IR tests and fault finding - any pointers?

[sparkybird]

Hi everyone

Tested existing lighting circuit today - got IR readings as follows
P-N 500
P-CPC 500
N-CPC 0.3

However, on testing as 250 V as a do first, I got N-CPC as 200 mega ohms

Strangely the lighting circuit was already 30mA rcd protected and client not reported any tripping. However I can't leave it as it is and need to investigate.

My question (finally!) is : do the different readings at 250 and 500V give me any clue as to the fault I am looking for?

Any help much appreciated

SB

 

[ricicle]

There is either a fault of finite resistance which is showing up only at the higher voltage (self explanatory) or a component in a hidden piece of equipment (aeriel amp. ?) that is shorting down at the higher voltage.

 

[oharaf]

Interesting set of readings.

On that lighting circuit did you remove every single bulb?

Also don't ever assume that nothing else is in circuit.

Check for extractor fans, smokes any fused connection unit is worth taking out of the circuit.

Check for any posibility of contact between N & CPC at ceiling roses (you know how often people like to not sleeve earth!!!)

Either way you need to chase that fault down, it needs to be >2MOhms.

Trying breaking the lighting circuit in half at a ceiling rose and chase it that way using Cont. & IR tests.

Good luck, it can take absolutely ages

 

[RF Lighting]

Not meaning to patronise you, but ,

Did you do this testing with the CU main switch?

It's maybe the reading is through the incommer where the N&E are tied together, possibly through a borrowed neutral scenario?

Not sure why the reading would vary with the increased voltage if that were the case though.

 

[Spark123]

Was the reading steady or fluctuating?

 

[sparkybird]

Thanks for suggestions everyone, some stuff to ponder...

In reply:

The lights are GU10 recessed lights, all on dimmers. Tested circuit without switch line (ie dimmers off and didn't remove lamps). However, the CPC-N test would test through to the fittings.

The P-N 500MOhm reading seemed to indicate that there were no connected loads surely??

The N-CPC reading was fluctuating-yes -up and down at 250V. I didn't keep my finger on the button at 500V as I could see straight away it was a low reading.

All cables tested out of board RF

This circuit in new colours, other lighting in old ones, but I guess doesn't rule out borrowed N-more for me to investigate.... Overnight I popped this circuit on the non RCD side, but put the other two lighting circuits on the RCD's (dual). If borrowed N then surely there would be tripping?

TV is cable.

Does this help?

Thanks
SB

 

[Spark123]

Fluctuating can indicate arcing.
I'm not a fan of having dimmer switches in there when doing IR testing either, normally if they had to stay I'd put the L&N together and test to earth - recording it as a limitation.
I take it you did similar i.e. switch off at switch for things like outside lights and bathroom light?

The RCD doesn't trip as the N & E are about the same potential, there is not enough potential there to push enough current down through the fault to trip an RCD.
Even with a pd of 230v a fault of 300K will only allow a fault current of 0.77mA to flow, (I'm not suggesting that it is correct to leave it by any means!).

 

[ColJack]

the dimmers wiill have filters built into them, try removing the wires from the dimmers, not just switching off..

 

[sparkybird]

Thanks again guys for your late night replies. I'll do as suggested and remove the dimmers from the equation. Although as there is no N at the dimmers, not sure if this will affect the N-CPC reading? I'll find out I guess.

I can see how at 250 the varying readings would go to a low one at 500V, yes - the current is just too much for the 'fault'.

It'll give me welcome respite from the sun as I'm doing outside lighting at the same job! Must hunt out that factor 50....

Cheers

SB

 

[Lectrician]

Thanks again guys for your late night replies. I'll do as suggested and remove the dimmers from the equation. Although as there is no N at the dimmers, not sure if this will affect the N-CPC reading? I'll find out I guess.

I can see how at 250 the varying readings would go to a low one at 500V, yes - the current is just too much for the 'fault'.

It'll give me welcome respite from the sun as I'm doing outside lighting at the same job! Must hunt out that factor 50....

Cheers

SB

When testing, the 'neutral' will be at the dimmer via the lamps, as you left these in.

 

[sparkybird]

Hi folks

An update....

The fault exists between the CCU and the switching for the lights, which unfortunately will be done via a jb or two under the floorboards but not accessible from the holes for the lights. GRRR

In the area where I'm pretty sure this is located, floorboards have been replaced recently and screwed down, so I wonder if the fault could be a screw into a cable which is OK at 250V but not at 500?

Will have to take up the carpet/boards to investigate further.

The 'loads' ie wiring between the lights is all OK.

Thanks again for suggestions and ideas

Regards

SB

 

[Adam_151]

Screws into cables can give funny readings, if a slither of insulation is all that remains, then it can do funny things, I had one that read 24Mohm, which just seemed too low, which I eventually traced to a screw in a cable, its not beyond possibility that the slither to be so thin that 500v flashes over it, yet 250v does.

It goes without saying, but if a screw has hit it, then chances are cables are in notches... so careful with your circular saw when lifting boards

 

[sparkybird]

Thanks Adam. I don't have a circular saw, but use my Fein Multimaster which has a depth stop which neatly avoids this problem!

I "heart" my Fein!! Sad, eh?

SB

PS I think the client screwed down the boards, so he may have gone in the middle rather than the edges - I'll find out when I lift the carpet....

 

[ColJack]

the exakt saw looks like it would be great for floorboards...
fine blade, adjustable depth stop and plunge cut action..

 

[sparkyspike]

I ♥ my Makita Lion Skillsaw!