Intermittentshort?

drew35 · Jan 7, 2011

I’ve extended several radial circuits for a customer whilst the house is being renovated. I initially sunk all the boxes, cut the chases, run cables under the floor, and left cables curled up in the back boxes, and then left the job until the plasterer had been in.

Now I have just been back and connected all my 2g sockets, tested the circuits and everything checked out ok. When I energized the circuits one radial just trips in a big way, definite short circuit. F word! Because the floor boards had all been up for the plumbers and I was picturing a four inch screw through a cable, because that’s what they used to fix the boards back down!

There is absolutely no question of anything else being on the circuit, and nothing was plugged in. So I disconnect in the CSU and check for continuity between L-N, L-E, N-E, of the problem circuit, nothing. So I IR test, nothing. After removing all the face plates I can see nothing wrong, no trapped wires, no damaged sheathing. So I disconnect all the sockets, do all my test again on the cables between each socket, nothing. So I put it all back together and tried again, this time the circuit energizes ok?

I go around and test each socket and I can’t fault anything? So several hours later I finish up with the radio having been plugged into the problem circuit all the time without a problem. The next day I get a call to say the MCB has tripped again and won’t reset, just a big spark from inside the MCB. The customer had already disconnected everything that was plugged in and it still won’t reset.

I’m due back tomorrow, now the customer supplied the 2g sockets, could there be a fault within a socket? Could there be a screw that’s just touching a cable and only intermittently making contact? If the faults not there when I test I’m left at guessing. I could replace all the sockets, if that fails there must be cable damage? Anyone had anything like this before?

The problem is also that it takes out the RCD main switch when it trips the MCB.

SimpleSimon · Jan 7, 2011

You say you've extended the radials as oppose to doing it new, I take it you checked the cable on the old circuits just in case? What were these radials supplying? Was the trip instant?

IQ Electrical · Jan 7, 2011

I've never seen such a fault on a socket outlet, I'd be amazed, my money would be on the screw theory-4" screws?.

It could just take someone to walk across the floor and slightly move one of those screws, causing a short.

Even if you get the MCB to reset, you're going to have to find this fault as it's flashed out twice now and you can bet that that's caused some damage somewhere on that circuit.

Start at the beginning again, R1+R2, IR tests, be methodical, you'll get there eventually, we always do!

Darkwood · Jan 7, 2011

At least the fault is permanent now, so your not chasing ghosts, just devide the circuit up until it tests clear for fault then you should be able to narrow it down and locate which devided section its in.

nickblake · Jan 7, 2011

disconnect the circuit in question from the board , remove all appliances etc from the circuit and i mean remove dont just switch them off , carry out your R1R2 as IQ has said then do your 500 volt IR if all ok then re do th IR but with 1000 volt test this should and see what the readings are using a higher voltage may well show a fault but dont normally tast at this high voltage ,and be very carefull using this voltage

Octopus · Jan 7, 2011

Do you know the route of the offending circuit? If so you could "split" the radial in half and test each half, then split the offending half etc, etc.

Just a thought but is there a correlation between the tripping and it raining?

drew35 · Jan 8, 2011

The reason I extended rather than replaced was because its relatively new cable its just that each room has its own circuit, and each room/circuit only had one, maybe two 2g sockets at most.

So the offending circuit is limited to one room only, now four 2g sockets. I had already removed all the sockets the first time, and them tested between each one without fault. I'm due back there on Tuesday now, no hurry as its vacant for a few weeks yet. Thanks to nickblake though, I didn't think to bump the test up to a 1000v at the time. I have since spoken to the customer and he has managed to reset the trip, even though I told him to leave it alone. The ghost is back! My money is on the screw touching a cable when somebody walks on it like IQ said. I will let you know on Tuesday when I've tracked it down. Cheers all.