GFCI's do not need an EGC (moved from another thread)

[quogueelectric]

I did not post this thread

I did not post this thread

And I am being completely misrepresented by this thread so have a ball whoever opened it maybe they can answer some of these questions.

 

[mivey]

Is there some mixing of the two conductors in this discussion, one that is intended to carry current and one that is not?

I did not think we were talking about the grounded conductor to start with.

Maybe quogue will let us know exactly what he is talking about as I guess it could be several things.

 

[roger]

And I am being completely misrepresented by this thread so have a ball whoever opened it maybe they can answer some of these questions.

If you will read the posts in this thread you will see where your questions/confusion/myths have been answered or busted.

Roger

 

[mivey]

And I am being completely misrepresented by this thread so have a ball whoever opened it maybe they can answer some of these questions.

I suspected as much. Even after much prodding you have not piped up. Since you have been the focus, could we get the real story?

 

[roger]

I suspected as much. Even after much prodding you have not piped up. Since you have been the foucs, could we get the real story?

Mivey, read this old thread it was explained to Quo then.

Roger

 

[480sparky]

OK everyone.... here is a PowerPoint slide show about how GFIs work.

Anyone care to tell me where in this show it requires a GFI recep or breaker to have the grounding conductor to function as designed?

Yes, some images show the ground wire, but the presence of it does not preclude proper operation.

 

[BackInTheHabit]

For the sake of argument:

If a ground was needed for a GFCI to work, then why does the NEC allow you to replace a non-grounding receptacle with a GFCI?

See NEC 2008 406.3(D)(3)(b)

 

[electricmanscott]

best thread ever.....:grin: :grin:

 

[mivey]

Mivey, read this old thread it was explained to Quo then.

Roger

Thanks roger. I think I see where quogue is coming from. He understands how they are supposed to work. I also think he gets that the device may have some let-through current that can hurt you before it trips and that the let-through current may not be of sufficient strength and/or duration to cause the device to trip.

Quogue's problem appears to be that he had a case where it worked better after the ground was connected. That was the mystery.

Since we know (assuming a good device) the added ground did not change the way the GFI circuitry worked, the only remaining part of the circuit is the part outside of the device.

I can only surmise that in the process of connecting the ground, quogue was able to reduce the environmental impedance. The box may not have had a good bond to start. Connecting the ground to the yolk lower the environmental impedance and/or the moving of the ground wire improved the connection to the box.

By lowering the impedance, a subsequent fault to ground would cause the device to operate faster.

[edit: he still would have gotten hurt, but I doubt the follow-up tests included a human test subject so the "hurt factor" probably was not available for comparison]

 

[roger]

Mivey, you are truly a diplomat and I appreciate that. :grin:

Roger

 

[crossman]

In order for an imbalance of current to occur, and be detected, the difference current must flow from one circuit conductor after the CT to the other conductor before the CT.

Larry, I think we both understand completely how the GFCI works and I completely understand where you are coming from, and I think you are completely aware of what I am saying, so, sounds like we are good to go!:smile:

 

[dbuckley]

Why would the plumbing need to be bonded? Bonding is intentionally creating a ground path. If it's copper or galvanized pipe thoughout, there's going to be a path to ground, even if it occurs at the pipe coming in from the street. A structure with all-metal piping is bound to be grounded, whether effective )or intentional) or not.

The purpose of a GFCI is to detect imbalance of current, on the basis that there is a possibility of that current leaking through a person, and they need to be saved. For the current to flow through a person, the current has to go somewhere, Kirchoff's law applies. In any normal installation, there is a neutral / ground bond somewhere, and that will provide the path necessary for the leaking current to flow.

I don't understand what you mean by 'two hots and no neutral".

In SDS with no bond anywhere to ground, there is no neutral, just two wires, which are both normally called hots. With this sort of system you can hold one hot, and not get a shock, as there is no path through ground for current to flow. A GFCI would not trip in this situation.

All you need is a second circuit that is faulted to the plumbing. No ground bond needed.

I was very specific to say "single fault": There are many double fault situations where there is danger, including the situation that you suggest. This does not remove the illustrative purpose of the single fault situation.

 

[mivey]

you can hold one hot, and not get a shock, as there is no path through ground for current to flow.

I would not recommend it. I know you are just using it as an illustration, but you could still be coupled to ground as one must take note of Don's #40 about capacitive coupling (no helicopters, please).

I was very specific to say "single fault": There are many double fault situations where there is danger, including the situation that you suggest. This does not remove the illustrative purpose of the single fault situation.

I know. I was just trying to get away from saying one of the faults had to be an intentional ground.

 

[gar]

080720-2254 EST

Looking at the referenced thread
http://forums.mikeholt.com/showthread.php?t=88679&highlight=GFCI's
and post #27 I believe I see part of the problem.

A GFCI will only trip within 1/2 to 1 cycle if the current is large enough, and that is much larger than 5 MA. See the trip time curve I previously referenced at
http://cache.national.com/ds/LM/LM1851.pdf
The UL943 curve looks to extrapolate to about 400 MA at 10 MS. As I previously mentioned 5 MA is about 7 seconds. A GFCI is not designed to prevent a shock.

The current causing the shock is not determined by the threshold of the GFCI but by the applied voltage and the resistance thru the person with wet feet and hands.

I suspect that adding an ECG wire to the receptacle provided thru the three wire cord to the drill a path for the fault current to flow on the ECG and therefore no shock.

.

 

[don_resqcapt19]

...
I suspect that adding an ECG wire to the receptacle provided thru the three wire cord to the drill a path for the fault current to flow on the ECG and therefore no shock.

This is a very important point. With a correctly installed and functioning EGC, there is very little risk of shock except under high current fault conditions. The only shock hazard would be from the voltage drop on the EGC when it is carrying fault current. With small amounts of leakage current there is almost no voltage drop and no shock hazard. This principle is the main reason that most of the code GFCI protection rules apply to receptacles and not to hard wired equipment. It is much more likely that you would lose the continuity of the EGC on cord and plug connected equipment than on hard wired equipment.

 

[winnie]

I for one want to applaud quogue for continuing to stand up and point out a flaw in the conventional understanding of GFCIs. And I want to thank this community for continuing the discussion in a civil enough manner that this became apparent.

1) The GFCI trip element is designed to function without an EGC present. (common accepted knowledge)

2) The sensitivity/time-current-curve/let-through of a GFCI functioning to required specifications will permit a noticeable shock (not conventional wisdom, although it is right there in the specifications). I have head that this current level may be sufficient to injure a child.

3) An EGC will provide additional safety, by quite frequently providing a defined path for the current described in 2) that doesn't involve a person. The EGC may make the difference between a sensible shock that is not sufficient to trip the GFCI and either no shock (because the current doesn't go through the person) or tripping the GFCI (because enough current flows).

Bottom line: a GFCI will function for its specified purpose (limiting shock current to a specified generally acceptable value) without and EGC. But it will be far more effective at its accepted purpose (preventing shock) with an EGC present.

-Jon

 

[roger]

Bottom line: a GFCI will function for its specified purpose (limiting shock current to a specified generally acceptable value) without and EGC.
-Jon

Jon, I think we need to include the fact that a GFCI will not prevent severe shocks or even electrocution with out a second current path, this is with or without an EGC.

For example, if someone were to come in contact with the ungrounded and grounded conductor of a GFI protected circuit on a wooden structure and isolated from an EGC (if present) the GFCI will not trip and the subject would be froze the same as he would on a non-GFI circuit.

Roger

 

[winnie]

Roger,

You are quite correct. If the current doesn't go somewhere outside of the sensed portion of the circuit, then the GFCI won't trip.

I believe that this is yet another way in which a GFCI can function as it is designed, but where the GEC facilitates the desired protective function.

-Jon

 

[480sparky]

Roger,

You are quite correct. If the current doesn't go somewhere outside of the sensed portion of the circuit, then the GFCI won't trip.....

In other words, your body becomes the load, or one of the loads, on the circuit.

If you become part of the circuit, and whatever amperage that is flowing through your body goes back through the neutral (OK, technically the grounded) conductor, then the GFI simply sees you as it would any other load.

 

[gar]

080721-0925 EST

On further consideration of the problem described in
http://forums.mikeholt.com/showthrea...light=GFCI's
and post #27 I believe the GFCI would never trip.

The drill contacted a hot wire in the wall, I assume the contacted hot wire is not fed from the GFCI outlet to which the drill motor is connected, also assume the drill bit and drill motor are all conductive components, there is no leakage in the motor from either hot or neutral to the drill housing, then a person standing on earth and holding the drill motor will receive a shock and this will never trip the GFCI because this shocking current never unbalances the current thru the GFCI.

.