GFCI Questions

[inspectorwill]

I frequently run into situations with GFCI's that I would like to educate myself on. I understand the general workings of a GFCI but run into peculiar situations that I would like to understand. Here they are:

-Will a GFCI function when the hot and neutral are reversed? Testers will not trip but generally the receptacle test button will trip the GFCI.

-Three garage receptacles on same circuit and one is GFCI type. What has occurred in a circuit when the two non-GFCI garage receptacles do not trip the GFCI when tested with an external tester but when the GFCI is tripped with the test button it cuts power to the non-GFCI receptacles?

-What is the more reliable way to test a GFCI (test button or external tester)? I run into instances where testers will not trip the GFCI on a correctly wired and grounded receptacle but the test button will.

Thanks

[John Kogel]

I'm no expert, but if the external tester does not trip the GFCI, I call for a repair. It stands to reason that the same condition would occur in an actual ground fault.

In your garage scenario, the downstream outlets don't trip the upstream GFCI, maybe a poor ground to those outlets?

Thirdly, I don't trust the little cheapo 3-light testers, so even tho I use them, I will check with another one if I get strange results.

[inspectorwill]

Thanks John. I use both the three light testers and an Ideal Circuit Analyzer just to be sure. Do you know why the GFCI would not trip when the receptacles downstream have a poor ground? A GFCI device itself does not require a ground to function as it monitors the current on the hot and neutral. This has always confused me. Same instance as an older home with no grounded circuits and GFCIs installed in Kitchens and Bathrooms where the receptacles downstream do not trip the GFCI.

[Marc]

I think the topic of the OP is more about the tester than the GFCI receptacles, though you may not have intended it that way.

Marc

[John Kogel]

You can try to figure it all out from this schematic. I know I'm going cross-eyed from looking at it. []

It looks like a transistor is used as a switch. I'm not sure if the test button closes more than one switch on that diagram.

Click to Enlarge

22.5 KB

I believe the GFCI should trip for the tester, assuming the tester is working well, no matter where you plug it in to the circuit.

[Marc]

You can try to figure it all out from this schematic. I know I'm going cross-eyed from looking at it. []

It looks like a transistor is used as a switch. The test button closes two circuits, one of which goes to the ground pin I think???

Click to Enlarge

22.5 KB

I believe the GFCI should trip for the tester, assuming the tester is working well, no matter where you plug it in to the circuit.

Have you a clearer copy of that diagram?

Marc

[Brandon Whitmore]

It is my understanding that the only proper way to really test a GFCI is to push the test button on the receptacle/breaker itself. I don't know why some GFCI's won't trip with my tester, but if they trip with the GFCI button, I don't write them up.

Will a GFCI function when the hot and neutral are reversed?

I'm guessing it will, but you'll have to wait for an electrical guru to say for sure. Since the GFCI trips when there is leaky current, I doubt it matters which way the current flows through the device. The reason your tester will not trip power to the GFCI is that with reversed polarity, you are attempting to shunt power from the neutral to the ground instead of the hot to ground. On newer GFCI's, I don't think that the receptacle will work unless wired correctly.

[gpdewitt]

The manufacturers of GFCIs and UL, who do the testing and listing, both accept the test button *on the GFCI* as the valid way to test it.

Here's an interesting scenario for you - You use your handy 3 prong circuit tester with GFCI test button to test a receptacle that is *not* a GFCI type. You do this because this particular receptacle is next to the sink in the master bath on the third floor, and you know it's required to be protected. The receptacle responds by going dead. Now what? You spend however long it takes to find the actual GFCI on that circuit and reset it. What if it's behind a permanently installed cabinet? What if it's behind the owner's hutch filled with 400lbs of his families' fine antique china? What if, also, there is a freezer in the garage, buried under much of the owner's belongings and out of sight, that's filled with expensive meat, or a wine cellar that must be kept at a low temperature to protect thousands of dollars worth of fine wine? You're screwed.

If I can't find GFCI where there should be one, in spite of having tripped any visible ones and noting any GFCI breakers in the box, I report it as such and suggest an electrician check it out for safety.

[Brandon Whitmore]

You do this because this particular receptacle is next to the sink in the master bath on the third floor, and you know it's required to be protected. The receptacle responds by going dead. Now what? You spend however long it takes to find the actual GFCI on that circuit and reset it. What if it's behind a permanently installed cabinet? What if it's behind the owner's hutch filled with 400lbs of his families' fine antique china? What if, also, there is a freezer in the garage, buried under much of the owner's belongings and out of sight, that's filled with expensive meat, or a wine cellar that must be kept at a low temperature to protect thousands of dollars worth of fine wine? You're screwed.

I've had a few that were difficult to track down, but have never left a house without re- setting all GFCI's. (that I know of..)

[John Dirks Jr]

Your 3 prong GFI testers need the receptacle its plugged into to be properly grounded before it will trip any GFI on the circuit using the button on the tester. On the other hand, the GFI will still trip when its own test button is pushed.

In some older houses with ungrounded wiring systems, GFI's are sometimes installed to increase safety of said undrounded system. Using the test button on a 3 prong tester is usless in that case and you can only test by pushing the button on the receptacle itself.

Bottom line, your 3 prong GFI testers need the circuit your testing to be properly grounded to be an effective test method. However, the GFI will still trip and provide protection to persons because it measures difference between hot and neutral currents, something you 3 prong tester cannot do. The 3 prong creates an imbalance by using the circuit ground to cause a fault. It's not effective for testing a GFI if the ground is not there. Meanwhile, even without a grounded circuit, a GFI can still perform its function and trip when it senses an imbalance.

[John Kogel]

Thanks, John.

The test button shorts the hot pin to ground thru an 18 k resistor on the GB tester I have here. I just checked it with a DMM. It does not short the neutral to ground, so if hot and neutral are reversed, the 3-light tester will not trip the GFCI.

One thing the 3 light tester can do is test for grounding.

Sorry, Marc. That fuzzy pic is all I got, copied from some website and blew it up. I posted it so you can redraw the blurry parts for us. []

[Jim Katen]

I frequently run into situations with GFCI's that I would like to educate myself on. I understand the general workings of a GFCI but run into peculiar situations that I would like to understand. Here they are:

-Will a GFCI function when the hot and neutral are reversed? Testers will not trip but generally the receptacle test button will trip the GFCI.

In theory, yes, the GFCI will still cut power when the hot & neutral have been reversed. However, if the GFCI was made after about 2000, it will have a lockout that prevents it from providing power to either the faceplate or to the circuit downstream of the GFCI. Basically, if you wire it with reversed polarity, it won't work.

-Three garage receptacles on same circuit and one is GFCI type. What has occurred in a circuit when the two non-GFCI garage receptacles do not trip the GFCI when tested with an external tester but when the GFCI is tripped with the test button it cuts power to the non-GFCI receptacles?

The first thing that comes to mind is a problem with the handheld tester. Second possibility: bootleg grounds at the non-gfci receptacles. I have not found this to be a common phenomenon. Are you saying that you find this frequently?

-What is the more reliable way to test a GFCI (test button or external tester)? I run into instances where testers will not trip the GFCI on a correctly wired and grounded receptacle but the test button will.

The GFCI's built-in test button is the most proper and reliable way to test it.