GFCI Tester

[wirebender]

I was checking receptacles with one of these:

A receptacle (not GFCI) with a MWBC (12/3) made all three lights come on. If I touched my knee to the floor while holding the strap of the receptacle the red light went out.

Everything was properly wired.

Receptacles downstream showed to be okay.

Why the three lights?

This turned out to be loose ground connection in a j above the panel. Still doesn't explain the three lights or the downstream receptacles showing okay.

At a GFCI receptacle in another location my tester would not trip the GFCI or any downstream receptacles.

I only read 103 volts hot to egc at the GFCI.

What do I need to look for on this?
An open ground upstream?

This was a loose ground connection in a switchbox. Still doesn't explain why the tester didn't show open ground. Had there not been downstream receptacles protected by the GFCI to test, I would never had known there was a problem.
Maybe I need to rethink how I test installs.

Thanks everyone.

 

[e57]

This turned out to be loose ground connection in a j above the panel. Still doesn't explain the three lights or the downstream receptacles showing okay.

All 3 lights lit - showing a voltage difference big enough to illuminate the light - you putting your knee down while contacting the yoke grounded it enough to shut it off..... JMSO.

This was a loose ground connection in a switchbox. Still doesn't explain why the tester didn't show open ground. Had there not been downstream receptacles protected by the GFCI to test, I would never had known there was a problem.
Maybe I need to rethink how I test installs.

Thanks everyone.

Hmmmm. I don't get this one either.... Maybe grounded just enough to shut the light off through the GFI's internal circuit???? It is the same as the one above... The little neon light need a potential difference between the ground and neutral to light. In both cases it was just enough - but strange enough for you to notice it.... So I would say it did it's job - but not very well. :roll: It still clued you into potential problems - and thankfully not the ones I was thinking of, as the same could have happened with open neutral connections IMO.

 

[wirebender]

Thanks, e57.

 

[c2500]

For what it is worth, I had a circuit one time that the tester showed was fine, but when the test meter went to it, hot to ground was less than 100volts. It turned out the tester was lighting up, but dimly. I think if ground is present it lights, and the better the ground, the brighter it gets. I finally traced it down to a loose ground in a j-box. New large size wirenut and all was right.

c2500