"Earth" floating at 80-120V ... why?

Duncan Di Saudelli expressed precisely :

You really need to make all measurements from one reference point and that should be the consumer unit earth, so you probably need a long cable of some sort to do that.

The sort of numbers you are quoting suggests the voltage is simply induced from the live running adjacent to the earth. I would probably be prudent to check the resistance of the earth from all lighting switches and lights, to gain a better idea of what is wrong.

If the wiring dates back to the 1960's, it was not unusual for the lighting circuits not to be provided with an earth. Later mods and improvements to the circuits might mean metal boxes installed, but with no earth.

D'oh! Yes, an obvious contributor. Thanks - I shall pick up on your other points (common reference via long lead etc.) and start afresh tomorrow. I tyake your point regarding 1960s wiring and I shall look into the earthing of the lighting rings.

thanks for the feedback. Time for some sleep!

DDS

Another thing to consider is the sensitivity of the volt meter, The more ohms per volt, the more you are going to get confused.

Dave

For a start, lighting circuits ain't rings - they are radials.

How old is the house? Earths haven't always been required on lighting circuits. Even although TW&E may have been used to wire them.

Most likely cause is it either isn't connected or is broken somewhere. Especially where a ceiling rose has been replaced with an alternative fitting which doesn't have the required terminals.

Rather than lift floorboards etc, you can do plenty of investigation with a few simple tests first.

The lighting circuit will be a radial - with each lighting position connecting on to the next (usually at the ceiling rose). The last one will only have two sets of cables at the rose - one in, and one to the switch.

To check the earth continuity you can disconnect the live at the circuit breaker or fuse, and temporarily join it to the earth wire for the circuit. The using a DMM on ohms range measure the live to earth resistance at each lighting position (or switch). You should soon be able to find where in the circuit you have a disconnection. Chances are it will be at a ceiling rose somewhere - leaving the remainder of the circuit with a floating earth.

Which would still leave the neutral connected and the neutral connected to earth at the sub-station. One or more lights in the circuit left on and the live will have at least some sort of connection to earth - enough to cause some confusion when using that method.

There were two standard methods to wire the lighting circuits using T&E. The current method is from the Dis-board then lopped into each ceiling rose in turn, then down from the ceiling rose to its associated switch. The older method was similar to the above, but a joint box was used as the central connection point for each section of the circuit.

If wired using the later joint box method it could well be that the break in the earth is at one of these under floor joint boxes.

Yes, sorry I did not explicitly say - turn the CU off at the main switch.

(personally I would disconnect all the circuit wires and twist them together - then using a low ohms meter one can do a number of other checks as well as establish the circuit layout a bit quicker)

There are plenty more than two out there! Loop in is the most common, but JBs and looping through the switch positions is are also fairly frequently found.

Could be - but you should still be able identify in what section the fault is from tests on what you can get access to at the light and switch positions.

Thanks everyone for your advice and a few fault-finding pointers. I did quickly realise that the voltage measured was probably sourced from a high impedance and that measuring with a DVM with a 20M input Z would giuve unrealistic readings - hence my decision to measure the current and see it as only a few hundred uA i.e. a "leakage" current.

I'll start working round the ring - sorry, the radial wiring, thanks for putting me straight - and we'll see where we get to. Hopefully I'll find a floating "ned" which can then be tied down.

Thanks all.

DDS

Although more focussed on socket circuits, you may find some of this useful:

DDS

Would need to be tied down if it was floating face down in the Clyde, best place for it.

Guess what job I am doing later in the week?

I got a phone call off a bloke complaining the metal backbox in the new bathroom shaver point that he has fitted lights up his neon screwdriver. His further tests showed that the same happens in one of his bedroom lightswitch metal backboxes.

I am glad that he has taken his girlfriend away for a long weekend as his words "You do not get a shock when you poke the backbox with your finger" did worry me a little.

My first test will be an earth loop test on the new shaver socket.

My second test will probably be a resistance reading between the shaver backbox and any available earth (eg a screw at socket front)

Cheers

Adam

If he doesn't use the proper tool for one job ...

Owain

I thought these days one was supposed to use insulated sleeves when poking boxes of any orientation.

Bob

But they dramatically reduce the sensitivity of your probing equipment... ;-)

In message , ARWadsworth writes

So did you ask if she did ?

In message , ARWadsworth writes

I had something vaguely similar a couple of years back on my fathers home, 200 miles away from me and in the middle of nowhere, overhead power lines.. He was getting tingles of anything that should have been earthed. Kitchen sink, water radiators, metal clad switches etc. It was "interesting"

A rather fast 200 miles later........

Measuring between live and earth in the CU showed only a couple of volts.

At the CU there was a rather old earth leakage trip, one where the earth passed trough a separate unit to a local earth stake and any earth current caused the trip to break the incoming supply. This had failed and was happily still supplying the house with power. ( now replaced with something more substantial, looked like a token gesture and probably once relied more on the lead water pipe, also new CU with rcd now fitted )

If it had been the old days when he had a lead water pipe then it would have either blown a fuse or warmed up the water!!

The reason for all of this was that in his green house he had trapped a cable under a propagating frame and the VERY old rubber cable had disintegrated inside and the live was short to the earth. There was no separate earth to the green house it relied on the house earth via the SWA that connected the two. The couple of volts measured at the CU were I assume the drop in the SWA.

The moral of this is that a) equipotential bonding works! Fine in the main house, not so sure about the green house though with all its water and wet floors. b) don't trust elderly, 90 year old, fathers to look after their own electrics. c) there is a God out there looking after gardeners.

It was an interesting experience..................................

Glad you wrote that, just remove any ambiguity.

Dave

ROTFL

Nice one

Dave