Thanks, I was hoping it would be something like that. I know that sometimes testing with a digital meter can be misleading, but felt it would be too optimistic to assume it was down to that.
The meter suggests that the metal faceplate earth is connected to the earth at the ceiling rose, and also the faceplate on the other side of the room. I think I'll inspect a few roses, and see if anything suggests itself.
Say hello to Mr Hoppy for me.
If you make up a long test lead extension, you should be able to test from one light fitting to the next (you will have to guess at a likely wiring order), but you may be able to narrow it down to one pair of light fittings where the earth connection is open circuit between them. Needless to say, any other fittings downstream of the break will also appear to be disconnected from earth even if they are connected to each other, creating an island type effect.
It's more likely that the leakage is capacitive, resulting from the capacitance between the line (L) conductor in the cable and the floating earth/CPC.
Mr. Hoppyyyyy!
It's actually a very common feeling - but not on mains switches etc.; that is odd. As others have said, suspect the earth connection.
Any time there's a tiny amount of capacitive coupling from the mains, e.g. via one of those 'wall wart' adaptors, to something that isn't earthed because it's supposedly insulated, you get that feeling.
The amount of current that flows is very small and wouldn't hurt you (unless some serious failure occurred). But it might well be enough to light up one of those neon-in-a-screwdriver testers.
I think that if the earth on the switch box was good, but you were standing on something which wasn't at earth potential as it should be (damp stone floor in contact with a damaged cable, for example) you might get the same effect. Or worse.
It does, but it dims after about a second.
On Monday 03 June 2013 21:25 Ivan Dobsky wrote in uk.d-i-y:
OK - you have earth faults with your system, somewhere...
I would strongly recommend a PIR (installation test) - not hugely expensive in the grand scheme of things. That will discover all manner of problems that are hard to find by other means.
Or you could have a look behind the plates (power proven off at CU of course!) and see if there's anything obvious like the earth wire's fallen out or is missing.
That's just capacitive leakage because to have no earth. However, if a live wire falls off inside and touched the screw or backbox, you'd then have a lethal situation.
The nature of the unexpected voltage in this case is less significant than the fact it has highlighted that you have a disconnected earth. So all in all a much better way of finding out that by it not working as an earth when really required!
This is measuring the resistance or continuity? So more than one light switch has no earth connection, relative to a nearby socket? Is there just one lighting circuit in the house, or more than one? Can you measure between two lighting switches by extending a test lead on the meter? It won't be 100% accurate, but will show if there is a connection or not.
Sadly, finding out just where the break is can be very time consuming. I'd start by opening up the CU to see if the earth is actually connected to the lighting circuit there.
I was measuring resistance between nearby faceplate screws - I don't know another simple way to test continuity. The two nearest the CU do seem to be earthed (there is no resistance between them, and nearby mains socket faceplate screws). After that (I think I know what the sequence is, and it doesn't appear to be a real loop, since another rose near the CU doesn't has only two cables), they do all seem to be connected to each other, but not to the last known earthed lighting faceplate, or any nearby mains faceplate.
If I discovered that one of the earth wires in the rose for that last known earthed faceplate had come out, then that would explain it perfectly. But I won't be able to check till I get home later.
Lighting circuits aren't a 'loop' or ring like the sockets. They are a radial circuit. So a break in the earth will remove it from all downstream of that break.
I dunno if you know the history of this house since it was originally wired. Does it use the sort of ceiling rose with all the connections within them (loop in loop out)? If so, I'd first look for where one is missing - replaced with perhaps a fitting mounted on the ceiling. Many of these are of non UK origin, with no provision for an earth.
Ah, I didn't know that. The roses do have the loop connections (like terminal blocks) in them, and judging by the fact all except one have three cables in them, I'd guess that my lighting circuit is in a C shape. I hope it is - it would make the break easy to trace.
House was built in the 80's so these things are modern-ish.
The problem is hopefully solved. My downstairs lights seem to be daisy-chained together in a C shape. I opened up the rose of the last one that had a good earth, and although there were three earth wires in the sheath, only two were screwed into the terminal. Just being in the sheath was not enough to make good contact (even though I'm sure there must have been at least a little contact). So, the rest of the fittings down the chain were not earthed.
Anyway, thanks all for your help and suggestions. It just shows that you don't know what's lurking even in your own house.
So the topology sounds like standard loop in. Power feed goes from the MCB in the CU to the first ceiling rose, then from that to the next and so on. Each rose has an additional connection to the switch. So three cables to each rose except the last, which has no "out" cable.
Chances are you will find the break with a visual inspection in the ceiling roses, however if not, repeat your test with the meter, looking at the resistance between one rose's earth terminal and that of the next one.
Result. Probably been like that from new, and not tested thoroughly.
A nail through a cable which breaks the earth only can take much more finding. ;-)
I found it, quite early in the loop. Thanks for your help!
Probably would never have thought to even look for it, if it hadn't been for this switch problem.
Sounds like fun.
Well done!
Doesn't suggest that to me. If it was live, a static finger should feel it, and there would not be a grain direction effect. Sounds to me like a mechanical phenomenon. Easy check would be a DVM (on AC volts setting) between the plate and an earth. Typical DVM impedance should be high enough to show some voltage if the leakage current is enough to feel.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.