GFCI dilemma resolved

For decades I've wondered why a GFCI will trip if you touch ground to neutral. Both ground and neutral are bonded together to the same point in the load center. GFCI trips when it detects a difference in current between power and neutral. Then today while talking to myself, because no one was around and I wanted to listen to someone talk, Self had an AHA moment. Even though there isn't anything plugged into the GFCI, the GFCI is using power. So, there is current in the neutral and power wire. Shorting the ground to neutral allows some of the neutral current to find it's way back on the ground wire.

To confirm all this I put a short 3 wire cord on a GFCI. Then plugged it into a device that measures watts. Sure enough, 2.6 watts. Not sure of the accuracy of my meter, so you may try the experiment and find something different.

Better not tell tree huggers that GFCIs are using that much power. A billion GFCI's will consume 2.6 billions watts. Is it possible that all this wasted power is causing GOBAL WARMING?
George

(quoted from post at 19:23:22 11/06/14) For decades I've wondered why a GFCI will trip if you touch ground to neutral. Both ground and neutral are bonded together to the same point in the load center. GFCI trips when it detects a difference in current between power and neutral. Then today while talking to myself, because no one was around and I wanted to listen to someone talk, Self had an AHA moment. Even though there isn't anything plugged into the GFCI, the GFCI is using power. So, there is current in the neutral and power wire. Shorting the ground to neutral allows some of the neutral current to find it's way back on the ground wire.

To confirm all this I put a short 3 wire cord on a GFCI. Then plugged it into a device that measures watts. Sure enough, 2.6 watts. Not sure of the accuracy of my meter, so you may try the experiment and find something different.

Better not tell tree huggers that GFCIs are using that much power. A billion GFCI's will consume 2.6 billions watts. Is it possible that all this wasted power is causing GOBAL WARMING?
George

Click to expand...

ot where I am! Been cold all day. Still am.

It's not the power drawn by the GFI. Use a volt meter and there will be a voltage difference between the ground and neutral at the GFI. When you touch the ground and neutral together current will flow in the ground due to the difference in potential. Now the GFI detects more than 5ma difference between line to neutral and pow, off she goes.

Don't know how this pertains,
but as an added note,
If yer place suffers a severe lightning strike,
all GFCI's will fry and never work again,
along with a lot of other stuff...
Don't ask me how I know, I wish I didn't...

We lost a lot more expensive stuff than a few GFCI's
when we had a lightning strike nearby!

About 30 years ago, after a lightning strike
hit my home, I went to Menards and purchased
GFCI/surge protector. When there is a lightning
strike near, it trips the GFCI. I can't buy them
anymore. I have them on all my TV's and
microwave.

When I installed an expensive kitchen aid
dishwasher with electronic controls. I wired the
dishwasher to a power cord. Then I purchased the
largest strike protector I could find, the type
one would use to protect compters. Installed it
under sink. Glad I did, lightning wiped it out,
saved the dishwasher.

Then I purchased a strike protector from the
power company. It goes between the meter and the
meter base. If I have lightning damage, they are
suppose to pay. Haven't hand any lightning
damage.

Hoping someone else will measure the power used
by a GFCI.

2.6 watts seems unlikely. I would believe 2.6 VOLT-AMPS at near-zero power factor, so maybe half a watt.

At any rate, it's quite unlikely the parasitic drain of the GFCI's internal circuitry has anything to do with it tripping when you short neutral to ground. First, I'd expect all that circuitry to be upstream of the differential current sensor. Second, as I said before, the GFCI is SUPPOSED to trip when there's a ground fault on the neutral side of the protected circuity.

Like B&D said, the GFCI trips because neutral and ground are not at the same potential, either due to voltage drop in the neutral or voltage induced in the neutral-ground loop.

There is a way to verify your theory: Take a short piece of wire and short the GFCI's ground and neutral screws together. Now they are at the same potential; if the GFCI still trips when you short the neutral to ground on the protected receptacle, your theory is correct.

PLEASE MEASURE POWER, POWER FACTOR AND MEASURE THE PD. REPORT YOUR FINDINGS.

Okay gang, lets do some of the math and look at how the GFCI operates and respond to Georges post: This is kind of an Engineering analysis, it gets a little deep and technical, bear with me, but its just how I am "wired" after engineering school and 40 years:


If George is right?????? (I'm NOT saying he is or isn't, just using his figures, its his post, his info which I take at his word, and his question)


1) If the GFCI is drawing (per George, NOT me) 2.6 watts at 120 volts, that means its drawing 2.6/120 = 22 Milliamps just sitting there with its sensing electronics at work. NOT ME SAYING IM USING GEORGES INFO and if my math is right NO WARRANTY so don't anyone have a calf, I may well be wrong and likely am, so there lol

2) ME SAYING How the GFCI works is that if there is 5 milliamps less current flow in the Neutral then the Hot line conductor (5 milliamps of fault current being returned OTHER THEN the Neutral) IT TRIPS OPEN. So if George is right??? (using HIS info, not me saying) SURE if there's a current flow of 22 Milliamps in the Ground that's NOT in the Neutral, THE GFCI TRIPS.

BUT HERES COMES THE RUB

HOW AND MOST IMPORTANT WHERE DO YOU GET THAT 22 MILLIAMPS TO FLOW IN THE GROUND INSTEAD OF THE NEUTRAL??????????????

LISTEN

If you short Ground to Neutral BEFORE the current flows through the GFI's Torroidal Coil THAT HAS NO AFFECT because the 22 Milliamps flowing out the hot is still returned by the Neutral BEFORE it may be split between Ground and Neutral where you make the jump/short.

It looks to me like if you jump the GFCI's Ground terminal with the GFCI's Neutral terminal on its INPUT (NOT the download downstream feeder) that short is BEFORE any current flows through the GFCI's Torroidal Coil, so the Neutral that's flowing through the sensing coil still sees all that's coming out the hot AND THAT SHOULDNT BE THE CAUSE OF A TRIP.

HOWEVER if the short (Neutral to Ground) is AFTER current flows through the Torroidal Coil, SUCH AS IF YOU SHORT THE DOWNSTREAM OUTPUT TERMINAL ON THE GFCI TO GROUND, then theres a parallel current path for that 22 milliamps of return current so the Neutral current (through torroidal coil) could possibly be 5 milliamps different then the hot current AND IF SO (IM NOT saying yes or no) THE GFCI COULD TRIP.

SUMMARY OF WHAT IM SAYING

If you short Neutral to Ground BEFORE Neutral current flows through the GFCI's torroidal coil THAT DONT AFFECT OR TRIP THE GFCI because all current out the hot (through the sensing coil) still gets returned by the Neutral (through the sensing coil) and theres NO IMBALANCE, NO TRIP.

If you short Ground to the Neutral output terminal that's downstream (current has flowed through sensing coil) and if 22 milliamps is flowing THAT CAN TRIP THE GFCI.

NOW LETS TALK ABOUT A VOLTAGE DIFFERENCE IN NEUTRAL TO GROUND AND ITS AFFECT

At the main panel Neutral is bonded to Ground so you can se there's sure NOT much voltage difference between those two wires there......

Okay as the wires run out from the panel, due to capacitance, and they are BOTH running in parallel with the hot, the voltage on the ground can be higher then mother earth for sure since the earth bond is way back at the panel HOWEVER HOWEVER HOWEVER since Hot and Neutral are BOTH in parallel with the Hot and BOTH are the same length and near same capacitance to hot, the voltage difference between the Neutral and Ground can sure be somewhat different BUT IT AINT MUCH and so little it would take an electrostatic high quality voltmeter (IE extreme high input impedance) to even measure it.

CONCLUSION I just dont envision enough voltage and enough energy difference between hot and ground to cause a 5 milliamp current draw enough to trip a GFCI BUT THATS NOTTTTTTTT BASED ON ANY CALCULATIONS OR ACTUAL MEASUREMENTS OR DATA WHATSOEVER SO NO FREAKING WARRANTY. If you have actual verifiable data to prove theres enough voltage and enough energy between Neutral and Ground to cause a 5 milliamp current flow THEN WHAT I ENVISIONED IS WRONG AS RAIN AND NO PROBLEM. I HAVE NO data, so if you have some and some sound engineering proof I will certainly believe it IE show and prove theres enough of a voltage difference and enough energy between hot and neutral to cause a 5 milliamp current flow....

NUFF SAID it got so deep I sort of lost interest already lol and am not gonna spend much more time on it, I may be right or wrong, it don't matter much..........but it takes some real data and measurements and not guesses or speculation which is what a lot of the above is, as I have no real data OTHER THEN I believe it takes 5 Milliamps of current difference between Hot and Neutral passing through the GFCI's Torroidal Coil to trip it I THINK AT LEAST THAT PARTS RIGHT!!!

John T

JohnT
"If you short Ground to the Neutral output terminal that's downstream (current has flowed through sensing coil) and if 22 milliamps is flowing THAT CAN TRIP THE GFCI."

That part of your answer is exactly what I'm saying. Short the two output terminals, ground and common, it will trip. The only logical conclusion is there has to be a minimum of 5 ma flowing in both the power and neutral input wires when GFCI is powered up.

As for me knowing what I'm talking about, the jury is still out. The verdict when be in if someone with accurate meters posts their results.
George

OK George, I found a GFCI receptacle in my electrical junk drawer and decided to waste an hour of my time this morning testing out your theories. I ended up moving into the house because all the outlets in my shop are protected by a GFCI breaker, which I kept tripping. Here are the results.

#1. Does a GFCI trip when you short neutral to ground on the protected circuit? You betcha! But as everybody has said, that's expected behavior. A GFCI that can't detect a fault on the neutral is no GFCI at all. No points awarded.

#2. Does a GFCI receptacle draw over 2 watts with no load on it? My Kill-a-Watt can't measure less than 10 milliamps (and I'd be suspicious of that), but it did seem to lock in on a power factor of around 0.3. My Fluke DMM reported a current draw of 7 milliamps. So that means the GFCI draws 120 x .007 x .3 = .25 watts. That's really nothing. Even if the Power Factor is unity, it would still be less than a watt per GFCI. Point to me.

#3. Is there significant voltage between neutral and ground? According to my DMM, there's 30 millivolts between ground and neutral in my shop. Note that my shop is 125 feet from the meter, with separate neutral and ground the entire distance. So your mileage may vary. Point to me.

#4. Is the voltage differential between neutral and ground responsible for the GFCI tripping when there's a ground fault on neutral? (This is the test I suggested you do.) I shorted the GFCI's source neutral to ground, thereby reducing the voltage between the two to zero. The GFCI would still trip. So B&D and I were wrong: Although the difference may be significant, it's not what trips the GFCI. My personal opinion is this behavior is BY DESIGN. (You seem to think it's a design fault.) At any rate, POINT TO GEORGE!

It looks like a tie to me. Ciao.

Mark,
Thanks for your effort.

Not sure I follow how your point system works.

It sounds like you did measured your GFCI using 7 ma. I knew my numbers were likely to be off, that's why I wanted someone with better equipment to measure it.

You confirmed that shorting neutral to ground trips the GFCI and the tripping is not caused by a voltage difference between neutral and ground., but by sending the 7 ma to ground.

What I don't follow is your point system works:
POINT TO GEORGE! It looks like a tie to me. Ciao.

So does this mean the jury of public opinion has reached a verdict, or is the jury still out?
Thanks,
George

(quoted from post at 17:04:55 11/08/14) Mark,
Thanks for your effort.

Not sure I follow how your point system works.

It sounds like you did measured your GFCI using 7 ma. I knew my numbers were likely to be off, that's why I wanted someone with better equipment to measure it.

You confirmed that shorting neutral to ground trips the GFCI and the tripping is not caused by a voltage difference between neutral and ground., but by sending the 7 ma to ground.

What I don't follow is your point system works:
POINT TO GEORGE! It looks like a tie to me. Ciao.

So does this mean the jury of public opinion has reached a verdict, or is the jury still out?
Thanks,
George

Click to expand...

If there had been not difference in potential the 7ma would not have flowed.

The way I look at it, according to Mark, there is
7 ma flowing through the common in side and power
in side. Connectiong the ground wire to the
common out side causes at least the minimum
current needed to trip the GFCI to flow in the
common out, but there is no current in the power
out side.

Look at Mark's experiment, didn't he short the
common in and the ground together, so there
couldn't be a potential difference between common
and ground?

George, by my score keeping I won 2 to 1. But I figured you'd appeal #1, so giving you a point for that makes it a draw.

I'm not convinced that the current draw of the GFCI's circuitry is what causes it to trip when neutral is faulted, but I accept that this is a reasonable explanation. This behavior is by design, though, not just a side effect. If you check out the PDF below, it refers to "grounded neutral detection" in several places, with the implication that there is dedicated circuitry just for that purpose. Also, page 32 says a GFCI "must trip with a 2 ohm grounded neutral". So it is a specific UL requirement.
Understanding GFCIs from NEMA

Mark,
Sounds like we have a hung jury.
#1 definately is my point.
#2 may be mine too. I said I wasn't sure of my equipment accuracy, but you confirmed GFCE use power with your better meters.
SOOOO SCORE 3 PTS TO GEORGE.
Thanks for the chat.
George