Mixed signals when testing GFCI outlets

The scenario in response to the OP was a “bootleg” ground. This will still cause a current flow and trip the GFCI because the testers (andGFCI) can not distinguish between ground and the grounded conductor (neutral).

Easy enough to demonstrate:


So the GFCI is supposed to trip when it detects a leakage of current, but the tester in the bootleg scenario is simply redirecting that leakage back to the neutral. How does the GFCI actually know there is a leak? Is it simply because the leakage is being directed around the GFCI on it’s way to the neutral? And a follow up, would the tester trip a bootlegged receptacle on a GFCI breaker?

Lets just talk about the test mode. When you press the GFCIs test button a connection (and current flow) is made to neutral (the grounded conductor). That is why the GFCI will work with NO ground. No ground installations should have the supplied “No equipment ground” sticker applied. A three light tester button will make a connection to ground to trip the GFCI. That is why an ungrounded GFCI will not trip with a three light tester button but will trip with the GFCI built in test button. A GFCI does not know if it made a test connection with ground or neutral, it doesn’t care and will trip either way. Downstream outlets that are wired properly to the GFCI will trip and test exactly the same (obviously the three light tester will make a connection to ground or bootleg). The GFCI senses a difference in current flow with a connection to either a ground or the grounded conducter.

How GFCIs work

1 Like

So this discussion prompted me to dig deeper because I have also seen weird results with my 3-prong, thus why I am asking questions. If the test button on the GFCI shorts load to neutral, where is the leak happening that the GFCI is supposed to respond to? In the picture below, of a much simplified GFCI diagram and very good explanation, they show the leak happening at the hedge trimmer, thus less current returning than what left the receptacle. But when the test button is pushed on the GFCI, the current is shorted directly to the neutral. Isn’t the current still equal in both conductors? Where is the leak? Is it because there is no load? Why would that matter? I have a feeling that it is because the short is made to the left of the “transformer core” when the test button is pushed (relative to how the picture shows it), but then wouldn’t the current through the core still be equal at 0?

When depressing the test button the current is sent to the EGC not the neutral. You’re correct sending the current to the neutral would do nothing.

I keep it at KIS&S. Keep it simple and simple.

" The GFCI was showing correct circuit readings when tested but the receptacle did not trip when tested. This may be due to a faulty receptacle or faulty wiring. Further evaluation by a qualified/licensed electrician is recommended." And then move on…

Actually, I only had to read a bit further in the document to find my answer. The test button connects the hot on the left side of the coil, to the neutral on the right, thus creating an imbalance. This also explains why a 3-prong will trip a bootlegged GFCI. Same scenario as the test button on the GFCI, except the short is made from the hot on the right side of the coil, and the neutral on the left side.

If anyone want to access the document these pics came from, here is a link. I think it is the best description of a GFCI I have ever read. It also has tear down pics of an actual GFCI.

What do you guys do when the house is a mint condition 1960-65, all original house, and there are no 3 prong outlets at all? Bunch of 2 slot outlets, no gfci’s. How do you go about testing those?

I used one of these and stuck one prong into a slot and pinched the other between my finger and thumb. If it lit up, the prong slot was hot. You can find reverse polarity with this method, too.


Then add GFCIs as needed.

Finger and thumb? I thought you licked it and stuck in your ear! :upside_down_face:

1 Like

It really does work well. :cowboy_hat_face:

1 Like

If that were the case then how does it still trip when there is no EGC?

You’re right Bob, that is not the case. The graphic I posted in my above post shows how the test button indeed shorts to neutral, and how it creates the appearance of leakage in order to trip the disconnect solenoid.

Short answer it doesn’t trip.

What would be a reasonable recommendation to correct the outlets… … …

If the Load and Line wiring are backwards the GFCI has power but doesn’t trip.

It should trip with the test button on the GFCI but will not trip with a 3-light.

Post can’t be empty…does it look empty?

1 Like

Your short answer is wrong - it does trip with no EGC.and the test button on the GFCI, but will not trip with a three light tester with no EGC.

1 Like