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

I hear ya…and as i said I liked the Video. I however, do not think that HI’s should partake in the effort and they should simply check the operation of the GFCI’s in accordance with their SOP only.

I also did not state that it was not concerning, I simply do not believe that HI’s will beheld liable for such failures as long as they follow their SOP correctly is all I am saying. When someone goes above and beyond the SOP, as many are doing, they open themselves up for issues.

Hey I am all about preservation of life, kinda why I serve on CMP’s as the many transcripts of my arguments at the NFPA hearings can attest too…The HI in my opinion (and can be the wrong one in your eyes, I am ok with that my friend) needs to stay within their scope of practice and stick to testing that has been standardized in their industry.

Do I think the GFCI Manufacturer should respond to what you believe if a concern, Yes but that is not the HI’s mission at this point.

Thanks, your recognition means a lot to me :slight_smile:

If the SOP involves test and reset while a plug in tester is in place I would call that sufficient.

Id say we have a problem… What I am hearing is that when people truly engage in through inspection that can save lives they are open to punishment. That is truly disheartening.

Could it be that such testing needs to be re-evaluated? Bear in mind everyone knew GFCIs were not being tested monthly long before UL943 mandated self testing.

It should be terms of double checking power has been cut. I am not the only one who has come across this which leads me to think this is a common concern.

Ok…here are my responses in RED.

You have missed all the posts where someone questions why it does not work on a 2 wire circuit.

I do not doubt the life saving abilities of GFCIs, in fact I have long believed GFP/GFCI should originate at the panel board for each circuit because of the protection it offers there, protection which actually goes beyond electrocution.

However, how do we know this one receptacle is just a one in a billion defect? If we have a failure mod which is actually masked by the manufacture’s recommended testing, something must be done to rectify that.

HIs can start by ensuring receptacles are truly dead. This gives piece of mind to the buyer, and simultaneously exposes more of these failures. I know you mentioned Mike Holt, FWIW I started a thread on it. You are welcome to chime in (I hope you do). We are debating if I should send this GFCI into Leviton or take it apart myself:

http://forums.mikeholt.com/showthread.php?t=178511

I know they gripe, but its all fueled by their feelings regarding AFCIs. Besides that they appreciate your input, at least I do. I think its solid. We may never agree on AFCIs, but I certainly do not have the code knowledge you do hence my deep respect for you.

Well Thanks Fella…but as always I am nothing special and am 100% sure the folks on MH Forum will agree…lol…and I am sure their are many things I have no clue about so with that said…I respect you and yours position on AFCI’s and always welcome the chats about it. However, nice to hear someone respects me at least…:wink:

Actually getting serious for a minute, I would not take it apart as the first line of defense of a manufacturers is to say you did while taking it apart or something similar…I would first document all your procedures and get me something that spells out your concern and I will see that it gets in the hands of folks who will have to provide answers…I can get it in the hands of the GFCI Section and so on.

Thanks :slight_smile: But again, I have learned a lot from you regarding code and the manufacturing process :mrgreen:

I did nothing special, just discovered that when I plugged in a plug in tester it glowed as open neutral when I pushed the test button and latter discovered that if I take a 100 watt lamp line to ground it lights up when the GFCI is tripped. I documented that in the video.

My concern is that while the GFCI test and reset normally, even with the LED going out and line to neutral loads going out, it still passed current line to ground.

If you can get a hold of the GFCI division I can mail them the GFCI in question.

Ok I will try to make some connections and see who might be interested on the manufacturers side to review the videos first and possible detect their level of interest. While I can’t promise anything I will indeed make the effort.

Anything is good. If you would like you are welcome to send Bryan Holland an Email as he might be interested in this is well. Either way, thank you, much appreciated :mrgreen:

Ok, I have sent the information and videos to some folks higher up the food chain than the NEMA Rep Bryan at this point. They will hopefully pass it to the right folks for comment. Only time will tell at this point.

Just for reference…do you feel the 100W lamp is relevant to the test. In other words, when you press the test button on the GFCI did you also detect a reading on a volt meter from the grounded mounting screw to the ungrounded terminal.

If you detected voltage on this application (at the least via the volt meter) then you might have a noninvasive test that could be incorporated into a inspection procedure. However, the rigging up of a 100 W Lamps device is not something that would ever make it into a procedure for HI’s.

So if your same test would yield detection from the ungrounded to a grounded screw after the test button is pressed then that is something that can be incorporated in general practice. However, did you actually try that or is the 100 W load imperative to the test itself.

Just trying to help express your issue if they contact me…and wanted to make sure you had all the answers for their probably questions.

Awesome, thank you :slight_smile:

I feel that in my case the 100 watt incandescent lamp is relevant. I choose that specifically in my video to prove a point the GFCI was passing power. I did not use a volt meter during the lamp test.

Most likely not, but a plug in tester would. Again, I choose a lamp not so much to advocate that HIs do the same, but rather to prove a point that I had a GFCI experiencing a failure mode where the hot terminal remained live despite passing the manufacturer’s recommended test.

The bulb was just to prove a point, a real world test/verification of proper functioning would be far less crude and more professional in nature.

Let me know how I answered them. I can answer any question need be.

Well since you have the “defective” device I would recommend you do the test with a volt meter and simply show one lead in the ungrouded plug opening and the other at the grounded screw as you have done with the lamps. Showing voltage readings (and subsequently current) may be a more test friendly approach. Theoretically speaking…(and I have done it here at all our GFCI’s)…if the test button trips you should get nothing between the ungrounded plug port and the grounded screw…showing a voltage reading on a circuit that is supposed to be DEAD to me would be better…while I liked the arc’s and so on…just saying data speaking of course.

The GFCI is already out of the wall, so I can’t do voltage readings. I did a continuity test and got 0 ohms line screw to line slot.

Ok…The video’s are in the Manufacturers Hands as we speak. They are aware of it and I am hoping they comment on it, either at the video location or via direct email. I will post accordingly to their response for you when or if it arrives.

Thank you, and please keep me in touch :slight_smile:

Ok, wow, awesome news and let me say Thank You all over again! :mrgreen: I sent the GFCI into Leviton and they emailed this over to me:

http://forums.mikeholt.com/attachment.php?attachmentid=15792&stc=1&d=1475165157

The burned out coil is indeed from me because I kept holding the bulb in there to see if anything would happen until I smelled burning plastic.

Thanks Martin and Paul for sharing, and going the extra mile on this.