Thread: Why you should never trust the test/reset buttons on a GFCI

Rather eye opening experience:


https://www.youtube.com/watch?v=MGQXsCc11qc


https://www.youtube.com/watch?v=XACCFgc1q-A



How its supposed to be:


https://www.youtube.com/watch?v=M8phlIi42qY

Similar Threads:

• Reset for bathrom GFCI is in garage
• Gfci trips on test but reset button is frozen?
• Testing GFCI circuit breakers with a Sure Test
• ungrouded GFCI test

Originally Posted by Mbrooke

Rather eye opening experience:

Indeed , I'll be following this thread.

A lot of inspectors use a 3-light tester. While not a foolproof device, it should have shown this problem. It would have been nice if mbrooke had shown the same test with some kind of standard testing device instead of the jury-rigged light fixture.

Unfortunately the newest requirements for AFCI operation weren't put into effect until rather recently. That means there are many years worth of the devices floating around that will reset when wired wrong. The new ones come tripped and won't reset till wired correctly.

The spark when connecting the ground and hot to a load indicates to me there is probably an open neutral somewhere and that the light bulb is providing another pathway back to the panel for other loads. This coupled with a (probably) incorrectly wired GFCI could cause the shown issues.

Personally I use a meter and Wiggy type tester to check wiring and the built in buttons to test the AFCI. I don't have time to play to duplicate what's happening in the video but think the open neutral is probably pretty close.

Originally Posted by Gunnar Alquist

A lot of inspectors use a 3-light tester. While not a foolproof device, it should have shown this problem. It would have been nice if mbrooke had shown the same test with some kind of standard testing device instead of the jury-rigged light fixture.

Its a pig tail, and for those who understand electrical theory does it not prove a point for the sake of the argument?

- - - Updated - - -

Originally Posted by Bill Kriegh

Unfortunately the newest requirements for AFCI operation weren't put into effect until rather recently. That means there are many years worth of the devices floating around that will reset when wired wrong. The new ones come tripped and won't reset till wired correctly.

The spark when connecting the ground and hot to a load indicates to me there is probably an open neutral somewhere and that the light bulb is providing another pathway back to the panel for other loads. This coupled with a (probably) incorrectly wired GFCI could cause the shown issues.

Personally I use a meter and Wiggy type tester to check wiring and the built in buttons to test the AFCI. I don't have time to play to duplicate what's happening in the video but think the open neutral is probably pretty close.

That open neutral had to be in the GFCI. Circuit was ok and the GFCI was indeed wired correctly.

Originally Posted by Mbrooke

Its a pig tail, and for those who understand electrical theory does it not prove a point for the sake of the argument?

Proper test equipment is needed to show actual results.

Even the simplest and rudimentary test setups would have been more beneficial than using that light/pigtail/etc - take a simple three-wire plug, wire in three pigtails with lights, and allow the viewer to see what is going on with each conductor at-the-same-time.

I found following your videos a bit disconcerting because of the discontinuity of what you were trying to show with all the changing around of the probes ... basically ruined what you were trying to show.

A three-light tester would have served a purpose similar to a plug with three pigtails with lights in them ... not a visual, but sufficient.

Better yet would have been three meters wired together (see photo) to show what was going on when the plug was plugged in and the GFCI test button and reset button were operated.

The photo shows three meters, the top meter shows the voltage between hot & neutral (top two slots), the lower right meter shows the voltage between the hot & ground (two right side slots), the lower left meter shows the voltage between the neutral & ground (two left side slots) - this tells ... and shows ... what is going on with each ... and all at the same time for real-time results.

Originally Posted by Jerry Peck

Proper test equipment is needed to show actual results.

Even the simplest and rudimentary test setups would have been more beneficial than using that light/pigtail/etc - take a simple three-wire plug, wire in three pigtails with lights, and allow the viewer to see what is going on with each conductor at-the-same-time.

I found following your videos a bit disconcerting because of the discontinuity of what you were trying to show with all the changing around of the probes ... basically ruined what you were trying to show.

A three-light tester would have served a purpose similar to a plug with three pigtails with lights in them ... not a visual, but sufficient.

Better yet would have been three meters wired together (see photo) to show what was going on when the plug was plugged in and the GFCI test button and reset button were operated.

The photo shows three meters, the top meter shows the voltage between hot & neutral (top two slots), the lower right meter shows the voltage between the hot & ground (two right side slots), the lower left meter shows the voltage between the neutral & ground (two left side slots) - this tells ... and shows ... what is going on with each ... and all at the same time for real-time results.

Ok, so tell me then how this light bulb missed the failed GFCI and how in depth testing would have caught it instead.

Originally Posted by Mbrooke

Its a pig tail, and for those who understand electrical theory does it not prove a point for the sake of the argument?

For some reason - at least for me - the little three-light tester is easier to get the full picture than with the pigtail. Maybe it's just because I am attuned to using it at inspections. It has those nifty little lights that can be read at a glance. Press the button and the tester is still lit. Easy-peasy.

In my own small-minded, average Joe kind of way, part of me wondered if the pigtail was the problem, since it looks jury-rigged.

Originally Posted by Mbrooke

Ok, so tell me then how this light bulb missed the failed GFCI and how in depth testing would have caught it instead.

Huh?

Read again, then show me where I said that.

Originally Posted by Mbrooke

I
That open neutral had to be in the GFCI. Circuit was ok and the GFCI was indeed wired correctly.

How was the light on with an open neutral?

Originally Posted by Jim Port

How was the light on with an open neutral?

Exactly my point, it was not in the circuit and not when the GFCI was reset.

- - - Updated - - -

Originally Posted by Jerry Peck

Huh?

Read again, then show me where I said that.

Right here:

Even the simplest and rudimentary test setups would have been more beneficial than using that light/pigtail/etc - take a simple three-wire plug, wire in three pigtails with lights, and allow the viewer to see what is going on with each conductor at-the-same-time.

How would it have been more beneficial? Anything to light a light bulb will kill you 1000 times over.

Originally Posted by Mbrooke

Ok, so tell me then how this light bulb missed the failed GFCI and how in depth testing would have caught it instead.

Originally Posted by Mbrooke

Exactly my point, it was not in the circuit and not when the GFCI was reset.

Right here:

How would it have been more beneficial? Anything to light a light bulb will kill you 1000 times over.

You need to read what I wrote and what you wrote - then think about what we each said (our actual words, not what you are apparently trying to turn them into).

Hint: "more beneficial" does not equal "missed"

Originally Posted by Jerry Peck

You need to read what I wrote and what you wrote - then think about what we each said (our actual words, not what you are apparently trying to turn them into).

Hint: "more beneficial" does not equal "missed"

You don't need complex numbers to understand what happened here. A GFCI could trip, reset and denergize line to neutral loads. Yet it still remained live on the hot side. The whole point is to never trust the test reset button alone. Always rely on your plug in tester to verify that that lights have gone out.

Originally Posted by Mbrooke

Always rely on your plug in tester to verify that that lights have gone out.

Always did - that's how I started finding GFCI'S which would reset torreverse polarity, but which checked correctppolarity at first.

I eventually figured out that when the reset button was not pushed in at the center of the button, I could reset to reverse polarity most of the time.

A Leviton representative came to one of our local inspector meetings, I showed him (the GFCI was a competitor brand).

Yes, always use your tester first, trip the GFCI, reset and verify.

Originally Posted by Jerry Peck

Always did - that's how I started finding GFCI'S which would reset torreverse polarity, but which checked correctppolarity at first.

I eventually figured out that when the reset button was not pushed in at the center of the button, I could reset to reverse polarity most of the time.

A Leviton representative came to one of our local inspector meetings, I showed him (the GFCI was a competitor brand).

Yes, always use your tester first, trip the GFCI, reset and verify.

What did the Lev rep say?

Originally Posted by Jerry Peck

Always did - that's how I started finding GFCI'S which would reset to reverse polarity, but which checked correct polarity at first.

Jerry,

I saw that a few times myself. As a matter of fact, I had one switch hot/EGC. Totally freaked me out! I should have asked for the device and dismantled it to see how that could occur. Did you find out what would cause that?

Originally Posted by Mbrooke

What did the Lev rep say?

Originally Posted by Gunnar Alquist

Jerry,

I saw that a few times myself. As a matter of fact, I had one switch hot/EGC. Totally freaked me out! I should have asked for the device and dismantled it to see how that could occur. Did you find out what would cause that?

When I first told the Leviton representative what was happening, his response was "That's impossible. It can't reset reverse polarity."

Several of our local inspectors were finding the same problem, with all the brands of GFCIs. We brought this up at our state meeting, others started checking it and found the same thing happening (this was before we met the Leviton representative).

When I wired up the known "bad one" I had been given, the Leviton representative saw for himself the it would reset to reverse polarity. He took that one with him back to their engineers for them to take apart.

As I recall, he came to another local inspectors meeting (we had monthly meetings) to let us know what they found. He said it was something in the switching mechanism as I recall (that over 20 years ago).

Gunnar found something different than we found - probably a similar cause (the internal switch).

The important thing to take from this is to document what you find, check with other inspectors to see if they are finding the same thing, then get with manufacturers to let them know what you are finding.

The manufacturers redesigned the switching mechanism after finding out what we were finding.

They redesigned the switching mechanism a few years later to address the line/load miswiring Iissue home inspectors were finding - GFCIs now are made 'tripped' and don't reset to 'on' if line/load usrreversed. This eliminates line/load miswiring.

Originally Posted by Jerry Peck

They redesigned the switching mechanism a few years later to address the line/load miswiring Iissue home inspectors were finding - GFCIs now are made 'tripped' and don't reset to 'on' if line/load usrreversed. This eliminates line/load miswiring.

Longtime ASHI members can confirm this, as I recall, ASHI participated in a program to document GFCI line/load miswiring with ASHI inspectors documenting the frequency of GFCI miswiring issues. I know ASHI members in Florida did. I think it was in the mid 1990's.