Weird GFCI Behavior

[Bob Sisson]

older house with a mix of 2-wire and 3-wire...

They put GFCI's in the Bathrooms

The GFCI's trip with the test button on the GFCI but they DON'T trip with my pocket tester... (yes my tester showed a ground)

I wrote them up...but am wondering...I didn't try going between the ground tab on my tester and a plumbing pipe to truly test the GFCI since I suspect the ground was really a bootleg ground...

How do you test a 3-wire circuit that is MARKED "ungrounded protected by GFCI" as the pocket tester doesn't have a ground leg to drain some current off to.... Do I need to carry a pig-tail with an aligator clip

[Jim Luttrall]

Bob, the proper test is to use the test button on the GFCI device, not the pocket tester. Two-wire GFCI protected outlets will NOT trip with an external tester.

[Jerry Peck]

How do you test a 3-wire circuit that is MARKED "ungrounded protected by GFCI" as the pocket tester doesn't have a ground leg to drain some current off to....


With the best and preferred tester ... the "TEST" button on the GFCI device itself.

No ground is required for that to test the operation of the GFCI.

[Bob Sisson]

Hum... one showed a ground, but went red on the tester and didn't trip...

How do we test/verify outlets MARKED as No Ground, GFCI protected?

With no ground at the outlet, how do we create a ground "fault" safely...

[Jim Luttrall]

Hum... one showed a ground, but went red on the tester and didn't trip...

How do we test/verify outlets MARKED as No Ground, GFCI protected?

With no ground at the outlet, how do we create a ground "fault" safely...
With the provided test button.

[John Dirks Jr]

Hum... one showed a ground, but went red on the tester and didn't trip...

How do we test/verify outlets MARKED as No Ground, GFCI protected?

With no ground at the outlet, how do we create a ground "fault" safely...


The ground "fault" is referring to a failure in an appliance or whatever is plugged into the receptacle. The GFCI device is detects a "fault" in the device and trips the circuit. The house ground is separate from this process. The house ground is not needed for the GFCI to perform its function.

[John Dirks Jr]

I could have explained it better. So as not to cause confusion, or maybe to cause more :D here goes;

The GFCI is constantly measuring for differences in voltage between the hot and neutral of the circuit. If it detects differences, it "assumes" that voltage is leaking to ground, hence the name "ground fault".

It can be confusing since they call it a ground fault circuit interupt. Confusing because the receptacle doesnt even need to be grounded to do its job. It acts on and is triggered by the measurements between the hot and neutral. The house ground is not involved.

Some electricians install GFCI's on ungrounded circiuts in older houses as an alternative to running a new grounded circuit. After all, a GFCI will trip much quicker than a 15 breaker will.

[Vern Heiler]

I could have explained it better. So as not to cause confusion, or maybe to cause more :D here goes;

The GFCI is constantly measuring for differences in voltage between the hot and neutral of the circuit. If it detects differences, it "assumes" that voltage is leaking to ground, hence the name "ground fault".

It can be confusing since they call it a ground fault circuit interupt. Confusing because the receptacle doesnt even need to be grounded to do its job. It acts on and is triggered by the measurements between the hot and neutral. The house ground is not involved.

Some electricians install GFCI's on ungrounded circiuts in older houses as an alternative to running a new grounded circuit. After all, a GFCI will trip much quicker than a 15 breaker will.

John I know you have a good understanding of the GFCI, but we need to change the voltage in your explanation to current. Other wise someone will try to verify with a volt meter.

[John Dirks Jr]

John I know you have a good understanding of the GFCI, but we need to change the voltage in your explanation to current. Other wise someone will try to verify with a volt meter.

I stand corrected.;)

[John Kogel]

A couple of anomolys I know about GFCI's: They will sometimes stay hot after tripping, Downstream outlets are protected but the unit itself does not turn off. I still don't know if this is a wiring fault or an internal fault in the unit. It may be either.:confused:

If two GFCI's are daisy-chained, like in two bathrooms, the downstream one will not reset untill you go reset the outlet that supplies it. Only the plug-in tester will tell you if this wiring is present, BTW, because it will trip the remote breaker. Then you push the test button and ..... no reset. :confused:

I recently found a GFI breaker in a panel that was wired in reverse 20 years ago, and has never worked, yet it snapped off nicely with the test button. :confused:

In one house, I tested the first bath outlet and tripped a breaker somewhere. The two other bath outlets were dead but there was no GFI in the panel and the outdoor outlets were still hot. I eventually found the GFCI outlet behind some crap inside a cabinet next to the panel. :confused:

[Jerry Peck]

A couple of anomolys I know about GFCI's: They will sometimes stay hot after tripping, Downstream outlets are protected but the unit itself does not turn off. I still don't know if this is a wiring fault or an internal fault in the unit. It may be either.:confused:

The older GFCIs were know for being wired incorrectly with the line and load wires connected in backward, which resulted in just what you said.

That was such a common problem that the standard for GFCIs was changed, the newer GFCIs will not reset when wired in with line and load reversed.

All the other things sound like wiring problems.

The only thing I ran across which was not a wiring problem, which was within the GFCI devices themselves (and this has also been changed as I recall), many of them would reset "reverse polarity" when reset. I found out that if the reset button was pushed in from one side and not from the center, I could reset those "reverse polarity" almost every time. I believe that was taken care of 4-5 years ago, though.

[Lee Place]

Bob, I had a load side outlet that had reversed polarity and would not trip the GFCI.

[Bob Sisson]

These were all old enough that they needed a closer inspection or replacing... backfeeding or reverse pollarity was a definate possibility...especially considering all the other stuff I saw and wrote..

Thanks...