RCD Tripping

[Nick]

Hi

This group has helped me before. I hope you can shed some light on this.

We had our house rewired and a new CU fitted about a month ago. Since then the RCD for the sockets has tripped about 5 times. The electrician maintains that it is one of our appliances as he tested the wiring after he installed it and the wiring is OK.

It happened again while we were away over Christmas. The only things on the RCD protected circuit were the fridge and two table lamps an timers. So it's got to be the fridge? Can it be tested? Is it something that electricians can do, or is it more of a fridge repair shop thing? Could it be something other than the fridge?

Thanks

Nick

BTW - Ive googled the archives and found some good info about E-N shorts, but I suspect this is different.

[Jonathan Pearson]

Is it possible to use an extension lead, and plug in the fridge on another circuit for a few weeks ? To test whether it is indeed the fridge? Jon

[meow2222]

I hope its not a fridge freezer youre putting on an RCDed circuit

illogical

of course. Best thing would be to PAT test all your appliances and fixed loads. Next best thing would be to megger them, but I dont recommend that for anyone not sufficiently knowledgeable. Practical householder thing to do is get a digital multimeter and measure resistances E to L+N on every appliance. You may find a leaky one.

As always, you need to find the fault by deduction. Anything elss is of little use.

NT

[Nick]

Good idea. Unfortunately all the sockets are on the RCD protected circuit. I guess I could bodge it up to the lighting circuit...

Nick

[Nick]

Yes it is. And yes all the food in the freezer was ruined. I don't really have any choice as the guy fitted a split load CU and put all the sockets on the RCD side. I'd rather have had a CU with separate RCDs for each protected circuit but he didn't ask and I didn't think to tell him.

Yes the kettle or washing machine seem much more likely - but we were on holiday and the fridge was the only thing on. (the heating is on the non-RCD side.)

But it has to be an appliance that was on at the time or the new wiring

- which means the fridge or the wiring. Am I correct??

Re. testing. The only equipment I have is a multimeter. I'll measure resistance as you suggest but I guess it will only leak when the compressor is switched on and I'm assuming that it's microprocessor controlled and will be off if the power is off.

Thanks for your reply

Nick

[Jonathan Pearson]

Mmm?? - not sure if the lighting circuit would take the load, and could make an interesting insurance claim if anything went wrong? - If it all went down whilst you were away then it would point to the fridge, but it could be the sockets aswell - when we moved into a new build a few years ago, one socket kept tripping, although they checked everything - it kept tripping every so often, in the end they replaced one of the cables and all was sorted!

You could also get one of these

it would show if anything obvious was wrong with the circuits

good luck Jon

[Nick]

I suppose a better idea would be to move one of the socket circuits to the unprotected side of the CU and plug the fridge into that for a while.

it would show if anything obvious was wrong with the circuits

That's interesting as, until it went while we were away, I was reasonably sure the fault was with the socket the kettle is normally plugged into (not the kettle as an electric mixer tripped it too) but I've got a similar thing to that tester - it says there's nothing wrong. And, as I mentioned, the electrician tested the wiring and says it's OK. I've no reason to doubt him but I wonder if there is a class of fault that his tests would not see?

Cheers

Nick

[Dave Stanton]

Get him to come back and fit a dedicated feed for the freezer, from the NON RCD side of the cu.

DAve

[Stephen Dawson]

I am surehe will come back and fit dedicated socket as a variation to the original job.

[Tony Williams]

Things don't have to be On to be the instigators of RCD troubles. An appliance with an N-E problem can sensitise the RCD merely by being plugged in.

Moisture/carbonising inside the kettle switch, or a build-up of crumbs in the toaster are good examples.

[Nick]

Now that IS interesting. Thanks very much.

Nick

[Nick]

I'd slowly come to that conclusion myself. I'm going to do it myself this w/e - will be easy as we have no carpets downstairs, yet!

Thanks

Nick

[Jim Gregory]

Agreed!

We have 2 freezers, a fridge, a dishwasher, a washing machine, a combi boiler, a kettle, a pop-up toaster, a hinged grill-pan, a m'wave oven, a Kenwood mixer/blender (and other, occasional hand appliances) plus TVs and PC gear connected to the 2 rings via RCD-protected MCBs, but not necessarily running together. I suppose this is a common domestic scenario.

No trips in 6 years caused by these loads, except mystery ones which turned out to be caused by a faulty (intermittent neutral-to-earth short) in an rarely used double-ganged socket in the lobby upstairs. Took a bit of sleuthing to pinpoint it! The 7kW shower is allocated its own 45A MCB off the RCD half of our CU. The radial to fan-oven unit has its own 32A MCB.

Jim

[Les Desser]

In article , Tony Williams Thu, 5 Jan 2006 09:54:15 writes

Considering the number of things that can trigger an RCD, I have never understood their use across several circuits, with the resultant problems.

I have always used RCDs on individual circuits as appropriate. The extra cost of about £25 per circuit is well worth it and a drop in the ocean in the context of an average project.

I would replace the relevant MCBs with separate RCDs and remove the master RCD.

[Andrew Gabriel]

Likewise. If you are DIY-ing it, this is an area where you can usefully splash out some of the money you saved by DIY-ing it.

You mean replace the relevant MCBs with separate RCBOs (combined MCB and RCD). Note that in the case of a TT system (own earth rod), it is likely all circuits will require RCD protection. However, in this case, the master RCD can be a 100mA (or more) time delayed type.

[Les Desser]

In article , Andrew Gabriel Sun, 8 Jan 2006 10:00:55 writes

Sorry - slip of the pen.

Why do you need a master RCD? Couldn't that still cause you some problems?

[meow2222]

unsatisfactory installation then.

so fit RCBOs, or run a new non-RCD supply for f/f and immersion.

again, test and see. Its a waste of time guessing.

no, off appliances trip RCDs, so do sockets, switches, cables, etc

possiby, so unplug it while the compressor is running.

Theyre normally bimetal thermostat controlled.

NT

[Tim S]

Because earth rods have a relatively high impedance path back to the substation. There's no guarantee that an earth fault will pass enough current to trip and MCB or fuse in a short enough time (remember a 32A MCB or fuse doesn't trip at 32.1A, it's in fact very slow for minor multiples of the rated current).

So hence the 100mA RCD.

That's why it must be time delayed to ensure discrimination - that is a fault on a non time delayed 30mA RCD protecting a socket must take out the

30mA RCD, not the 100mA one. RCDs aren't good at mutual discrimination, so it needs to be built in as an explicit time delay.

Tim

[tony sayer]

In article , Les Desser writes

In all of our rented houses we have everything on 30 ma trips. Over say the least 5 odd years only one duff cooker ring and immersion heater element have tripped the RCD's.

However on at least Three known occasions tripping was caused by accidental human contact with a live conductor!.

Of course if any where being newly installed these days the RCBO option would be used.......

[Les Desser]

In article , Tim S Sun, 8 Jan 2006 11:34:12 writes

[.....]

Thanks for the explanation. Living in London, I have not come across the need for an earth rod.