RCBO question

But did they all have a neutral?

Did you do a continuity test?

Yes, they all work and after I reconnected the neutral the situation returned to normality. I only removed one socket.

I thought this was not allowed because it involves entering the consumer unit.

You could have done a minimal test (assuming you have a multimeter) between the two neutrals you'd separated at the socket, would tell you if there is a hidden neutral break somewhere else ...

This is good knowledge. Thank you. Would you expect interrupting the neutral to activate the RCBO or is this a speculative question?

I wonder if you have something with the neutral bonded to earth, or connected to earth through say a capacitor?

If you have a break in the neutral and a low resistance load (heater or incandescent lightbulb), applying power to the live side could cause the neutral to go high voltage via the load. No current would flow due to the break, but the isolated neutral would become at live potential.

In another device there's a capacitive path from neutral to earth, perhaps an RF snubber in a SMPSU. Capacitor starts to charge, taking current from the live that isn't returned on the neutral at the RCBO, so the RCBO trips.

It would have to take some number of mA to trip the RCD (less than 30mA but could be say 10 or 15) which is a moderate power (2.4W at 10mA for some ms), but of course there's no guarantees this component didn't cook itself in the process.

Theo

N/E bonding (other than upstream of the CU) would trip the RCBO

EMI suppressors in appliances with non-trivial motors consist of a delta arrangement of capacitors (a class X across L/N and two class Y, one each from L and N to E) but if they get old and 'leaky' they'll tend to cause an RCD/RCBO trip, not prevent one ...

This is a ring main so every socket will have a neutral connection even if the ring is interrupted. It is now functioning okay (with the socket reconnected) so it seems highly likely that the problem was caused by removal of the socket leaving the live connected and neutral disconnected. Indeed, twisting the neutral wires together solved the problem for the time the socket was removed so it seems to be something to do with the neutral.

Could it be that two paths for the live and one path for the neutral confuses the RCBO?

Do you know whether the RCBO actually tripped for current imbalance, and it wasn't some spurious mechanical interaction tripping it, such as vibration from operating a nearby main isolator?

nib

That can be a problem when there are lots of similar filters on a circuit, but I can't see a broken neutral on a ring changing that behaviour particularly. Assuming there is only *one* break, then all the sockets still have a neutral connection. (hence other's comments about a continuity test)

A circuit can become "sensitised" when the normal leakage is approaching the trip threshold of the RCD - then it does not take much extra disturbance to push it over the edge.

Yes I do because the power was disconnected while I removed the socket and as I turned it back pm (to make a cup of tea) the RCBO immediately tripped. When the two neutral leads were reconnected the problem went away.

The RCBO (well its RCD section specifically) only really cares about current imbalance between L & N. A broken N on a ring should not change the current balance, although you potentially have a higher resistance neutral path due to the halving of the effective CSA of the wire available, and possibly longer wire run distances if the socket was close to one of the ends of the circuit.

Do you know what brand and type of RCBO you have?

Some makers RCBOs (like MK Sentry range) may include an additional "functional earth" connection flylead (often white / cream coloured).

This allows the RCD part to detect fault conditions like a loss of neutral connection to its *input* (a situation that would normally leave the circuit neutral sat at or close to line voltage since it would be "floating" and connected to live indirectly through any loads plugged into the circuit. The functional earth will allow the RCD to sense this and also give it a return path to draw enough current though to operate its trip mechanism)

As far as I can think (which may not be far enough) the only way that could affect an RCBO is if there is an unexpected extra path from neutral, to create an imbalance between L current and N current which the RCBO can detect ... do you get nuisance trips?

If I was swapping a socket and a minor works certificate is needed then I would actually do the end to end resistance readings at the socket and not at the CU. The results would be the same.

The only way forward in this investigation is to determine if the neutrals are are actually a ring.

This is genius. On the basis that both cables are linked at the consumer unit I could start at the socket for a continuity test. If the live and the neutral are both continuous and have similar resistance, I have the answer to my question. As the old saying goes, it's not what you know, it's who you know!

Are we certain that we haven't got a Earth - Neutral connection somewhere that can cause odd "ish" problems and what was the loop resistance should be of the order of .5 to .7 ohms 2.5 ring main conductor average three bed semi?...

Earth neutral bridges can certainly cause random tripping. Although if you have a broken neutral and a earth neutral short in the ring that would be more likely to increase the neutral to earth resistance - making the current diverted via the earth lower rather than higher.

Having said that I have found it is easy to cause a RCD trip on an neutral to "earth" fault when that earth is not actually connected to earth at all. A big enough lump of metal will normally have enough free electrons to look like an earth briefly. Had a cable with its fuse removed at the MCB cause a RCD trip by virtues of the neutral making contact with a metal stool base that was sat on a carpeted suspended wood floor.