GFCI Receptacle Troubleshooting - A short video

[JFletcher]

There are a lot of times here at MH forum where a poster poses a problem and asks for help in diagnosing it. The old saying is "A picture says a thousand words", yet rarely are pictures included, let alone video.

I decided to shoot a short video (2:32) of the GFCI receptacle tripping in my hotel room. Since I'm on vacation, I don't have the standard tools with me, however I was able to test it anyway. My basic troubleshooting logic is as follows:

1) Identify problem:

In this case a GFCI Receptacle that is intermittently tripping.

2) Attempt to duplicate problem (in multiple ways if possible):

Here, a nightlight and an AC adapter were used. They were tested upon insertion, removal, and switching on and off. I was able to duplicate the problem in more ways than one.

3) Figure out cause of failure:

Since I used two different loads on the receptacle, it's unlikely both of them have ground faults, which i proved by plugging them both into the bathroom GFCI receptacle and had no issues with tripping. The GFCI is likely bad, tho it could be a loose wire which breaks contact on the slightest movement of the receptacle.

4) Figure out how to prevent future failures:

GFCI receptacles simply dont last forever. Aside from using quality parts, here is nothing here that can be done to stem future failures. If it were an outside GFCI with an old flip cover, rather than a new bubble cover, I could replace that as well as make sure the gasket is in good condition and installed correctly.

5) Fix the problem:

Maintenance will be by shortly to replace the receptacle. If it were me I'd make sure none of the wiring in the box is damaged, make sure the box is bonded (if metal), check polarity, line/load, all connections are tight, replace the receptacle, and check its operation.

Back to the video, I encourage MH members to post them more often as they can be invaluable when others are trying to help. It's much easier to visualize a problem if there is a video.

Apologies for the somewhat shaky and out of center video, it's the first one I've shot on this cell phone.

Here's the video:

 

[mbrooke]

Thanks for the vid


If you really want to see a jaw dropping failure look at this



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



Just want to say Leviton resolved the issue promptly with a replacement GFCI free of charge.

 

[JFletcher]

I remember your topic on it, and that video. The inside contacts were welded together causing the no-trip condition under the induced ground fault.

Videos like yours are great in helping everyone understand/visualize the numerous ways GFCI receptacles can fail. That Leviton got involved in yours was great, and they probably wouldnt have w/o the video.

 

[mbrooke]

I remember your topic on it, and that video. The inside contacts were welded together causing the no-trip condition under the induced ground fault.

Videos like yours are great in helping everyone understand/visualize the numerous ways GFCI receptacles can fail. That Leviton got involved in yours was great, and they probably wouldnt have w/o the video.

Yup- Leviton even analyzed the GFCI and found the contacts welded shut. Still very pleased with the response, most other companies would have looked the other way.

GFCI can fail in many ways, sometimes seemingly bizarre.

 

[gar]

180117-2034 EST

JFletcher:

I agree that "a picture can be worth a thousand words". However, the picture may require 1,000,000 bytes of storage vs 1000 * 5 = 5000 bytes.

On the other hand a video may require a hugh amount of storage or bandwidth.

Can I effectively convey or teach something better with video (moving pictures) than with written material and slides. In most cases I think not.

A moving picture is needed when there is importance to the movement, but otherwise my experience is that still pictures with written information are better. Easier study, better random access, and easy of comparing different time periods.

The above videos did not discuss how a GFCI works, what the experimental conditions were, what the test equipment was, and the characteristics of the components in the experiment.

A GFCI has a pair of input terminals, and a pair of output terminals at a minimum. This is all the terminals required for the GFCI function. The current to and from the two input terminals pass thru a differential current sensor to a DP switch to the output terminals. Power to operate the electronic circuit in the GFCI is obtained from the input terminals. The electronic circuit determines if there is an excessive current difference thru the current sensor. If there is, then the double pole switch is opened. If tripped, then the unit remains tripped until manually reset.

If the GFCI is a GFCI and a duplex socket, then the output (load side) neutral and hot lines of the outlet are the same as the output terminals. There will also be EGC pins in the socket and a straight pass thru bus from the input EGC terminal to the output EGC terminal. This EGC bus is solid and not switched. The EGC bus is in no way used or connected to the electronics. There will be 100s of megohms of leakage between the power wiring and the EGC. Thus, virtually no current flow.

In your video the loads applied were two terminal devices and appeared to be in contact with relatively high resistance material. Air for the night light. So nothing in the loads likely to produce an unbalanced load to the GFCI that would trip the GFCI.

I have created a lot of video, some for instruction where motion was of importance or produced an interesting effect.

I also have some video where I show up for a few seconds along with other classmates when I was in first grade. One or more large arc lights were used for illumination with power from a gas driven generator on a truck. Large heavy wires. I believe the original was on 35 mm. I made a copy from a 16 mm copy that my teacher had. The original is in the National Archives.

I made extensive use of video I took of each machine on an assembly line for timing and functional analysis. With this study I was able to improve the net thruput from less than 100 parts per hour to over 120 on average. Sometimes we got to close to 130 parts per hour for a full hour. That was close to absolute saturation. Here the motion and timing information was extremely important. Time study people got pretty much the same timing information, but did not contribute to the thruput improvement.

.

 

[LarryFine]

180117-2034 EST

If the GFCI is a GFCI and a duplex socket, then the output (load side) neutral and hot lines of the outlet are the same as the output terminals. There will also be EGC pins in the socket and a straight pass thru bus from the input EGC terminal to the output EGC terminal. This EGC bus is solid and not switched. The EGC bus is in no way used or connected to the electronics.

A correction, if I may: I used to think that the load out terminals were directly tied to the receptacle slots, but when the GFCI trips, they are separated, at least with the brands I've explored this on.

By the way (and not a correction), the EGC is used by the test button on plug-in GFCI testers. It has to be used because the plug-in tester does not have access to the neutral ahead of the differential current sensor to mimic a shock.

 

[JFletcher]

"The above videos did not discuss how a GFCI works..."

Quite correct. Nor did I remove or disassemble the faulty GFCI for analysis. Making a classroom quality video and doing an NTSB level failure analysis would be a long video, or possibly a term paper length report. I chose brevity over completeness for the sake of getting information across as quickly as possible. I also made the video solely for members of the forum, electricians, who (usually) understand the basics of GFCI receptacles and operations.

"In your video the loads applied were two terminal devices and appeared to be in contact with relatively high resistance material. Air for the night light. So nothing in the loads likely to produce an unbalanced load to the GFCI that would trip the GFCI."

Yet plugging them in caused repeated trips. I also checked with a crock pot and a toaster, both also 2 prong, and got tripping. What I might conclude from this is that older/damaged GFCI can fail (Trip) even with a balanced load. This is a failure mode i have not seen before, thus i figured I'd video it. Even if the video is useless to anyone else, I now have a visual record of what occurred that wont deteriorate with time or memory.

The usual procedure for any GFCI failure is device replacement. It's only rare cases where RF or problems on other circuits are the culprit. Since 99% of GFCI problems can be solved with device (receptacle or breaker) replacement, a $15-40 part, it doesn't make economical sense to spend a lot of time into the troubleshooting. Sure, I could have saved the receptacle and pulled it apart later for my own edification, however as the parts are non-rebuildable, there is really no point imo.

While my video may have been unnecessary in describing this problem, I find many cases here where at least some pictures would be helpful. For example, no matter how well someone describes what they see in a panel, if they were to ask "are there any violations here?", a picture is better than all their words. Seeing a video of a problem, like LED strobing, or watching the readings live as someone uses a DMM (or even where they are testing), or hearing a motor that doesn't sound right can be much more helpful than reading about it.

 

[jumper]

Electricians are wierd.

Goes on vacation and proceeds to rewire hotel room.

 

[JFletcher]

Electricians are wierd.

Goes on vacation and proceeds to rewire hotel room.

:lol:

I did resist temptation in fixing it myself. I'm paying enough here for someone else to do it. Only reason I shot the video and made these posts is snow snarled up everything around here so our plans for the day were cancelled.

 

[jumper]

:lol:

I did resist temptation in fixing it myself. I'm paying enough here for someone else to do it. Only reason I shot the video and made these posts is snow snarled up everything around here so our plans for the day were cancelled.

And when the guy shows up to fix it you will spend a hour BSing with him while ignoring your wife, who will be sitting on the bed muttering “they are both nuts”.

 

[gar]

180117-2429 EST

LarryFine:

In the Leviton that I have looked at it is actually has a 4 pole switch on the output with the input being 2 poles. Can't presently access my photos of the device. When latched the receptacle neutral and hot are fed by a set of neutral and hot contacts, and the output terminals by another set of terminals.

Plug-in GFCI testers are not valid testers for this reason. The only approved test method is use of the internal resistor, and that does not require EGC. Therefore, a GFCI receptacle could be used on a two wire circuit.

My point about the GFCI EGC connection is that it is not required for the GFCI to work.


JFletcher:

To make a good teaching or training video requires good equipment. Meaning at least a tripod for a steady image, adequate lighting, a focus on the important elements (video is not high resolution), and good audio without unnecessary background noise. General broadcast TV does a very poor job with good equipment. They just don't care.

Then one needs to plan a story, develop an outline and script, be concise, and do a lot of retakes and editing. To do a good job I would estimate 10 to 50 hours possibly more per hour of output. This is not easy. I run away from most random video on the Internet. You can not tolerate watching it.

To record something for your future reference do whatever is easiest.

The following is quite correct.

While my video may have been unnecessary in describing this problem, I find many cases here where at least some pictures would be helpful. For example, no matter how well someone describes what they see in a panel, if they were to ask "are there any violations here?", a picture is better than all their words. Seeing a video of a problem, like LED strobing, or watching the readings live as someone uses a DMM (or even where they are testing), or hearing a motor that doesn't sound right can be much more helpful than reading about it.

.

 

[LarryFine]

Yet plugging them in caused repeated trips. I also checked with a crock pot and a toaster, both also 2 prong, and got tripping. What I might conclude from this is that older/damaged GFCI can fail (Trip) even with a balanced load.

It seemed to me that the tripping occurred every time something was unplugged, suggesting a mechanical defect, not an electrical/electronic one.

 

[Little Bill]

JFletcher:

To make a good teaching or training video requires good equipment. Meaning at least a tripod for a steady image, adequate lighting, a focus on the important elements (video is not high resolution), and good audio without unnecessary background noise. General broadcast TV does a very poor job with good equipment. They just don't care.

Then one needs to plan a story, develop an outline and script, be concise, and do a lot of retakes and editing. To do a good job I would estimate 10 to 50 hours possibly more per hour of output. This is not easy. I run away from most random video on the Internet. You can not tolerate watching it.

To record something for your future reference do whatever is easiest.



.

Com on, the OP was just trying to provide an aid to his description of a problem. He was not trying to make a full length production for sale to a college training course. Maybe he should have rented a scope, cleaned off the furniture in the hotel room, and set up a test bench.

I appreciate anyone going the extra mile to help understand a problem, or at least helping to convey what they are trying to describe.

 

[JFletcher]

It seemed to me that the tripping occurred every time something was unplugged, suggesting a mechanical defect, not an electrical/electronic one.

Possibly so. Still, the course of action is the same: GFCI replacement

Com on, the OP was just trying to provide an aid to his description of a problem. He was not trying to make a full length production for sale to a college training course. Maybe he should have rented a scope, cleaned off the furniture in the hotel room, and set up a test bench.

I appreciate anyone going the extra mile to help understand a problem, or at least helping to convey what they are trying to describe.

Thank you.