GFCI issue, ghost fault

[Justin@valkyrie]

I'm troubleshooting a GFCI circuit for a customer. Garage circuit, includes outside receptacles, door opener receptacles on line side of GFCI.
I'm not able to measure any faults on the load side of the GFCI with my T5 (no continuity ground-to-neutral or hot-to-neutral) but GFCI won't reset. I have replaced the device with new ones and have the same issue (I have tried three different brands).
I suspect moisture maybe, but have not been able to trace it down. I'm going to try isolating the issue by process of elimination by working my way around the circuit disconnecting legs of it, but I'm kinda stumped and feel it's making me look bad to my customer.

Any help is appreciated.

 

[Dennis Alwon]

You are on the right track. Separate whatever you can and slowly put things back once the fault is cleared. There is no magic to it...Process of elimination

 

[ptonsparky]

Outside receptacles are one of the bigger offenders, especially those with loose fitting or missing covers.

Had a snake wrapped up In one last year.

 

[LarryFine]

One relatively expensive, but effective option would be to install a GFCI receptacle at every outlet, all line-wired.

 

[mopowr steve]

The door opener receptacles at this point now should be on the load side of GFCI. By the way, the live wires are going to the line terminals right?

 

[Justin@valkyrie]

I chased the trouble down to a UF cable that had been installed poorly by the homeowner and had been compromised. Still, it must have been amperage leaking through the insulation but not creating a measurable short from the neutral to ground (not measureable with a Fluke T5 at least).

 

[Dennis Alwon]

Remember that a gfci is measure in milliamps. It will trip at a 6 ma leak.... hard to measure that with a standard meter

 

[Justin@valkyrie]

Remember that a gfci is measure in milliamps. It will trip at a 6 ma leak.... hard to measure that with a standard meter

Exactly! I had returned with my more robust Fluke meter, but had discovered the problem by process of elimination before clamping on the meter (I should have done it anyway to see what the reading was).