“Phantom” Short on radial circuit

[Tom281193]

Hi,

I’ve been recently working on a living trailer for a friend, when they have the supplies connected (240VAC 16A) they have nuisance tripping on the RCD. I’ve traced through and found an outside waterproof socket to be at fault, short circuit between live and earth. Removed the socket and tested, fault was not on socket, short still within the wiring. So I traced the cable back to a small joint box about 3M away. When I disconnect the cable from the box, then test of both sides of the connection box, there is no short. But as soon as I wire the cable back up the short comes back. I’ve completely removed the connection box and used inline crimps (to eliminate the connection box for testing) still the short is there. But when all is disconnected, each cable is fault free. Only when everything is connected will the fault reappear?

Any ideas?

 

[pc1966]

Could it be that moving the cable is causing/breaking a short elsewhere (e.g. hole without grommet, etc)?
[automerge]1597144464[/automerge]
Also are you testing with an insulation resistance sort of meter (i.e. a "Megger" or similar that uses 500V) or a multimeter?

 

[Tom281193]

Could it be that moving the cable is causing/breaking a short elsewhere (e.g. hole without grommet, etc)?
[automerge]1597144464[/automerge]
Also are you testing with an insulation resistance sort of meter (i.e. a "Megger" or similar that uses 500V) or a multimeter?

I thought it was the cable, but was testing it while moving it round and especially around the gland areas, no luck.
Was using a multimeter at first, then IR tested it, no problem on either end while disconnected >999

 

[Strima]

TBH I'd replace the cable, could well be insulation damage that only shows when in certain positions.

 

[buzzlightyear]

IMO to find the fault you need to start from the beginning of the circuit.
Has the others say if the IR down on the cable replacement is needed..

 

[static zap]

Do check you are doing tests with cpc present fo vehicle chassis.
(Thinking Live to chassis short if not actual cable) ?
..Must read posts better..!

 

[Lucien Nunes]

The fault may be from L in one leg of cable to CPC in the other, so each cable tests good by itself. As Static points out, the CPC in one leg may not be connected to the chassis. If the L on that leg is faulted to the chassis only, it won't show up on its own CPC but it will via the chassis to the CPC of the other leg. Test all to chassis and each other, not just each cable separately.