testing gfci's

Status
Not open for further replies.
J

jem.nola

Member
Location
New Orleans, Louisiana
I'm working on a home built a year ago and am now doing remodeling to it. I'm using my volt meter to simulate a ground fault by putting one lead on hot and the other on metal faucets and it's not tripping the gfci. I've even went from hot to water and I'm getting 120volts and it's not tripping. I've checked the grounding from the main and subpanel which are side by side. subpanel wasn't grounded and water line at the main wasn't grounded. (how it passed an inspection makes me wonder) Anyone has any ideas where I can look or something I can try would be much appreciatted. Thank you.....

James NOLA
 
What type of voltmeter? If it is a solenoid type (wiggy) then there may be a problem. If it is an electonic/digital VOM type then it does not create a current path and so it won't trip a GFCI.
 
The resistance of the meter is high enough that it doesn't provide a 4-6ma imbalance on the GFCI device. Try an old "wiggy" style tester. That should do the trick. This point is completely moot if the problem is that the GFCI is not functioning properly.
 
At first, I had used a fluke digital meter, then I had used a square d wiggy and I still haven't been able to trip out the gfci. If it was one circuit I would try inspecting the other devices and wiring, but it's through the whole house. The kitchen, 3 1/2 bathrooms and garage. They are doing the same thing I feel that this a serious problem and I can't and won't ignore it. I want to simulate a ground fault and don't want to rely on the test button on the gfi and one of those 3 prong test buttons. Any clues on what I check throughout the house, any grounding or bonding of any sorts. Any advice would be taken greatly and thank you for your time and knowledge for those of you that reply.
 
With a house that new, I wouldn't believe that the water pipes are conductive.

If you don't trust a GFCI tester plugged into a GFCI receptacle, then go get a "grounded plug cheater", you know, an adapter, that one plugs a grounded cord cap into, that permits one to plug into a two wire ungrounded receptacle.

Connect a long piece of #12 or #14 to the adapter ground and run the wire back to the electrical service and put it under a screw.

Plug the GFCI tester into the cheater. Pressing the GFCI tester test button puts an exact resistor between the hot and the ground. The ground the GFCI tester test button is now forced to use is the #12 or #14 and, in turn, the connection at the electrical service.

Never mind the water pipes.

If a GFCI tester is not good enough for you when plugged into the GFCI receptacle, then using a voltmeter to "simulate a ground fault" is much less acceptable. There is no control over the fault current in the voltmeter, because very few people deal accurately with the input impedance of the various voltmeter manufacturer's makes and models. You would need a variable resistance bridge to put either in series or in parallel with the voltmeter to create a precise 5 milliamp fault current.
 
infinity said:
... The resistance of the meter is high enough that it doesn't provide a 4-6ma imbalance on the GFCI device. ...
Click to expand...
The high impedance of the DMM is what I also suspected ... but I was curious ... so I just tried a little experiment myself as I have two Fluke DMMs.

I connected two DMMs in series, with one measuring VAC, the other amps AC.

The result was that the voltmeter only drew 0.12 milliamp when reading across 122 VAC. (and conecting across 244 VAC drew 0.24 milliamp)

Obviously, a Fluke 87 will never come close to tripping a properly operating GFCI.

BTW, this is my first post here, although I've been "observing" for quite some time. Great forum!
 
Last edited:
You said that you have a "Wiggy". Run an extension cord to another receptacle and test with the Wiggy from the GFCI hot to the extension cord neutral.
 
You could easily build a test device that will test the GFCI and the grounding wire. Take a table lamp and cut off the cord cap. Rewire it to a 3-prong plug and put the grounded wire (the one that goes to the screw spart of the lamp base) on the grounding pin instead of the neutral prong (the neutral pin will be unused).

If your grounding system is good, turning on the lamp will trip the GFCI. This will only test the grounding system up to the GFCI let through current which isn't very high. But at least you'll know you have some sort of wire there making some kind of connection.

Instead of a lamp, you could just put a resistor in the plug body and label this "GFCI Test plug". You would need about a 20K ohm resistor for a 120V circuit to trip a GFCI. A 1 watt resistor would be ideal (or two half watt ~40K resistors in parallel) or a push button momentary switch in series with the resistor, as you don't want a small resistor sitting energized on a receptacle that didn't trip.
 
This may not sound as intellectual as fabricating your own device, but why not just buy a UL listed GFCI interrupter for about $22. The time saved on troubleshooting why your meter or Wiggy's wont trip the GFI will pay for itself. You will also save time testing all the load side recpts. I would imagine it's quicker to just plug something in rather than worry about placing your meter on the floor and sticking in the leads. I've done over 200 units in the past 4 years and haven't had a problem with one yet. Just my 0.02
 
burga said:
This may not sound as intellectual as fabricating your own device, but why not just buy a UL listed GFCI interrupter for about $22. The time saved on troubleshooting why your meter or Wiggy's wont trip the GFI will pay for itself. You will also save time testing all the load side recpts. I would imagine it's quicker to just plug something in rather than worry about placing your meter on the floor and sticking in the leads. I've done over 200 units in the past 4 years and haven't had a problem with one yet. Just my 0.02
Click to expand...
well i must agree on this one. I use the interrupter myself and it does save time..
 
The only true way to test a GFCI is it's on board test button, exterior trip units work in most cases but are not truly recognized as a fool proof test method.

Roger
 
dingokangaroo said:
thats what the tester is...push the button and it acts like a short. hence tripping the gfci
Click to expand...

Not a short, that would not test the GFCI.

What I am asking is why are we talking about using wiggys to a water pipe to trip the GFCI when they come with built in testers and are correct way to check a GFCI.

When checking a normal outlet supplied by a GFCI the GFCI tester will do the job in a modern building (one with EGCs)
 
iwire said:
Not a short, that would not test the GFCI.

What I am asking is why are we talking about using wiggys to a water pipe to trip the GFCI when they come with built in testers and are correct way to check a GFCI.

When checking a normal outlet supplied by a GFCI the GFCI tester will do the job in a modern building (one with EGCs)
Click to expand...
correct. as for the wiggy...i have no clue why.
 
"Not a short, that would not test the GFCI."
Actually yes its a short. The test button inserts a resistor between the hot and neutral to create the imbalance or ground fault. The plug in testers create the imbalance between hot and ground and if it does not trip then there is no ground at the GFCI, hence the correct (see 110.3(B) and NEC compliant way is to use the GFCI's test button
 
roger said:
The only true way to test a GFCI is it's on board test button, exterior trip units work in most cases but are not truly recognized as a fool proof test method.

Roger
Click to expand...

As Roger and later Tom have stated the test button is the only approved way for testing a GFI.

In a small presentation from Mr. Joe Ross (who sits on both a MA panel and NFPA panel) he has represented manufacturers in advocating the test button as the only approved method. He stated that any other method will lead to premature failing of the GFI device.

He informed us that he did presentations for over 2 years trying to convince inspectors to use the test buttons only, with a success rate of less than 10%. At least the inspectors were honest with him.:lol:
 
j_erickson said:
...He stated that any other method will lead to premature failing of the GFI device...
Click to expand...

Now that statement brings out my skeptical nature. As I understand it, the test button shunts a small current to the line side grounded conductor, while a plug in tester shunts roughly the same small current to the equipment grounding conductor. The imbalance seen by the sensing coil should be roughly the same by either method. I can't see how a plug in tester could cause premature failing of the device any more than use of the test button. :roll:
 
Status
Not open for further replies.
Top