Best way to establish a short

[John Borman]

Hi, After having rewired a lighting circuit using the loop in method, (Upstairs of which works fine) downstairs seems to trip a MCB. When all light switches are switched off the MCB is fine but as soon as one is switched on its trips.

I'm pretty certain I haven't cut through any sheaths but I "may" have damaged the cable when pulling some of it though.

What kind of short does this signify? (I know that if you do the classic of wiring all reds and blacks then you tend to get the light on and the MCB blowing when you switch off)

Ok I have checked all ceiling roses and all seem fine, but how do I go about testing the circuit to establish where the short is.

Thanks for any help J

[Stephen Dawson]

Does it happen with all the light switches or just one??

[Ian Stirling]

N-E short? When no current is flowing, there is no voltage across the N-E connection, so no current flows.

When the light is turned on, the current flowing in the N wire partly goes back through the E wire. This means L and N have different currents in, so the ELCB trips. It may well not trip if you remove the light bulbs, and turn on a light.

(assuming (as is the case in many homes) that E and N are bonded at some point)

[John Borman]

[Stephen Dawson]

What is it tripping ?? 6A MCB or RCD

[John Borman]

Its the RCD actually thats tripping not the MCB sorry I did say MCB meant RCD

[Stephen Dawson]

Oh dear, this is not going to be easy. You are looking for a short circuit between N-E. What work have you been doing ??

[John Borman]

[Owain]

"John Borman" wrote | After having rewired a lighting circuit using the loop in method, | (Upstairs of which works fine) downstairs seems to trip a MCB. | When all light switches are switched off the MCB is fine but as | soon as one is switched on its trips. ... | Ok I have checked all ceiling roses and all seem fine, but how do | I go about testing the circuit to establish where the short is.

The circuit should have been tested with a megger before being connected ot the mains. Disconnect and make safe the circuit cable at the CU. With all switches closed (ie on) (i.e. you will have to repeat the test for each permutation of two-way switches) and all lamps removed, megger the circuit (i.e. L-N, L-E, N-E all open circuit). This should tell you whether you are looking for an L-N or an L-E fault. N-E would trip an RCD but not MCB unless you have live and neutral reversed.

If it fails the initial meggering, which it should as it's faulty, go to the mid point of the circuit, which should be fairly easy on a radial lighting circuit, break into the circuit by separating the loop-in and loop-out and test each half separately (again repeating for each permutation of two-way switches). Repeating ad infinitum/nauseum will eventually get you down to a single cable.

An alternative and possibly quicker method is to open (i.e. off) all switches and remove all lamps, then megger the switched live (L1) against N and E at each switchplate. E will be present at each switchplate through the circuit protective conductor, but for the N (if it is an L-N fault) you will need a long lead that you can wander round the house with, connected to circuit N at some point

Owain

[Stephen Dawson]

You will be looking for a low resistance between the N-E conductors. If you can disconect the cable that you have pulled through and then test this you may confirm where your fault is. You did connect up the cable correctly did'nt you ??

[larry]

also make sure if its a split board that you have terminated the neutral to the right (correct) neutral terminal strip.

[Stephen Dawson]

Very true. Forgot that one.

[Dave Stanton]

Not to be nasty John, but you make up your mind wether you are going to top post and piss people off, or bottom post and make most happy.

Dave

[Steven Briggs]

In message , Stephen Dawson writes

Probably the most common error when wiring a split load box. Lights shouldn't (in general) be on the RCD protected side anyway.

[BigWallop]

The loop through must be coming from the consumer unit and going to each ceiling rose in turn, so open them up and disconnect between each. Test the links with a multi-meter to see which one is giving problems.

If the links between the consumer unit and first ceiling, and then each ceiling rose in turn down the line are OK, then turn your attention to each of the switches. Between the switches and ceiling roses should be tested to see if any of them have short circuit condition on them.

If the links between switches and ceiling roses are testing OK, then you turn your attention to the flex from the ceiling rose to the lamp holder.

[Lurch]

On Fri, 24 Dec 2004 19:56:15 +0000, Dave Stanton strung together this:

Actually, I've given up with this thread now, too confused with the random top\bottom and no no snipping.

[N. Thornton]

In message cqhlur$83m$ [e-mail] From: John Borman ( [e-mail])

There is. Disconnect suspected wires, all conductors, at both ends. Now test with a multimeter: each core should not conduct to anything anywhere. If it does, its insulation is gone. Not 100% but will pick up most such faults.

However it could still possibly just be miswiring: I'm sure we've all had the experience of staring a fault in the face and not seeing it

1st 2nd or 2rd time.

NT

PS top post, bottom post, intersperse.... Helps to snip though. Ya cant please em all!

[Dave Plowman (News)]

I've only got limited experience compared to some here, but I've never found an insulation fault that a decent DVM won't pick up. Although I do have a Megger.

[John Rumm]

Is this a split load consumer unit? If so which side is the lighting circuit connected to?

If not split load is there just one RCD for the whole CU?

[John Rumm]

as an aside, what are the current thoughts on on meggering lighting circuits that have modern 12v transformers (i.e. SMPS ones), and dimmer switches?