Lights not fully off when off ??

We recently had the upstairs rewired and I am now fitting new lights. I had a bit of a shock (not literally thankfully) last night though and am puzzled as to what could be going on.

I'd turned the upstairs lights off at the CU and undid the ceiling rose. I checked the voltage between each of the 3 terminal blocks and earth to be sure and they were all 0 V.

I then started to remove the wires from the rose. However, the first one sparked as it came out and the light came on, dimly, but enough to scare me to death.

I measured various voltages between the terminal blocks and earth and these ranged from 230V, to 53V, but kept varying. I noticed some of the other upstairs lights also glowing dimly and the hall lights (that are on a different circuit that I hadn't turned off) were flickering but still fairly bright. Reconnecting the wire I had removed put things back to normal with the upstairs lights off and 0V on all the terminals, the hall light stopped flickering and was full on.

I removed the wire again and same thing happened. I turned the hall light off and the strange effects stopped.

Does anyone have an idea what might be happening. I remember when they came to do the re-wire at first they complained it wasn't straightforward as various upstairs and downstairs circuits were seemingly connected.

The rose in question had the following wires connected:

Neutral block: One old style black wire (this is the wire I disconected). One new style blue cable wire. One blue flex to the lamp.

Common block: Two old style red wires, one new style brown wire.

Live block: One old style black wire. One brown flex to the lamp.

What I am guessing is that the neutral from the hall light circuit is connected to the neutral from the upstairs circuit, but I can't understand how a current flows in the upstairs lamps as the live cable is switched off at the CU.

Either way I'm a bit unhappy that when I turn off at the CU, the lights suddenly come half alive when I disconnect a neutral wire.

Advice appreciated.

Yes something is very wrong.

My immediate instinct is that you have a burrowed/bridged neutral from another circuit.

I'd recommend you get the lighting circuits professionally tested ASAP.

Who did the partial rewire?

The people who did the rewire were electricians working for a large electrical contractor company. They did it "unofficially" on a Saturday for cash, however I was under the impression it was a proper job as they were recommended by someone and are obviously professionally qualified.
The rewire was actually due to major alterations upstairs, moving rooms, bathrooms, etc. Their brief was to install several new sockets and lights and to remove some (that were in the new bathrooms). Not really a rewire as such. Also they were to provide a new CU with RCD protection.

They quoted for the work, but when came to do it found problems with the existing wiring (one thing shown to me was a switch that had more wires running to it than it should and that was linked with downstairs). This meant they went away to requote as the job was more complex than they has envisaged.

Upstairs now we have mostly new light roses, all of which seemed perfectly OK when I fitted lights to them. The one I was doing last night was an existing rose that has a mixture of old and new style wiring.

I will probably get another electician in to check, but out of interest how could a bridged neutral cause this. I can see that I might get a voltage appearing on the neutral side, but how can this flow anywhere through the upstairs lights ?

This is hard to explain. However if both circuits are sharing the neutral you disconnected the only alternative path would be for the current to flow backwards through the lamp, into the live conductor to the next lamp and then through it back to a neutral on the other side of the break. You have lamps in series, hence dimmer.

if the neutral is connected to the upstairs, the the circuit is still live. even tho the voltages may read 0, there could still be current flowing for an upstairs light what is on.

I would agree that the neutral from downstairs is connected to the neutral of the upstairs ceiling rose.

By removing one of the neutral wires from upstairs ceiling rose, it would remove the fixed "O volts" and force the downstairs light to run in series with the upstairs light.

Which electrical installation test would pick this fault up?

It would only be evident if all bulbs were in place and you measured resistances between both lighting circuits.

thats why its called testing AND inspection

testing is mainly to find degredation or faulty components (insulation resistance rcd test) or to find important numerical parameters for the installs complliance (earth loop impedences)

inspection is to find damage and bodges i would think you haven't done a proper inspection if you haven't looked for the very common borrowed neutrals in a up/down lighting setup

dmccormick said:

The people who did the rewire were electricians working for a large electrical contractor company. They did it "unofficially" on a Saturday for cash, however I was under the impression it was a proper job as they were recommended by someone and are obviously professionally qualified.

Click to expand...

Just not very good....

pegasus said:

I would agree that the neutral from downstairs is connected to the neutral of the upstairs ceiling rose.

Click to expand...

That's not a borrowed neutral. The neutrals are all connected together back at the CU anyway, so a parallel connection between 2 ceiling roses wouldn't make any difference here. Usually what's happened is that someone has gone for the 2-wire method of 2-way switching (see the second diagram here: http://www.tlc-direct.co.uk/Technical/DataSheets/MK/WiringDiagrams.pdf ), and they've got a 2-gang switch at the end where the live feed comes in. They put a strapper between the two COMs, run TE from L1/L2 on one switch to L1/L2 on the other, and then a single, or one core of T/E to the rose for the light. And then they use the local neutral.

So what you then have, effectively, is this (switches not shown for simplicity)