Mcb tripping

[JohnW2]

It's the first time I've seen this diagram and I confess my first reaction was to read it upside down, I thought the RCD was individual to the outgoing circuit. Then when I realised my error I didn't find it easy to follow the words and diagram together. Ironically it is completely correct.

Indeed.

Even more ironically this is a situation I have encountered several times and even witnessed 2 CU's being replaced due to exactly this problem, which did not cure the problem of course.

Yes, I don't think it is all that uncommon. Depending on the situation (wiring, loads etc.) it is not necessarily that more common for an N-E fault to result in an RCD trip when a load is applied to the circuit with the fault than to some other circuit (on the same RCD).

The much rarer (at least theoretical) phenomenon, which is a fair bit more difficult to get one's head around (and probably only happens, if ever, with heavy loads and an appreciable distance between cutout and CU) is for an N-E fault to result in a trip of an RCD which is not serving the circuit with the fault, when a heavy load is applied.

Kind Regards, John

 

[ericmark]

I have two old Wilex fuse boxes fed from two RCD's and it is common when one RCD trips when resetting it the other one will trip. I will assume switching causes a spike which trips the other RCD? Same when fuse box switched on with main switch, often trips RCD, but switch on one MCB at a time it's OK. The RCD's are around 1996 and no electronics the modern RCD is far better.

However in this case think we are looking for a neutral earth fault. As said since many switches only switch the line, switching off does not help. Also I have found even going around with a multi-meter the 9 volt or less used by the meter often can't detect the fault. My insulation tester went missing, however found I could get a cheap replacement at £35 so not too bad.

As to DIY well anything which can unplug then easy to test, but if it will unplug then no real need to test, just unplug it. Since the fused connection unit (FCU) with a switch often the switch is double pole then again switch off, no real point removing fuse.

Although we have earths to lights, most bulbs do not have an earth connection, so first thought is no point removing bulbs. However experience says do remove bulbs, I have found a bulb full of water, seems bath had leaked, leak cured, and lights left off for 3 months waiting for me to come home. (Working in Falklands at the time.) There was a hair line crack and bulb was full of water. So yes remove all bulbs.

If this still does not allow the kettle to be used then it likely needs some disconnection inside the consumer unit to locate fault, I personally feel if that is required it needs some one with skill, and test meters, it is beyond DIY. Since the main switch turns off both line and neutral with the main switch off it is quite easy to test all neutrals to earth, so likely one at a time removing and testing neutral wires is not that long of a job. Once you know 100% which circuit however, the tracing can then become a slow process.

I had it with daughters house, one by one I was moving socket to socket isolating parts of the ring, it turned out the fixing screw in one socket had caught the neutral, in the bedroom, and just walking in the room would cause it to make and break contact so lucky to find it. Also seen where spiders or mice have caused the problem. It took me around 3 hours to find the fault, it's just luck as to if first or last socket removed has the fault. But it did need the meter to find it, and so I would say once all unplugged and all bulbs removed, then time to hand over to an electrician.

 

[John D v2.0]

The much rarer (at least theoretical) phenomenon, which is a fair bit more difficult to get one's head around (and probably only happens, if ever, with heavy loads and an appreciable distance between cutout and CU) is for an N-E fault to result in a trip of an RCD which is not serving the circuit with the fault, when a heavy load is applied.

I'm not sure I can understand that. Perhaps you can go into more detail.
If a circuit is not faulty there shouldn't be any leakage downstream of the RCD to earth. If a heavy load does cause an imbalance, then that circuit surely is the one with the fault?
I can imagine some kind of capacitive or electro magnetic coupling having effects on the RCD itself perhaps.
Are you saying the RCD would legitimately trip on an imbalance in that circuit, or just be induced to correctly trip? And would it be removing a danger by tripping?

 

[JohnW2]

I'm not sure I can understand that. Perhaps you can go into more detail. If a circuit is not faulty there shouldn't be any leakage downstream of the RCD to earth. If a heavy load does cause an imbalance, then that circuit surely is the one with the fault?

I forget who it was that described this theoretical issue (here) but it certainly took me a long time to get my head around it at the time, and I have forgotten the precise details. I try to find the details and/or work it out again for myself.

As I said, as I recall it is very unlikely to be seen in practice, since it relies on very high currents flowing and a substantial distance between cutout and CU - I suspect that it may also only be possible with TN-C-S.

From what I recall, the theoretical/potential situation arises if currents being drawn by the installation are so high (and the length of the 'tails' etc. so long) that there is an appreciable VD in the neutral conductor between the CU and where (in a TN-C-S) it is connected to the MET at the cutout. However, I can't currently remember where the story goes from there! Watch this space.

Kind Regards, John

 

[JohnW2]

I'm not sure I can understand that. Perhaps you can go into more detail. ... If a circuit is not faulty there shouldn't be any leakage downstream of the RCD to earth. If a heavy load does cause an imbalance, then that circuit surely is the one with the fault?

I've managed to find the notes I made 'back then', and it transpires that what I recently wrote misrepresented the situation I was thinking about - and (although some people might still be surprised) what I should have been writing is more straightforward, and much easier to get one's head around. What I recently wrote (which, understandably, got you confused - or 'incredulous' !) was:

The much rarer (at least theoretical) phenomenon, which is a fair bit more difficult to get one's head around (and probably only happens, if ever, with heavy loads and an appreciable distance between cutout and CU) is for an N-E fault to result in a trip of an RCD which is not serving the circuit with the fault, when a heavy load is applied.

... whereas what I should have written was

The much rarer (at least theoretical) phenomenon, which is a fair bit more difficult to get one's head around .... is for an N-E fault to result in a trip of an RCD protecting the circuit with the fault when a load is applied to a circuit which is not protected by the same RCD.

As I said, some people may initially be surprised by that (I think I probably was), but the explanation is simple enough. If, now that I've explained what I should have been talking about, you still need an explanation, please just ask!

Apologies for having confused things a bit!

Kind Regards, John