I just helped my friend change an outlet in the garage of a home he just moved into. The outlet was an old duplex that I replaced for a GFCI. The installation went fine, outlet tested properly and we used it to power several power tools and shop lights around the garage. The problems started when we went to plug in the post lamp that is located in a garden bed right next to the garage door and has a 12/2 direct burial conductor running from the post through a hole in the floor. The previous owner put a two prong non-grounded connector (I know, going to change that) on the end of the conductor. Now when we plug in the lamp it comes on, no issues except for the fact that after being on for about 10-15secs the GFCI cuts out. If we unplug the lamp, the outlet will reset and remain on until we plug in the lamp again. I didn't have my multi meter on me so I couldn't check but from what I described. Being that there wasn't a ground on the lamp connection...Would you assume the connection somewhere between the connector and the post lamp is having a:

  • Line to ground short (going to GRN at the post or in the earth) where its causing an imbalance on the return (neutral) which would explain the delay in the GFCI turning off?
  • Neutral to ground with a similar situation as the above scenario?

If so would doing a resistance check with my DVOM figure that out? Any other ideas? Thanks

  • if the cable is long, inductive coupling to earth can result in a few milliamps of phase-shifted current which the GFCI will treat as a ground leakage. this because any common current passing through the current transformer in the GFCI will induce that current to the detection circuitry. this can happen even on normal circuits in the home if they are very long (80 ft or more). the type of cable doesn't matter.
    – Skaperen
    Aug 6, 2019 at 5:46
  • @Skaperen no one has ever been able to explain to me where the electrons go for siesta when that happens. For that to trip the GFCI there must be many fewer electrons in the copper wire and lamp, or many more. This is a lot of electrons. Aug 6, 2019 at 14:04

1 Answer 1


It could be either one. The point is the loop is leaking to earth, and it shouldn't do that, and you should go through it and look at all the connection points.

It's unlikely to be the cable unless he used indoor rated cable.

DVMs are inconclusive. Obviously a low voltage reading is a problem, but some ground faults act like VBOs - they don't actually suffer insulation breakdown until a certain voltage is reached. Or they may require a certain operating temperature, etc.

Your starting point is to give all wire endpoints and connections a good once-over. Do that first, because you'd feel pretty foolish doing an exhaustive technical troubleshoot only to trace it to the lamp and find a mouse nest at the terminals lol...

  • 2
    I don't even answer when harper's around..lol
    – JACK
    Aug 6, 2019 at 12:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.