RCD tripping problem

[fido]

For the last week we've had our RCD trip two or three times a day. We can switch it back on immediately so we can't identify which circuit the fault is on. We're not doing anything when it happens so we presume it's something being switched on by thermostat. At present we are trying to find out what is doing it by a process of elimatation. Where possible we are unplugging things, keeping in mind not all 13amp sockets have double pole switches, and the fault could be neutral- earth. However it's like trying to prove a negative. Has anybody any ideas of any other way of finding out what it is, and why does the fault disappear so quickly? The consumer unit is fairly new, and the RCD covers everything.

Richard

[Owain]

(a) unplug everything and megger the fixed wiring to eliminate installation faults (b) test the RCD - it may be unduly sensitive

Whole-house RCDs are contrary to Regulations.

Owain

[Lurch]

On 18 Feb 2007 04:34:49 -0800, "fido" mused:

Could try testing everything, with the correct meters. May need to find yourself an electrician for this bit.

Think that may need attention, depending on what type and why.

[Andy Wade]

Appliance faults are more likely than fixed wiring faults. The usual suspects are immersion heater, washing machine and dishwasher heaters, fridges and freezers. If you can, try isolating those things in turn and see if the problem goes away.

Oh dear, single 30 mA whole-house RCD? That's no longer considered acceptable practice.

[fido]

Thanks for all the answers. It's rented property and we don't want to get the landlord involved until we can find out what it is in case it's one of our appliances. If it's the immersion heater then it's their problem.

Richard

[meow2222]

The logical thing is to check all your plugin appliances with a multimeter, which will pick up close to 100% of earth leakage faults, if not quite all. Test for any conduction from E to L&N.

NT

[Dave Liquorice]

Not quite true. They are required when the earth loop impedance is too high to allow overload devices to trip quick enough. In which case the whole installation RCD is 100mA time delayed rather than a normal 30mA "fast".

The OP just states "RCD" he doesn't give any further information.

[Andy Wade]

I was looking at a fridge-freezer yesterday: insulation resistance according to Fluke 77 multimeter ~ 6 Megohm, according to 500 V Megger <

100 kohm. (The leakage turned out to be in the defrost heater.)

This is nothing like as rare as you're trying to suggest.

[fido]

yes... my wife got caught out yesterday evening with the torch in the wrong place stil eliminating items

[Owain]

Sounds *very* painful; I do hope she's better soon, and I'll look out for the full details in the Sunday Sport.

Owain

[meow2222]

I didnt say what you describe is rare, I said failure to detect leakage is the exception. When testing appliances with a mulitimeter, write down the figures and zero in afterwards on the main suspects.

One always need be aware that leakage often worsens as V rises, perhaps I should have mentioned that rather than assume.

NT

[fido]

it is a 30ma trip and I'm still having problems