RCD headscratcher

[Deleted member 9648]

Finished up for Christmas now but every year last knockings, just when you think you are on the home straight a teaser throws a spanner in the works. This year said spanner came in the form of a large house with external lighting fed from a remote garage. The offending circuit fed a number of lights on the garage walls,two either side of the large gated entrance,and a number of bollards down the side of the long drive to the house.These are all controlled by a PIR on the garage,an external 2w switch on the garage,and a 2w switch in the house around a 100m distant from the garage.I noted on arrival that a new 30ma RCD main switch DB had been installed in the garage,the client said the lights often tripped the RCD but insisted they had worked ok since the DB change up until recently.It quickly became apparent that the RCD only tripped when the garage 2w switch switched the lights off,switching the lights on,by any other means including the garage switch and they worked fine,switch them off on the garge switch and the RCD tripped every time.
IR testing threw up some low readings,which were rectified fairly quickly,but I could not improve things above 0.7Mohms.Breaking the various legs out to each set of lights gave good readings,mostly 1+mohm....and the 2w cable to the house,a buried swa gave readings >200Mohm.The 0.7mohm overall reading was a result the combined resistances and I'm convinced not an issue.Existing RCD ramped fine,and other circuits in the DB all clear.
Next obvious port of call then was the garage 2W switch..IR tested the 3c to this,clear. Swapped switch for a temp plate switch....RCD still pops on switch off. Connect plate switch direct to JB in garage...RCD holds!...3c to garage 2W is barely 3m long,but partially buried in the brickwork,so this has to be the problem,as bypassing this 3m of cable cures the issue,perhaps there is some hidden damage? Rewire the 3m of cable from JB to ouside switch...reconnect...result?RCD pops every switch off *&%$&&*!!!! Reconnect switch direct into JB eliminating the rewired cable...RCD holds every time.
Ashamed to say after an entire afternoon of frustration I left the job with the issue remaining. Had a think about it that night and decided it had to be some sort of induced voltage/capacitive effect due to the long cable runs.So next day I returned,took the circuit off the DB and fed it from a dedicated 100ma S type RCD with a 6a MCB......no more problems,RCD holds every time.
Thoughts anyone?...I'm still baffled.

 

[plugsandsparks]

Would be interested in any running leakage when the cct lamps are all lit and if there is any meaningful leakage try and reduce it bit by bit by disconnecting certain parts of the cct, hopefully its only 1 fault

 

[Deleted member 9648]

Would be interested in any running leakage when the cct lamps are all lit and if there is any meaningful leakage try and reduce it bit by bit by disconnecting certain parts of the cct, hopefully its only 1 fault

But that wouldnt explain why it only tripped when the lights were switched off from the garage 2w?...It did this every single time,and never when any of the other switching devices were operated,on or off.

 

[Silly Sausage]

What's happening at this JB?
What cables are there and where are they going?

 

[plugsandsparks]

I know, but switching off can raise the volts and increase leakage, but I would have expected to at least see some at 240V when on load, that's all. Hell only a few days ago I had a motor overload tripping, but only when it switched off, did my head in until I pulled the contactor out and spotted the arcing residue across the now burned through phase barrier on the Supply side of the contactor which was upstream of the overload, go figure.

 

[Marvo]

But that wouldnt explain why it only tripped when the lights were switched off from the garage 2w?...It did this every single time,and never when any of the other switching devices were operated,on or off.

Maybe that particular switch was slow to operate with excessive arcing causing noise and spikes above the clamping voltage of the surge protection in the fittings.

 

[Deleted member 9648]

Maybe that particular switch was slow to operate with excessive arcing causing noise and spikes above the clamping voltage of the surge protection in the fittings.

But I swapped the existing switch for a plateswitch which didnt change a thing.But disconnecting the short 3c to this switch at the JB and connecting a 2w switch via short links from the JB terminals and the RCD holds!...just dont make any sense at all. Gotta admit this is the first time I've ever been completely stumped by an RCD tripping problem.

 

[telectrix]

it's a 2 way switch.could it be leakage from the strapper that has no voltage when light is on, but then gets the switched L when you turn the 2 way switch over.

 

[Deleted member 9648]

it's a 2 way switch.could it be leakage from the strapper that has no voltage when light is on, but then gets the switched L when you turn the 2 way switch over.

Could be,although regardless of which strapper is live it still tripped every time the garage switch switched the lights off. Think I'm going to just accept this is unfathomable and leave it on the S type,doesnt need 30ma protection anyway.

 

[Richard Burns]

OK this is an off the wall thought, one of the light fittings is failing slightly and causing a small peak in leakage current on switching off, however when switching off from the house the capacitance of the house cable is sufficient to damp this leakage over time and avoid the RCD trip, however the short cable for the garage does not have this capability and so the RCD trips. (I am assuming the supply line and switched line are fed to the JB and that there are separate three cores going from the JB to each switch)

 

[Engineer54]

But that wouldnt explain why it only tripped when the lights were switched off from the garage 2w?...It did this every single time,and never when any of the other switching devices were operated,on or off.

That would depend on what type of lamp fittings have been installed, if most or all are CFL or metal halide or the like, that could well produce a combined capacitive effect that could take out a 30mA RCD when being switched off, rather than when being switch On. An S type RCD wouldn't see a problem as the time factor will exceed the so-called fault duration, and is also proving, that the fault is momentary.

It could well, be that one or more of these external fittings are showing signs of age internally, in that control gear (if any) is getting tired.

So how many light fittings and of what type, are actually on this what seems to be, very long circuit?? These type of job's can end up being a very drawn out process of elimination, before you find the actual problem. It seems that it's been an ongoing problem too, so I wouldn't take too much notice of said new RCD curing the problem for a while. But again, you can try temporarily replacing that unit with a spare you may have on the van....

It may well turn out to be the overall design/configuration of the circuit itself!!

I'm in holiday switched off mode right now, so could well have missed something in your descriptions of the problems... lol!!