Link between one light fitting and socket circuit?

[uksparks]

Hi,

I changed a CU this morning, old 3036 for split board.

I had tested it prior, IR etc, few Zs readings and the like, continuity of rings etc, nothing abnormal, even checked for borrowed neutral on light circuit which there was.

Anyway, changed it this morning, put lights on one rcd and the sockets on another.

Tested everything, R1+R2, IR etc on all circuits and other tests, all absolutely fine, tested every point on each circuit, all readings all ok.

Turn it on after completing all tests...

Everything seems ok, thats good then, walk round the house, switch lights on upstairs so see if they all work as expected, yep, all fine, all fine downstairs, but when it comes to the hall light (not stair light) i turn it on and it trips both RCD's? hmm thats odd. Take the bulb out and its fine.

So i put my meter on it on line neutral and earth, 230 L-N, 230 L-E, 0 N-E, ok so I would imagine the bulb would work.., put it in switch it on again and it trips both.

Anyway, I narrowed it down, it appears there is a link between the socket circuit and that light fitting.

That light fitting is definately on the light circuit.

I even then tested N-E from the socket circuit, so see if theres any reading (dead) at the light, nothing, I tried all combinations, I get zero voltage at the light fitting when the light is off and also when the light MCB is off.

So somewhere, and for some reason that particular light is connected, but cannot seem to do any tests that suggest it is.

Unless I have missed the obvious.

Anyway, so resolve it until monday, I just moved the lights over to the same RCD as the sockets, then on monday will look at it as have got to pull the floor up to add more sockets in anyway.

I was happy to a degree that it was safe to leave for the weekend as I could not get however much I tied any voltage on that light fitting when the lights were off at the CU.

The point to this I guess is, if I had put them both on the same RCD to start with which I may have done, it was just chance I didn't, then I would never have found this problem.

Any thoughts? I can only see that it is connected on the neutral, but can't understand why as when you disconnect the socket circuit at the board, the light still works fine...

Bit of a weird one, I have not had this happen before.

 

[telectrix]

my guess would be that someone has fitted an outside light, possibly from hall light switch, and found that there was no neutral at the switch, so pinched one from the easiest point, maybe a socket.

 

[uksparks]

There are no outside lights, and only one socket in each room upstairs and minimal downstairs, none at all in the hall way.

I just cannot understand it, but the main thing that is different about that light, there is two cables in the switch, a twin and earth, one of which is cut off and only one conductor being used, and the other is a single, these are both red, and the one thats cut is red. And at the light fitting, there's only one cable, black and red...

Above the ceiling there will be a joint at a guess.

Its just chance that I even have this problem as I may not have ever known if I had put them both on the same RCD to start with, I can't really think of a way to get around avoiding this as I have tried everything I can think off.

 

[telectrix]

maybe trhat dissed wire was L feed to some wall lights, since removed, but their N came from another circuit and is still connected?

 

[Leesparkykent]

Have you had all the socket fronts off to have a look?

 

[Octopus]

I did a CU change recently and discovered that the upstairs lighting circuit was connected to the immersion circuit - literally connected via 3 cores! Fortunately on the "load" side of the (redundant) FCU!

Very odd!

 

[uksparks]

yes I have looked at the sockets in the front of the house up and down which I guess would be the obvious place.

Im going back next week anyway so I am sure ill find it.

Thanks.

 

[davesparks]

I love a good game of guess the fault!
My vote goes to borrowed neutral from the socket circuit.

 

[Octopus]

If its in the hall then look "above" in all the obvious places inc the airing cupboard, because thats almost certainly off the socket circuit.

 

[uksparks]

That's my guess, I'll update you when I get to the bottom of it.

 

[uksparks]

If its in the hall then look "above" in all the obvious places inc the airing cupboard, because thats almost certainly off the socket circuit.

The airing cupboard just has one outlet for the boiler on a radial, most annoyingly the rooms all have laminate down and are full of furtniture, it's going to be fun!

The biggest concern Is that I could have never found it if on the same rcd, so would I be at fault? I'd say probably not as I follow dead testing to a t and do the whole job and still didn't pick this up.

 

[HappyHippyDad]

It does sound very odd! If it was a borrowed neutral with that socket circuit then why does the light still work when that socket circuit is disconnected at the CU? Sorry to state the obvious Tom, but when you say disconnected, you did mean the neutral as well?

Perhaps that light is connected to both the neutral of that lighting circuit AND the socket circuit, hence the tripping, but working once the sockets are disconnected!!

You MUST update us Tom.. I cant stand it when we dont get to hear the end of the story. :smile5:

 

[Leesparkykent]

I would drop one of the legs of the ring out at the board and split the legs on a socket at roughly midway point and repeat the process on the sockets until I found what leg the borrowed neutral is connected to.

 

[telectrix]

I would drop one of the legs of the ring out at the board and split the legs on a socket at roughly midway point and repeat the process on the sockets until I found what leg the borrowed neutral is connected to.

far to sensible for a friday evening. STOP IT!.

 

[HappyHippyDad]

I would drop one of the legs of the ring out at the board and split the legs on a socket at roughly midway point and repeat the process on the sockets until I found what leg the borrowed neutral is connected to.

But why does the light still work when the socket circuit is diconnected from the CU?

 

[7029 dave]

If the circuit feeding the sockets has been disconnected at cu, and the light is still working it can't be linked.

 

[davesparks]

Did you try it with all of the other lights switched off and just the hall light switched on?

 

[Leesparkykent]

But why does the light still work when the socket circuit is diconnected from the CU?

I must of missed that, where does it say that?

 

[Octopus]

More to the point why were you changing a CU when you should have been in church?

 

[davesparks]

More to the point why were you changing a CU when you should have been in church?

In church? I doubt very much he'd have a clue where to start on an MICC installation. I'm not sure you'd be able to work in a church today anyway, isn't there some sort of religious nonsense going on this weekend?

 

[richy3333]

More to the point why were you changing a CU when you should have been in church?

Perhaps it's the vicarage he was working in?

 

[telectrix]

if the church has a graveyard, maybe it's a ghost voltage.:49:

 

[7029 dave]

I would isolate, leave the socket circuit disconnected at cu, and do an continuity test from light to said socket circuit to prove, could be doubled up.

 

[uksparks]

Yes the light still works with the sockets fully disconnected at the CU, it's like its just connected for absolutely no reason.

 

[Knobhead]

So you’ve left it like this until next week?

 

[davesparks]

So you’ve left it like this until next week?

Of course he has, don't you remember this from the last time he found a fault on a CU change. I think it was a cable with a hole drilled through it he found and then left live for a while.

 

[Knobhead]

I though this sounded familiar.

 

[HappyHippyDad]

I must of missed that, where does it say that?

2nd to last paragraph in the OP.

 

[HappyHippyDad]

It's just a matter of doing all your tests now UK, interesting though. Look forward to the update :smile5:

 

[123]

Of course he has, don't you remember this from the last time he found a fault on a CU change. I think it was a cable with a hole drilled through it he found and then left live for a while.

Recall that thread, don't know if it's been fixed yet! But the install is safer now with the RCD's and MCB's so all is good apparently!

 

[uksparks]

Not that old chestnut again, that cable got fixed, see photo below, was very minimal damage to be honest, but still repaired it.

 

[uksparks]

It's just a matter of doing all your tests now UK, interesting though. Look forward to the update :smile5:

The thing with this is that I nor anyone else would have picked this fault up unless they had the light switch on with a bulb in that lamp holder and did an IR between the sockets snd lights. There's a limit as to how much you can test and that is one that I didn't do.

i won't bite at this but as I couldn't seem to get a voltage reading at the light however much I tried with the light circuit off, I'm not bothered about leaving it until next week. It's been like it probably since the house was built in the 80's anyway.

I was 20 miles from home, and had no means to rip floor boards up, nor was I in the mood to be doing that, especially as its all laminate anyway.

 

[Knobhead]

Not that old chestnut again, that cable got fixed, see photo below, was very minimal damage to be honest, but still repaired it.

Yes, that old chestnut again because you left a job in a dangerous state rather than fix it.

There’s an echo follows you around, it’s faulty/dangerous, I’ll do it next week.

 

[uksparks]

Did you try it with all of the other lights switched off and just the hall light switched on?

Yes various combinations.

 

[uksparks]

Yes, that old chestnut again because you left a job in a dangerous state rather than fix it.

There’s an echo follows you around, it’s faulty/dangerous, I’ll do it next week.

And you fix everything thete and then do you? the installation is no worse than when I found it, if anything it's safer, I'm not entering into am argument about this.

 

[uksparks]

If the circuit feeding the sockets has been disconnected at cu, and the light is still working it can't be linked.

It could be, it worked with it disconnected and connected but only when on same rcd. It's like its wired correctly but with an additional neutral link which is why it works both ways.

 

[uksparks]

But why does the light still work when the socket circuit is diconnected from the CU?

Because it's from what I can te two normal circuits with an additional link on the neutral. I'll find out next week when I go back.

 

[davesparks]

And you fix everything thete and then do you? the installation is no worse than when I found it, if anything it's safer, I'm not entering into am argument about this.

Generally speaking I will fix, or at least diagnose and temporarily fix faults there and then. What I definately don't do is go telling the whole world about any little bodge I may have to do on the spur of the moment. Sometimes I phone my mentor who taught me the trade for a bit of advice or a second opinion.

 

[davesparks]

The thing with this is that I nor anyone else would have picked this fault up unless they had the light switch on with a bulb in that lamp holder and did an IR between the sockets snd lights. There's a limit as to how much you can test and that is one that I didn't do.

i won't bite at this but as I couldn't seem to get a voltage reading at the light however much I tried with the light circuit off, I'm not bothered about leaving it until next week. It's been like it probably since the house was built in the 80's anyway.

I was 20 miles from home, and had no means to rip floor boards up, nor was I in the mood to be doing that, especially as its all laminate anyway.

I agree it is something that I would not have found with any standard testing, it would have showed up pretty quick though once the new cu was fitted as I fit SPSN RCBOs for all circuits as standard.

You shouldn't be ripping floorboards up whatever mood you are in! A tradesman should be carefully lifting them so as to cause the minimum possible damage. Ripping people's houses apart is what demolition contractors and plumbers do.

 

[uksparks]

I accept that, but it's not being left, it's going to be fixed.

you keep your bodged or temp knock ups etc to you and your friends down the pub and I'll continue to post the odd thing on here, I knew I'd get stick for bringing it up.

 

[uksparks]

I agree it is something that I would not have found with any standard testing, it would have showed up pretty quick though once the new cu was fitted as I fit SPSN RCBOs for all circuits as standard.

You shouldn't be ripping floorboards up whatever mood you are in! A tradesman should be carefully lifting them so as to cause the minimum possible damage. Ripping people's houses apart is what demolition contractors and plumbers do.

I hsve to say this has caught me out as didn't pick it up, and it's only by chance it's to come to light as they were on different sides of the board but could very easily have been on the same side then I would never have known.

Ive got three more boards lined up for this month so will look a bit deeper at the pre testing to try to eliminate anything like this again.

as forge comment about ripping the floor up, it was just an expression.

 

[davesparks]

You'll not catch something like this on pre-change testing. Not unless you are going to spend more time devising bizarre tests than actually doing the job. You just have to suck it up sometimes and hope you've allowed enough contingency in your price. If not then you just have to accept that some jobs go wrong and make a loss, shyte happens sometimes and you just have to accept it.

 

[uksparks]

It's something I'm now aware of and will think about it next time.

im going to sleep now, night