puzzeling electrical problem

My son lives in a 100+ yr old house with wiring brought up to mid 60's code. About 8 years ago I added another outlet to a dedicated circuit for a freezer. The freezer is long gone and that space is used for different stuff. I put a GFI outlet on the added box as it it within reach of the kitchen sink (if you have long arms). Recently, the GFI tripped and would not reset. My son replaced the GFI thinking it was toast. Still no change. I looked at it and realized he had hooked it up wrong, I put another one in. Same thing. Now I have figured out that only pops when the outlet is touching the metal box. The ground is attached to the outlet, so the box is only attached to the wall. Anyone have any ideas? I am about to start checking back up the line to see if anyone has re-wired something.
This on has me stumped.
Tim in OR

GFI's can be a real pain in older houses.

Might want to get a volt/ohm meter, check each wire and the box for proper connections and stray voltage. Try checking it under load, see if you get a differential voltage between the ground and neutral, could be a weak neutral connection.

Also make sure the insulation hasn't cut through the clamps inside the box. Make sure the ground is really grounded.

> Now I have figured out that only pops when the outlet is touching the metal box.

Are you saying it trips when nothing is plugged in? Is there a separate ground wire to the GFCI? Are you sure the ground wire is actually tied to ground? Are you feeding any outlets downstream of the GFCI?

If the GFCI and box are both properly grounded, then it shouldn't make any difference if the GFCI is in contact with the box or not. My assumption is one or the other is not actually tied to ground.

GFCIs work by sensing a difference in voltage between the hot and neutral conductors. You can have current running through ground all day without tripping the GFCI, as long as that current isn't originating in the protected circuit. Somehow, when you touch the GFCI to the receptacle box, you're completing a path to ground. I'll bet you will read voltage between the box and the GFCI when they're not in contact.

Mark,
I discovered, by accident, that a GFCI will also trip if the ground wire comes in contact with the neutral wire. Try it, jump a wire from ground and put in the neutral side. Perhaps you can explain why there is a small potential difference between ground and neutral.

I've made water sensing devices that trips a GFCI, turning the power off to a 110v water pump. The water sensor has a small switch that I've wired to short ground to neutral. It works, turns power off to well. It has saved me from having a flooded basement many times. Got 2 wells wired up that way.

I would recommend running the ground wire to the GFCI and see what happens.
George

(quoted from post at 00:46:54 11/04/14) My son lives in a 100+ yr old house with wiring brought up to mid 60's code. About 8 years ago I added another outlet to a dedicated circuit for a freezer. The freezer is long gone and that space is used for different stuff. I put a GFI outlet on the added box as it it within reach of the kitchen sink (if you have long arms). Recently, the GFI tripped and would not reset. My son replaced the GFI thinking it was toast. Still no change. I looked at it and realized he had hooked it up wrong, I put another one in. Same thing. Now I have figured out that only pops when the outlet is touching the metal box. The ground is attached to the outlet, so the box is only attached to the wall. Anyone have any ideas? I am about to start checking back up the line to see if anyone has re-wired something.
This on has me stumped.
Tim in OR

Click to expand...

As previously stated somebody has wired neutral current through the ground system somewhere.

>I discovered, by accident, that a GFCI will also trip if the ground wire comes in contact with the neutral wire.

Well, that depends. If you're talking about shorting the neutral feed to ground, that's shouldn't do anything. But it will definitely trip if you sort neutral to ground at a GFCI receptacle or downstream of a GFCI. You've given the neutral current an alternate path so some current will naturally follow the ground conductor. That throws the hot vs. neutral current out of balance, and it only takes a few milliamps of imbalance to trip the GFCI.

Didn't you swear off responding to my posts?

Actually a GFCI trips by sensing a CURRENT (NOT voltage as stated) DIFFERENCE in the Hot Ungrounded Line conductor "current" and the Neutral Grounded Conductor "current". There's a torroidal coil encircling the Line and Neutral and as some know current passing through a coil induces a voltage. HOWEVER if the incoming line and outgoing neutral currents are the same they cancel, so there's zero net current so no voltage gets induced into the torroidal coil. ITS THE "CURRENT" not Voltage DIFFERENCE (Hot and Neutral) THAT TRIPS A GFCI........

HOWEVER if theres a fault leakage whereby all the current flowing in the hot is NOT all being returned by the Neutral the current difference (something like 5 milliamps) sends a signal which trips the GFCI breaker.

NOW BOTH the metallic receptacle or junction box PLUS the green grounding terminal on the GFCI (or any) receptacle must be grounded.

Sooooooooo if the GFCI trips by connecting its Grounding receptacle to the Box THERE MUST BE A FAULT/SHORT by which some current flowing in the hot isn't all being returned by the Neutral.

Soooooooooo where's that short/fault?? Maybe a wire has bad or stripped insulation and is touching the metal box?????????? I think I would check all the wires coming into and exiting that box as well as the circuit that goes to that box or exits it..The fault circuit isn't completed until the box comes into the equation.

Other problems such as mixing and matching or crossing Neutrals and Equipment Grounds can come into play, but Id start looking at that box and all its wiring first. Have one of those three prong testers that verifies proper wiring???

Best I can do not being there

John T

I didn't proofread close enough, John. Of course the GFCI senses the difference in current, not voltage. You WANT to have voltage between hot and neutral!

Hey, Im the KING of not proofreading lol Im often in a hurry and run off at the mouth then hit the post button lol

John T

Mark,
Swear, I don't swear.

Please try taking a wire putting it in the ground, round hole of GFCI, then put the wire in the neutral side of GFCI. All of mine will trip. Yea, they aren't suppose to. Please just try it. Then tell me why. It's way above my pay grade.

Like I said, I have both of my wells, 110v pumps, wired to a GFCI. I have a water detector that will short ground wire to neutral, causing GFCI to trip. shutting the power off to well. It works. I was surprised to discover that a GFCI will actually handle a pump, but they do.
George

The GFCI trips because you created a ground fault, and that's what they're supposed to do when they sense a ground fault. It doesn't matter whether the fault is on the hot or cold side, because the GFCI doesn't really distinguish between hot and cold. All it cares about is whether the current through the two conductors is exactly the same or not. When you short the neutral to ground at the receptacle, you've created another path for current to follow. As a result, there's less current traveling through the neutral conductor than through the hot conductor. Ground fault!

This probably makes sense as long as there's a load on the GFCI, so there's actually current traveling through it. Now I think what's puzzling you is you've seen it happen even if there's nothing plugged in. But when you connect the neutral and ground together, you've made a nice loop which will pick up induced current from all the current flowing in adjacent conductors. When you induce current in the neutral but not in the hot conductor, the GFCI says "Ground Fault!"

I was thinking either induction or capacative reactance or both may be playing a part. Only a guess. Just came form pole barn. I measured 0.1 vac difference between ground and neutral right out of the 2 different GFCI with no loads. Came to the house and no difference. Still when neutral touches ground, on all 4 GFCI I tested tonight, trip.

So back to the original post, I think the guy should ground his GFCI first, not using the box as ground and see what happens.

Very similar experience at my FILs vacation home several years ago. Troubleshooting results drove me around the bend to the point I called an electrician. Turns out when he had the exterior stuccoed years earlier, one of the long screws holding on a sheet of styrofoam pierced the insulation on a run between one outlet and another. It didn't start acting up till a nearby lightning strike took out a tv on my watch.

Is there any wiring down stream of the GFCI? If so disconnect it from the Load Side and see if the GFCI trips when the outlet is installed.

Thanks for the Help. Things kinda went sideways so I won't be able to get over there till the weekend. But now I have a new resolve and some ideas where to look. Thanks again,
Tim in OR