Odd Fault

[securespark]

Installed a high integrity board today. Only one lighting circuit. Wired onto the RH RCD bank.


Lighting circuit IR acceptable. RCD operation OK.

Neutral connected to the correct 'bar.

Problem: two lights trip the RCD when switched on, but only when their lamps are in.

Take the lamps out and the RCD stays in.

One is an old Thorn Arrowslim flu in the bathroom, the other is a pendant with 60W GLS in the bedroom. All other lights operate OK. Loading does not seem an issue as you can have all the other lights on the circuit on without tripping.

Any ideas please guys?

 

[Spark123]

Borrowed neutral springs to mind.

 

[spark1]

Problem: two lights trip the RCD when switched on, but only when their lamps are in.

Do both lamps have to be in for the rcd to trip?

 

[matt1e]

problem on neutral obviously
borrowed from somewhere maybe? since you have done IR tests
check that the neutral for the two faulty lights isnt somehow connected to the rcd neutral bar via some other connection

edit
damn spark123 beat me to it ,the wine is slowing me down

 

[ricicle]

Got to be borrowed or interlinked neutral somewhere as it only happens when there is load current ie lamps in....

 

[Spark123]

Or an illegal TN-C system, IR showing clear if the lamps are out/switch open.

 

[securespark]

Problem: two lights trip the RCD when switched on, but only when their lamps are in.

Do both lamps have to be in for the rcd to trip?

Interesting. Not tried that. I was thinking of a borrowed neutral, but cannot (so far anyhow) find any link.

Wondered if there could be any other weird and wonderful cause...

I shall return on the morrow for more banging my head against a brick wall.

I found one NE fault caused by a deliberate link in a socket outlet (see separate thread. What dolt would do that??

 

[RF Lighting]

What about a neutral and CPC swapped in a junction box somewhere.

I once had this. Took me forever to find. It worked fine for years, using the CPC as the return path, but obviously takes out the RCD straight away when it was installed. It IR tests ok as the lamps were out for testing.

To make things worse, mine was a TT supply with no rod or RCD and relying on the water supply pipe as earth

Have you tried an IR test with the lamps in?

 

[securespark]

Nope! Food for thought here...

 

[mfarrow]

When you say high integrity, do you mean 2 banks protected by 2 RCDs, or one bank and RCBOs on the other?

If the former, does the fault still occur/light work with the L/H RCD switched off? If it doesn't, and that neutral busbar becomes live, then you have a borrowed neutral; else a swapped earth as RF suggests.

 

[securespark]

What about a neutral and CPC swapped in a junction box somewhere.

Give the man a biscuit!

Both lights fed off a JB well hidden in the loft.

The netrals for both the lamps were indeed connected to the CPC.

Now those faults along with the deliberate NE link on a socket outlet makes me wonder about the mental process that arrived at these decisions.

mf:

The board is a 5/4/1 Hager. From the right:

main switch, RCBO, RCD bank of 4, RCD bank of 5.

 

[1john]

Are you working on the AM2 testing rig by any chance?? I had a little problem like this on a radial socket circuit once, someone had changed a single socket to a switched fused spur and sonnected a 4 way trailing lead to it, and got their neutral and CPC connected the wrong way round. Bloody DIYers!!

 

[securespark]

No, not a rig.

 

[RF Lighting]

What about a neutral and CPC swapped in a junction box somewhere.

Give the man a biscuit!

I still haven't received my biscuit yet Simon.

Any idea when I might expect it?