An interesting fault finding job!

[HappyHippyDad]

1pm, quite happy doing my planned job. Builder then asks if I could go out to another job where the top lights have all tripped out.

I like peace and quiet when fault finding, plus I like to plan it in advance and always first thing in the morning as you know it can get pretty complicated and I like to have a clear head, so I'm already off to a bad start. Plus, it turns out to be a noisy and busy work site.

6A MCB is tripping (not the RCD). IR test at 500V shows 0.01Mohms L-E. I then decide to use a smaller easier to use continuity tester (my big error) and get 80ohms between L-E, so basically I've got a short. Easy track down I think.

Narrow it down (using continuity tester) to a single piece of cable. The faulty cable has a joint in it (wago's). I disconnect the wago's to test either side in order to see which side of the cable has the fault.... and the fault has disappeared. I assume it was something to do with the wago's or the cable going into the wago's. I strip back, use new wago's, connect up, fault reappears. Disconnect, fault has gone from both sides of cable! I then connect up again and the fault has gone. It seems to come and go. It's 5pm. I leave the customer with all top lights working bar bathroom (I have disconnected the cable that sometimes tests faulty).

Go back today with a clear head. Continuity test again, all clear. Within a few seconds I am thinking why am I not IR testing? I'm just sticking to an easy continuity test as I got a few hits from that to begin with. IR test at 500V is a bit odd, changing from 5 or 6 Mohms to 0.3Mohms during testing. Still nowhere near a short circuit.

I test at 1000V (which isn't really on my radar) as i'm thinking burnt out section and it ranges from 5 Mohms to zero Mohms.

I pull out the faulty piece of cable from inside the wall (drawing new cable in). It looks absolutely fine. No visible damage at all and I have run my hand along every single inch of it, slowly and carefully.

I then cut in half etc etc and keep testing as i'm intrigued. I eventually find an incredibly thin stanley knife slice in it. I strip back the outer sheathing (which looks fine) and find this...



The line is almost completely severed.

I realise you guys may have come across this lots, so I'm a little embarrassed at being so excited about it (even more embarrassed at not testing correctly to begin with), but I'm still feeling chuffed at finding it

 

[James]

Don’t be embarrassed,
intermittant faults like that are always hard to find.

sometimes, you fix it based on a theory that is the only logical thing that works in your head and the fault is gone but you never manage to find the faulty item or cable.

well done for actually locating and confirming the damage to the cable.

 

[Dustydazzler]

Reminds me of the times you get customers with intermittent faults when a cable its pinched under a floorboard nail and the mcb only trips when the tread on the exact spot on the floorboard and the nail bridges the line and cpc

 

[littlespark]

I’m in the middle of finding a fault on a lighting circuit. Already found a leg from “joint box” to a light had be nibbled through by rodents…
(Not really a joint box… 1 gang pattress with blank plate with 5 t&e’s going in.

Remaining faults. Lighting circuit has two outgoing cables, both for downstairs lights. One is shorting L-E, other is Shorting N-E. Going back Monday after customer has moved a wardrobe.

Expecting to find a deceased Mickey Mouse straddling the two cables with its jaws.

 

[Lucien Nunes]

There were two interesting clues in your description. One was that the fault seemed to disappear when the wagos were disconnected. That immediately suggests to me that there is both an insulation breakdown and a severed or intermittent conductor that prevents the tester seeing the breakdown other than via some parallel route. From this side of the internet I can't tell whether that was the mechanism at work, or whether for example disconnecting the Wagos changed the tension on the cable enough to mechanically disturb the fault. But this finding would make me want to test from L of one side to E of the other and vice versa. There is another possibility that is not applicable here, being an insulation breakdown between two separate cables, e.g. both hit with one nail.

The other was that only the MCB tripped, yet you registered an L-E fault. MCBs can be faster than RCDs but if the fault is bad enough to trip one, the RCD is very likely to follow. I am not sure that any useful inference can be made from this here (other than that it's worth testing the RCD) but it's interesting nonetheless.

Over the years I've spent a lot of time on electronic and electrical faults some of which have been elusive. I could list a dozen subtly different flavours of frustration with various developments along the process, and another dozen subtly different flavours of excitement when some success, temporary or permanent, is achieved. One of the interesting aspects of working on old, complex equipment is that it might have many minor issues and discrepancies, only one of which is the main cause of the present breakdown. One has to take care not to get lost chasing 101 things that are not exactly right, but not seriously wrong either.

 

[Vortigern]

I guess you replaced the cable and did a few tests such as Zs? What is the PFC? Just wondered if the MCB is now knacked as Icn was exceeded? And I am suspicious of that damage. Almost look like an act of sabotage. From the picture it seems that the action was between L + N as in a dead short that blew apart the seam of the cut and the earth was incidentally vapourised thus did not make a connection for the RCD to trip??? Just sleuthing a little for entertainment. But, Yes you found the fault and that always sends a wave of endorphins through the ventricles of the mind, often a good feeling.

 

[Markyd]

That was a great find well done

 

[timhoward]

Well done, not an easy one.
It’s already been said but the RCD not tripping would pique my interest further and worth testing it.
I had a fault feeding a socket spur once where the appliance would cut out during vibrations. Nothing tripped. IR was consistently about 5M. The only clue was that R1+R2 was intermittently open circuit depending on the route I walked to the socket! The cable ran under a beautiful wooden floor.
In the end I disconnected the spur and ran another cable from upstairs.
It wasn’t possible to remove the original cable to see what had actually happened but I suspected a very clean break in the line.
(The customer was gutted as two weeks earlier they had thrown away a £3K electric piano that kept resetting itself when they played low notes loudly after a service technician couldn’t fix it. )

 

[ferg]

It's easy to feel daft after the event with the benefit of hindsight.
Fault finding has a knack of making experienced and capable people feel very daft very quickly.

 

[Dustydazzler]

There is also an element of luck when fault finding

I remember way back helping a couple older sparks on a massive Eicr (pir as it was then) on a mansion in London.
One of the upper floor rings had no continuity of the N. They had spent hours trying to solve the issue.
I pulled a chest of drawers to one side found an old brass socket and unscrewed to as I did so the 2 N cables literally were flapping in wind and not in the terminal.
They had spent about 4-5 hours trying to find the issue I found it in 15 mins as it was the only socket they hadn't found