I have just spent 3.5 days chasing down a prevailing fault on a lighting circuit.
The 1st one I found was a low IR between L & E which I traced to a woodscrew through a feed cable traveling through a joist at a 40' angle to emerge into the light fitting. The hanging bracket from which the fitting was suspended had been removed by the plasterer before a skim being applied, the plasterer must have replaced the hanging bracket but used longer screws. Anyway that was fault no 1 sorted - a quick reconfiguration where the loop in/out and switch drop/return were consolidated into a wago box with only a switchwire and neutral going out to the light fitting. This one was odd because there were 2 lives and 2 neutrals leaving the rcbo, both of which fed the downstairs lighting exclusively. The place had been altered and a 2 storey extension added, although I never did find where the old original wiring and the new extension wiring married up. My guess is under the floor somewhere. The other red herring was that they had had a rodent infestation which got the alarm bells ringing.
The next fault was confined to 8 x downlights in the kitchen area. Still part of the same circuit but which tripped the rcbo whenever they were switched on. I took all 8 GU10 downlight fitting down and visually checked their respective connections - all good. I then traced the switchwire and neutral leaving the light switch and IR'd them to each other and to earth - again, all good. The kitchen lights can be switched from 3 different locations via 2 x 2 way and 1 x intermediate switch. I began to think that there was a fault between line/earth in the switch lines when the switches were in a certain configuration but no, nothing.
At this point I was beginning to tear what was left of my hair out and was wondering how to justify the thick end of 3 days fault finding on the invoice.
One other thing I noted was that I could refit 1 x LED lamp into the 1st fitting in the chain from the switch and it would work ok, but when I fitted a second one the trip went. I got to thinking that maybe there wasn't enough current draw to create enough fault current to trip the OPD with only one lamp, but when I added a second 5w LED load via fitting a 2nd lamp, the thing drew enough current for it to fault to earth or neutral somewhere. Having found nothing in terms of a 'concrete fault' I decided to change the LED lamps for new items. Low and behold, everything worked ok and nothing tripped. I added the new lamps one by one until all 8 were back in their lampholders and worked as expected when the switch was operated.
So, having never come across this before, and still being at something of a loss, is it possible for LED lamps to fail in a mode whereby it trips the opd? The lamps in question were Cromptom 5W LED GU10s and I have never had any problems with them before. Indeed, these are the only LED GU 10s I am happy to fit for my customers because they seem to be the most reliable.
Looking back, my first action should have been to replace all of the lamps. This would not have solved the woodscrew through the cable in the lounge light fitting, but it would have prevented what I believe to have been a wild goose chase and a few red herrings.
