Switching OFF bathroom lights trips the RCD

[KevDIY]

Switching OFF my bathroom lights sometimes trips the consumer unit's RCD, but it never trips when switching them on or during them being on and it never trips the circuits MCB. Its only an occasional thing, happens perhaps every few weeks or so, but I've finally had enough and decided to try sorting it.

They're five 12v MR16 bathroom spotlights each with its own transformer above the ceiling, running 35w halogen bulbs. The switch is the common cheap pull-cord type of thing. It can happen any-time (including the middle of the night!), not just when theres steam etc. about.

Since all is well when they're actually on, I'm a bit stumped as to what to look for... Does anyone have any experience of this?

Cheers
Kev

 

[delmel]

odd only when you turn off ?.

I would have expected any problem to be when turning on. More common are the LED ones now.

What transformer have you got ?

 

[KevDIY]

Yes, it only happens as they turn off - its a new one on me

The transformers were what came with the lights, supposedly good for 20w to 50w bulbs although they seem to flicker with just 20w bulbs in so I went with 35w. Unfortunately I forget the brand but I could go up and have a look tonight if its important. IIRC they are double insulated (so no earth connection) and obviously the plastic pull-cord switch is too, so it seems even stranger to me that the RCD is somehow detecting leakage to earth as the current is turned off.

Yeah, they pre-date the LED type lights coming in unfortunately, but I've decided not to update for the moment; I'm currently broke, and these lights just aren't used enough to save much £energy.

Cheers
kev

 

[Taylortwocities]

I suggest you have a good look at the wiring to the switch, and also above it (if you have access) just in case there's a rogue wire or connection that moves when the switch is operated.

Change the switch too.

If that all works, assuming you do not have an insulation tester, try eliminationg one light at a time by diconnecvting the LIVE AND NEUTRAL before the transformer. That may help with narrowing it down.


With RCDs, dont forget that the RCD will trip if there is an earth leakage on the line OR the neutral. Light switches only break the line.

 

[KevDIY]

Ah, I'd not thought about the switch vibrating, many thanks. It would be good if it is that; cheap and easy to fix. As would be replacing the switch.

Yeah, if that fails I will try eliminating the lights one by one. It could be tedious due to the intermittent nature of the fault, and some of them are hard to reach, but short of replacing £everything I can't think of a better way.

Thanks
kev

 

[deadshort]

Hi, i have had that issue once with bathroom lights checked the wiring not the problem, a over sensitive Crabtree RCCB was the cause, switching off my Neff oven had the same effect! Took weeks to eliminate the problem

Kind regards,

DS

 

[JohnW2]

Hi, i have had that issue once with bathroom lights checked the wiring not the problem, a over sensitive Crabtree RCCB was the cause, switching off my Neff oven had the same effect! Took weeks to eliminate the problem

Yes, I've seen something similar, complicated by the fact that the 'over sensitivity'/'trigger-happiness' of the RCD was itself very intermittent, so wasn't apparent on testing the RCD. I do wonder if people sometimes spend far too much time searching an installation for causes of unexplained sporadic RCD trips before spending the few minutes it usually takes to replace the RCD, even if only temporarily, as a 'diagnostic' measure.

Kind Regards, John

 

[KevDIY]

Ah another possibility, thanks! The consumer unit is split with two RCDs, so (ignoring part P for the moment) I could potentially swap them over and see if the sensitivity moves with the RCD or not.

At the moment I'm suspecting the lights, as nothing else seems to trip the RCD - including the garage which has fairly haphazard wiring and runs stuff like a big welder and saws etc. But actually, nothing else on the circuit has these electronic transformers, so perhaps they can exploit weaknesses.

TBH though, whilst I can envisage a magnetic field collapsing in a transformer when switched off and dumping some back EMF unhelpfully, I'm still a bit stumped as to how it would find its way to earth in a double insulated setup.

Cheers
Kev

 

[JohnW2]

Ah another possibility, thanks! The consumer unit is split with two RCDs, so (ignoring part P for the moment) I could potentially swap them over and see if the sensitivity moves with the RCD or not.

Hypothetically speaking, yes, that could be done.

At the moment I'm suspecting the lights, as nothing else seems to trip the RCD - including the garage which has fairly haphazard wiring and runs stuff like a big welder and saws etc. But actually, nothing else on the circuit has these electronic transformers, so perhaps they can exploit weaknesses. ... TBH though, whilst I can envisage a magnetic field collapsing in a transformer when switched off and dumping some back EMF unhelpfully, I'm still a bit stumped as to how it would find its way to earth in a double insulated setup.

These situations in which RCDs trip when things are switched off usually defy most attempts at rational analysis - or, at least, stimulate lengthy debates.

Kind Regards, John

 

[Taylortwocities]

TBH though, whilst I can envisage a magnetic field collapsing in a transformer when switched off and dumping some back EMF unhelpfully, I'm still a bit stumped as to how it would find its way to earth in a double insulated setup.

That may be a decent enough theory if you had traditional transformers that had nice meaty iron core, but your average 12v halogen "transformer" is really a bunch of electronic components. Really a switched mode power supply.

There are some around here that will assert that these do not meet the dictionary definition of "transformer". But they do have the "T" word written on the device!!

Hides behind settee

 

[EFLImpudence]

Are they nominal transformers?

 

[Taylortwocities]

Are they nominal transformers?

Off-nominal, surely?

 

[JohnW2]

Are they nominal transformers?

They could well be

Kind Regards, John

 

[radioredcat]

I had a problem with our lights the other week I switched a light off and heard a click from the CU and the lights went out.
The MCB had tripped taking the light switch off and the Live wire was blown apart this was caused by the single back box screw which was too big for the hole making it protrude by a few mm and the edge had cut the insulation.

In our case our lights are on the RCD side of the board so if was then the RCD would have tripped so I chamfered the hole so the screw was flush with the back box.

Andy

 

[KevDIY]

Heh, yes okay 'transformer' may not be accurate, its just whats written on the box. I believe switch modes have a small coil, but no idea if these halogen jobbies are the same. I was just searching for a possible 'how', since it seems odd to me that the RCD goes only with these lights, which are all double insulated, and only when they're being switched off.

I guess that I don't really need to know though, as long as I can find/replace the culprit.

Andy, thats not quite the same effect as mine but its a good reminder that it could be something very simple, if a bit unusual. I'll venture into the loft at the weekend and have a look around at the wiring.

Cheers
kev