Half a fault

Arthur Fault? Bloody good bloke! Yip yip! Nah, I’m just pullin’ your Christmas cracker pal; this is a video about us not finding a fault. Well, we find half a circuit the fault isn’t on, but then Christmas comes along and we clock off. Should we survive the booze, food and Omicron variant, maybe we’ll post an update in the new year to show why this RCBO is tripping.

This one is less well explained, so I've thrown in some visual aids to show what we’re testing and what we’re seeing. Nonetheless, it may be a bit confusing as it’s us thinking on our feet out in the world where there’s not always a nice clear fault to easily be found.

The tripping RCBO has three circuits connected. The HomeServe chap who came before us has disconnected the circuit that has a low insulation resistance of 0.3MΩ, something we verify before we start dicking around. This is shown at the nine-minute mark.

We then restored power to the other circuits so we could see what was working and what was dead. The exterior lighting was not powered, so we started there.

At the fifteen-minute mark, we thought we’d traced the fault. An IR test on what we think are exterior bollard lights was at zero, so with those disconnected, we returned inside to test back at the board. To our dismay, the IR result was now even worse – a straight zero indicating a resistive short.

To prove a resistive short, the Metrel was switched to continuity mode at 15:47 and a test was performed between L+N and the earth bar. A reading of 1.08Ω was returned and, surprisingly, the live RCBO tripped off even though our circuit under test had been disconnected and was supposedly isolated from everything else. Naughty naughty: we should really have powered off the CU while dicking around dead testing.

The energised RCBO tripping on a continuity test for a circuit that's supposedly disconnected and isolated can only happen if our tester has found a common path to neutral i.e., if our faulty circuit’s neutral was cross connected to an energised neutral on another circuit, as that would provide a path that was resistively shorted between N-PE (they being connected together at the head).

At 16:55, I test for a neutral cross connect by continuity testing between the disconnected circuit neutral and the neutral bar. A reading of just 0.90Ω is returned where there should be no continuity between individual circuits. This explains why they shared an RCBO as splitting them would have caused tripping issues when the lights were turned on. It goes to show one shouldn't assume a disconnected circuit is always isolated, especially where it's evident a moron has been fiddling with the wiring.

AT 17:40, we figure the discrepancy between the IR figures obtained at nine minutes and fifteen minutes is because the first test was undertaken with the main switch off, so the tester recorded only the circuit earth fault which has a higher resistance and bombs out at 0.03MΩ, whereas the second test was performed with the other circuits energised, thus providing a path via a neutral cross connect, through the RCBO, through the main switch, through the connection between N & E at the head (which is a low resistance connection compared to the circuit fault), then back through the main earthing to the tester. The 1.08Ω recorded earlier was the resistance of this path.

Make sense? Crikey, try picturing it all in your head when it’s the last job before Christmas and you just want to get down the bloody boozer!

To show that the first IR reading was down to the fault and the second was down to the N-PE path through a cross-connect, that path is closed at 18:10 by knocking off the main switch and the test then repeated. The tester is forced to pass through the circuit fault again as that’s the only path now available, and this time we get a similar number to that of the nine-minute mark, 0.04MΩ.

Fault finding and correction is made much more difficult here because the electrical installation is gash and obviously the work of an amateur. I’ve said it before, but while an amateur may well get something working, that doesn’t mean it’s safe, has any longevity or that the likes of us won’t have to rip apart the walls and ceilings should it all go wrong.

The bonnie situation:
The hall, kitchen, exterior lights and ventilation are back on. The cabling tests good going into the outside bulkhead, then reads bad going out. From here it goes on to serve the two bedrooms, so they’ve been left disconnected. Partial success to date then, although we still need to find why the bedrooms are borked and that’ll likely involve opening the ceilings.

Follow up vid:    • Half a fault 2 : The odd switcheroo  

Don't worry, Nigel will get over his hissy fit and be back for more abuse in 2022.

A previous video where he and I were trapped in a dank cellar with no lighting and had to rub each other off is… on OnlyFans…, but another one where we had to actually fix the lights can be found here:    • Three-fault Friday