RCD nuisance trip

[Andy Burns]

2 year old split load MK CU with 30mA RCD, PME supply, lighting circuits and outbuilding (with own CU/RCD) on non-protected side, everything else on protected side, no history of tripping.

Just had the RCD trip out, went down to investigate, it tripped straight out again, twice, so initially appeared to be a fault remaining somewhere. To identify which circuit I switched off all MCBs on protected side, the RCD then reset OK, so switched on MCBs one at a time expecting RCD to trip when I reached the offending circuit, but they all came back OK.

Of course now I wish I'd done it the other way round, switching *off* MCBs one at a time attempting RCD reset after each. Still seems strange behaviour though, will see if it re-occurs, any possible explanations for the sequence of events?

Luckily I was at home when it happened, if nothing else this serves as a reminder that I really ought to have a non-protected circuit using a spare way on the CU for fridge/freezer (perhaps central heating too) and to fit a rechargeable battery back in the radio alarm clock.

[meow2222]

couple of sections useful:

[Andy Burns]

general information being consolidated there, I indicated I already knew I'd be better served by an extra non-protected circuit, also it doesn't seem to have mention causes of RCD trips (nuisance or otherwise) I suppose it could mention damage to wiring, lamps blowing, damp ingress, EMI suppression filters, etc.

I was really questioning why simply opening and closing all the MCBs had allowed the RCD to reset when it initially wouldn't, everything else was scene-setting.

Thanks anyway.

[Andy Burns]

requirement to create an account to do so, so I suppose that's a clue, but is it considered good form to confer with FAQ section maintainers, or just contribute in good faith and let corrections get made as and when required? So far I've restrained myself to typo-swatting.

[John Rumm]

I think at this stage in its life, let rip and see how it goes. I would prefer edits were not made anonymously, but that is only my personal preference.

[Andrew Gabriel]

I don't think there's any generally accepted definition of a nuisance trip. I can think of several possible definitions spanning a very wide range of causes.

Possibly some part of the earth leakage pathed has been burned out now. Maybe the leakage is only marginal against the breaker's trip level.

[meow2222]

I was thinking of the discussion of RCBOs. Fitting them really is the best approach to these situations. Imho anyway. Article explains why IIRC.

NT

[Andy Burns]

Yes I re-read it while correcting some typos and realised it was discussing multiple RCBOs vs one or more RCDs.

[Dave Plowman (News)]

Snag is an MCB only isolates the line and an RCD will trip if a fault is in the neutral. Of course it gives you a 50/50 chance of finding out which circuit it is.

[Andy Wade]

Or perhaps something that got wet and has now dried out?

Sensitive "earth leakage" clamp meters like this

now being promoted as a good way of tracing faults causing RCDs to trip.

[Roger Mills]

Fascinating! Can you hire them - 'cos you could replace quite a lot of potentially faulty kit for the £300 cost of buying one?

[Tony Williams]

I made a cruder version of that a couple of years ago, when trying to trace an RCD trip. It had very little measuring accuracy, just an indication whether the unbalance current was larger or smaller than 15mArms.

Started off with one of those RCDs built into a 13A plugtop. Stripped all comps off the internal pcb except the Current Transformer and the two mains wires looping through it. Then terminated the secondary of the CT with a pair of diodes and a burden resistor (10k to start with afair). Then brought out a couple of wires for connection to a digital multimeter, ( on the 200mVrms range).

I wired a 15k resistor across L-E in a 13A plug and plugged that into the modified RCD as a 15mArms calibration current and noted the reading on the DMM.

It worked ok. I was surprised at the very low unbalance current for most items.

BTW: It did not help in locating the RCD trip because the fault was actually an N-E short that only produced a tripping unbalance during the transient startup surge of a heavy load, (say a vacuum cleaner, or computer, etc).

[Andy Burns]

I realise that things could have changed during my reset attempts, either by an appliance that I've not used since remaining on standby, or the wind changing direction, or self heating/drying from the leak itself, but the sequence of events was all within just a few seconds, and it definitely wouldn't initially reset with the MCBs on, but then it would reset with them off and didn't trip as they were all turned back on.

Useful to know, but excessive to trace a single episode ...

[John Rumm]

Ohh nice... handy they numbered it the same as the Honeywell prog stat to save having to remember another number. ;-)

I suppose the slightly cruder way I use to enhance the resolution of my clamp meter could equally well be applied in these circumstances...

For doing accurate(ish) low current measurements with my "ordinary" true RMS meter, I have a special short extension lead with about a meter of outer insulation stripped from it in the middle, and each of the conductors cable tied into ten turn coils. You can then clamp round one of the coils and read ten times the actual current flow.

One could knock up a jig that allowed a flyling lead to be taken to the CU, and the original leg of the power circuit taken to a junction box on the test jig, and it provides a number of pre-wound coils to clamp round.

[terry]

. Sounds similar to the reason that GFI (so called Ground Fault Interrupter) outlets (or the equivalent equipped circuit breaker) are not recommended here (Canada) for certain appliance circuits. They are recommended for wet and 'outside' locations. Reason given being that GFI operate on unbalance of current in the live and neutral wires Some appliances (fridges/freezers etc.) may have unbalances when starting. Nothing worse that going away and coming back to a freezer full of rotten food!