RCD fault finding pointers

studentspark · 19 Aug 2014

Small flat only 4 circuits

Installed dual RCD 6 way board (3-3)

Lighting on it own RCD but will not set

Split lighting circuit and got it to hold, so I thought I has isolated faulty section....

But then when a socket circuit is used (Separate RCD) lighting RCD trips

I know I have no crossed neutrals at the CU

IR's are OK pre install

Neutrals on the right side

Any ideas appreciated

JohnW2 · 19 Aug 2014

studentspark said:
But then when a socket circuit is used (Separate RCD) lighting RCD trips ... I know I have no crossed neutrals at the CU ... Neutrals on the right side

If loading a circuit on one RCD causes the other RCD to operate, then (despite what you say/imply) I can't see any possible explanations other than 'cross-connections' (presumably of neutrals) either within the circuits or due to the wrong neutral bar being used for something in the CU.

Have you tried disconnecting both the lighting and sockets circuits from the CU and checking for continuity between the neutrals of the two circuits?

Kind Regards, John

studentspark · 19 Aug 2014

Thanks

Yes, that is one of my first jobs in the morning.

ericmark · 20 Aug 2014

My first thought is neutral - earth leakage. As load increases the voltage difference between neutral and earth also increases so will no load RCD will not trip and with load on any circuit even if not supplied with that RCD it will trip.

bernardgreen · 20 Aug 2014

Neutral_Earth_trip

bernardgreen
16 Jan 2011
48

JohnW2 · 20 Aug 2014

ericmark said:
My first thought is neutral - earth leakage. As load increases the voltage difference between neutral and earth also increases so will no load RCD will not trip and with load on any circuit even if not supplied with that RCD it will trip.

Maybe I'm missing something but, as I implied in my response above, I can't see how that mechanism can result in tripping of an RCD which is not protecting the circuit with the N-E fault - unless there is some connection between the neutrals of the two sets of circuits.

Kind Regards, John

JohnW2 · 20 Aug 2014

bernardgreen said:
<diagram showing how an N-E fault can result in tripping of an RCD protecting the circuit with a fault even if the SP MCB of the circuit is open>

As I've just written to eric, I don't see how that mechanism can result of tripping of an RCD which does not protect the circuit with the fault - unless there is some link between the neutrals of the two circuits. Am I missing something?

Kind Regards, John

Spark123 · 20 Aug 2014

I'd start with checking the N-E IR on the lighting circuit, I'm assuming it is a TN-CS supply? If it wasn't TN-CS then the lighting circuit RCD would never hold up.
A N-E fault is effectively a parallel path to the neutral measuring coil of the RCD hence when a load is applied on the system that results in 30mA ish missing the RCD it trips out, this current being from the potential difference between N & E caused by the impedance of the meter tails or upstream cabling.

JohnW2 · 20 Aug 2014

Spark123 said:
A N-E fault is effectively a parallel path to the neutral measuring coil of the RCD hence when a load is applied on the system that results in 30mA ish missing the RCD it trips out, this current being from the potential difference between N & E caused by the impedance of the meter tails or upstream cabling.

I must be missing something! As I keep saying, I really don't understand how an N-E fault on a final circuit, downstream of the RCD protecting that circuit, can cause a different RCD to trip - unless there are some unintended connections between neutrals in circuits protected by the different RCDs.

Kind REgards, John

Spark123 · 20 Aug 2014

If I am understanding correctly the one that trips is the lighting RCD, it trips when something is used on this circuit but also, when nothing is switched on on the lighting circuit this lighting RCD trips when something is being used on another circuit.

They aren't downstream as such, it is on the same supply.
Any pd between n & e causes a current to flow through the neutral leg of the faulty circuit, through the n-e fault and back down the CPCs etc to the suppliers cutout (where n&e are combined in a TNCS)
As it is only in the neutral leg the RCD on this circuit is imbalanced hence trips.

The n-e potential difference at the board is caused by current flowing in and the impedance of the upstream cables to the cutout. This current can be from any circuit in the board.

JohnW2 · 20 Aug 2014

Spark123 said:
If I am understanding correctly the one that trips is the lighting RCD, it trips when something is used on this circuit but also, when nothing is switched on on the lighting circuit this lighting RCD trips when something is being used on another circuit.

It's not totally clear (at least, not to me). The OP wrote:

studentspark said:
Lighting on it own RCD but will not set ... Split lighting circuit and got it to hold, so I thought I has isolated faulty section.... But then when a socket circuit is used (Separate RCD) lighting RCD trips

I'm not sure what he means by "split lighting circuit", but this seemed to indicate that he had arrived at a situation in which the lighting RCD was not tripping (Until sockets circuit was loaded). Maybe I'm wrong. Whatever ....

Spark123 said:
They aren't downstream as such, it is on the same supply. Any pd between n & e causes a current to flow through the neutral leg of the faulty circuit, through the n-e fault and back down the CPCs etc to the suppliers cutout (where n&e are combined in a TNCS). As it is only in the neutral leg the RCD on this circuit is imbalanced hence trips.

Fair enough.

Spark123 said:
The n-e potential difference at the board is caused by current flowing in and the impedance of the upstream cables to the cutout. This current can be from any circuit in the board.

I see what you're saying, but, given that you're talking about TN-C-S, the N-E pd at the board is surely going to be minute, unless the board is very heavily loaded? ... I suppose you must be suggesting is that if the impedance of the path of the N-E fault is very low, then even that minute pd will be enough to result in a 30mA (or whatever) current to flow through it - is that correct?

Kind Regards, John

Spark123 · 21 Aug 2014

Yep, that is what I am saying. A minute pd is all it takes.
For example take a lighting circuit with a n-e short up the line giving a n-e resistance of 1 ohm (cable resistance only). V=IR therefore you only need 0.03v pd to drive 30mA through it.

JohnW2 · 21 Aug 2014

Spark123 said:
Yep, that is what I am saying. A minute pd is all it takes. ... For example take a lighting circuit with a n-e short up the line giving a n-e resistance of 1 ohm (cable resistance only). V=IR therefore you only need 0.03v pd to drive 30mA through it.

True, but as I said, there would have to be a substantial load on the board to achieve even that. By my reckoning, with 1m tails, you'd need around 21A with 16mm² tails or around 34A with 25mm² ones. There might, of course, be other loads on other circuits (which would then 'tot up') but you would not get those sort of currents by ('alone') applying a single load to a sockets circuit, which is presumably what the OP was doing.

However, as I said before, I do now understand (and agree with) the concept of what you are suggesting. It would presumably be easy for the OP to determine whether this was the explanation for what he was experiencing, since applying a substantial load to any any circuit (immersion? cooker?) ought to trip the RCD protecting the lighting circuit if your hypothesis is correct?

Kind Regards, John

studentspark · 22 Aug 2014

Yes I found a few faults
A neutral /earth short in a kitchen cabinet supplying under cabinet lights.
Earthed screw pinching a neutral
IR readings did pick up a N/E reading but it was about 1.9 Mohms ??

Yes it was a TN-CS,

Found this as well !!!

View media item 81513

JohnW2 · 22 Aug 2014

studentspark said:
Yes I found a few faults ... A neutral /earth short in a kitchen cabinet supplying under cabinet lights. Earthed screw pinching a neutral ... IR readings did pick up a N/E reading but it was about 1.9 Mohms ??

Interesting. 1.9 MΩ obviously could not, in itself, result in an RCD trip

studentspark said:
Found this as well !!!View media item 81513

Ooh - my goodness! That, alone, would obviously result in the RCD tripping as soon as that light was switched on.

Kind Regards, John

diynot

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

RCD fault finding pointers

Similar threads