I’ve just started a new job and my new role looks like it will involve more reactive work, first job today was an earth fault on a Kentec syncro. 

 

As I approached the job something occurred to me - I really don’t have a methodical way of tracing these faults. 

 

I know this may sound like an obvious question, is probably second nature to many, but I really don’t know the quickest and most efficient way of finding an earth fault.

 

Originally wired in two loops, now joined in the panel to make it one loop.  The fault is on the newer of the two loops. I think it’s about 16 devices.

 

I can think of various things I’d do which I would muddle my way through and eventually reach the source of the fault, but I want a methodical way of doing this. 

Difficult when you aren't familiar with the wiring route or mid-way point of the loop.

 

In simple terms, you need to half the loop, multimeter to find which leg has the earth fault, half again until you narrow it down to a device or a leg of cable.

 

First of all I'd check all devices on the loop for the environment there in - certain ones may stand out, water ingress is a common one from my experience. Rear of chillers for example when the cable entry hasn't been sealed correctly causing a difference in temperature and the condensation sets causing the device to get filled with water. It may save you time with a bit of common sense first of.

 

 

As above.

halfway point and use your meter, before I do anything I look for external devices, eg sounder, call points are my first target.

with only 16 shouldn’t take too long even if you looked at them all.

 

had to find one on a 120 loop once, took about 6 hours, mixture of old MI and botched in 4 cores ect, real treat it was, traced it to an adjacent private building, he just got a new dog who was chewing an old cable in the porch!!!

 

the bloke thought it was funny... until he got the bill