We have an extensive, professionally installed, garden lighting system which has recently started tripping its dedicated RCD. I have eliminated all the bulbs, luminaires and accessible cables which, sadly, means the problem is almost certainly in one of the many 'waterproof' underground connection boxes. How on earth does one trace a fault like this? Dig up the entire garden?
Richard.
In message , Richard Russell writes
I think the answer is "Yes!".
My son's house has an RCD in the garage that feeds the cooker, an outside plantroom (that feeds some garden lights, pumps and other things), a summerhouse with leaking roof, some street lights up his drive and the filter and pump in the fish pond.
Every few days, in the middle of cooking a meal, the RCD would trip. I took my old wind up megger down there, but was met with contempt.
Disconnecting the feed on from the plant room cured the problem so he re-wired the summerhouse - no change. Replaced and rewired the outside box feeding the pond. No change. Bought and installed a new submersible pump. No change. He bought and installed some awful cylindrical in-line connectors between the new box and the pond pump and separately the filter. It was the filter. Dismantled it and discovered water had leaked past a seal. Dried out and re-gunged and so far it has been OK.
This took us weeks to find, and at least he now has more knowledge of the distribution (we still don't know how the street lights are wired) and some of the installation is good. He did the wiring. I went along to be a second pair of hands and to do the semaphore during testing.
I must find out what the in-line connectors were so that I am never tempted to use them.
What would be the acceptability, under the regs, of fitting an isolating transformer? No chance of tripping the RCD then, leakage or no leakage!
Richard.
Apart from the lack of the required RCD protection?
Well if you know the route it takes one could start at the furthest one and disconnect, then reconnect if it still trips and move closer etc. the snag occurs if its branched ie several circuits or you do not know what goes where first!
Brian
Ah and then just watch for the smoke when the leak gets worse, brill. Brian
Divide & conquer. Simple as that. Its going to mean undoing many connections as you megger it all bit by bit.
If its a practical option to feed it off more than one rcd, that would ensure half of it stays on in a fault.
NT
Every single connection is in a buried waterproof box - there are a lot of them - so such an approach would involve lots of digging (in a mature garden with pretty much the entire ground surface covered with plants and/or weed membrane and gravel). It's not very practical!
Richard.
I don't know what the letter of the regs says, but if an isolation transformer (with a 115-0-115 secondary) is considered safe for on- site power tools I dont really see what the objection would be to using one for garden lighting.
Richard.
On Tuesday 18 June 2013 23:03 Richard Russell wrote in uk.d-i-y:
That's 55-0-55 aka 110V you're thinking of and the 55V (relative to earth) counts as ELV.
Fair point, but isolation transformers with a 230v secondary are sold as 'safety site transformers', which is what I was meaning. For example these from RS:
Our house is of a vintage (built six years ago) when no RCD protection was required on the lighting circuits, so in that context I can't get myself too worked up about garden lighting with no RCD protection, especially if it's only running at 115 volts wrt ground.
Richard.
On Wednesday 19 June 2013 00:03 Richard Russell wrote in uk.d-i-y:
I will have to check chapter and verse - but even if fully floating, relying on it is a bad idea IMO.
"Only 115V"? I think you *should* worry about that - outside is a particularly special case as you both have enhanced shock risk (wet ground/rain/etc) *and* you are more likely to expose a member of the public to risk (eg visitors).
If its really not doable, I'm struggling to see a better option than going
55-0-55.NT
I can't say I agree. AIUI, the requirement to have RCD protection at all o n garden lighting circuits is a fairly recent one. In addition, garden lum inaires are far better protected from accidental contact than a domestic (b ayonet or screw) light fitting, because of the weatherproofing requirement.
To gain access to the bulb, let alone remove it to expose a live terminal, typically requires unscrewing a watertight cover - something which I know f rom experience is not easy even when you want to do it! The idea that a vi sitor might unscrew one of these covers, remove the bulb (presumably lit) a nd poke their finger into the exposed socket - mostly GU10 so that's not ea sy - seems wildly implausible to me.
As it happens the point is moot anyway. As an experiment I bypassed the RC D so I could power up the circuit for a few hours. The result is that (may be through some local heating drying out the moisture) the leak is now redu ced below the trip threshold and it's running fine again with the RCD re-in stated. The problem may recur, of course, but I'll worry about that if and when it happens.
Richard.
Think it's any circuit "exported" not just lighting. B-)
But given the choice between a traditional L 230 V wrt ground and N bonded to ground circuit and a fully floating nominal 115V L and N wrt to ground but 230 V l to N I'd prefer the latter. If you touch either it drops to essentially ground you'll still get a belt but nothing like the one you'd get from touching the L in the first case. Problems come when there is a fault to earth on one side of the fullly floating circuit and you touch the other...
Design failure, there is no such thing as a "waterproof box" unless you fill it with something to stop water getting in, be that a potting compound or hard setting tar.
If there is a void water will get in, either by capillary action through any seals or as water vapour and then condensation through the seals or even the box material itself.
They should have been left accessable in some way or marked, At the very least an accurate map of cable routes and joins made.
RCD so I could power up the circuit for a few hours. The result is that (m aybe through some local heating drying out the moisture) the leak is now re duced below the trip threshold and it's running fine again with the RCD re- instated. The problem may recur, of course, but I'll worry about that if a nd when it happens.
it inevitably will.
Converting to 12-0-12v is another option
NT
Not especially no. Outside circuits in general require RCD protection.
So lets say the cause of the fault is that you have a live to case short on a lamp and the case is not earthed, and only under certain circumstances is there enough leakage to earth to cause your trip. Still fancy leaning against that lamp post with the RCD disabled?
Implausible yes, but not relevant either.
Ah, the ostrich algorithm ;-)
In message , John Rumm writes
I've got the answer. Bypass the RCD, insert a series diode, let it run until a period of rain, then walk round the garden with a candle.
If you go the binary chop route, you can eliminate 3/4s of the circuit by digging up just two of the junctions...
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.