GFCI question

[Andy Jarchow]

Hello,

I just inspected a 50+ year old home today. I tested the receptacles with my $10 Sperry gfci circuit tester and the second it made contact the gfci kicked out. I didn't get a chance to puch the button! I tried my other little cheap circuit tester made by GB Electrical and it showed the receptacle was wired correctly. I also have a Sperry two wire tester for ungrounded receptacles and that showed correct polarity and also showed a ground. I started to think my gfci tester was bad.

When I arrived home I checked it with my gfci receptacles and everything worked.

Can anyone explain what was going on?

Thank you for your help!
mk

Similar Threads:

• Lights on GFCI for bathroom
• Sewage ejection pump-GFCI?
• AFCI and GFCI Question
• Testing GFCI Outlets
• Kitchen area GFCI

[Jim Luttrall]

Retire the tester. I had one a couple years ago that the button wore out and would do that intermitantly.

[Roger Frazee]

Next time you do this see if you can plug a power cord appliance into the gfci, ...hair dryer etc and see if it trips on contact. Leviton several years back had gfci's out that would trip for no apparent reason when plugging something into the receptacle.

As for the sperry tester it simply has a 15K resister in series with the ground pin of the tester and places an approx. 8 milliamp leak from the 120 volt source to the equipment ground of the premise wiring. Just use ohms law using 15,000 ohms and solve for amps.

Also when testing the actual gfci for proper operation it is not necessary to use the tester .. the best way is to use the test button of the gfci. If it trips the gfci and you are able to reset then the gfci is working properly. Use your tester for receptacles that are protected by the gfci.

[Corn Walker]

Next time you do this see if you can plug a power cord appliance into the gfci, ...hair dryer etc and see if it trips on contact. Leviton several years back had gfci's out that would trip for no apparent reason when plugging something into the receptacle.

I must have a dozen of those. Drives me nuts, but not enough to prompt me to change them out.

[Jerry Peck]

We keep telling you guys ... those three light testers make --- good night lights ... but you keep buying and using them as they are are actually good for something else.

[Andy Jarchow]

Next time you do this see if you can plug a power cord appliance into the gfci, ...hair dryer etc and see if it trips on contact. Leviton several years back had gfci's out that would trip for no apparent reason when plugging something into the receptacle.

As for the sperry tester it simply has a 15K resister in series with the ground pin of the tester and places an approx. 8 milliamp leak from the 120 volt source to the equipment ground of the premise wiring. Just use ohms law using 15,000 ohms and solve for amps.

Also when testing the actual gfci for proper operation it is not necessary to use the tester .. the best way is to use the test button of the gfci. If it trips the gfci and you are able to reset then the gfci is working properly. Use your tester for receptacles that are protected by the gfci.

I did try my other similar tester without the gfci button and it showed the receptacle was fine and it didn't trip. Not sure what was going on. Is this what you ment by plugging in a hair dryer or something? Or do you mean plug something in that would have a load on it (turn hair dry on before plugging it in)? Thank you

Sorry it took me so long to comment. Things are finally getting busy in the inspection business!! I hope it stays that way but I have my doubts.

Thanks again!
mk

[Andy Jarchow]

Thank you all!!!
mk

[Roger Frazee]

I did try my other similar tester without the gfci button and it showed the receptacle was fine and it didn't trip. Not sure what was going on. Is this what you ment by plugging in a hair dryer or something? Or do you mean plug something in that would have a load on it (turn hair dry on before plugging it in)? Thank you

Sorry it took me so long to comment. Things are finally getting busy in the inspection business!! I hope it stays that way but I have my doubts.

Thanks again!

mk

Hi Mike

Yep that would probably do the same thing. Sounds like you may have an issue with the tester. Does that tester trip other gfci's or was the issue just with that one gfci ? Also if the gfci was not a Leviton then that problem I talked about would not be related.

These three light testers with gfci trip buttons are low end instruments so when business gets better you might consider a upgrade to a ideal 'sure test' or similar tester.

I'm not an inspector but using the test button is considered the only way to insure proper operation of the gfci receptacle or breaker.

I actually use a voltage tester called a 'wiggy" on receptacles protected by a gfci. It's a solenoid type voltage tester and it works great ... as soon as the solenoid is actuated it will trip any protecting gfci. You just put one probe in the hot terminal slot of the gfci and the other in the ground terminal hole. The wiggy requires a small current to actuate the solenoid. Since the gfci it doesn't see that current on the neutral it trips.

[Scott Patterson]

Everyone does know that the proper way to test a GFCI is to use the test button on the outlet or the breaker, don't you? That is why they put them on them!

[Vern Heiler]

Everyone does know that the proper way to test a GFCI is to use the test button on the outlet or the breaker, don't you? That is why they put them on them!

Scott, if the GFCI trips when tested with the cheap-o 3-light tester, is there any chance the GFCI is not functioning as intended? Could it still be bad?

[Andy Jarchow]

Hi Mike

Yep that would probably do the same thing. Sounds like you may have an issue with the tester. Does that tester trip other gfci's or was the issue just with that one gfci ? Also if the gfci was not a Leviton then that problem I talked about would not be related.

These three light testers with gfci trip buttons are low end instruments so when business gets better you might consider a upgrade to a ideal 'sure test' or similar tester.

I'm not an inspector but using the test button is considered the only way to insure proper operation of the gfci receptacle or breaker.

I actually use a voltage tester called a 'wiggy" on receptacles protected by a gfci. It's a solenoid type voltage tester and it works great ... as soon as the solenoid is actuated it will trip any protecting gfci. You just put one probe in the hot terminal slot of the gfci and the other in the ground terminal hole. The wiggy requires a small current to actuate the solenoid. Since the gfci it doesn't see that current on the neutral it trips.

OK
I'm not sure what kind of gfci they were I didn't pay attention to that but I will in the future.

Remember when you were helping me understand the function of a ground and a grounded conductor. I asked about the ground and grounded conductor being put together in the main panel. This was my post and your answer- This is a question unrelated to my original post but similar topic. If current travels on the hot wire through the appliance and then back on the neutral to the service panel. In the panel this same neutral wire is hooked to the neutral bus bar that ground wires also connect to. Why doesn’t the current flow back on these ground wires to the equipment?
[FONT='Verdana','sans-serif'][/FONT][FONT='Verdana','sans-serif']That's a very common question. The neutrals and grounds bond together only at one place ...the service equipment. The service equipment is the enclosure where the first means of disconnect for the dwelling is located. It can be in an enclosure by itself or it can be in a panelboard as a "MAIN " breaker. They bond in the service equipment because there is only one low impedance path for neutral and ground fault current to get to the utility transformer center tap at that point. That path is the service neutral. You can see this in the diagrams I posted. We also do not want any neutral to ground connections load side of the service equipment. This creates alternate paths for current on metal and wires that is dangerous and unwanted as we have been discussing in this thread.[/FONT]
[FONT='Verdana','sans-serif'] [/FONT]
[FONT='Verdana','sans-serif']My new questions-[/FONT]
[FONT='Verdana','sans-serif']Is this the same reason a sub-panel must have a floating grounded conductor and be separated from the ground until they reach them main panel? And what are the specific dangers if it is not done like this? [/FONT]
[FONT='Verdana','sans-serif'] [/FONT]
[FONT='Verdana','sans-serif']Is it ok to run a wire from the main panel to an unattached garage say with a #12 2 with a ground on a 15 amp breaker? [/FONT]
[FONT='Verdana','sans-serif'] [/FONT]
[FONT='Verdana','sans-serif']Thank you sir[/FONT]
[FONT='Verdana','sans-serif']Have a great night[/FONT]
[FONT='Verdana','sans-serif']mk [/FONT]

[Scott Patterson]

Scott, if the GFCI trips when tested with the cheap-o 3-light tester, is there any chance the GFCI is not functioning as intended? Could it still be bad?

Honestly I have no idea. I just do not trust the 3-light testers. My guess is that the tester tripped it for who knows why.

I have tripped a GFCI with a 3-light tester before and then found that I could not reset the GFCI and the outlet was still live!

[Vern Heiler]

Honestly I have no idea. I just do not trust the 3-light testers. My guess is that the tester tripped it for who knows why.

I have tripped a GFCI with a 3-light tester before and then found that I could not reset the GFCI and the outlet was still live!

I look at it from the other direction. I know what the 3-light tester did when the button was pushed. I don't know what the button on the GFCI outlet does without a diagram of the specific model under test. I have had GFCI outlets smoke when tested with the 3-light. I don't know what the result would have been from the built in test button but the smoke was an obvious fail of test.

[Matt Fellman]

Both methods test it just fine in most cases. The button on the outlet often gets favored because it's what the manufacturer's say is the only reliable way to test it. Obviously, that's because they don't want to take responsibility for some other piece of equipment. It doesn't mean the other ways don't work too.

To me all these testers are just tools for gathering information. I never trust any one of them as the end all in documenting a problem. They're just a way of quickly finding where some closer looking is needed.

[David D. Whitt]

I test the GFCI device with the test button. If it doesn't trip, write it up. I like the three light testers to identify if there are other protected outlets downstream from a GFCI, like a kitchen circuit or a second device in the bathroom or outside.
I had a light bulb out on one of those things, and I wrote just about every device as being deficient. Embarrassing.

[Corn Walker]

Is it ok to run a wire from the main panel to an unattached garage say with a #12 2 with a ground on a 15 amp breaker?

Presuming that cable is listed for the application and enclosed in conduit listed for direct-burial or exposure, you could do that. You'll need to verify that the length and installation conditions don't derate the cable to less than 15A though.

But if you're going to go through the trouble, you're probably better off running 2-2-2-4 SER with a 60A breaker. In case you ever want to install, say, an air compressor or some power-hungry tools in that garage.

[Roger Frazee]

My new questions

Is this the same reason a sub-panel must have a floating grounded conductor and be separated from the ground until they reach the main panel? And what are the specific dangers if it is not done like this?

Mike

First thing to understand is location of the panel. Is it in the same dwelling/structure that the service equipment is in or is it in a detached building like a garage or shop etc...?

If it is in the same structure with the service equipment a 4 wire (H-H-N-Grd) feeder must be run to serve that panel. With 4 wire feeders neutral and ground are separated so that a parallel path for current is not created. If we were to bond the neutral and ground at the panel the neutral current would use both the feeder equipment ground and the feeder neutral to return to the service equipment. Remember that neutral current will take any path available to it to get back to the source transformer. See the diagram attached below.

Where a neutral to ground bond is made and metal conduit like emt is serving as the equipment ground this becomes very dangerous as metal conduit is often touched by humans. You never want metal or equipment grounding wires that are part of the effective ground fault path energized with system current. The diagram shows two improper neutral to ground connections. One is where they have landed both the feeder equipment ground and the feeder neutral to the same terminal bar.

The other is where they added a grounding bar then landed the feeder neutral to the neutral bar and the feeder equipment ground to the grounding bar as should be but then the bonding means (green bonding screw) was improperly installed bonding the neutral bar to the metal of the panel. This effectively connected neutral to equipment ground and the neutral current finds a parallel path by using the bonding screw to the metal of the panel then the metal of the panel to the grounding bar where it can then return to the source over the feeder equipment ground. Remember in the diagrams that the neutral bar is not bonded to the metal of the panel as it sets on insulated standoffs. Installing the bonding means is the only way to bond it to the metal of the panel. If a grounding bar is added and is also bonded to that metal you have effectively by installing the bonding means connected them together by way of the panel metal. Also in this example the metal of the panel is energized...not what you want..as you have to touch that metal panel to reset a breaker or other maintenance.

Is it ok to run a wire from the main panel to an unattached garage say with a #12 2 with a ground on a 15 amp breaker?

Yep as long as the breaker is not over 20 amps and the wire/cable rated for the application . Code requires small conductors size 14 awg, 12 awg, and 10 awg to be protected at not over 15 amps, 20 amps or 30 amps respectively. In other words you cannot use a 20 amp breaker on 14 awg or 30 amp breaker on 12 awg. But you can use the reverse 15 amp breaker for 12 awg or 20 amp for 10 awg. There generally is no good reason to do that however. Of course there are a few exceptions but that is another story. Voltage drop can enter into this decision as an example.

[Andy Jarchow]

Presuming that cable is listed for the application and enclosed in conduit listed for direct-burial or exposure, you could do that. You'll need to verify that the length and installation conditions don't derate the cable to less than 15A though.

But if you're going to go through the trouble, you're probably better off running 2-2-2-4 SER with a 60A breaker. In case you ever want to install, say, an air compressor or some power-hungry tools in that garage.

The garage was wired like this on one of my inspections last week. It was in conduit but it was not an underground wire that I was able to see coming out of the conduit near the ceiling. It was regular 12-2 with a ground. It fed two lights with 14 ga wire and three receptacles including the garage door opener.
Thank you
mk

[Roger Frazee]

The garage was wired like this on one of my inspections last week. It was in conduit but it was not an underground wire that I was able to see coming out of the conduit near the ceiling. It was regular 12-2 with a ground. It fed two lights with 14 ga wire and three receptacles including the garage door opener.
Thank you
mk

Mike

Conduit underground is a classified as a wet location. If the cable was 'romex' or more properly nm cable with ground or without ground it is a violation as nm is not wet location rated. In general with a few exceptions nm cannot be outside.

Be sure that it is not UF (underground feeder) cable as it is rated for wet locations.

[Andy Jarchow]

My new questions

Mike

First thing to understand is location of the panel. Is it in the same dwelling/structure that the service equipment is in or is it in a detached building like a garage or shop etc...?

If it is in the same structure with the service equipment a 4 wire (H-H-N-Grd) feeder must be run to serve that panel. With 4 wire feeders neutral and ground are separated so that a parallel path for current is not created. If we were to bond the neutral and ground at the panel the neutral current would use both the feeder equipment ground and the feeder neutral to return to the service equipment. Remember that neutral current will take any path available to it to get back to the source transformer. See the diagram attached below.

Where a neutral to ground bond is made and metal conduit like emt is serving as the equipment ground this becomes very dangerous as metal conduit is often touched by humans. You never want metal or equipment grounding wires that are part of the effective ground fault path energized with system current. The diagram shows two improper neutral to ground connections. One is where they have landed both the feeder equipment ground and the feeder neutral to the same terminal bar.

The other is where they added a grounding bar then landed the feeder neutral to the neutral bar and the feeder equipment ground to the grounding bar as should be but then the bonding means (green bonding screw) was improperly installed bonding the neutral bar to the metal of the panel. This effectively connected neutral to equipment ground and the neutral current finds another parallel path by using the bonding screw to the metal of the panel then the metal of the panel to the grounding bar where it can then return to the source over the feeder equipment ground.Remember in the diagrams that the neutral bar is not bonded to the metal of the panel as it sets on insulated standoffs. Installing the bonding means is the only way to bond it to the metal of the panel. If a grounding bar is added and is also bonded to that metal you have effectively connected them together by way of the panel metal.

Yep as long as the breaker is not over 20 amps and the wire/cable rated for the application . Code requires small conductors size 14 awg, 12 awg, and 10 awg to be protected at not over 15 amps, 20 amps or 30 amps respectively. In other words you cannot use a 20 amp breaker on 14 awg or 30 amp breaker on 12 awg. But you can use the reverse 15 amp breaker for 12 awg or 20 amp for 10 awg. There generally is no good reason to do that however. Of course there are a few exceptions but that is another story. Voltage drop can enter into this decision as an example.

Thank you Roger,

My thinking was right on the garage wiring question.

I need to read the first part of your answer a few more times and let it bounce around a bit.

As always I appreciate your thorough answers and diagrams!!!
[FONT='Times New Roman','serif']mk [/FONT]

[Andy Jarchow]

Mike

Conduit underground is a classified as a wet location. If the cable was 'romex' or more properly nm cable with ground or without ground it is a violation as nm is not wet location rated. In general with a few exceptions nm cannot be outside.

Be sure that it is not UF (underground feeder) cable as it is rated for wet locations.

My guess is that it was not underground feeder because it was nm coming out of the conduit near the ceiling. UF is normally black right?

Thank you
mk

[Andy Jarchow]

Thank you all for your help!

Good night

[Roger Frazee]

My guess is that it was not underground feeder because it was nm coming out of the conduit near the ceiling. UF is normally black right?

Thank you

mk

Never go by color most modern UF is gray in color but color is no way to identify wire for a wet location rating. You need to be able to tell the difference between UF and NM or find the writing on the outer jacket.

[Andy Jarchow]

Never go by color most modern UF is gray in color but color is no way to identify wire for a wet location rating. You need to be able to tell the difference between UF and NM or find the writing on the outer jacket.

Ok good info

Thank you

[Roger Frazee]

Mike

Posting a few more diagrams because I want you to see the difference in 4 wire feeders and a 3 wire feeders. First 3 wire feeders are not allowed to panels installed and fed from the service equipment that is located in the same dwelling/structure with that panel. You must have a 4 wire feeder in this case.

3 wire feeders are allowed to panels in detached buildings/structures fed from the service equipment located at the dwelling/structure separate from the detached building.. There are restrictions that must be met in order to run 3 wire feeders to detached buildings. Those restrictions if not met require a 4 wire feeder to be installed. Some local jurisdictions required 4 wire feeders period to detached buildings as an amendment to the NEC. In 2008 the NEC now requires 4 wire feeders no 3 wire feeders unless already existing before 2008 and no local amendments requiring 4 wire feeders at the time the feeder was installed. So you need to know what local code says and what NEC code cycle was being enforced at the time of installation and what code cycle is presently being enforced so that you know what 'new' construction requires.

Ok the first diagram is a 3 wire feeder to a detached building from the service equipment in the house. The 3 restrictions that all must be met to run a 3 wire feeder to the detached structure are
1.) NO equipment ground is ran with the feeder
2.) NO other metallic paths bonded to the grounding system between the dwelling with the service equipment and the detached building. This would be for example a metal water pipe connecting the two structures and you bonded the water pipe to the grounding system in both structures. This would provide a parallel path for neutral current if using a 3 wire feeder.
3.) Ground fault protection of equipment (GFPE) has not been installed on the supply side of the feeder.

Notice that in 3 wire feeders you must bond neutral and ground because the feeder neutral must share duties as the system current carrier for neutral current and the effective fault path for any ground fault. So in this case the bonding means is installed through the neutral bar so that the panel metal and any added grounding bars that have equipment grounding terminations are provided with a low impedance path to the feeder neutral so fault current can open a breaker by returning to the source over the feeder neutral. With out the main bonding means installed at the panel with a 3 wire feeder there will be no way for an effective fault path to be established and your circuit breakers will not open on ground fault.

In contrast a 4 wire feeder must have the main bonding means removed and not installed because the effective fault path is now being carried by the feeder equipment ground (the 4th wire) instead of the feeder neutral.. If we install the neutral bonding means we now give neutral current a path to the feeder equipment ground and system current will flow on both the metal of the panel and feeder equipment ground along with the feeder neutral.....a low impedance parallel path with objectionable system current on the feeder equipment ground. We do not want that.

On the 3 wire feeder diagram showing an added grounding bar visualize that you have removed the green screw or bonding means from the neutral bar. By doing this you have removed the 'bridge' that allows any fault current on the equipment grounding wires to reach the neutral bar and the feeder neutral to return to the source there-by preventing any circuit breaker from opening a ground fault in a branch circuit. Since the breaker cannot open... the panel metal and all bonded metal involved with the fault will come to line voltage and pose an electrocution hazard to anyone coming in contact with that energized metal.

Also understand that the added grounding bar is not necessary with a 3 wire feeder... you can just install the bonding means and land both egc's and grounded conductors to the same bar .. same as you do with service equipment. The bonding means in this case simply connects the panel metal with the effective fault path.

The last drawing is showing a violation of a 3 wire feeder install where a metal water pipe has been ran between the structures and bonded to the grounding system in each structure. Neutral current will flow on the water pipe back to the service equipment and then to the source. The water pipe causes a 4 wire feeder to be required. Same thing if communication/data cable is ran between structures and are bonded to the grounding system.