Fault Find - advice please!

[MikeKing]

Team - I'm after some help on a fault find that I was at this week and to work out if there's something obvious that I've forgotten, overlooked. I've only been working as a spark for a couple of years, done my city & guilds 2365 - L2&3 and done my AM2, just wanting to finish my portfolio for my NVQ - (just ome back ground to get me out of the DIY forum).

Customer reports RCD trip on down stairs ring circuit, MK split load board, half RCD protected, and half isn't. After a lot of messing about working out what was and wasn't on the circuit, eventually removed all the loads (found a few hiding) and IR tests commence. RCD is a MK 80a - 3oma, fault appears to be intermittent but there is no issues when the down stairs ring is switched off - so I'm happy that is where the problem lies.

L-E 530 MΩ, N-E 383 MΩ, L-N >999 MΩ- I measured this at the board and at a socket on the down stairs ring.

There is a brake in the ring on the line conductor, which I think is a historic problem, contained within one of the spider web of junction boxes underneath floorboards upstairs, which are underneath a very expensive floor. I have explained the need to trace it. – however the customer has sold the house to a developer who is only a couple of months away from levelling the place.

RCD test - trips @ 1/2 rate when testing on that circuit. So I disconnected the down stair ring and ramp test it from upstairs - trips @ 19ma. I thought it was a bit sensitive, so swapped it out for a new one. Same story, however the RCD holds now and does so for some time. 4 days in one instance.

But when the loads start to be plugged back in – The only thing I had in was the fridge freezer and the TV (which is in three way adapter thing) virgin media box. I’ve checked the fridge freezer cpc with a clamp meter on an extension lead – no issues, trips whilst I’m there. I take the freezer out and fridge and trips 5 mins later. The only thing in is the tv / media thing. Checked that with a clamp meter and no issues. I’ve left the client after a good few hours and I started to struggle to see the wood for the trees by this point. So I told him to take that out and slowly put things back in at 30 minute intervals – thinking could it be an intermittent fault with an appliance like a compressor on a fridge that isn’t constantly running?

So my questions are this – why would the rcd trip at half trip current on that circuit only, with no loads, despite good IR tests - is it still likely to be fixed wiring?

Could the TV / media thing have a component that is intermittently working and faulty? – When I suggested the tv he did say his son had had the back off and fitted his own speakers to it……. (my ears did ----- up at that)

Testing with a new megger mft 1711 by the way.

Anything I’ve forgotten – if it’s obvious – please don’t shoot me, it’s been a long day!

 

[Octopus]

To be honest these guys have an c&g in electricians banter, I've struggled past that one too!

C&G

Some of them have First class degrees .....

 

[MikeKing]

Hi Mike - once there are circuit faults you will only get a clear result if you do the RCD test with circuits removed as per #22. Then if RCD tripped at 15mA or less it's out of spec, as it should not trip at 15mA. Since it's not resolved, I would look for N-E fault in any circuit of the installation. Good hunting

Really appreciate that, thanks

 

[johnduffell]

C&G

Some of them have First class degrees .....

You're right, with a student tuition fees loan behind you you'd need a sense of humour

 

[Bob Geldoff1234]

Is the downstairs a solid floor or floorboards? You say the downstairs sockets are a ring.Does each socket downstairs have at least 2 twins in each socket or is the ring upstairs via joint boxes and just legs run down to each socket(if it's solid floors)?

If the floor is floorboards then it should be wired from underneath.In which case you need to start taking off sockets and disconnecting each leg of the ring and testing back to the mains.If that leg is totally clear then move onto the next leg and disconnect and test.My guess is the lack of continuity on the lives is probably where the fault is also.
Don't forget that just because the insulation resistance is reading clear on your meter doesn't mean that there is an arc somewhere within the ring and your tester isn't powerful enough to bridge the gap.If the fault is in the upstairs joint boxes then your going to have a hard time finding it without lifting the floor.

Good luck!

 

[Richard Burns]

When testing your IR to earth did you include the earth bar, i.e. were the cpcs connected to the earth bar at the time of test or were they disconnected cables.
If they were disconnected cables then there maybe a fault from a live conductor to a different earthed part, so a repeat IR test from the live conductors to the earth bar may give a faulty result.
Presumably because the RCD is almost holding at times the fault sound like a progressive fault i.e. it has been getting worse and worse until it just starts to trip the RCD. This would then tend to indicate something like slow water/moisture build up or a partial breakdown of an electronic circuit.
Just some thoughts.

 

[Toneyz]

Mate it's difficult to grasp the tone in text and when the last few posts were critical and personal rather than helpful I hope you can see why I may have misinterpreted your post, if it was just banter, then my apologies for flying off the handle.

Keep on the forum Mike their not all that bad they do warm to you.

 

[Michael Sands]

I know I left after my first x
Unwarranted of course with no explanation as to why everyone does the same jobs ther own tried and tested way but finding out how others do it is sometimes helpful and can help others still not sure and apprehensive

 

[MikeKing]

Hi Mike - once there are circuit faults you will only get a clear result if you do the RCD test with circuits removed as per #22. Then if RCD tripped at 15mA or less it's out of spec, as it should not trip at 15mA. Since it's not resolved, I would look for N-E fault in any circuit of the installation. Good hunting

Really appreciate that, thanks

Is the downstairs a solid floor or floorboards? You say the downstairs sockets are a ring.Does each socket downstairs have at least 2 twins in each socket or is the ring upstairs via joint boxes and just legs run down to each socket(if it's solid floors)?

If the floor is floorboards then it should be wired from underneath.In which case you need to start taking off sockets and disconnecting each leg of the ring and testing back to the mains.If that leg is totally clear then move onto the next leg and disconnect and test.My guess is the lack of continuity on the lives is probably where the fault is also.
Don't forget that just because the insulation resistance is reading clear on your meter doesn't mean that there is an arc somewhere within the ring and your tester isn't powerful enough to bridge the gap.If the fault is in the upstairs joint boxes then your going to have a hard time finding it without lifting the floor.

Good luck!

Thanks Bob - my thoughts exactly. It's lift the floor or chase my tail around a bit more. I appreciate the thoughts on the arcing though. I had split the ring and tested back towards the board, but IR tests came back satisfactory. But I think your right, the brake in the line conductor is likely to be the location of both faults.

 

[MikeKing]

When testing your IR to earth did you include the earth bar, i.e. were the cpcs connected to the earth bar at the time of test or were they disconnected cables.
If they were disconnected cables then there maybe a fault from a live conductor to a different earthed part, so a repeat IR test from the live conductors to the earth bar may give a faulty result.
Presumably because the RCD is almost holding at times the fault sound like a progressive fault i.e. it has been getting worse and worse until it just starts to trip the RCD. This would then tend to indicate something like slow water/moisture build up or a partial breakdown of an electronic circuit.
Just some thoughts.

Thanks Richard, just to answer your question, I did test with the earths connected. I took the neutrals out, but the earths were connected.

I thought the fault might be a component in an appliance that isn't running constantly like a fridge compressor or a heating element in the washing machine etc.

Really appreciate the feedback - thanks

 

[LankyWill]

Fault finding is all a process of steps, it takes a while to learn the shortcuts that can help speed things up. Forget all the leg pulling they’re just teasing, plus don’t take what you did in your am2 as gospel. They don’t like you working in live boards hence the rcd rest remotely, as a spark you will inevitably have to work on/in live boards.

Ramp the RCD at the board with loads connected/Not and note the difference, this is your leakage.

Next do a global IR with all circuits on(mcb not power) and slowly turn them off to find out which has the short.

Then it’s a process of halfing the circuit until you narrow it down, remember the leakage might be accumulative across multiple circuits.

Good luck

 

[DPG]

Happy hunting! Keep us posted with what you find.

 

[Richard Burns]

Thanks Richard, just to answer your question, I did test with the earths connected. I took the neutrals out, but the earths were connected.

I thought the fault might be a component in an appliance that isn't running constantly like a fridge compressor or a heating element in the washing machine etc.

Really appreciate the feedback - thanks

OK then it might be worth testing IR: line on your side to the other neutral bar, or if there are too many loads just between neutral bars with the RCDs out of circuit and MCBs On, main switch off, to see if there is a connection across circuits. This would allow some current to flow down a different path and trip the RCD, however if the circuits are all well separated and the other RCD is not tripping then this is less likely.
If the RCD is tripping with any load then it is likely to be a circuit issue rather than an appliance.