Gfci won't trip

[jkrauss]

I installed a new type gfci outlet on a two wire circuit in a residence and it won't trip . I had a old style gfci and it will trip with a gfci tester. Any ideas why the new gfci won't trip ?

 

[GoldDigger]

I installed a new type gfci outlet on a two wire circuit in a residence and it won't trip . I had a old style gfci and it will trip with a gfci tester. Any ideas why the new gfci won't trip ?

Your GFCI tester relies on an internal connection to the ground/EGC pin to carry the test current. Your ungrounded receptacle cannot supply that.
Use the test button built into the receptacle instead.

 

[augie47]

Unless I'm behind the curve, GFCI testers will not trip a GFCI installed on a two-wire circuit.
I have no idea why the old one tripped. (The old one did not have the neutral bootlegged to the ground terminal, did it?)

 

[jap]

You did terminate the wiring to the "Line Side" terminals correct?
lot of the new GFI's will not let you mess up and wont work if they're wired wrong, but, some of the old ones you could land the wire on the load side terminals and completely bypass any gfi protection.

JAP>

 

[roger]

Your GFCI tester relies on an internal connection to the ground/EGC pin to carry the test current. Your ungrounded receptacle cannot supply that.
Use the test button built into the receptacle instead.

I agree, the onboard test button is the only way to test a two wire connected GFCI.

Unless I'm behind the curve, GFCI testers will not trip a GFCI installed on a two-wire circuit.
I have no idea why the old one tripped. (The old one did not have the neutral bootlegged to the ground terminal, did it?)

I would suspect the bootleg connection too.

Roger

 

[mhanson]

I've heard and read many times that the only UL recognized test method for my GFCI receptacle is the on board test button, but I have not been able to locate the UL statement, or section number.
if someone is able to describe the UL section I would appreciate it.

Thank You
M Hanson

 

[iwire]

I've heard and read many times that the only UL recognized test method for my GFCI receptacle is the on board test button, but I have not been able to locate the UL statement, or section number.
if someone is able to describe the UL section I would appreciate it.

Thank You
M Hanson

NEC 110.3(B) and the instructions that come with the device.

 

[kwired]

I've heard and read many times that the only UL recognized test method for my GFCI receptacle is the on board test button, but I have not been able to locate the UL statement, or section number.
if someone is able to describe the UL section I would appreciate it.

Thank You
M Hanson

NEC 110.3(B) and the instructions that come with the device.

Sorry Bob, but he asked for UL documentation, you gave him the NEC section that indirectly points us back to what he was asking for.

 

[iwire]

Sorry Bob, but he asked for UL documentation, you gave him the NEC section that indirectly points us back to what he was asking for.

LMAO

I gave them a legitimate code cite for it.

I don't believe there will be any available UL document that spells this out.

But apparently you do so how about posting it?

 

[mhanson]

LMAO

I gave them a legitimate code cite for it.

I don't believe there will be any available UL document that spells this out.

But apparently you do so how about posting it?

Thank you, I appreciate the input.
I am aware of NEC 110.3 (B)
In reviewing previous posts I see UL being referenced as the authority on field testing. However the posts are not providing a UL reference number.
I would simply like to substantiate the claim.

Thank you again
M Hanson

 

[dfmischler]

I agree, the onboard test button is the only way to test a two wire connected GFCI.

I'm sure the test button is the only sanctioned test because it is safe to run. But, to me, it does not feel very much like an independent test.

I can think of a real test that does not depend on the test button, but I am not sure what you would have to do to make it safe to perform. If you ran a wire from a real (panel) ground through a 10K ohm resistor to the hot side of the protected outlet, then that should draw about 12 mA in an unbalanced manner. If that does not trip the GFCI then it is defective.

 

[K8MHZ]

I'm sure the test button is the only sanctioned test because it is safe to run. But, to me, it does not feel very much like an independent test.

I can think of a real test that does not depend on the test button, but I am not sure what you would have to do to make it safe to perform. If you ran a wire from a real (panel) ground through a 10K ohm resistor to the hot side of the protected outlet, then that should draw about 12 mA in an unbalanced manner. If that does not trip the GFCI then it is defective.

I have used a 3 prong adapter and connected the grounding tab to a known good ground, usually a pipe in the kitchen, but sometimes all the way back to the panel. Then I plug an ice cube type GFCI tester in the adapter. So far, it's worked the handful of times I have used it.

As far as safe, I thought so for electricians, but maybe not for the general public.

 

[iwire]

I'm sure the test button is the only sanctioned test because it is safe to run. But, to me, it does not feel very much like an independent test.

But if you really look into how the test button does it's job you will find it is a reliable, real test of the unit. It does not simply drop out the relay, it shunts current via a resister around the CTs causing an current imbalance resulting in a trip it the device if working.

 

[iwire]

Thank you, I appreciate the input.
I am aware of NEC 110.3 (B)
In reviewing previous posts I see UL being referenced as the authority on field testing. However the posts are not providing a UL reference number.
I would simply like to substantiate the claim.

Thank you again
M Hanson

UL is quoted as it is UL that is typically used as the listing service for items like GFCIs.

Unless you can find a GFCI with directions that tell you that you can test the device using an external tester doing so would not count as a 'test' if you have reason to prove they are tested. As you might on a job with OSHA compliance.

 

[dfmischler]

But if you really look into how the test button does it's job you will find it is a reliable, real test of the unit. It does not simply drop out the relay, it shunts current via a resister around the CTs causing an current imbalance resulting in a trip it the device if working.

I remember looking at this quite some time ago, and being impressed at the quality of the test circuit design (i.e. I could not find any failure case that would show a wrongly successful test, unless the resistor is way out of spec). But I really like the fact that you can take an external tester and verify that a ground fault in the downstream load wiring will trip the unit. That kind of test gives me the feeling of confidence that I want.

I have used a 3 prong adapter and connected the grounding tab to a known good ground, usually a pipe in the kitchen, but sometimes all the way back to the panel. Then I plug an ice cube type GFCI tester in the adapter. So far, it's worked the handful of times I have used it.

As far as safe, I thought so for electricians, but maybe not for the general public.

Your way sounds pretty good to me. I was originally thinking in terms of a piece of wire with a resistor in it, and if you connected the hot end first then the end that is supposed to be grounded could potentially provide a shock with dangerous current. I would have no problems using that setup under laboratory conditions, but it doesn't seem safe for the workplace. I had concluded that a momentary switch would be needed, but wasn't really sure if that was enough.

 

[roger]

I'm sure the test button is the only sanctioned test because it is safe to run. But, to me, it does not feel very much like an independent test.

I can think of a real test that does not depend on the test button, but I am not sure what you would have to do to make it safe to perform. If you ran a wire from a real (panel) ground through a 10K ohm resistor to the hot side of the protected outlet, then that should draw about 12 mA in an unbalanced manner. If that does not trip the GFCI then it is defective.

You can go through all kinds of acrobatics to perform tests but it is unnecessary, the following is an excerpt from this EC&M article

Testing GFCIs on nongrounding circuits.

Testers should not be used to test GFCIs installed as replacements for two-wire receptacles on nongrounding circuits. By design, GFCI testers will not test a GFCI protecting a 2-wire circuit and can expose the user to a potential shock.

The test button integral to the GFCI applies the test current between hot and neutral. This is not the case with GFCI testers; the test current in these devices is applied between hot and the equipment ground. Therefore, if there is no equipment ground, no test current will flow. If there are any exposed metal parts connected to the receptacle grounding contact (such as a metal face plate or a weatherproof cover), they will be energized by the test device. Since some of the test devices apply up to a 30mA test current; using such a tester on a 2-wire circuit while touching a metal cover plate could result in an uncomfortable shock.

Keys to testing GFCIs.
* Use the GFCI integral test button.
* Check for line-load reversal using the GFCI integral test button supplemented with a lamp or appliance.
* Be sure that any GFCI tester used is listed to the current UL standard.
* Do not use GFCI testers on two-wire nongrounding circuits.

Roger

 

[kwired]

LMAO

I gave them a legitimate code cite for it.

I don't believe there will be any available UL document that spells this out.

But apparently you do so how about posting it?

Solenoid type voltage testers and many low impedance volt meters draw enough current to trip a class A GFCI if you test from protected hot to any unprotected component of any differing potential (does not necessarily have to be ground).

This does introduce some shock hazard potential. You can pull the GFCI out and test from line side hot to load side neutral or load side hot to line side neutral with solenoid tester or low impedance meter and by doing so you are putting current in nearly the same path the test button on the device places current, just through your meter instead of the on board resistor and test switch.

 

[kwired]

LMAO

I gave them a legitimate code cite for it.

I don't believe there will be any available UL document that spells this out.

But apparently you do so how about posting it?

Maybe I should have followed with a smiley, as I was playin with ya more then anything here. I am not proficient at all on UL document numbers or methods of organization of them, so I have no idea where to find such a document if one exists.

 

[dfmischler]

You can go through all kinds of acrobatics to perform tests but it is unnecessary, the following is an excerpt from this EC&M article




Roger

So how do you satisfy yourself that you didn't inadvertently miss an outlet somehow if you can't test them all?

 

[roger]

So how do you satisfy yourself that you didn't inadvertently miss an outlet somehow if you can't test them all?

I don't understand the point of your question, you could miss an outlet in any testing scenario no matter how complicated you make it.

Roger