Four RCBOs trip at the same time - IR is good on all circuit

[martinxxxxxx]

I have been back today and tightened all the neutrals up, after giving everything a good wiggle. With two us on site we were not able to replicate the fault before touching any screws with wriggling wires at the incomer and at the service head.
I would not think that a neon could cause a problem in any way.
I have left the test light on and the other circuits with a light on, I will return in the morning.
We routinely check the IR between separate circuits and there wer not any problems there

 

[martinxxxxxx]

I have an update if anyone is interested......

I have been back and rechecked the IRs. Especially as the building was split in to two flats at some time. So with a long lead I checked every outgoing isolated conductor of the top flat agianst every one of the bottom flat and found......................

they were all good.

But since I had the board apart I also rechecked the top ones against each other and found a 0.6mohm reading between the lighting neutral and the socket circuit neutral.

So i went in search for it on Friday, this invloved crawiling around a rat infested loft (still living the the dream) I could not see any damage up there but it was getting mighty hot. So I decided to split the ring and get a clue as to where to look.

I had it narrowed down to the sockets in th on one side of the kitchen and another one in the hall and one in a bedroom all had conductors showing a low IR WRT the lighting N. So I took all the sockets apart and found that all the kitchen ones are wired in............(dramatic pause).,.......1.5mm TE!

At this point I ran out of energy and had to go to the pub.

Unfortunatly when I returned on Monday morning the IR had improved everywhere to a wopping 23Mohms.

So now I am back at square 1.

I put back the lighting circuit and it tripped again at 10pm (rats getting up time? perhaps)

So any clarvoiyants out there I need your help

 

[securespark]

So does this mean you have isolated the fault to the lighting circuit and you suspect a NN fault?

Or could it be a fault on the sockets, given they are wired in US cable?

I can understand this tripping 2 RCBO's, in the same way a borrowed neutral connecting two lighting circuits will trip the two associated RCBO's, but why 4?

 

[martinxxxxxx]

I can only promise you that I try to fault find in a logical and methodical manner. However this is really difficult when the readings change without warning. We have not had a drop of rain.
Those white tails out of the rcbo's might be the problem, say something is spiking the earth connection. As I understand it (waits to be shot down in flames) the white wires on the rcbo serve to provide an eath for the test button (it does not work if they are disconnected) so I have now as a test (and I know some of you are going to object to this) replaced the main switchfor an RCCB and put mcb's in across the board. This will at least show if the probem lies with the earth bar or a phase to earth fault. ie if it trips over night then I shall know that there is a fault to earth occuring and not a problem a)with share neutrals or b) with the earth bar spiking. if it does not trip then think I will have to look very carefully (ie rip out) the kitchen.

 

[JohnW2]

As I understand it (waits to be shot down in flames) the white wires on the rcbo serve to provide an eath for the test button (it does not work if they are disconnected)

I've never heard of that. The standard story is that the functional earth enables the RCBO electronics to still have power (so the device would still work - in the event of a 'vanishingly improbable' co-incidence of faults!) in the event of a 'lost neutral. AFAIAA, the test button usually just puts a resistor between one input and 'the other' output, nothing to do with earth.

Kind Regards, John

 

[martinxxxxxx]

As I understand it (waits to be shot down in flames) the white wires on the rcbo serve to provide an eath for the test button (it does not work if they are disconnected)

I've never heard of that. The standard story is that the functional earth enables the RCBO electronics to still have power (so the device would still work - in the event of a 'vanishingly improbable' co-incidence of faults!) in the event of a 'lost neutral. AFAIAA, the test button usually just puts a resistor between one input and 'the other' output, nothing to do with earth.

Kind Regards, John

OK why does the test button not work with the wire disconnected then? Sorry sounds sarci when typed but not meant to be
Also if the resistor was across the two outputs then surely there would just be a load and not an imbalance?

 

[JohnW2]

AFAIAA, the test button usually just puts a resistor between one input and 'the other' output, nothing to do with earth.

OK why does the test button not work with the wire disconnected then? Sorry sounds sarci when typed but not meant to be

I wasn't aware that test buttons didn't work without the functional earth lead connected to earth - in fact, I assumed they did. Indeed, I was under the impression that connecting them to earth was regarded by many as 'optional'! I'm off to do an experiment!

Also if the resistor was across the two outputs then surely there would just be a load and not an imbalance?

That's not what I said - see above. I said the test button connects a resistor between "one input and the other output" - i.e. between supply-side L and load-side N (which I think is the usual way) or vice versa. IIRC some devices have a diagram on them, which shows what the test button does.

Kind Regards, John

 

[JohnW2]

I wasn't aware that test buttons didn't work without the functional earth lead connected to earth - in fact, I assumed they did. Indeed, I was under the impression that connecting them to earth was regarded by many as 'optional'! I'm off to do an experiment!

Experiment done ... this obviously only relates to the one make of RCBO which I had to hand, but the test buttons of Wylex NSBS16-B/1 and NSBS6-B/1 RCBOs both work fine with the functional earth lead not connected to earth (not connected to anything).

Kind Regards, John

 

[martinxxxxxx]

I wasn't aware that test buttons didn't work without the functional earth lead connected to earth - in fact, I assumed they did. Indeed, I was under the impression that connecting them to earth was regarded by many as 'optional'! I'm off to do an experiment!

Experiment done ... this obviously only relates to the one make of RCBO which I had to hand, but the test buttons of Wylex NSBS16-B/1 and NSBS6-B/1 RCBOs both work fine with the functional earth lead not connected to earth (not connected to anything).

Kind Regards, John

Try the MK ones..I think they are different I need to do more experimenting too.

 

[JohnW2]

Try the MK ones..I think they are different I need to do more experimenting too.

Unfortunately, I don't have any MK ones to try, but maybe you do - it only takes a couple of minutes.

If what you are saying is correct (that the test button in some makes of RCBO use a 'leak to earth'), I wonder why. The test button functionality I described is obviously totally 'established' for RCDs (where one has no choice), so I wonder why some manufacturers (clearly not Wylex) chose to do it differently for RCBOs? As I said, I've never really thought about this before, and simply 'assumed' (I know!) that they all used the 'RCD method' for their test buttons. Some (i.e. Wylex) clearly do, and some RCBOs don't even come with functional earth leads, so they must as well.

I look forward to hearing anything you discover.

Kind Regards, John

 

[bernardgreen]

I did ask earlier if this was a TT installation ( earth from a ground rod ) or a PME type ( earth from the Neutral ).

I wondered what would happen if the electronics in the RCBO saw there was a potential difference between the "functional earth" and the Neutral. It would be sensible to trip the RCBO if there was a potential difference of more than a few volts as that would indicate a faulty Neutral.

If the potential difference at which the RCBO trips is a few volts and if a TT system then it may trip when the Neutral is not close to true ground potential

 

[JohnW2]

I wondered what would happen if the electronics in the RCBO saw there was a potential difference between the "functional earth" and the Neutral. It would be sensible to trip the RCBO if there was a potential difference of more than a few volts as that would indicate a faulty Neutral. ... If the potential difference at which the RCBO trips is a few volts and if a TT system then it may trip when the Neutral is not close to true ground potential

An interesting thought - but, if they did trip under such circumstances, I think we would be hearing many reports of RCBOs being 'unusable' in many TT installations (maybe even some TN-S ones) - and I've never heard that reported/suggested even once.

Kind Regards, John

 

[martinxxxxxx]

I did ask earlier if this was a TT installation ( earth from a ground rod ) or a PME type ( earth from the Neutral ).

I wondered what would happen if the electronics in the RCBO saw there was a potential difference between the "functional earth" and the Neutral. It would be sensible to trip the RCBO if there was a potential difference of more than a few volts as that would indicate a faulty Neutral.

If the potential difference at which the RCBO trips is a few volts and if a TT system then it may trip when the Neutral is not close to true ground potential

Sorry I have not responded to that question. I missed it
It is a a TN-S lead sheath on the lower foor looped up to the unstairs flat.

I think you are on to something here, because in the two days since I replaced the mainswitch with a 2 pole RCCD ther have not been any reports of outages.
SO I think we have either a intermittant connection of some sort between neutrals of the lights and the socket circuit (cause a complete mystery to me, and I have crawld around the loft inspecting cables and junction boxes.
or
Your idea is correct, having said that the flat below that the supply is looped from had exactly the same CU put in at the same time and they have had no problems at all. I have checked and have had tightend all of the connections inbetween. So I am not sure that it is possible? But I am open to ideas as my next action is a bit radical.