This discussion is locked.

You cannot post a reply to this discussion. If you have a question start a new discussion

Induced voltage or dodgy wiring?

A domestic Flat has 1 * 6A lighting cct, 1 * 32 RFC, and 1 * Ckr cct. Supply is TN-S.

A 2G plate switch in the kitchen (which has a flat roof / attic space above other rooms) is fed by a twin and earth - one core feeds the commons of the two switches, while on one switch the other core supplies two downlights. A sheathed single core from the second switch feeds two wall lights. Each downlight is fed by a single T&E. One wall light has a T&E loop in / loop on, the other light just a T&E in. All wiring is concealed in walls.

At the downlights, 244V (L-N) appears with the switch turned on (using a MFT). However, with the switch off, approx. 23V still shows. With the switch off and a lamp installed in either fitting, while testing at the other, V=0. The 23V does not appear at the wall lights.

With the CB’s turned off on the other ccts at the CU, the 23V remained, as when the N’s were disconnected also. V=0 with the lighting CB off.

L/N – E IR test with all switches on is > 300MOhms.

Before hunting about for things like dodgy connections in JB’s in a grotty/dirty loft space, any ideas what may be going on please?

Thanks

F

Parents

0 mapj1 over 5 years ago

I thought for a moment you were going to say the neutral for the wall lights comes back via the sockets. I presume for now it does not, but if it did, that would be bad, but may not show on your tests.

If this voltage is serious or not rather depends on the input impedance of your volt meter -- if you have a megger, you could measure the input resistance of your MFT on the volts range, and the voltage of the megger at the same time.. If not then the data sheet page of the handbook may reveal.

I suspect it is very high (some millions of ohms) and you are not reading 23V from some odd faulty connection, so much as the effect of a very small capacitive coupling between parallel cables.

If the line going out to the lights before and after the switch run close by one-another for a few metres, then a (very poor) parallel plate capacitor is formed , with the two 'plates' being the live and switched core respectively. This may be a few hundred picofarads or so per metre of adjacent cable (small parasitic capacitors are formed in this way between lots of other things too, but they do not normally trouble us or show up on the meter. This sort of thing only really matters when you need to consider operation at higher voltages or higher frequencies)

The clue is that makes me thing this is that the 23V falls to near zero when any current is drawn.

If you want to be sure, and are feeling brave, (and at 23V you can be, but please do not be brave with significantly higher voltages) even a wet finger across the same place as meter will cause a voltage from such capacitive pick-up to dip quite noticeably. A better and safer test uses a few K of resistance or a small capacitor of 100nF or so on croc clips.

If the voltage does not droop, it is not a capacitance effect.

I suspect in this case it is capacitance and if so all it is telling you is something about the cable layout in the wall, and nothing is wrong at all.

EDIT, OK, I'm too slow, the others beat me to it while I was composing. But we agree.
Cancel
Vote Up 0 Vote Down

Cancel

Reply

0 mapj1 over 5 years ago

I thought for a moment you were going to say the neutral for the wall lights comes back via the sockets. I presume for now it does not, but if it did, that would be bad, but may not show on your tests.

If this voltage is serious or not rather depends on the input impedance of your volt meter -- if you have a megger, you could measure the input resistance of your MFT on the volts range, and the voltage of the megger at the same time.. If not then the data sheet page of the handbook may reveal.

I suspect it is very high (some millions of ohms) and you are not reading 23V from some odd faulty connection, so much as the effect of a very small capacitive coupling between parallel cables.

If the line going out to the lights before and after the switch run close by one-another for a few metres, then a (very poor) parallel plate capacitor is formed , with the two 'plates' being the live and switched core respectively. This may be a few hundred picofarads or so per metre of adjacent cable (small parasitic capacitors are formed in this way between lots of other things too, but they do not normally trouble us or show up on the meter. This sort of thing only really matters when you need to consider operation at higher voltages or higher frequencies)

The clue is that makes me thing this is that the 23V falls to near zero when any current is drawn.

If you want to be sure, and are feeling brave, (and at 23V you can be, but please do not be brave with significantly higher voltages) even a wet finger across the same place as meter will cause a voltage from such capacitive pick-up to dip quite noticeably. A better and safer test uses a few K of resistance or a small capacitor of 100nF or so on croc clips.

If the voltage does not droop, it is not a capacitance effect.

I suspect in this case it is capacitance and if so all it is telling you is something about the cable layout in the wall, and nothing is wrong at all.

EDIT, OK, I'm too slow, the others beat me to it while I was composing. But we agree.
Cancel
Vote Up 0 Vote Down

Cancel

Children

No Data

Induced voltage or dodgy wiring?

Community Rules & Guidelines