Weird GFI problem - wired correctly but not tripping

[blueman]

I have a Leviton GFI in my bathroom which I happened to test today with my little 3 prong/3 lamp tester. The socket shows as being wired correctly (2 yellow lights, no red) but pressing the 'trip' button on my tester failed to trip the GFI (note when the trip button is depressed all 3 lights (yellow-yellow-red) light up on the tester). The test button on the GFI works fine though.

Still figuring that something must be wrong with the GFI, I replaced it with a new one. The physical wiring is correct and again the tester shows it is wired correctly. Also, the little green LED on the GFI is lit presumably showing it is working. But again the trip button failed to trigger the GFI even though again the manual test button on the GFI worked.

My GFI tester worked fine triggering the other 13 or so GFI's in my house.

Any idea what could be happening here? Why would both the old and new GFI show as being wired correctly and yet fail to trip?

I am stumped...

[blueman]

CORRECTION, when I hold down the test button the lights show: which in a normal situation would be signalling a Hot/Neutral reverse but is probably just an artifact of the test button shunting current.

[Seth Goodman]

Your GFIs, both old and new, are fine. On a GFI, the tester built-in to the receptacle works by shunting some current from hot to neutral.

Your hand held tester works a little differently - it shunts some current from hot to ground. Either there is a grounding problem on that circuit, or the hand held tester itself is faulty.

[John Grabowski]

*Pull out the GFI and use a pigtail socket and bulb to confirm that you have a functional ground.

[stan]

Suggestion/question?????. GFIs trip when there is a certain amount of (milliamps) unbalance between current flowing in the live and neutral conductors; true? If the tester itself use say LEDs which take a very small amounts of current maybe there is not enough (total) current flowing to unbalance the GFI? Not familiar with the circuit arrangement of the tester. Or as suggested the ground is imperfect.

[Tony Hwang]

Hmm, I'd measure Ohmage between ground and neutral with a meter on that outlet. Reading should be very close to zero Ohm.

[Mike rock]

A GFI's operation has nothing to do with the ground. It monitors the current flowing throught the hot and neutral.

[Doug Miller]

If you'd read the entire thread, you would have seen that the OP stated that the GFCI trips normally when he presses the test button on the GFCI, but fails to trip when he presses the test button on his plug-in tester -- and you would have also seen an explanation of why this is so: the plug-in tester shunts current to *ground*, and *cannot* trip the GFCI unless there is a functional ground at the outlet.

[blueman]

BINGO - no *functional* ground.

Though not clear why the gfi tester didn't read it as an open ground unless there was some "inducted" current flow from ground to neutral in the cable sheathing. Also, interestingly, a digital (not analog) ohmeter read 120v between hot and ground again maybe consistent with inducted current. But as always resistive loads (i.e. bulb on a pigtail) tell the truth.

This all does make me worry though about the accuracy of the low-end GFI tester I have -- i.e., how many open grounds are there lurking somewhere in the house that the tester has failed to detect...

I fixed the problem by tracing upstream where I found that the ground wire had broken off in the box. Problem solved & thanks for the help.

[blueman]

Actually it is kind of ironic that while the GFI tester didn't detect a floating ground in its normal mode, it (indirectly) signalled a bad ground by failing to trip the GFI when the test button was pressed.

Do better quality GFI testers do a better job of testing for

*functional* grounds?

Also this led me to experiment and I noticed that if the ground and neutral pin on the GFI tester are both wired to neutral then the GFI tests ok which in some ways is electrically understandable since ultimately the neutral and hots are bonded at the service entrance. However, it is not a code ground.

Do better quality GFI testers have a way of testing for functional ground vs. neutral used as ground? (perhaps as a proxy they could measure resistance between neutral and ground???)

[blueman]

Yes - that now explains almost everything. But how does the GFI test button work? Does it really just test if the breaker works or does it have a way of testing that the current monitoring part works? (I'm not sure how it would do that since without shunting current to ground, it really has nowhere else to shunt it and create a current imbalance between hot and neutral).

[Wayne Whitney]

The GFI uses a current transformer to detect a downstream imbalance between hot and neutral current. The test button has access to the hot and neutral both before and after this current transformer. So it shunts a small amount of current from the hot after the transformer to the neutral before the transformer (or vice versa, I don't know), which creates an imbalance in the current transformer.

Cheers, Wayne

[Stormin Mormon]

Much safer, now. Good job.

[stan]

Thanks for posting the outcome. A broken ground; glad the OP found it.

BTW: I agree with the person who posted; Quote: "If you'd read the entire thread, you would have seen that the OP stated that the GFCI trips normally when he presses the test button on the GFCI, but fails to trip when he presses the test button on his plug-in tester -- and you would have also seen an explanation of why this is so: the plug-in tester shunts current to *ground*, and *cannot* trip the GFCI unless there is a functional ground at the outlet.".

It seemed obvious the question was why didn't the 'test' feature of the 'tester' work as expected!

Interesting thread and as often makes one think (and learn!). Thanks for posting the original question.

[blueman]

Your welcome :) And I learned a lot too -- it is always sobering to learn the limitations of your test equipment (or any other limitations).

I wonder how many professionals (e.g., electricians, home inspectors, city inspectors) let alone DIY'ers realize that:

1. A GFI tester can show the circuit being fine even though there is no functional ground. Similarly a (digital) multimeter can read the full 120V hot-to-ground. Both presumably due to induced currents.
2. The test button on a GFI can work also without a functional ground.

It all makes sense retrospectively, but probably not something that most people think about...

[John Grabowski]

*I usually use my Wiggly to test a GFI. It will trip between hot and ground. Using a wire to short between neutral and ground will also trip the GFI.

[Gary H]

Less incompletely, hot on the LOAD side to neutral on the LINE side.

[jeff_wisnia]

I don't think that "induced" is quite the proper term to use for what's happening there. "Current flow through capacitive reactance" would be more correct.

Jeff

[blueman]

Can you explain what that means...

I guiess I was (mistakenly?) thinking that it was just basic E&M where an alternating current in one wire induces a similar current in an adjacent wire.

[blueman]

You must have either a very thin wiggly or special wide 'hot' slots ;)

Why would a short between neutral and ground cause the GFI to trip?