ceiling rose completely to split the ring in half and only power one side and then the other if the first trips to narrow down the fault?

 

[AndrewCollis]

Thanks EFLImpudence. I have managed to find a Fluke Made robin 1620 Multi-Function Tester Plus 3010 & 4116 Testers all for £150 which does all the tests you said. It should be with me Saturday so I will have a new toy to play with

Thanks John, I will split it in the morning and see how it goes

All the best
Andrew

 

[EFLImpudence]

Good price.

 

[JohnW2]

I didn't say disconnect one end of that ring at the CU. I said disconnect one side and see what happens. In response to

Fair enough - I may have (in the absence of quoted context) slightly misunderstood what you meant by "disconnect one side". However, unless you also (as Andrew suggested) disconnected one end of the ring at the CU (which, as you stress, you didn't say), it would achieve nothing (since both 'sides' would, in the absence of a break in the ring, be connected at the CU).

Kind Regards, John

 

[EFLImpudence]

Well. It wouldn't be disconnected if he didn't, would it?

 

[JohnW2]

Well. It wouldn't be disconnected if he didn't, would it?

I'm also having difficulties following you My point is that, in order to achieve what is required, and as the OP knows, he has to disconnect two things - one 'side' of the ring somewhere along it's length (e.g. at a rose), to split the ring AND one end of the ring at the CU (so that only one 'half' of the ring is 'in circuit'). That's all I was saying

Kind Regards, John

 

[EFLImpudence]

I know you were saying that but you were mistakenly saying I was not saying that.

How can you disconnect one side of a ring without disconnecting two things?

completely to split the ring in half and only power one side

disconnect one side and see what happens.

So, what justification is there for then saying:

if the OP were to do as you say and disconnect one end of that ring at the CU, that would make absolutely no difference

 

[JohnW2]

I know you were saying that but you were mistakenly saying I was not saying that.

I could explain why I was confused by what you wrote, but it's only a question of semantics (and what was, and was not, implied), so, given that we both understand each other, I don't think anything is to be gained by perpetuating the discussion (I'm sure that neither of us want to emulate BAS!)!

The interesting thing will be to hear what the OP reports of what happens when he uses one 'half' of the lighting ring at a time, the other 'half' being totally disconnected (at both ends).

Kind Regards, John

 

[AndrewCollis]

Good Morning Everyone

As discussed, this morning I disconnected half the upstairs lighting ring. Switched on the mcb and checked to see which lights were working. The bathroom and landing light both came on which just left 2 bedrooms which did not. This did trigger a thought which is an extractor fan in the bathroom wired to the light but I thought I will just leave it and see what happens. Within 30 minutes the RCD tripped. I opened up the extractor fan and removed all 3 wires (live, switched live and neural) It is a new extractor so less likely to be a fault with the unit I would have thought. I switched the mcb on again to check it wasnt the wiring to the extractor. Again within 30 minutes the RCD tripped. I went to the cu and removed that half ring and put the other half in. So far, nothing has tripped.

Confession time. I cant get the extractor out of my head. I fitted it and wired it. I will try and put a picture of the rose wiring below..... one minute.... ok, that took longer than a minute and the picture is huge.

I wired the rose in the normal way then for the fan I I used some 3 core & earth. I wired brown to loop, blue to neutral and grey to switched live. The fan instructions said its double insulated so doesnt need an earth so I put the earth in the rose as usual and cut the earth off at the fan end. I am almost certain this will be the problem as the rose is in the ceiling with no access from above and I am was about to plaster the ceiling....

What do you think?

All the best
Andrew

 

[EFLImpudence]

I wired the rose in the normal way then for the fan I I used some 3 core & earth. I wired brown to loop, blue to neutral and grey to switched live.

That sounds alright.

The fan instructions said its double insulated so doesnt need an earth so I put the earth in the rose as usual and cut the earth off at the fan end.

Naughty. Never cut unused cores; they may be needed one day.

I am almost certain this will be the problem as the rose is in the ceiling with no access from above and I am was about to plaster the ceiling....

Not sure what you mean - but:
If you are satisfied this is where the fault is:
Disconnect at the fan terminals, reconnect the light ring,
if it trips disconnect the fan cable at the rose and replace it,
if it doesn't trip replace the fan.

 

[JohnW2]

....I opened up the extractor fan and removed all 3 wires (live, switched live and neural) .... I switched the mcb on again .... within 30 minutes the RCD tripped. I went to the cu and removed that half ring and put the other half in. So far, nothing has tripped. .... Confession time. I cant get the extractor out of my head. ...

If I understand the above correctly, you appear to have eliminated the extractor itself as the source of the problem, since the RCD still trips when the extractor is totally disconnected. However, your finger still seems to be pointing at the half of the lighting ring which feeds the extractor.

Is there anything else connected to this 'half ring'. For example, are there any dimmer switches or suchlike involved?

Kind Regards, John

 

[AndrewCollis]

What I meant EFLImpudence is its just my luck for the fault to be in the bathroom as its the only wiring I cant access without taking the ceiling down.

I did eliminate the extractor John by removing the wires completely from the extractor and the RCD still tripped. The image of the rose was one I took from before I put the ceiling up. I did this as a reference in case I needed it later on. The rose is now in the ceiling.

The only things on the disconnected half of the circuit are the bathroom light and extractor with a bathroom pull switch and a landing light with a twin switch. The twin switch will turn the upstairs and downstairs landing lights on and both can also be turned on and off from downstairs.

All the best
Andrew