outside path lighting on RCBO

Bit of a long story, but I am having constant issues with a set of path lights. I installed these 2 years ago, and they were fine for about a year and a half then started constantly started tripping RCBO. The lights are mains fed in SWA to buried boxes by each light which have been gel-filled. The lights are LED with drivers in each unit. 4 out of 7 drivers have failed already and been replaced (these were not cheap units but seem to have a driver batch issue). Now I've had all the boxes up, dug out gel and re-filled them - I have found the drivers, if duff, give poor insulation test results, so the only way to fault find is disconnect everything - right PITA. With everything disconnected IR for CCT is good. Needless to say customer is peed off. I thought I had resolved the issues a few weeks ago, he turns up (its a holiday home), trys the lights about a week later, and they trip again....FFs. Now I have 2 points I would like opinions on:-

1/ I suspect voltage spikes may be destroying the drivers, but I can't think of a way of proving / disproving that without monitoring gear which I don't have.

2/ I am considering changing RCBO for MCB, at least on a temporary basis. Given that cable is SWA do I need RCBO protection at all? Earthing system is TN-S. No part of cable run prior to exiting the building falls into requirement for cables hidden less than 50mm in walls, and I've seen plenty of other outdoor lighting on MCBs. The RCBO is more for the gardeners, just in case.....but its SWA cable which is buried along a path edge where they shouldn't be digging and reasonably protected / deep.

RCBO trips at about 23mA, bit low, but it shares no other ccts. I also suspect damp in the lights may be an issue, ie lack of use is not driving damp off albeit everything is IP rated supposedly. 

23ma trip current shouldn't be an issue. I know of numerous brand new out of the box RCD's and RCBO's that test at anywhere from 20ma to 25ma. Excluding the internal wiring aspect, I cannot think of any requirement for RCD protection to buried SWA or light fittings. That is of course assuming your Zs values are all good to meet disconnections times if you did happen to short out Live-Earth. Did the manufacturers instructions say that RCD protection is required?

Doc H. 

Nope manufacturers instructions where 'fit lights' pretty much! LOL. Zs values are fine.

I was trying to think of any requirement for RCD protection of external lighting, couldn't find anything conclusive in BS7671, and RCDing everything seems to have become the norm, so I wanted to double check. 

If it's the rcbo tripping how do you know it's earth leakage rather than a live to earth, or live to neutral fault?

I would temporarily swap the rcbo for a mcb to see what happens...

Binky I sometimes wonder about that sealer-gel stuff  .  We inherited a garden light job  done in SWA   ,  buried , stops at each light with a Galvo tee box lid , gasket and filled with the thick green sealing stuff ,  like pyro compound .    Flex from tee to the fitting . 

They were constantly tripping ,  we found the sealing compound was saturated with condensation within the boxes  so we cleaned them out , dried the boxes , taped up the connectors , fitted new gaskets  but left out the sealing compound .

As far as I know  they've been OK since.    

I need to go back and re-test is the answer to that @Murdoch. I am planning to replace RCBO with MCB. I think I have another failed driver, but its hard to tell when none are working. Last time I was there I had the lights on for over an hour and all was well!

id swap for an MCB, cant see any reason it would require RCD protection

binky said:

I need to go back and re-test is the answer to that @Murdoch. I am planning to replace RCBO with MCB. I think I have another failed driver, but its hard to tell when none are working. Last time I was there I had the lights on for over an hour and all was well!

Click to expand...

But short of taking all the drivers out you won't really find out much - you really need to do a proper 3 core IR test. and L&N to E isn't really going to help IMHO

Evans Electric said:

Binky I sometimes wonder about that sealer-gel stuff  .  We inherited a garden light job  done in SWA   ,  buried , stops at each light with a Galvo tee box lid , gasket and filled with the thick green sealing stuff ,  like pyro compound .    Flex from tee to the fitting . 

They were constantly tripping ,  we found the sealing compound was saturated with condensation within the boxes  so we cleaned them out , dried the boxes , taped up the connectors , fitted new gaskets  but left out the sealing compound .

As far as I know  they've been OK since.    

Click to expand...

that's a bit interesting, the product I used is the 2 part 'Magic Gel', judging by the stuff that gets all over your fingers it's a silicone based product, so should be waterproof? Maybe I need to check the spec.

spec says fully immersible = IP68 or better

Murdoch said:

But short of taking all the drivers out you won't really find out much - you really need to do a proper 3 core IR test. and L&N to E isn't really going to help IMHO

Click to expand...

I did that last time I was there, right PITA digging out connections from potting compound. I checked LN-e before leaving with wet compound, having refilled boxes, and got over 10meg Ohm.

I hope lots of DIYers read this thread - I'm at the point where I'll refuse to bury any connections unless they pay the the absolutely best kit...

I wish I hadn't followed the trend for 'gel' - it makes fault finding such a pain, but there was no option for above ground connections in this location. Shouldn't have let the customer pick the lights either  :slap

In my short  experience  with LED lighting , majority of problems seem to come from drivers .

Evans Electric said:

In my short  experience  with LED lighting , majority of problems seem to come from chinese products .

Click to expand...

Edited that for you.

Exactly right there Murdo . 

Was watching a documentary on North Korea where the majority of new buildings in  Pyongyang have been paid for and build by the Chinese although  many are empty do to the lack of services; many are also empty because they are unsafe for habitation.

That's in north Korea where much of the country is unfit for habitation.

:C

We recently had a recall on some garden lighting, been in a few years. GU10 spiked lights in the flowerbeds with black rubber flex running back to a junction box, tons of them. After trying everything else we realised it was the manufacturers flex that had a low ins resistance on every single one. On talking to their tech we were told, 'oh we dont do them  like that now, changed the cable but yours are out of warranty so tough' 

Cant remember the make, will get back to you.

a lot of the external light fittngs I've seen lately seem to come with HoRn flex, ie natural rubber - it seems to be considered suitable for burial?

Biggest problem I had on one job was Ants dragging soil up into the base of the lights

each light had been screwed to a small flagstone, hole drilled in slab for factory fitted flex to,go through to,a remote JB

every fitting was full of soil...bit of rain...soil,soaked it up...trip city!