Identifying shorted bulb in a series of 70

[pcdj]

I have 35 stepping stones placed around my garden, each with two 1W uplighters embedded into them shining upwards.

Occasionally these units will get wet and short out, requiring them to be replaced.

How can I efficiently figure out which of the 70 bulbs is causing the short given the moment I turn them on the MCB trips in the distribution board? They're all 240V LED units.

 

[oldbutnotdead]

Short answer- assuming they are all hardwired in parallel -you can't.
If you have any easily accessible couplers or joints between lamps then (starting from the far end) disconnect a block of 10, try it, repeat until the breaker stays in then check the last 10 you disconnected.

 

[pcdj]

Sadly all the wiring's buried under concrete and all I have access to are the tails.

I have easy access to the ends of this circuit coming into the distribution board so I had the idea of disconnecting it and passing 12V down the line and using a multimeter to check which light's metal housing is energised, but I take it with this many bulbs and what's probably 100m of cabling I guess this won't work?

 

[SimonH2]

I have 35 stepping stones placed around my garden, each with two 1W uplighters embedded into them shining upwards.

70W of LEDs, do you have aircraft mistaking them for runway lights I know a complaint from a significant number of people is the amount of light pollution which effectively makes the night skies "disappear". I have a couple of strings of coloured LED GLS festoons in my back garden - at full brightness that's a total of 160W and "fairly bright", but I normally run them off a transformer so they give a nice glow without the National Lampoon's Christmas Vacation effect on the neighbours.

Occasionally these units will get wet and short out, requiring them to be replaced.

How can I efficiently figure out which of the 70 bulbs is causing the short given the moment I turn them on the MCB trips in the distribution board? They're all 240V LED units.

MCB or RCD (or RCBO) ?
And are the integrates units, or "bulbs" plugged into a separate unit ?
How are they wired - do you see two cables come up to all but the last light, or are there underground joints and you just see a single tail to each light ?

I would be surprised if a light caused an MCB to trip at the first sign of water ingress - my experience is that everything else keeps going, and one (or more) light goes dim. Mine are non-earthed hence don't trip an RCD. So I'm thinking that you have an RCD tripping if the lights are earthed.
I'm struggling to see an easy way of testing - it reminds me of the old routine of trying to make the fairly lights work each Christmas in the days before the (series connected) bulbs went short-circuit when they blew.

 

[pcdj]

70W of LEDs, do you have aircraft mistaking them for runway lights

So they've actually all got "hoods" - I don't really know the technical term - the light shines sideways, rather than directly upwards, hopefully diminishing some of that light pollution.

MCB or RCD (or RCBO) ?

You're right; sorry, it's an residual current circuit breaker.

And are the integrates units, or "bulbs" plugged into a separate unit ?

They're not a bulb but a single unit of bulb + small circuit board (transformer?) all packaged into an inch diameter tube - pretty much the same as the picture I linked to for the "hood" above.

How are they wired - do you see two cables come up to all but the last light, or are there underground joints and you just see a single tail to each light ?

So at each stepping stone I can see a pair of tails reaching each light, and at the distrubution board I can see a pair of wires where this circuit goes into the circuit breaker. Everything else is buried.

I would be surprised if a light caused an MCB to trip at the first sign of water ingress - my experience is that everything else keeps going, and one (or more) light goes dim. Mine are non-earthed hence don't trip an RCD. So I'm thinking that you have an RCD tripping if the lights are earthed.

Answered above; you're correct in that it's a residual current circuit breaker.

 

[SimonH2]

Right, so a single 2 core cable at each light ? That means there's 69 buried joints which is a "poor" system design IMO.

I'm afraid my suggestion is very much "I wouldn't start from here" I assume it would be impractical now to split the system into multiple sets of lights, each separately cabled back to the house (or to a weatherproof box) ? Having the ability to disconnect (via a 2 pole switch) all the lights and switch them back on in small groups would massively speed up the diagnosis.

As it is, the only thing I can think of would be to connect both wires together and to a DC supply - say 250V or 500V, current limited, and with the other side earthed. Then use a "wander lead" and a high impedance voltmeter - prod the damp stone/concrete next to each light and see where you get a flicker from the current flowing to earth.

 

[333rocky333]

An old way of fault finding was to put an old lighting choke or a 60 watt lamp, temporarily in series with the live supply, this limits the current, thus not blowing the fuse, any lamps before the short would then glow, or the voltage measured along the run, thus indicating roughly where the short is, however as this lowers the voltage it worked with tungsten mains voltage lamps, so not sure if the lower voltage may damage led drivers so be wary.