How is it possible for a GFCI cord to fail in one outlet but not another?

[BlueRaja]

I have an outdoor electrical subpanel with a nearby outlet. We've plugged an extension cord with a built-in GFCI into it , but the GFCI trips every time, with any device plugged in. When we plug the same GFCI/device into a different outlet near the house, it works fine.

• We've replaced the GFCI/extension cord, devices, outlet, wiring between panel and outlet, and breaker. Problem still happens.
• Extension cord(s) and GFCI(s) are rated for 15-20 amps, weather-proof, and cold weather. It's 30°F (0°C) out.
• New outlet and wiring are 20-amp, weather resistant.
• We've tried multiple devices, ranging from 250W to 1500W. All work when plugged into the side of the house, but trip the same GFCI cord(s) when plugged into the other outlet.
• The outlet measures 120V, and gets an "ok" from an outlet tester.

From my understanding of GFCIs, this situation should be physically impossible. Since the GFCI just compares the current coming in and out, even if the outlet were delivering too low a voltage or something, the GFCI still shouldn't trip at one outlet but not the other.

Does anyone have any ideas for what could be causing this, or further troubleshooting steps I could take?

 

[pc1966]

I don't know much about the USA style of GFCI, but in some cases our equivalent (RCD) will trip if it detects something wrong with the supply socket like L-N reversal.

I would suggest getting a professional in to check the socket as those testers are not very reliable and the quality/performance can vary a lot between brands and models. If they tell you there is a fault it is almost certainly true, but if it tells you it is OK it might be missing something as they are poor at detecting open earth (ground).

Some look better than others, for example:

https://www.homedepot.com/p/Klein-Tools-AFCI-GFCI-Outlet-Tester-RT310/308709791

But I don't see (from a quick search) any that offer a measurement of the ground impedance that the better ones you get in the UK offer (costing ~$75).

What tester do you have?

 

[BlueRaja]

I replaced the outlet and all the wiring, so I don't think it's that. I've also tested the resistance between neutral and ground at the subpanel, and measured 2 Ohms. I don't have anything to test the inductance, though, so I suppose there is a small chance that it's that.

 

[pc1966]

The sort of inductance you get in any sane installations is not going to be a problem, and it seems you have independently verified the N-E impedance is good enough.

Assuming that socket is therefore good, does the GFCI chord trip if it is indoors and just plugged in to this socket? I'm wondering if it is something to do with the cable/load being outdoors that is causing this.

 

[BlueRaja]

When I plug any of the GFCIs into any outdoor outlet attached to the house (none of which are on the subpanel circuit), they work correctly, so I don't think it's the cold. It's only when plugged into the only outlet on that subpanel that the GFCIs trip.

 

[pc1966]

If you use your tester on the sockets that don't trip, what results do you get? Anything at all different?