Weirdest lighting problem ever

[llykstw]

I challenge anyone to solve this problem! I'd be grateful for any input any of you have - I think that it's safe, but it'd be a comfort if somebody could confirm for me.

I've put a new light into an upstairs bedroom - nothing complex involved, simply remove the old fitting and install the new one. Whilst I was at it, I decided to remove the two switches and check the wiring was ok (it's a two-way set-up, with pull cord switch as well as a normal rocker-type switch).

The circuit isn't exactly ideal - it's not a ring, nor even a radial, but simply seems to be a chain circuit. Once I'd removed the light fitting from the bedroom ceiling and disconnected the three sets of twin-and-earth cable coming into the ceiling rose, the bathroom light next door stopped working, and only one of the three twin-and-earth cables coming into the ceiling rose showed live using an electrical screwdriver on the live core. If I connected the mains feed twin-and-earth cable to one of the other twin-and-earth cables (live to live, neutral to neutral and earth to earth) then the bathroom light started working again. If I connected that mains feed twin-and-earth to the third twin-and-earth cable at the rose, then the live core at the pull-switch showed up as live. Therefore, at the ceiling rose, there was a twin-and-earth cable carrying the mains feed, a twin-and-earth cable running to the bathroom light, and a twin-and-earth going to the pull switch. A three-core and earth cable then ran from the pull switch to the wall rocker switch. All straightforward so far?

When I connect the system in the normal way, it works fine. Also, the RCD circuit breaker on the consumer unit definitely works - touching the live core of the mains feed to the neutral core or earth core on the same cable trips the circuit.

BUT - if I connect only the live core from the mains feed to the live core going to the pull switch, but connect no other cores (i.e. the neutral and earth cores going to the pull switch are connected to nothing at either end), then, using an electrical screwdriver, both the neutral and the earth cores going between the ceiling rose and the pull switch test as live using an electrical screwdriver. The same is true if I connect the live core of the mains feed cable to the live core of the cable going to the bathroom - the associated neutral and earth cores both test live using an electrical screwdriver. The same happens if I connect the live core of the mains feed to the live core going to the pull switch, and at the pull switch connect that live core to the live or common core going to the wall rocker switch. All disconnected cores also show live using an electrical screwdriver.

My first reaction was to assume that I'd shorted out the cables somewhere - if they're connected to nothing at either end, then how can making one of the cores live make the other two live as well? However, testing with a multimeter showed the following:
1. The voltage across the live and neutral cores, and the live and earth cores, of the mains feed cable is 235v (as one would expect).
2. If you connect the live core of the mains feed to the live core going to the pull switch or to the bathroom light, but no other cores, then although the electrical screwdriver tests the neutral and earth cores at the pull switch as being live, there is no voltage drop between the neutral and earth cores in the pull switch or bathroom cables and the neutral or earth cores of the mains feed cable. Furthermore, if you connect the neutral and earth cores that show live to the mains feed earth or neutral cores, there is no short, and the RCD does not trip.
3. Testing for continuity, there is definitely no short across the cores of any cable.
4. Although the disconnected neutral and earth cores show live using an electrical screwdriver, if you then connect them to the neutral and earth cores of the mains feed cable, then they no longer show as live with the electrical screwdriver.

All of this has me totally confused. How on earth does an electrical screwdriver test disconnected cables as live? Any advice would be gratefully received!!!

 

[breezer]

simple, electrical tester screwdrivers are great...........as screwdrivers, nothing more.

as you have found out they lie, this is probably because you get induction from the live cable to the other, a tester screwdriver needs very little to make it operate.

i used to have one years ago, i was sad when i lost it as it was great for undoing a particular sized set screw, oh the guts, soon as i got it took them out and trew them away

 

[BR]

That’s some post!

A few points:
The lighting circuit should not be connected to an RCD.
You should NEVER test an MCB/RCD by shorting out the load.
Your "electrical screwdriver". As a youngster, an electrician took one off me and as we were passing a bridge, dropped it in! I suggest you do the same with yours!

Your problem with the cabling, I think we could narrow it down to two things. You have not properly identified the cables, you are dealing with a different cable at each end

OR

Your 'electrical screwdriver' is faulty, (depends on the type you have) Unlikely in the case of a phase tester. One things for sure, you don’t have a live cable, connect it to something and suddenly it's dead. Highly unlikely in a domestic installation anyway. Bear in mind that the current required to operate a phase tester is pretty minimalist, it is possible that a thing called induction is causing your phase tester to light.

 

[BR]

Sorry for steping on your toes breezer, my page, yet again was not not up to date. I got called away to do something and then came back!

 

[altec]

Are phantum voltages and induction the same thing, what are both??

 

[breezer]

BR, no worries two opinios are better than one.

--------------------------------------------------------------------------------

Altec
in electrical terms a moving magnetic field indueces a current in a wire it is next to.


basic example

take a transformer for example.

it has two coils of wire one next to the other (no electrical connection)you connect one to the mains which generates an alternating magnetic filed, this magnetic field induces a current in the 2nd coil.

depending on wire size and length of each coil, and input voltage determines what you get out

 

[AdamW]

Are phantum voltages and induction the same thing, what are both??

Phantom voltage is a way of powering a device via a signal line. The best example is a microphone. A typical microphone lead has 3 conductors. 2 signal conductors (pins 1 and 2) and a ground conductor. This is known as balanced line. At the mixing desk end a DC voltage (usually 48V I think) is applied to pins 1 and 2, so they are both at +48V with respect to the ground. The microphone can then draw current and thus power from this. The resulting voltage drop is equal in both conductors. Then the simple cleverness of balanced line kicks in:

As pin 2 carries the same signal as pin 1, but at negative voltage with respect to ground (i.e. a mirror image with ground in the middle), the mixing desk just inverts it to positive and adds the two signals together. The 48V dc then disappears as you are adding +48V to -48V and you get 0. However the signal remains.

I hope that helps.

 

[AdamW]

Actually, thinking about it I didn't describe a Balanced Line there. A balanced line still holds a phantom voltage in the way I described but instead it uses one signal wire to carry the signal and the second one is connected to ground and just picks up the interference. This interference signal is then subtracted from the "actual" signal.

I have to say I prefer the signal line I described first of all, but that isn't the one that is used in industry so there must be a downside to it.

Just didn't want someone else to point out my error there!

 

[llykstw]

That’s some post!

A few points:
The lighting circuit should not be connected to an RCD.
You should NEVER test an MCB/RCD by shorting out the load.
Your "electrical screwdriver". As a youngster, an electrician took one off me and as we were passing a bridge, dropped it in! I suggest you do the same with yours!

Your problem with the cabling, I think we could narrow it down to two things. You have not properly identified the cables, you are dealing with a different cable at each end

OR

Your 'electrical screwdriver' is faulty, (depends on the type you have) Unlikely in the case of a phase tester. One things for sure, you don’t have a live cable, connect it to something and suddenly it's dead. Highly unlikely in a domestic installation anyway. Bear in mind that the current required to operate a phase tester is pretty minimalist, it is possible that a thing called induction is causing your phase tester to light.

Thanks for everyone's advice above!

Sorry - in my post, read "MCB" for "RCD". The only RCD's I have in the house are one running to the exterior sockets in the garden and one running to power the whirlpool bath. Lighting circuits come from the consumer unit that uses MCBs.

I also agree about not deliberately shorting to test the MCB. In my defence, I was desperate Being concerned about the integrity of the circuit, it seemed prudent to make sure that the circuit would at least trip the MCB if it was shorted, and I couldn't think of another way of testing it. It's not something I'd do routinely though

Cables are definitely correct - I tested pretty exhaustively. My first thought was also induction, but I've never come across a situation where inducted voltage was sufficient to activate an electrical screwdriver bulb - I'm very reassured that you people have I had a similar problem wiring a light elsewhere on the same circuit, so maybe the cable that was used on wiring the circuit gives rise to induction on neighbouring cores for some reason.

Still, I've learned to always use a multimeter rather than the electrical screwdriver when wiring!

Thanks for all the helpful advice.

 

[AdamW]

Just a thought: if this was caused by induction then you would get electrical test screwdrivers glowing whenever you touched the earth wire. Has anyone ever found a correctly wired twin and earth where the earth is live? I found a live earth in my parents' house in a twin pole light switch but that was traced to the previous owners getting in cowboys to do the wiring in an extension.

With a long straight wire (as in this situation) the resistance of the copper would be enough to render any induced voltage in the earth conductor as effectively 0. So I doubt it is induction. If induced voltage was significant in this case then the neutral conductor would suffer from the same problem and the voltage across live and neutral would be reduced to near zero. Also running two cables right next to each other could cause interference resulting in voltages anywhere between 0 and 480V. And running a speaker cable parallel to a mains wire would result in your speakers catching fire instead of the mildly annoying 50Hz hum that can result.

So to sum up: I will stake my electrical test screwdriver on this NOT being induction! And not just cos I plan to throw it away and buy some proper kit after reading what others have written about them!

 

[securespark]

Why not eliminate the screwdriver by testing the same cables with your new multimeter to see if you get a voltage in N & E?

 

[ban-all-sheds]

...The circuit isn't exactly ideal - it's not a ring, nor even a radial, but simply seems to be a chain circuit. Once I'd removed the light fitting from the bedroom ceiling and disconnected the three sets of twin-and-earth cable coming into the ceiling rose, the bathroom light next door stopped working, and only one of the three twin-and-earth cables coming into the ceiling rose showed live using an electrical screwdriver on the live core. If I connected the mains feed twin-and-earth cable to one of the other twin-and-earth cables (live to live, neutral to neutral and earth to earth) then the bathroom light started working again. If I connected that mains feed twin-and-earth to the third twin-and-earth cable at the rose, then the live core at the pull-switch showed up as live. Therefore, at the ceiling rose, there was a twin-and-earth cable carrying the mains feed, a twin-and-earth cable running to the bathroom light, and a twin-and-earth going to the pull switch. A three-core and earth cable then ran from the pull switch to the wall rocker switch. All straightforward so far?....

Perfectly straightforward, and perfectly conventional and perfectly acceptable.

Why do you say it's not exactly ideal?

 

[securespark]

Better turn my hearing aid down in the dinning room!!!! Only kidding!