Negative to earth fault on lighting circuit at consumer unit

[Thomas Lawson]

The RCD on the master switch where the supply enters the house tripped, completely out of the blue, with no work or alteration’s having taken place.

I have traced the fault to the ground floor lighting circuit, specifically the feed from the consumer unit to the lights. My multimeter registers continuity across negative and earth on the wires exiting the consumer unit at a junction box immediately after the consumer unit, but only when there is a supply from the master switch. If the master switch is off, there is no continuity. Sometimes the continuity pulses on and off, when there is power to the consumer unit but other times it is continuous. I also detect c7.5v across positive and negative on the same cable at the junction box when the supply is on.

The fault is not consistent; having had the lighting MCB off overnight I switched it on this morning and nothing tripped for at least one hour, and this has been the same a few times now. Previously, the master RCD would trip as soon as the lighting MCB was switched on, and that is what happens now.

The circuit has been fine since first wired more than 15 years ago, so I am puzzled as to what is going on. Can anyone suggest what may be causing the fault, please?

 

[terryplumb]

There is no positive and negative in AC.
There will often be continuity between neutral and earth when main switch is on at CU.
Do you have any external lights on the circuit in question ?

 

[Thomas Lawson]

Thanks for your reply,

No external lights, just one cluster of 3 led spot lights off a one way switch , one led spot off a one way, and one cluster of three led spots off three switches, all indoors in the same room

 

[terryplumb]

Are they the only light fittings on the circuit ,none in any other room ?
No bathroom or kitchen extractor fans ,burglar alarms , smoke alarms ,anything else on the circuit at all ?

 

[EFLImpudence]

I have traced the fault to the ground floor lighting circuit, specifically the feed from the consumer unit to the lights. My multimeter registers continuity across negative and earth on the wires exiting the consumer unit at a junction box immediately after the consumer unit, but only when there is a supply from the master switch. If the master switch is off, there is no continuity.

That is how it should be.

 

[Harry Bloomfield]

You need to disconnect both L and N, at the consumer unit end, then retest, for your N to E fault. You may find it disappears.

 

[Thomas Lawson]

Yes, no other lights on that circuit.

Are they the only light fittings on the circuit ,none in any other room ?
No bathroom or kitchen extractor fans ,burglar alarms , smoke alarms ,anything else on the circuit at all ?

That is how it should be.

What about the pulsing continuity

 

[Thomas Lawson]

You need to disconnect both L and N, at the consumer unit end, then retest, for your N to E fault. You may find it disappears.

I think I was sent of track by finding continuity between negative and earth at the junction box, which replies have told me is expected and normal. I have a hunch that the fault that is tripping the RCD is between the junction box and one of the single lights, as with this disconnected the others didn’t trip the RCD,but maybe that’s also a red herring, as the fault is not always immediate. My next step could be to isolate this light at the junction box and see what happens.

 

[terryplumb]

Are all of the downstairs lights connected into this junction box ? That would be odd .

 

[Thomas Lawson]

Are all of the downstairs lights connected into this junction box ? That would be odd .

Yes they are

 

[terryplumb]

Show us a picture of the wiring in the J/box please.

 

[Thomas Lawson]

Show us a picture of the wiring in the J/box please.

 

[Murdochcat]

Are all of the downstairs lights connected into this junction box ? That would be odd .

Why would that be odd? Very common in 1960s homes

 

[terryplumb]

Why would that be odd? Very common in 1960s homes

Odd that such a set up still exists 6 decades later.

 

[Harry Bloomfield]

If you are convinced there exists a fault on that circuit, that JB - best thing to do, is disconnect all of those wires and test each twin and earth separately, to determine which of the four exhibits the fault. No need for any notes as to what goes where - all the reds together, all the blacks, all the earths.

The usual warning applies - of turning the power off, before touching it.