New GFCI tripping when downstream neutral attached

[Slowbuilder]

I apologize if this has been covered before, but I didn't see this specific issue in a search - please point me to a thread if it's been covered.

I'm finishing up the wiring in my shop extension. I've added a couple of 20a 110v circuits for outlets in the new shop. Both are protected by individual GFCI's as the first outlet downstream from the panel, and the remaining outlets are fed from the load side of the GFCI. All outlets are in metal boxes, with THHN in EMT for all runs.

One circuit works just fine; with the other, the GFCI triggers as soon as it has been reset. If I remove the downstream feeds (hot and neutral) from the offending GFCI, it will reset and stay that way. As soon as I connect the downstream neutral to the load terminal of the GFCI (whether or not the downstream hot is attached) the GFCI will not reset - it immediately triggers.

As I understand GFCI's, they sense a difference in current between the hot and neutral legs. I'm not sure how that can happen if the hot isn't connected.

It was late when I finished up last night, but as I recall there is no continuity to ground on either the hot or neutral downstream legs. Is this a defective GFCI? What should I test?

 

[rlitman]

...As I understand GFCI's, they sense a difference in current between the hot and neutral legs. I'm not sure how that can happen if the hot isn't connected...

Current is flowing from neutral to ground.

 

[wyliesdiesels]

Current is flowing from neutral to ground.

But if that were the case, then connecting the hot leg only, as he did, should trip the GFCI since there would be an imbalance.

 

[American Locomotive]

But if that were the case, then connecting the hot leg only, as he did, should trip the GFCI since there would be an imbalance.

There wouldn't be an imbalance. If the downstream "hot" is all good, the GFCI won't even know that it's connected. If the downstream "neutral" is leaking current to ground (it's not unusual for the neutral to be at a slightly higher potential than earth ground), then there would be an imbalance when it was connected.

 

[Milton Shaw]

Pull all the downstream outlets back out, you will probably see the problem with wiring on one of the outlets. If that doesn't fix it then connect only the first one, try it and then keep adding outlets until you find the one that is faulty.

 

[mm08822]

With downstream wires disconnected from gfci, check for continuity between neut/grd of the downstream hardware. Look for a neutral connection touching ground in each box. It could be a nicked wire in the box.

Disconnect the last half of the downstream recepts and re-test. If it still trips, problem is in the first ½. If no trip, then it’s in last ½ of disconnected recepts.

Could even be a nail driven into cable between recept outlets. If so then check continuity between neut/grd in the cable.

************
typing same time as milton!

 

[Slowbuilder]

Current is flowing from neutral to ground.

But if that were the case, then connecting the hot leg only, as he did, should trip the GFCI since there would be an imbalance.

Not sure I've tested with the downstream hot connected and the downstream neutral lifted...

There wouldn't be an imbalance. If the downstream "hot" is all good, the GFCI won't even know that it's connected. If the downstream "neutral" is leaking current to ground (it's not unusual for the neutral to be at a slightly higher potential than earth ground), then there would be an imbalance when it was connected.

This is my guess - there's a short between the downstream neutral and ground, but I don't see continuity there.

Pull all the downstream outlets back out, you will probably see the problem with wiring on one of the outlets. If that doesn't fix it then connect only the first one, try it and then keep adding outlets until you find the one that is faulty.

That was my plan (visual downstream outlets for mis-wiring add them back one at a time). I was wondering if there was something obvious that I missed.

Thanks all for your quick responses!

 

[Slowbuilder]

With downstream wires disconnected from gfci, check for continuity between neut/grd of the downstream hardware. Look for a neutral connection touching ground in each box. It could be a nicked wire in the box.

Disconnect the last half of the downstream recepts and re-test. If it still trips, problem is in the first ½. If no trip, then it’s in last ½ of disconnected recepts.

Could even be a nail driven into cable between recept outlets. If so then check continuity between neut/grd in the cable.

************
typing same time as milton!

Thanks, MM! It must be neut/grnd connection somewhere downstream; your approach should nail it down quickest. I must be mistaken about the continuity between neut/grnd - the neutral does float a few volts above ground.

No nails, it's all in EMT, surface mount (and no nailing since the install).

 

[wyliesdiesels]

Not sure I've tested with the downstream hot connected and the downstream neutral lifted...

Ok i misunderstood your earlier comment then.

 

[mm08822]

Thanks, MM! It must be neut/grnd connection somewhere downstream; your approach should nail it down quickest. I must be mistaken about the continuity between neut/grnd - the neutral does float a few volts above ground.

No nails, it's all in EMT, surface mount (and no nailing since the install).

My bad, skimmed OP too quickly.

Is there any load plugged in on the same ckt upstream of the gfci?

 

[Zeke]

Pull all the downstream outlets back out, you will probably see the problem with wiring on one of the outlets. If that doesn't fix it then connect only the first one, try it and then keep adding outlets until you find the one that is faulty.

Does it sound like a hot/neutral reverse to you? That's what I thought.

 

[DieselNut88]

How many are downstream of the gfci? I think more than 5 will cause issues. I am not positive on this.

 

[Slowbuilder]

My bad, skimmed OP too quickly.

Is there any load plugged in on the same ckt upstream of the gfci?

Nope. <panel>----<gfci>----<outlets>

Does it sound like a hot/neutral reverse to you? That's what I thought.

Downstream? Could be (anything could happen), but there is nothing plugged in to any outlet.

How many are downstream of the gfci? I think more than 5 will cause issues. I am not positive on this.

There are 9 duplex outlets (4 boxes with 2 duplex outlets each, and an additional duplex outlet in the box with the GFCI) on each of the 2 circuits, downstream of the GFCI on both circuits. I certainly don't expect to be using all or most of them at the same time.

 

[charmin35]

As a DIY'er I was wiring a branch circuit with a GFCI, to a switch, to lighting outlets. I assumed with the switch turned off and no hot current running to the remaining outlets to wire (tested for no voltage) that it would be no problem. But I found that touching both neutral and ground at the same time with a tool would trip the GFCI. Derp.

Others have already pointed that out but just throwing in my story as well. So possibly a ground wire is touching a tiny bit of stripped neutral, maybe just happened when stuffing the outlet into the box after wiring. Or nicked neutral is contacting the grounded conduit (it's grounded right), or a grounded box.

Did you judiciously adhere to strip gauges (and maybe even wrap electrical tape around receptacle screws after wiring)?

 

[American Locomotive]

Some GFCIs can nuisance trip with extremely long runs of wires due to the capacitance of the conductors.

 

[wyliesdiesels]

How many are downstream of the gfci? I think more than 5 will cause issues. I am not positive on this.

nope

 

[red]

I've run into this with a GFCI and using metal boxes.
Don't remember which wire (hot or neutral) but it was shorting out to the metal box from the outlet being pushed into the box.

Divide and conquer using what Milton said "Pull all the downstream outlets back out, you will probably see the problem with wiring on one of the outlets. If that doesn't fix it then connect only the first one, try it and then keep adding outlets until you find the one that is faulty." Except start disconnecting in the middle of your downstream boxes.
Could also be a defective gfci . . .

 

[Slowbuilder]

Thanks all for you input and comments. It turned out to be an easy one.

First, I pulled the load wires off of the GFCI, and sure enough, neutral was shorted to ground on the downstream side. Doing the binary search pattern suggested by MM, I opened up a box, pulled the cover and outlets out, checked for wiring errors (none found), and rechecked the short. On the third box, the short went away when I pulled the cover off.

I think I didn't have the ground bundle pushed deep enough in the box, and it was contacting the terminals on the outlet. A little cleaning up of the wire dressing, and after buttoning things back up, all is well. Both GFCI's reset and stay that way.

Thanks again all for your timely and helpful input!

 

[ard]

I think it is a neutral to ground fault. Check your wiring.





(I kid, I kid.... Just seems like a totally GJ thing to post.... )

 

[Norcal]

Did you use a bare grounding conductor?

 

[Slowbuilder]

I think it is a neutral to ground fault. Check your wiring.





(I kid, I kid.... Just seems like a totally GJ thing to post.... )

Great idea! I'll get right on checking that

Did you use a bare grounding conductor?

I didn't - I used the same size THHN as the circuit (12), but I did use the bare copper sleeve connectors like https://www.homedepot.com/p/Gardner-Bender-14-8-AWG-Copper-Crimp-Connectors-50-Pack-10-311C/202520223, which does leave bare copper ground in the box.

 

[sparky 1971]

I didn't read the entire thread so maybe this has already been suggested. Check the downstream boxes. It has happened more to me than I care to admit that when I shoved the device back, the ground wire bent around and laid over the screws on the neutral. That will trip a gfci or an afci faster than you can push the reset button.

 

[fitter30]

Did all pipe get reamed? Was pulling lube used and all pulls hard? A megger could be used to fest the wiring and if bad enough a ohm meter.

 

[Slowbuilder]

I didn't read the entire thread so maybe this has already been suggested. Check the downstream boxes. It has happened more to me than I care to admit that when I shoved the device back, the ground wire bent around and laid over the screws on the neutral. That will trip a gfci or an afci faster than you can push the reset button.

Sparky, you are correct, an insufficiently dressed ground wire touched a neutral terminal in one of the downstream boxes. Dressed them back in the box, all is well.

Did all pipe get reamed? Was pulling lube used and all pulls hard? A megger could be used to fest the wiring and if bad enough a ohm meter.

Every conduit end was reamed, and pulling lube was used. Turned out to be simple errant ground wire issue.

Thanks to everyone who replied. This is what I love about GJ - so many folks willing to jump in and lend their expertise!

 

[justsam]

It was late when I finished up last night, but as I recall there is no continuity to ground on either the hot or neutral downstream legs. Is this a defective GFCI? What should I test?

The above test, which you did, should have shown this fault, and I would have been replacing the GFCI. Perhaps it was a little too late at night! Glad all is now resolved!