Changing light fitting: electrical advice reqd, please

[Matt Barton]

Perfect, thank you.

Rightho.

Sounds rational and sensible. Thanks very much.

[Dave Plowman (News)]

?

The description is pretty sparse, I'm afraid. What to look for is continuity buzzer.

I'm afraid I'm not well up on what's available as these things change near daily, and I'm not in the market for one.

A decent DVM is a very useful tool which will have many uses other than the immediate one, so I'd choose carefully.

I'd look for one which can also measure current up to 10 amps. And also temperature - very useful for balancing a central heating system, etc.

Ebay 'test equipment' usually has a selection both new and used.

[adder]

It's good to work out how things work but can't you simply wire up the new light exactly the same as the old one? i.e a switched live feed, a neutral and earth if necessary.

[Dave Plowman (News)]

You'd be amazed how many simply disconnect everything in sight when removing the old fitting without making any note of where things went. Even those with plenty experience. And then have to test for which ones are the switch pair. Trust me. ;-)

[Owain]

"RichardS" wrote | Yes. Absolutely. Definitely. And check that the voltage between live and | neutral is zero on both cables before you touch it as well. | I think that there is a safety procedure that you ought to go through when | checking that a circuit is in face isolated, and it goes something like: | switch off the circuit, check the light doesn't operate. remove covers and | fittings so that the connections can be reached, don't disconnect the wires | though. Check voltage is zero. Switch circuit back on again, VERY | CAREFULLY test the voltage again, check that you can get a 240 (or | thereabouts) reading (making sure that the DVM is on the appropriate AC | voltage scale!!!), now switch circuit off again and make sure the reading is | zero. I'm sure someone will correct me if this procedure is wrong. | Rationale is that you need to know that you have truly measured a zero | voltage and it isn't just that your meter isn't working....

The rationale is correct,but the procedure is incorrect. Your last test "now switch off circuit again and make sure the reading is zero" - you could have a failure of the test meter and not know it. If the switch also fails (or you turn off the wrong one second time around), you could have a zero meter reading and a live circuit, and not know it.

The correct procedure is:

Switch off circuit. Test meter on known voltage for live. Test circuit for dead. Test meter again on known voltage for live.

I find DVMs a fiddle for electrical work. Far better are the two-probes-joined-by-wire with a ladder of LEDs in, (I think mine is made by plasplugs) that show AC/DC with polarity in 6-12-24-48-110-240-415V steps. Does vehicle wiring, telephone battery, and mains up to 3-phase.

Owain

[Martin Angove]

Not picking on you specifically, but *please* can we avoid using the term "ring main" (or more correctly "ring final circuit" as I'm sure one of the network bods will point out) for anything except the ring-topology sockets circuit common in British houses?

In maybe 999 cases out of 1,000 lighting circuits are *not* a ring; they are a radial. They have no need to be a ring since common lighting cable at 1mm2 is capable of carrying over twice the normally fused limit for lights of 5 or 6A, unlike the 2.5mm2 used for rings which can carry at most 27A and yet is fused at 30/32A.

Two cables on a light fitting *usually* mean that this is the last fitting on the radial - one cable supplies the power, the other goes to the switch. Other lights on the circuit would have three cables; power in, power out, switch.

This misunderstanding of the "ring" has caused me immense problems recently; having to sort out the completely bodged work of a jobbing builder who had installed two new lights (yes, just two lights), wiring each one separately from the CU with *two* runs of 1mm2 cable in order to form a "ring" ("if you put everything on a ring then you can have as many fittings as you like" he apparently told the housholder).

As he had left it there were *seven* "live" cables in the 5A fuseway supposedly supplying the lights. It was actually fused at 15A and three of the seven wires wouldn't fit under the terminal and so had been layed in the bottom of the fuseholder socket. Of course one of the rooms in question - a bathroom - was completely tiled and so the cables were absolutely inaccessible.

Oh and the builder disappeared several weeks ago leaving no address or home phone number... of course.

Much as I'm against it normally, this is a classic case for the registration of all electricians :-/

Hwyl!

M.

[Ian Stirling]

Sorry, it was sloppy terminology on my part, I wasn't thinking hard.

I don't know about registration, but in this case, neutering would probably be appropriate...

[usenet]

I often use my really cheap analogue meter in preference to my moderately cheap digital one simply because it's better for "go/no go" testing on mains. If the analogue meter needle says there is 240 volts there then there *really* is 240 volts of the sort that will bite since it doesn't give readings due to capacitive coupling.

It's also better for contuity tests compared with the resitance range on the DVM simply because a swing of the needle is much easier to see quickly than reading a number from the DVM. The continuity buzzer on the DVM is OK but will buzz on quite a high resistance which may or may not be 'good'.

On the other hand the DVM is excellent for many other things that the analogue one is pretty useless for.

[Matt Barton]

Thanks everyone for your advice. I managed to get the light fitting changed today with 0 hassle and didn't blow myself up.

Mrs. is now happy with her new light, which means my ears will be happy. Thank you all very much.