Strange RCD fault any ideas?

[flyingsparks]

Recently changed a consumer unit for dual split RCD unit. Also fitted a DP switch to isolate an exernal socket where they have a pond pump and filter plugged in. It was spurred straight off the ring from a socket.

Anyway everything meggered ok so off I went.

Few days later they phone complaining that on a few occasions the rcd has tripped when operating the dp switch. Go back and find that it appears the pond pump is faulty so disconnected it. Outside socket and wiring meggering abut 8M. Also checked to see if there was a slight nick in the neutral behind the switch catching on the galv box when operated, all looks ok. Also put the meter on the circuit whilst operating the switch and it never picked up a short.

They bought a new pump plugged in and now after a few days it is tripping RCD again. It appears to only do it when switching the dp switch OFF. they have also operated the switch numerous times without the pump plugged in, all ok. I am going back on Saturday to take another look, but at bit of a loss now s the cause. I'm hoping the new pump is faulty, but just wondered if I could be missing something?

Cheers

 

[flameport]

The cause could be the DP switch itself - not being faulty but just the fact it is a DP.

Wire the switch as SP only (link the neutrals permanently), and see if that eliminates the problem.

 

[flyingsparks]

Cheers flameport. I guess your thinking if it is breaking one conductor before the other you could get a slight imbalance?

 

[JohnW2]

Cheers flameport. I guess your thinking if it is breaking one conductor before the other you could get a slight imbalance?

It's by no means the first time that I've seen such a suggestion, but I don't really get it. Unless you're going to invoke stray capacitance routes and/or the inductance of the cable (both of which would almost certainly be trivial), it is surely true that 'a circuit is a circuit' - such that interupting one conductor stops the current flow in both, simultaneously. Apart from anything else, if a DP switch breaking one conductor very slightly before the other was going to result in a problem (which I find hard to believe), a SP switch breaking just one conductor (and never breaking the other) would surely result in at least as big a problem?

What am I missing?

Kind Regards, John

 

[flameport]

The explanation, from the BEAMA RCD Handbook, which can be downloaded here:
http://www.beama.org.uk/en/utilities/document-summary.cfm/docid/3FA319A6-4159-4816-87188608E386D98D

9.3 Double Pole Switching

Double pole switching within the fixed wiring is known to produce a strange phenomenon
whereby switching OFF a double pole switch supplied through an RCD can cause the RCD
to trip. Single pole switching does not produce this effect, and it is known that changing
over from double pole to single pole switching can overcome the problem, where such
replacement is permissible and safe.

The phenomenon is explained by the fact that while capacitance between neutral and earth
will exist in all installations, the earth leakage through this capacitance will be negligible due
to the low (almost zero) potential between neutral and earth.When the neutral pole of a
double pole switch is opened, the voltage across this capacitance will suddenly increase,
with a subsequent increase in neutral to earth capacitive earth leakage currents. This
increase will be at a maximum if the neutral opens first and will be aggravated by arcing at
both switch contacts which will cause high frequency voltage spikes to cause even higher
neutral to earth leakage currents.The effect may be aggravated further by the slow break
feature of the switch often found in domestic a.c. switchgear.

Closing a double pole switch may also trip an RCD (as may closing a single pole switch),
but in this case any tripping is unlikely to be caused by neutral to earth capacitance since
the fast-make characteristic of the switch reduces both arcing and the time that any
significant voltage might exist between the load-side neutral and earth. Any circuit, which
incorporates only single pole switching, will never experience this type of problem since
(under normal operating conditions) the neutral to earth voltage is so low that the neutral
to earth capacitive earth leakage is negligible.

The problem appears to occur most frequently in installations where a separate consumer
unit and RCD are installed. The opening of the consumer unit switch will allow a higher
voltage (possibly 230V) to appear suddenly across the neutral to earth capacitance of the
whole installation.This can result in sufficient earth fault current to trip the RCD.

 

[JohnW2]

The explanation, from the BEAMA RCD Handbook ...

The phenomenon is explained by the fact that while capacitance between neutral and earth will exist in all installations, the earth leakage through this capacitance will be negligible due to the low (almost zero) potential between neutral and earth.When the neutral pole of a double pole switch is opened, the voltage across this capacitance will suddenly increase, with a subsequent increase in neutral to earth capacitive earth leakage currents. This increase will be at a maximum if the neutral opens first and will be aggravated by arcing at both switch contacts which will cause high frequency voltage spikes to cause even higher neutral to earth leakage currents. ...

Thanks. Interesting - so perhaps the stray capacitance is not as trivial as I suggested! However, this explanation presumably only applies if there is a substantial run of wiring (or some other source of N-E capacitance) downstream (i.e. on load side) of the switch concerned - since it is only that part of the neutral wiring which will experience a sudden rise of N-E potential when the neutral pole of the switch opens?

I suppose that, if one were experiencing this problem, one could test the above explanation. If, as you suggested, changing to SP switching (of L) cured the problem, then (only temporarily!!) changing to SP switching of neutral ought to cause the problem to come back if the above mechanism is the explanation.

Kind Regards, John

 

[bernardgreen]

so perhaps the stray capacitance is not as trivial as I suggested! However, this explanation presumably only applies if there is a substantial run of wiring (or some other source of N-E capacitance) downstream

There is the capacity between motor windings and the earthed metal in the compressor. Also the inductance of the windings will slow the rise of the current prolonging the pulse of unbalance making the trip of the RCD more likely.

 

[JohnW2]

so perhaps the stray capacitance is not as trivial as I suggested! However, this explanation presumably only applies if there is a substantial run of wiring (or some other source of N-E capacitance) downstream

There is the capacity between motor windings and the earthed metal in the compressor...

Sure - that could sometimes qualify as "some other source of N-E capacitance" with a load like a motor.

Also the inductance of the windings will slow the rise of the current prolonging the pulse of unbalance making the trip of the RCD more likely.

I'm not so sure about that. I would imagine that one's only 'hope' of getting a 30mA imbalance by the postulated mechanism would be reliant on 'high frequency' transients and, as you say, if you bring inductance into the equation they will be diminished, thereby presumably reducing the chances of enough L-N imbalance to trip an RCD.

Kind Regards, John

 

[JohnW2]

The cause could be the DP switch itself - not being faulty but just the fact it is a DP. Wire the switch as SP only (link the neutrals permanently), and see if that eliminates the problem.

If that solves the problem (and the mechanism is as we've been discussing), but one wants to retain DP switching, one solution might be to change to a make of DP switch (such as MK) which has 'make neutral first, break neutral last' functionality.

Of course, if one already has such a DP switch, yet has the problem, then one would have to question the mechanism of what's going on!

Kind Regards, John

 

[flyingsparks]

The cause could be the DP switch itself - not being faulty but just the fact it is a DP. Wire the switch as SP only (link the neutrals permanently), and see if that eliminates the problem.

If that solves the problem (and the mechanism is as we've been discussing), but one wants to retain DP switching, one solution might be to change to a make of DP switch (such as MK) which has 'make neutral first, break neutral last' functionality.

Of course, if one already has such a DP switch, yet has the problem, then one would have to question the mechanism of what's going on!

Kind Regards, John

And MK is exactly the make of switch I have. So where would you go from here?!

 

[JohnW2]

If that solves the problem (and the mechanism is as we've been discussing), but one wants to retain DP switching, one solution might be to change to a make of DP switch (such as MK) which has 'make neutral first, break neutral last' functionality.
Of course, if one already has such a DP switch, yet has the problem, then one would have to question the mechanism of what's going on!

And MK is exactly the make of switch I have. So where would you go from here?!

I suppose that was inevitable! Do I take it that it's a fairly recent MK switch? - it's the current Logic Plus ones which say that they have 'make first, break last' functionality for neutral, and I don't know whether that was the case with older MK switches.

Where would I go from here? I suppose the starting point is what flameport suggested (temporarily changing to SP switching) - have you tried that yet? If that doesn't cure the problem, one obviously couldn't blame neutral-breaking-first in the DP switch, so would then have to think further. The mechanism we've been discussing would not seem to apply (at least, not to anything ike the extent) on breaking the L.

How reproducible is this problem? Does the RCD trip most times the DP switch is switched off, or only very occasionally? If the former, I think the next thing I'd probably do would be to try some load (ideally an essentially resistive one) other than the pump.

Kind Regards, John

 

[flyingsparks]

Yes John, it is a new logic plus switch. I suppose the switch may be faulty and switching neutral off before the live but probably unlikely.

It seems to trip about once every 3 times or so, so yes fairly regularly.

Haven't tried doing just SP switching yet I'm going back on Saturday as its not very local to me. I will try a resistive load see what happens should there still be a problem. Cheers

 

[EFLImpudence]

You could try reversing the polarity.

 

[JohnW2]

Yes John, it is a new logic plus switch. I suppose the switch may be faulty and switching neutral off before the live but probably unlikely.

Just about possible, I suppose. If (and I'm not holding my breath!) changing to SP switching gets rid of the tripping then, as EFLI has suggested you could try going back to DP but with L&N reversed, just to make sure it's not switching in the opposite order to intended.

It seems to trip about once every 3 times or so, so yes fairly regularly.

That's something - at least it means that 'experimenting' shouldn't be too difficult.

Haven't tried doing just SP switching yet I'm going back on Saturday as its not very local to me. I will try a resistive load see what happens should there still be a problem. Cheers

It will be interesting to see what you discover. As above, if SP switching cures it, then I'd be inclined to try DP with reversed L/N. If you have another identical switch, you could obviously try swapping it as well. I'll see if I can think of any other ideas before Saturday!

Kind Regards, John

 

[flyingsparks]

Thanks for all your replies! Thought it was an interesting one to get your brains working.... And mine!

Will let you know what I find on Saturday, but any more ideas in the mean time will be appreciated.. Cheers