Best way to locate a fault in garden 12v lighting

[captaincaveman]

Hi Guys n Gals,

I'm a little stumped on a job and was wondering if anyone could advise?

I've got 3 drivers going to 3 sections of garden lights. Two of the three sections are not working (about 10 lights). I presume it's two sections due to the number of lights and I can't find a third section anywhere else. All the cables are buried under the patio flags or within the cavity in the patio wall with no access to connections. None of the connections have enough length on the cables to pull the connections out (so I have no idea where the connections are).

I tried the good working section with the two drivers of the bad sections and the good section still worked so I'm assuming those drivers are ok. The chances of all 10 lights faulting at the same time is unlikely so I'm assuming there is a bad connection(s) somewhere (Although there doesn't appear to be any work done to disrupt them anywhere).

It's a nice patio and the customer is reluctant to have walls opened up and flags lifted so I'm just wondering if there's anything I can do before suggesting that's the only way?

Cheers in advance

 

[marconi]

It seems to me that with a series wired string of lamps, if a string does not light because a lamp has failed open circuit or high resistance then there is(can) only(be) one duff lamp fitting. Two or more lamps failing at the same time is much less likely. So I would take Adonis2's tried and tested idea - because pins are sharper and smaller than needle tip probes - and simply short out each lamp in turn. With power on it would be immediately obvious if there was a single lamp failure and no other defect cos the remainder would light up. Nice and quick and less faff to begin with.

 

[Simon47]

if it is possible to identify the fixtures of a duff string What about taking the lamps out of the working string and putting them in one of the defective strings?

Problem here is that the lamps are sealed units, and none of the connections are accessible - complete numpty installation designed to make fault finding as hard as possible.
I think the sewing pins suggestion is as good as anything.

If it can be identified where the cores are, it should be possible to stick two pins into the cable near the lamp fitting - then use a wandering lead to get voltage readings between there and the drivers. If it is a failed lamp then in principle it should be possible to splice a replacement in. If it's a bad cable or joint, then the customer has to choose between pulling up the patio or not having working lights.

 

[buzzlightyear]

@captcaveman it would take sledge hammer to crack a nut or walls and path ,some customer start putting their fingers in the ears when it comes to spending more moneys on the jobs after they have been done .

 

[captaincaveman]

It seems to me that with a series wired string of lamps, if a string does not light because a lamp has failed open circuit or high resistance then there is(can) only(be) one duff lamp fitting. Two or more lamps failing at the same time is much less likely. So I would take Adonis2's tried and tested idea - because pins are sharper and smaller than needle tip probes - and simply short out each lamp in turn. With power on it would be immediately obvious if there was a single lamp failure and no other defect cos the remainder would light up. Nice and quick and less faff to begin with.

Just to update on this, for what it's worth....
I returned the other week to try some of the suggestions but didn't really get anywhere. Not helped by the British weather!

Although I suspected the constant current drivers were ok anyway, as I had tested each one with the group of lights that were still working successfully, I tested them at source getting 83v for each (pic 3) and 0.3A at source with power to the lights (pic 2).

I tried continuity of the conductors with my circuit finder and the continuity was fine (it also helped me group the lights to each transformer). I tried measuring the voltage at various lights with the needle into conductor trick and got around 0.67V at each. I tried bypassing various lights by shorting across conductors with a needle (pic 1) with no luck.

I finally swapped a non working light with one I knew worked and it still worked so I suspect there are multiple lights out. I couldn't test the 'faulty' one where the good one was working as the cable slipped through my fingers and disapeared into the wall. Doh! After multiple attempts to retreive it and getting very wet in the mean time I called it a day. I've told the customer I'd think the situation over first but rather than spend more time on it I suggested it might be easier to just put new spkike lights in at the gravel trench in front of the wall (where connections can be accessed!). As the existing lights are £130 each to replace like for like and there are suspectedly multiple ones out it would probably be more finactially viable as well...

 

[timhoward]

suggested it might be easier to just put new spkike lights in at the gravel trench in front of the wall (where connections can be accessed!)

Sounds wise. As my mentor used to say - ‘always remember that you can’t polish a tur*’

 

[marconi]

Did you obtain access to the T connectors - see #8.

Have you taken each fitting and powered it alone from a known good psu?

 

[captaincaveman]

Did you obtain access to the T connectors - see #8.

Have you taken each fitting and powered it alone from a known good psu?

Alas none of the T connectors are accessable. Even the few that I can see are out of reach :/

I suppose I could take each fitting out to test by snipping each cable. I'd have to clamp all the cables to stop them retracting back into the wall (& have decent weather!). I'll ask the customer if they are prepared to pay £130 per fitting that's faulty and the labour charge on top (plus ££ for parts making new weatherproof joints)