RCD Conundrum

securespark · 8 Feb 2007

Imagine an installation with an intermittent RCD tripping fault.

The RCD tested alone trips fine and tests OK. The ramp test is OK.

The EFLI on circuits is fine. And, indeed, the RCD trips intermittently.

So would it be possible to perform tests that would normally trip the RCD and find that they don't?

NuttallsSpark · 8 Feb 2007

If you have tested the circuit and its fine, and the R.C.D operation is fine. Then it could be a faulty appliance causing an inbalance every now and then, or one i found a screw for a socket plate was close to the phase and when people plugged anything in or knocked the cable if anything was plugged in it was enough to make to C.P.C touch the phase

Only advise is to unplug everything and use one item at a time until the fine the appliance (If any) which causes the R.C.D to trip

sorry i cant help more. R.C.Ds are super safe but also a super hassle Lol

bernardgreen · 8 Feb 2007

securespark said:
So would it be possible to perform tests that would normally trip the RCD and find that they don't?

In a word Yes

It can be set up that the test button has to be held in to keep the RCD ON and when the test button is released the RCD will trip to OFF.

How to do it ?

The test button connects a resistor from LIVE after the sensor to NEUTRAL before the sensor meaning current in the resistor passes through the sensor on the LIVE but does not return through the sensor on the neutral.

A resistor external to the RCD between NEUTRAL after the RCD and LIVE before the RCD will create a fault current to trip the RCD. If that resistor is about the same value as the one inside the RCD then pressing the test button creates a second fault current through the sensor but in the opposite "polarity" to the fault currrent from the external resistor. No un-balance in the sensor so no trip while the test button is held.

Some recent RCDs may have a means to prevent the RCD being reset while the test button is held down so this won't work on them.

securespark · 8 Feb 2007

Yes, this RCD will not reset when the T button is depressed (what about when it is happy?)

The RCD has been changed. The only fault found on the old one was the test button would not operate the device.

I cannot understand why the RCD trips intermittently when in circuit, tests fine in isolation, yet does not respond to RCD tests conducted from several circuits, even though their loop readings are healthy.

ELZ4742 · 8 Feb 2007

have all the appliances been disconnected?

securespark · 8 Feb 2007

The RCD tripping is not my concern. I can deal with that no probs. I'm just mystified as to why the RCD does not react to tests done via the circuits.

securespark · 8 Feb 2007

Just had a word with a colleague who is suggesting that if the cpc is linked to the neutral somewhere in the circuit, it will not trip.

Will think about this to see if there's any mileage in it.

scousespark · 8 Feb 2007

When you say it trips intermittently do you mean
a) repeated tests on one circuit causes it to trip sometimes, or
b) some of the circuits on the RCD do not cause the RCD to trip?

hope it's not answer a

securespark · 8 Feb 2007

There are two problems: One is that the RCD is tripping intermittently due to an unknown problem.

Yes. When you connect test gear to some circuits that are RCD protected, and then carry out a test expected to trip the device, it does not.

Does that make sense?

Thinking about the neutral/earth short as a solution, I don't think this will work, as there are 3 circuits that do not trip the RCD. A fault on one would not affect the others.

scousespark · 8 Feb 2007

The occasional trip has a number of possible causes (faulty kettle, accumulation of a number of small earth leakages, and so on). I think John D has a stock answer with a comprehensive list. Have you been on site when the RCD tripped?

I'd go through the full test schedule on the non-tripping circuits, to check their integrity.

I'd suggest you sort the faulty circuits, then hunt the trip fault.
Hope this isn't teaching granny to suck eggs mate.

JohnD · 8 Feb 2007

If it won't trip on test instrument, I'm thinking a compensating fault N to E.

Does your instrument just introduce a P to E flow? Could you temp rewire the test plug to flow N to E?

Is the N to E resistance unusually low, even when you disconnect the earth link in the CU?

edited: I see you've already had this idea

securespark · 8 Feb 2007

No, don't be

John, the only problem I have with your idea is that assuming there was some kind of N/E fault preventing a tester from tripping,

a) it would not affect other circuits &

b) it would not allow the RCD to trip (on the affected circuit at least)

Would you agree?

JohnD · 8 Feb 2007

(a) don't know... if all the circuits come together before going through the RCD, surely a NE fault on one would compensate for PE test on another.
the simulated fault from the Test button is usually much higher than the rated trip surrent.

(b) I suppose it would depend... I'm wondering if length of cable and resistance between compensating fault and the test point on other circuit would affect results. On a split load, maybe not. I'm trying to picture the effects of odd small leakages that might be scattered about the installation, some P-E and some N-E; and maybe some are intermittent, e.g refrigeration and heating or cooking devices; some cancel each other out and some add together.

Hooray for RCBOs!

bernardgreen · 8 Feb 2007

securespark said:
There are two problems: One is that the RCD is tripping intermittently due to an unknown problem.

If this is a neutral to earth fault then it may be continuously just below the trip threshold and occasionally going over the threshold and tripping.

Yes. When you connect test gear to some circuits that are RCD protected, and then carry out a test expected to trip the device, it does not.

The fault current created by the test ( assuming this is live to earth ) will be opposed by the fault current from the neutral to earth fault.

Does that make sense?

Thinking about the neutral/earth short as a solution, I don't think this will work, as there are 3 circuits that do not trip the RCD. A fault on one would not affect the others.

That is a bit difficult to explain unless there is a connection between the neutrals of these three circuits somewhere in the building meaning the one neutral earth fault affects the RCD whenever any one of these three circuits is connected to the RCD,

mapj1 · 9 Feb 2007

Or it could be a fault that introduces a effect that is a variation of the early D-LOCK, namely a DC imbalance that saturates the ring core in the RCD, causing it not to respond to a fault. Gently winding up a low voltage DC power supply between N and E, until you have perhaps 1/4 A flowing will lock on many of the traditional designs of RCD, enough that you can then do the 20A L-E earth loop test.
Not all RCDs will let you do this, and some will only let you do it one way (I mean perhaps only N+ E-, or only the other way). You don't say but are there thyristor lamp dimmers/heating controls or anything electronic like that downstream of the RCD when it refuses to trip (I presume people only report it tripped when something is being used, just not always the same thing.)
Alternately large circulating RF currents can fool the electronics in the ones that use hall efect devices to avoid being fooled by this, but I guess you'd have mentioned a large antenna on the roof.

diynot

If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.

Select the supplier or trade you require, enter your location to begin your search.

Are you a trade or supplier? You can create your listing free at DIYnot Local

RCD Conundrum

Similar threads