GFCI tripping with no loads


  #1  
Old 04-23-19, 05:06 AM
G
Member
Thread Starter
Join Date: Mar 2018
Location: United States
Posts: 8
Upvotes: 0
Received 0 Upvotes on 0 Posts
GFCI tripping with no loads

I'm having problem with a GFCI circuit in my kitchen. It keeps tripping, even with nothing plugged into the outlets. I'm certainly no electrician, but I'm not totally clueless. I know the basics, and I understand how a GFCI circuit works. There's another circuit on the other side of the kitchen that wasn't GFCI protected (but should be according to code). I replaced the first outlet in the circuit with a GFCI, and it's working fine.

The problem GFCI is above the kitchen sink, in a wall with tile. The house was built in 97, but the tile was added later. The GFCI is the first outlet in the circuit, and it feeds two other outlets. One outlet is in the same wall as the GFCI outlet. The other one is on the end of a cabinet that forms a penninsula in the kitchen. The two outlets have only one wire pair going to each one, so there is (I hope) a junction box somewhere. I think it's under the cabinet, and might require removing a piece from the cabinet to gain access. The GFCI keeps tripping, even with nothing plugged into the outlets. Sometimes it can go hours between trips. Sometimes seconds.

Things I've tried:

- Replaced the 20+ yr old GFCI outlet. No change. The new one trips just as often.
- Disconnected all loads from the outlets. No change.
- Disconnected the load line from the GFCI outlet. The nuisance trips stopped.
- Measured the resistance of the load downstream of the GFCI outlet (nothing plugged in). 250 Kohms across the lines. Open from each line to ground. Shouldn't they all be opens?
- Measured the resistance of both non-GFCI outlets after disconnecting them. Measured an open across each line, and an open from each line to ground. So the resistance measured above isn't coming from the outlets.
- The previous owners installed tile on this wall, but didn't properly install spacers. I installed spacers in the outlets on the tile wall. These spacers are plastic, and are made for this purpose. When I pushed the GFCI outlet back into the wall, it shorted. The breaker tripped as soon as I closed it. Tile is ceramic, right? So it's not conductive, right?

What could cause this? Bad insulation? Can insulation deteriorate in this amount of time? If so, is there any way to fix it without tearing out a tile wall to replace wiring?
 
  #2  
Old 04-23-19, 06:20 AM
I
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 9,785
Upvotes: 0
Received 45 Upvotes on 43 Posts
- Measured the resistance of the load downstream of the GFCI outlet (nothing plugged in). 250 Kohms across the lines. Open from each line to ground. Shouldn't they all be opens?
They should all be open / infinite. Sometimes these measurements can be difficult to interpret as general-purpose meters are not reliable at the extremes. I do however consider this a good lead to investigate. An electrician could use a meter called a megger to get a conclusive determination if this cable has gone bad through some means.

Possible causes that come to mind would be something like a cable that is pinched too tight in a cable clamp or staple. Minor water infiltration into the cable, either direct cause like a plumbing, siding or roof leak or indirect like hot moist air creating condenstation. Cable damage in the wall from something like hanging a picture or a rodent chewing. Something hidden on the circuit you don't know about, pipe heat tapes for example.
 
  #3  
Old 04-23-19, 06:42 AM
E
Member
Join Date: Apr 2019
Location: USA
Posts: 109
Upvotes: 0
Received 17 Upvotes on 16 Posts
Did you measure the resistance downstream of the GFCI again after disconnecting the outlets? It's unlikely, but perhaps the terminals on one of the outlets is contacting the tile or other material that is at least partially conductive. The ceramic in the tile is an insulator but there might be some glazes that have a resistance low enough to be measurable.

When you placed a spacer behind the outlet, this moved the terminal screws on the outlet forward. Does the box there have a metal mud ring cover with a narrow opening, or is it a narrow "switch box"? If it's the first case then the terminals may have shorted on the narrow opening of the mud ring. The outlet needs to be centered in the opening, and wrapping the outlet with electrical tape to cover the terminals would be a good practice in this case.


​​​​​​



​​​​
 
  #4  
Old 04-23-19, 02:18 PM
CasualJoe's Avatar
Member
Join Date: Jan 2010
Location: United States
Posts: 9,871
Received 185 Upvotes on 166 Posts
I think I'd try pulling the downstream receptacles out of their boxes and testing the GFCI device again. It's very possible that all you have is a neutral and ground touching inside one of the boxes.
 
  #5  
Old 04-23-19, 07:07 PM
G
Member
Thread Starter
Join Date: Mar 2018
Location: United States
Posts: 8
Upvotes: 0
Received 0 Upvotes on 0 Posts
Sometimes these measurements can be difficult to interpret as general-purpose meters are not reliable at the extremes.
My meter is a lab quality meter. 250 KOhms is not at the extreme.

A meggar might be needed. I don't have one of those.

Did you measure the resistance downstream of the GFCI again after disconnecting the outlets?
No. I measured with one outlet disconnected. But I didn't measure with both disconnected. Both wall outlet boxes are plastic. The one in the cabinet is metal. I measured the lines with the cabinet outlet (from the metal box) pulled out. Still measured 250 KOhms.

When you placed a spacer behind the outlet, this moved the terminal screws on the outlet forward. Does the box there have a metal mud ring cover with a narrow opening, or is it a narrow "switch box"?
There's no metal mud ring cover. The box is all plastic. By "narrow switch box", do you mean, a single wide outlet box?

Thanks for all the suggestions.
 
  #6  
Old 04-23-19, 08:03 PM
PJmax's Avatar
Group Moderator
Join Date: Oct 2012
Location: Jersey
Posts: 64,928
Received 3,947 Upvotes on 3,540 Posts
There should be no continuity what so ever between the neutral and ground.
You must locate where that resistance is coming from.
 
  #7  
Old 04-24-19, 10:46 AM
E
Member
Join Date: Apr 2019
Location: USA
Posts: 109
Upvotes: 0
Received 17 Upvotes on 16 Posts
Try disconnecting the hot and neutral lines on the load side of the GFCI one at a time, and check if the GFCI trips. If only one side causes tripping then that could provide an additional clue.
For example, if it trips with just the hot side wire connected then the insulation may be breaking down somewhere with the application of 120V. This would not show up on an ohmmeter test. If it trips with just the neutral wire connected then it's more likely that an intermittent connection of neutral and ground metal conductors is happening, because any voltage on the neutral should be quite small (not enough to break down insulation).
Like Pete said the root cause for the 250K ohms across the conductors needs to be tracked down. By itself it would not trip the GFCI if the resistance maintained itself only across the hot and neutral wires, with no leakage to ground, when full line voltage was applied. However, this resistance should not be there and tracking where it's coming from is likely to point you to the source of the problem.
 

Last edited by engr3000; 04-24-19 at 11:49 AM.
 

Thread Tools
Search this Thread
 
Ask a Question
Question Title:
Description:
Your question will be posted in: