Mysterious Lighting Problem

[dingbat]

Okay, this has been one of those days. I'm busy putting right what somebody else has done terribly wrong. The usual stuff; loop-in lighting confused by DIY light replacement. Anyhoo, here's the one that's bugging me.

Bedroom, two lights, two-gang switch. Turned power off, identified both switch wires with continuity tester, verified that the second cable from the second light was indeed the loop from the first. So, therefore the third cable into light number one must be the feed. Power on, check, correct. Sorted. To summarise then, light one appears to be standard loop-in/loop-out plus switch cable, light two appears to be the final one on the circuit plus switch cable. With all wires separated, of the three red cores at light one, only one is live, at line potential with every other possible combination showing zero potential. Lovely.

But here's the snag: When I connect this incoming live to the switch red only, not only does the remaining red (which goes on to light 2) become live, but so does its associated black core... and its earth. Now remember, I can see both ends of this cable and they are not connected to anything. Testing reveals a variety of voltages with respect to the neutral on the feed; 25V, 130V and 240V. At the switch all four conductors plus earths are live regardless of switch position. (The switch works perfectly well on its own)

The only explanation I can think of is that somewhere in that ceiling is a horrible short situation. Anybody got any other ideas before I take up the floor above (yes, it's a three storey jobby... no nice easy loft to poke around in!)

 

[dingbat]

I might add that along the way I discovered lights wired up between what looked like live and earth, which actually worked... because they'd then b*gg*r*d about with the switch connections, yer classic earth-as-phase-conductor in two-way wiring between switches, lights permanently on, lights shorted, lights cut out of the loop. Like I said it's been a hell of a day. (The house has eight - count 'em - lighting circuits!)

I'm back in again tomorrow afternoon to try and finish the job. Any advice gratefully appreciated and cyber-beers all round if anybody can say, "I know exactly what's wrong!"

 

[BR]

Well I'm afraid I don't but do this before you start ripping up floorboards.

I can see both ends of this cable and they are not connected to anything.

are you sure that they are the same cable, they probably are but just be sure.

Try taking a continuity/resistance measurement between the three conductors at the ceiling rose and again at the switch. Get some body to shake/pull at the associated cable (where possible) while you make the test.

Certainly a range of voltages like what you describe would fit in with a short circuit / damaged cable.

 

[dingbat]

are you sure that they are the same cable, they probably are but just be sure.

Try taking a continuity/resistance measurement between the three conductors at the ceiling rose and again at the switch. Get some body to shake/pull at the associated cable (where possible) while you make the test.

Certainly a range of voltages like what you describe would fit in with a short circuit / damaged cable.

It is the same cable - I already identified it as the final leg of this circuit. It's reassuring though, that you think it may be a short. Good idea to get a 'jiggler' while I go through some more tests. Cheers.

 

[breezer]

two things

take the lamp OUT when testing

2 (i forgot)

 

[dingbat]

two things

take the lamp OUT when testing

2 (i forgot)

Fair point but, as I may have explained... there ain't no lamp there - just bare wires all over the place. It's one of those jobs where the customer got 'a friend' to help him out. I've had to reconnect about a dozen light fittings, all of them wrongly connected and all of them different. One of them was wired via an fcu and delivered 120V to the point of utilisation. When I asked, "Why the fcu" the explanation includede the possibility that there, "...might be a transformer in there somewhere...". I linked that one to the existing light circuit and disconnected the fcu.

Still, eh? It's all experience...

 

[breezer]

good luck

although not practicable a rewire may be quicker

 

[MANDATE]

It's quite likely someone put a nail through a floorboard and straight into a cable. That's what I found recently.
I agree a continuity tester across red/black- red/earth - black/earth will confirm the situation.

 

[the trician]

By the sound of things it would be quicker to rewire the thing. At least you then have the knowledge that it is intact and has been done properly .

Sounds like something is 'backfeeding' via a short to me

 

[dingbat]

Thanks for the responses, guys, they pretty much confirm my own thinking. .. which is nice. As the customer wants to remove the second light anyway I may well discover the problem when I try to draw back those cables. I'll let you know what I find!

 

[dingbat]

Okay, it's sorted. There was a low resistance earth-to-live path further back in the circuit, due to an earlier bodged attempt (not mine) to fasten a light to the ceiling and nipping the cable.

 

[breezer]

just out of curiosity how long did it take to find it?

 

[dingbat]

just out of curiosity how long did it take to find it?

About forty minutes, but I think I got lucky... there were a good few, less accessible lighting points further back in the circuit.

Almost all of the expensive yet incredibly tacky brass and glass 'chandeliers' were wired in by pulling the circuit cables out through a ceiling hole adjacent to the rose the lamp flex was supposed to pass through. The flex was cut short and connected via choc-blocks, so dangling next to each suspension chain was a lovely bundle of cables and masking (not insulating) tape. None of the fittings matched and although each had many lampholders, very few lamps were fitted.