Finding cause of RCCB tripping ?

I have a standard 4 bed house with an outbuilding which has power fed to it. The electrics are in "generally good shape" (i.e. no rubber wire etc) and I have had a new split load consumer unit fitted with each side on an RCCB main switch, with the lights not on a RCCB and the fridge / freezer on its own RCCB, so that another circuit tripping does not take them out with it.

Occasionally, like two or three times a year, one of the circuit trips out - it might do this two or three times (re-setting in between) over a period of a couple of hours then be fine again for another few months.

The problem is that it this occurs so infrequently that I can't build up a sample of tests to narrow down where the problem is occurring. If I had a selection of specialist test equipment I guess I could measure each circuit and item to test for earth leakage, but only have "basics" like a multimeter, scope (and a working but basic old uncalibrated PAT tester).

Can anyone suggest a course of actions and tests to help identify the problem with what I have got, or can I hire the necessary piece of kit ?

Thanks

Nick

Reply to
nick smith
Loading thread data ...

You may actually have no problem at all. RCD's seem to trip on quite small imbalances, and I spent ages trying to ascertain whgat was causing mine to go - In teh end I put i a 100mA one and its been fine since, except

- tyhe washing machine debveloped an earth short in the motor

- I left and extension plugged into an outside socket, and it rained...

Ther is alwys some earh leakage: Its damned hard to identify what it is.

There are things you can do, like check for earth neutral shorts. One of these somewhere in the house will cause random tripping on a seemingly arbitrary basis.

The way to check this is to unhook the erath wire (on a dead CU of course) from every circuit and check it shows no neutral connection - use a sensitive analog ohmmeter - e.g. cheap dial based meter.

If those check out OK, I duno what else to suggest. What one really wants is a very sensitive clamp on ammeter that you can pass both live and neutral through and measure current impbalance, and then switch everything on and off to try and work out what is causing it. If only RCD's hade meters on them...

A little damp in e.g. a cooker can be the culprit, or voltage spikes coupled with RFI filters on e.g. TV's etc.

Reply to
The Natural Philosopher

hi nick, sounds like you have a minor fault, first off you may as well PAT test all the appliances cus you got the kit, mind you this is not foolproof especially on things like washing machine, because a fault could be present only on a certain cycle of the programe.

the next thing to do would be to carry out an insulation test of the fixed wiring in the house, an insulation test cannot be carried out with a multimeter, it requires an insulation tester, these units test at 500 Volts thereby putting extra strain on the circuit being tested showing up even the slightest insulation break down and or earth leakage. i wont bother giving any more details on this as it requires a sparkie to do it anyway.

how worried about it are you ? regards bob

Reply to
burbeck

multimeter, scope

youve got all the kit you need there! I pretty much agree with Bob, except I'd say one can make good use of low voltage testing as well.

  1. PAT test appliances (500v), assuming you know how to handle the HV aspect of the PAT tester safely. If you dont, test them with multimeter instead. Low voltage testing, while less accurate, still picks up the majority of faults.

Test resistance L-E and N-E. Write a list of the results for all appliances - this is important, as its comparison youre after. When LV testing the actual figures you get are often not a reflection of what the figs would be at higher voltage.

LV testing is not perfect, but it does normally pick out the culprit. When testing wash mc, if it has a mechanical program timer, run it thru the clicks while testing it. With cooker, switch each ring etc on in turn, and so on.

Again, only if you know how, its easy to test L-E resistance of the ring circuit without undoing anything: the CU MCB just needs to be off and all appliances unplugged, one can connect to L and E at a wall socket.

N-E can only be tested if you open the CU and disconncet the right wires, and I gather youre not happy about doing that safely.

I wasnt clear about your terminology. Best option is to put each ring on its own RCBO, and replace the RCD with a plain switch, if and only if youre not on a TT system.

This has 2 advantages:

  1. Power out only affects one ring circuit
  2. Less chance of trips due to the leakage being divided over several RCBOs.

Make sense?

NT

Reply to
bigcat

It doesn't meet the regulations.

30mA is the maximum acceptable leakage current in a domestic installation.

You mean the windings fused? Well that'd do it - should have tripped the MCB too.

And rain will do it too.....

White goods like washing machines are leaky anyway.

It isn't unusual to see leakage in the order of 10mA from a typical washing machine. You can expect similar from tumble driers.

Well, you could, save for the fact that if you've a TN-C-S system the earth and neutral are the same conductor anyway at the incomer.

That won't cause an RCD to trip by itself.

?

Why would you want to do this, precisely?

Nope. A clamp meter is unlikely to tell you much either - above and beyond the fact that there is or is not a degree of current flowing - it isn't the correct tool for the job, however.

What you really need to do is to start with some facts, some knowledge, and then use the correct tool to identify the problem -

- which is probably, in all likelihood the RCD itself.

I was suggesting some facts - an RCD is rated at a given trip current, e.g. 30mA.

Manufacturing tolerances (this directly from MK) are such that so long as that RCD operates within + or - 50% it is acceptable.

In simple terms, your 30mA RCD probably trips at nearer 15mA than

30mA.

The proper solution is to fit an independent 30mA RCD, or to use an RCBO rated at 30mA, for the circuit containing the washing machine.

Under pretty well NO circumstances should you fit a 100mA RCCD as the main switch in a domestic installation.

How?

Reply to
Sugar Free

Nah. The spec is: (a) shall not trip at 50% of nominal - so should allow

15mA imbalance indefinitely; (b) shall trip within 200ms at 100% of nominal; (c) (if a 30mA device, being used for supp. protn agin direct contact) shall trip within 40ms - two full mains cycles - at 500% of nominal. (This directly from the OSG, referencing regs 713-13-01 and 412-06-02; I'll see your MK tech-sales desk and raise you an IEE pub ;-) That said,

this is very much the case - to meet the above specs it's only sensible for MK and others to build for an expected trip around 20mA, so few will be oversensitive (tripping at 15mA) but practically all will meet the

30mA rating.

Agreed - the more circuits fed by a single RCD, the more difficult it is to track down which one(s) have the small leak(s) causing a nuisance trip, *and* the more little leaks accumulate from multiple appliances (particularly anything with an interference suppressor with capacitors L-to-E and N-to-E, as most have).

Well... TT installations aren't *that* rare, expecially out in t'sticks, and 100mA-time-delay RCDs are just about always needed there, since it's really really really REALLY rare that you can guarantee a sub-one-ohm earth impedance with any economically feasible earth rod(s)...

Stefek

Reply to
Stefek Zaba

May not be relevant to your case but I've just pinned down an infrequent RCD trip to the washing machine. The motor brushes were very worn and the motor had a lot of graphite dust around the communtator, brush housings, and field coil connectors. New brushes and a thorough clean up seems to have cured it.

Chris K

Reply to
CJK

Not necessarily time-delayed; the main-switch-feeding-two-RCDs-in-parallel method has much to commend it.

On a public mains supply you might just as well write "NEVER" there. The impedance to earth on the supplier's side is allowed to be as high as 20 ohms (usually is much less of course) - giving a max. Ze value of

21 ohms when you add on a notional one ohm for the transformer and metallic conductor impedances [see sec. 1.1(iv) in your OSG].
Reply to
Andy Wade

HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.