Help Please - a bit of a puzzler!

[Electrifying]

Hi all,

Could do with some help.

At a house yesterday and came across a problem on the upstairs lighting circuit whilst testing.

It's a three bedroom, traditional semi.

TN-C-S

MK 12-way, twin RCD board.

With all other MCBs OFF, except for the lighting circuit, I attempted to take a ZS reading at the ceiling rose in the back bedroom.

It was to be the 3 lead, no-trip test with my Megger MFT.

I connected my earth clip to the CPC and touched my neutral probe to the neutral of the rose........the alarm sounded on my Megger with the >50v sign in the display.

Thinking this was weird I proceeded to test voltage between the conductors, getting the following readings:

All other lights on the circuit were switched OFF and the lamp removed from the rose that was under test.

With the light switch OFF.....

L - N 12v

L - E 2v

N - E 81v

With the light switch ON.......

L - N 239v

L - E 110v

N - E 101v

I'm having trouble getting my brain to work this out - in particular, how there can be voltage present on the Neutral and/or the CPC, at the levels given, without causing the RCD to trip.

I then went into the front bedroom to conduct the same test - and got the same results.

Taking the same readings at the CU, (on this circuit), gave the readings that you would expect to get.....i.e. L - N 230v, L - E 230v, N - E 0v, so something is happening somewhere.

Any suggestions as to what the cause could be or suggestions on how to find the fault.........I bet it's something simple but I just can't see it

I'm back there on Sunday to hopefully sort this.

Thanks in advance.


I've got to go out to work now but I'll check for any replies when I get back in.
Sorry, won't be able to give any more info until then.

 

[Spark123]

Sounds like an open circuit CPC, have you done R1+R2?

 

[JohnW2]

If the RCDs not tripping with those figures, the CPC can't having continuity to earth, can it?

Kind Regards, John

 

[riveralt]

Some questions first - what was the reason you focused on this particular circuit - was there something wrong with it in the first place?
What were your Continuity results when you tested the circuit dead?
Have you done a visual inspection of the wiring inside the CU for this particular circuit?

 

[Electrifying]

Sounds like an open circuit CPC, have you done R1+R2?

That's to be my first port of call on Sunday, along with an inspection of each other lighting point on the circuit.

Open circuit CPC wouldn't explain the N - E reading of 81v with the light switch off though - in fact it wouldn't explain why the alarm sounded on my meter when connecting the probes between N & E for the original Zs test.

 

[Spark123]

You can end up with a whole manner of weird readings, capacitive coupling with a high impedance meter.

 

[Electrifying]

If the RCDs not tripping with those figures, the CPC can't having continuity to earth, can it?

Kind Regards, John

RCD trips on difference of current between L & N, so would still trip without CPC continuity.

 

[Electrifying]

Some questions first - what was the reason you focused on this particular circuit - was there something wrong with it in the first place?
What were your Continuity results when you tested the circuit dead?
Have you done a visual inspection of the wiring inside the CU for this particular circuit?

Fitting a new light fitting, so using Zs to confirm CPC continuity.

Haven't tested dead yet, but, like I said, I will do now.

CU is spot on, no problems.

 

[JohnW2]

If the RCDs not tripping with those figures, the CPC can't having continuity to earth, can it?

RCD trips on difference of current between L & N, so would still trip without CPC continuity.

Your figures suggest a floating CPC with, as has been said, those figures resulting from capacitance etc. (with a high impedance meter). However, if the figures were 'real', then there would almost certainly be a consequent L-N imbalance in the RCD to trip it. Either way, it's unlikely that the CPC has continuity with earth.

Kind Regards, John

 

[Electrifying]

If the RCDs not tripping with those figures, the CPC can't having continuity to earth, can it?

RCD trips on difference of current between L & N, so would still trip without CPC continuity.

Your figures suggest a floating CPC with, as has been said, those figures resulting from capacitance etc. (with a high impedance meter). However, if the figures were 'real', then there would almost certainly be a consequent L-N imbalance in the RCD to trip it. Either way, it's unlikely that the CPC has continuity with earth.

Kind Regards, John

Hi John,

I think you've started me thinking in the right direction now - I need to find the break in CPC continuity and I need to find what else is connected to this circuit apart from standard light pendants - is that what you're saying?

Because, obviously, if it was just open circuit CPC, then I wouldn't get a voltage reading between L - E, nor between N - E.

 

[Spark123]

You get a reading as the makeup of the cables has the CPC near the live conductors with an insulating material between, the very makeup of a capacitor.
If the CPC was connected properly and at 101v with respect to neutral (remembering in a TN CS that before the DNO cutout they are the same wire) then you'd have oodles of current flowing, tripping the circuit breaker and RCD.

 

[JohnW2]

Hi John, I think you've started me thinking in the right direction now - I need to find the break in CPC continuity and I need to find what else is connected to this circuit apart from standard light pendants - is that what you're saying? Because, obviously, if it was just open circuit CPC, then I wouldn't get a voltage reading between L - E, nor between N - E.

Not quite. Yes, you certainly need to try to locate the suspected break in the CPC continuity. However, even if it has, say, simply fallen out of the CU earth bar (i.e. the whole CPC just floating), with a high impedance voltmeter you could well get exactly the sort of readings you did - with the CPC roughly midway between L and N, because of the effects of capacitance in the cable. If I'm right, I suspect that if you put a test lamp (or low impedance meter) between L&E or L&N at the fitting you would not detect any voltage, despite the apparent ~100V you've measured in each case with a high-Z meter!

I probably didn't express myself clearly enough, but my previous point was that, if the CPC were connected to 'earth', there is simply no way you could get 'real' voltage measurements such as you did - since if the earth had attempted to move even just a fraction of a volt due to a fault, it would have created enough L-N imbalance to trip the RCD before you had a chance to make the measurement!

Kind Regards, John

 

[THRIPSTER]

Were the readings taken, when the light switch was on, with luminaire fitted or without?

Regards