[ArcLazer]

Hey guys, new here. I'm a third year apprentice and as part of my quest to fully understand proper bonding and grounding I was hoping somebody could confirm or deny my thoughts on a potential problem with a GFCI.

Would a GFCI fail to trip if you have bonded twice upstream of it? For example if you bonded at the first means of disconnect at the meter, and within the panel supply power to a GFCI. My guess is that it wouldn't, but looking for some electrical theory here without having to physically test this. Thank you

 

[IslandGuy]

A GFCI does not need to be grounded to work properly.

 

[ArcLazer]

I understand the code allows for the replacement of a 2 wire outlet for a GFCI, however if you have a parallel path of your neutral to ground going back, aren't you essentially "tricking" it into thinking there is no ground fault? Am I really overthinking this? GFCI do not protect against phase to neutral faults.

 

[joed]

A gfci works by sensing the difference in current between the hot and neutral wires. It doesn't matter where the current goes if there is more then a 5mA difference the GFCI will trip.

 

[Jnaas2]

A gfci looks at the current going out and the current coming back on the neutral, That's why the neutral is connected to the breaker, If you have a fault between the phase and neutral it will not trip because its not going to ground, If you have 2 separate return paths for neutral that's against code and the gfi will trip because it wont see all the voltage returning on the neutral

 

[Oso954]

I fail to see how the double bond/parallel paths upstream of the GFCI, unbalances the current downstream of it.

You might want to diagram such a circuit, and study it.

I am also puzzled by your second post. With a 2 wire circuit (romex) and GFCI, how do you create the parallel path (neutral to ground) when there is no ground wire ?

 

[Tom738]

I fail to see how the double bond/parallel paths upstream of the GFCI, unbalances the current downstream of it.

You might want to diagram such a circuit, and study it.

I am also puzzled by your second post. With a 2 wire circuit (romex) and GFCI, how do you create the parallel path (neutral to ground) when there is no ground wire ?

re: 2nd post. My understanding is that sufficient current imbalance between its hot and neutral outputs triggers a GFCI. So even without an additional wire in the Romex for ground, if someone or something touches a hot and the *person* or other surface acts as a sufficient alternate path for electron flow, (I suppose whether as a path back to your ground prong, back to your nominal ground somewhere else in the house, or as an independent or unreferenced source or sink of adequate current), in theory it should trip.

I'd need to look at the circuit diagram and think about it more to verify that, but that's my 50,000-foot understanding.

 

[ArcLazer]

re: 2nd post. My understanding is that sufficient current imbalance between its hot and neutral outputs triggers a GFCI. So even without an additional wire in the Romex for ground, if someone or something touches a hot and the *person* or other surface acts as a sufficient alternate path for electron flow, (I suppose whether as a path back to your ground prong, back to your nominal ground somewhere else in the house, or as an independent or unreferenced source or sink of adequate current), in theory it should trip.

I'd need to look at the circuit diagram and think about it more to verify that, but that's my 50,000-foot understanding.

That's all good. To clarify my thinking on this. If you double bonded neutral and ground, it's as if it becomes 1 wire. In a normally code compliant install, you can trip a GFCI (downstream) if you touch neutral to ground. I don't see how it would trip in my hypothetical scenario, that's like touching the neutral to a neutral (?)

Also because you've tied neutral to ground twice, wouldn't any currently between the hot to "ground" would trip the breaker because it's now going back on it's "intended" path?

 

[Jnaas2]

Neutral and ground are only tied at the first disconnect in the panel, From there all neutrals and bonding are separated to keep from having 2 paths for the neutral back to the main panel. That's why a sub panel has the neutrals and equipment grounds separated in the sub panel

 

[Tom738]

That's all good. To clarify my thinking on this. If you double bonded neutral and ground, it's as if it becomes 1 wire. In a normally code compliant install, you can trip a GFCI (downstream) if you touch neutral to ground. I don't see how it would trip in my hypothetical scenario, that's like touching the neutral to a neutral (?)

By bonding neutral to ground, in two places, yes, they become the same wire (albeit a larger wire that accordingly has a lower voltage drop). As Jnaas2 points out, this is the precise reason why you don't bond the ground to the neutral in the sub panel--it would turn the ground from the main panel to the sub into a second neutral wire.

However, why would touching neutral to ground trip a downstream GFCI? I suppose it would result in a different current flowing through the GFCI because of the change in voltage..., but the current should be changed equally on both the hot and the joint wire that the GFCI thinks is the neutral, so the GFCI shouldn't trip. It would trip an upstream GFCI, at least if there was any current on said combined wire...

I'm only theorizing, but would think tripping a downstream GFCI in that scenario would be the contrived exception rather than the rule--if the GFCI is also an arc-fault device and the connection is noisy, for example, or if the initial bond between neutral and ground was not good and the new bond was the only effective one, and it happened to make an alternate path to ground from the downstream GFCI's hot leg effective. If you are talking about an intermittent touch, perhaps there is something about the noise on the line from the sudden change in voltage that the GFCI just doesn't handle correctly. But maybe some of the guys who do this a lot can weigh in on what happens in practice.

Also because you've tied neutral to ground twice, wouldn't any currently between the hot to "ground" would trip the breaker because it's now going back on it's "intended" path?

If there is a neutral to ground connection *downstream* of the GFCI protection, then connecting any load in between that neutral/ground wire on the downstream side and the hot wire will trigger a current imbalance between the hot and the neutral and should trip the GFCI Protection. But if the connection is *upstream* of the GFCI protection, then the GFCI protection should not trip because from its perspective all you've done is increased the size of the neutral wire feeding it.

That applies to GFCI breakers and GFCI outlets. My apologies for clarifying if it seems insulting, but you didn't specify you were talking about a GFCI breaker so I just want to be clear for any future visitors to the forum that ordinary breakers do not trip based on the path a current takes; they only trip if there is a sufficient current draw for a period of time that varies based on the amount of current drawn, which you can see in the trip curve chart for the breaker from the manufacturer.

 

[ArcLazer]

Jnaas2: An irrelevant answer. I know what the code says. This is a hypothetical situation. Please stay on topic.

Tom738: Thank you

 

[ddawg16]

The GFCI does not care how the earth is bonded upstream....if bonded at all....

As long as the current on the black wire is the same as the current on the white wire, all is good.

 

[Jnaas2]

Im gone

 

[brric]

And some people believe there are no stupid questions. :vs_worry:
A totally inane and useless discussion.

 

[Know A Little]

If a neutral is grounded down stream and the current flow on the downstream ground exceeds the trip level the GFCI will trip.

 

[brric]

If a neutral is grounded down stream and the current flow on the downstream ground exceeds the trip level the GFCI will trip.

The GFCI will trip immediately.

 

[Tom738]

And some people believe there are no stupid questions. :vs_worry:
A totally inane and useless discussion.

:smile:

Those people reminds me of a guy I know. One day he called me and I told him I was in the middle of fifteen things, all of them annoying, so he should keep it short. He unironically asked "what things?" Watching his brain try to function is really, really sad.

Still, even an impractical question can be valuable if it helps someone think about what's going on with a system they'll be asked to work on and diagnose problems with. I am certain time learning from a good mentor and master electrician is much more valuable, but also enjoy the asides that seem silly when you have decades of experience under your belt but where we lay folk have a chance to be useful! :vs_coffee:

 

[ArcLazer]

Oh man, if you think this is a stupid question, then you better skip all my future posts, this probably doesn't even rank top 100. Anyway I, and potentionally others found it to be beneficial. I take my work seriously and I'd rather discuss something and come to a solid answer, vs laying out the work and finding out the hard way it's wrong and needs to be redone. I'd ask a billion more stupid questions if it also meant saving somebody's life.

 

[brric]

Oh man, if you think this is a stupid question, then you better skip all my future posts, this probably doesn't even rank top 100. Anyway I, and potentionally others found it to be beneficial. I take my work seriously and I'd rather discuss something and come to a solid answer, vs laying out the work and finding out the hard way it's wrong and needs to be redone. I'd ask a billion more stupid questions if it also meant saving somebody's life.

Why aren't you being taught these things through your apprenticeship?

 

[Know A Little]

The GFCI will trip immediately.

Yes because you setup a parallel path and current flow exceeds the limit of the GFCI.