Fault finding a tripping RCBO

This happened to me, except it was the main trip. In the end I called in a professional, he found that it was a fault with the trip itself, replaced it and has been fine since.

When you disconnect the ring to fit the new RCBO, measure the resistance between all the conductors with a DVM. That will generally show a fault large enough to trip an RCBO, if you don't have a megger. It will also show if there's a load on the ring you've not found. ;-)

Remove a socket and disconnect the ring 'halfs' about half way round the ring and check the resistance of each leg again. That will hopefully show what half of the ring the fault is in.

Favourite appliances to give trouble are cookers, dishwashers, washing machines and kettles. Mainly heating elements. Make sure you have isolated both poles for each of these before ripping out sockets.

In message , Chris French writes

Well, it looks like it's the RCBO, or rather the connections to it.

There has been overheating at the outgoing live connection from a loose terminal screw, which has caused localised melting of the plastic, and presumably some sort of internal failure.

down to me I guess, as installed it, though I'm sure I would have triple checked all the terminals. all the others in the CU seem fine.

Noooooo! You blame the gremlins[1]. The buggers get everywhere.

Has the heating caused any carbonisation of the insulation? Could be conductive.

The was some damage to the cable insulation near to the end of the cable, I've cut back to new, undamaged insulation.

In message , "Dave Plowman (News)" writes

Hmm, well, so far resistance checks seem ok.

Resistance round the ring from the CU is about 0.9 Ohms on L, N and E. (cable is a mixture of old multi-stranded cable and more recent solid core)

Between the various pairs of conductors the meter reads high resistance (about 400 Ohms).

I thinking there might be a leakage to earth somewhere? (damp in socket connection or something)

anyway waiting for tomorrow and daylight.

Unlucky :(

You prompted me to go and check a 45A water heating circuit where the isolator has been running warm (filling a bath so 30 mins run time) for a bit.

One of my terminals, wot I did, was not exactly loose, but accepted another 1/4 turn rather more easily than it should. The one that's round the back near the wall, so not quite as easy to get a screwdriver onto.

I do not much like large CSA T+E - that's getting done next time[1] in singles in tube which is a little more manageable.

I did give my CU a cursory check a few weeks ago and all was solid - Hager seem good in that respect. 'twas Adam's recommendation :)

You could test to the all conductors to something metal that buried in the wall. I suggest not testing to any pipe or electrical box (unless the box is isolated from the supply earth) as with PME you might get some stray voltages between supply earth and true local earth which will play havoc with your multimeter readings.

Hmm - still running warm (feels like 40-45C right round the switch toggle) after 15 mins.

Heat's coming from the switch element I reckon.

I wonder if that's bad? Anyone know. I've not actually been in the habit of feeling high load isolators, strangely - just expected it to be stone cold.

Replace the bloody thing with a good old fashioned fuse. I never have these problems.

It only takes a tiny resistance at that current to start dissipating 20W or so... That's going to cause quite a temperature rise on a small bit of metal. (think how hot a laptop CPU would get without its heat pipe and sink)

This is the isolator. Not the breaker.

That's miles too low. Is that consistent between line and earth, neutral and earth and line to neutral? (This is with the ring totally disconnected from the CU at both ends.)

En el artículo , Chris French escribió:

My interpretation of that is that you're seeing ~400 ohms L-N, L-E, and N-E.

No one else's commented on it, and it's 5am so my mind isn't exactly clear, but that doesn't feel right to me. That would be expected between L and N if you haven't disconnected everything from the ring (i.e. you're seeing the various connected loads), but not between L and E?

I wonder if you have an N-E short somewhere. Dave's suggested binary chop would be the way to go here - split the ring approximately in half at a socket and test each half.

You're right John. Even at 1 milliohm, it would be kicking off nearly 2W in a plastic moulding. Anyway, it's not looking "cooked" so I guess it's fine.

I did. But asked for clarification as to what was being measured, and how.

Ie, was the ring being measured disconnected from the CU totally.

If it was, I'd expect to see many 10s of thousands of ohms. It's obviously not the best way to test - but can show up a simple fault if you don't have the correct test gear.

Again, something very wrong there. If you were talking pure DC, you'd be looking at a current draw of some 1.7 amps with no load connected. Just to remind myself, I've measured a near full 100 metre drum of 2.5 TW&E. No reading at all - so many megohms.

Yes.

My guess is there are still things plugged into (or wired to) the ring. If you didn't install it yourself, it might be feeding things you'd not suspect.