AFCI/GFCI Outlet Keeps Tripping

Reply

  #1  
Old 02-04-18, 06:02 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Question AFCI/GFCI Outlet Keeps Tripping

Short Summary: I added a new GFCI/AFCI combo outlet today (replaced an old 2 prong outlet) and the outlet keeps tripping. Here is what I did to troubleshoot:

- I tried 2 other outlets (same brand, brand new unopened box), they all tripped.
- I disconnected all of the known LOAD outlets (as in left all wires open with wirenuts on top) which are downstream of this particular outlet, it still trips. That being said, the circuit does have substantial amount of lights connected to it which are all in the LOAD of this AFCI/GFCI outlet.
- I disconnected the LOAD wires on the outlet itself, it does NOT trip. Which means the LINE wires are fine.

Please help me troubleshoot this. I would like to know if I were to hire you (assuming you here is a licensed electrician), how would you tackle this?

Detailed Summary: The outlet belongs to a circuit which I believe is the very first added to the house in 1950s. It's the first breaker in my breaker box. The circuit is labeled as "kitchen lights, 1st floor entry way, etc" and indeed it controls 2 outlets in the living room, entryway ceiling light, the first floor bathroom light and GFCI outlet, and kitchen lights, flood lights outside by the kitchen. Sounds good right?

Well, I found out today, it strangely has an outlet on the 3rd floor all by itself!. Our bedroom on the 3rd floor has 5 outlets and of those 5 outlets, this 1 outlet is on the corner by itself thats on this separate circuit which is 95% 1st floor. Once I disconnected that outlet, I realized it is the LINE of the entire downstairs circuitry with the EXCEPTION of the kitchen recessed lights (so strange). Once I realized that, I tried moving my AFCI/GFCI outlet from the living room to there which is when I encountered this outlet tripping all the time.

Question I have is, if I were to hire an electrician, how would he troubleshoot this? This is probably by far the longest circuit in the house as you can see spanning to the 3rd floor, then all the way down and all around the entry way of the house, kitchen, and the first floor bathroom.

I m obviously concerned that there is a short somewhere and I dont know how I can possibly troubleshoot this. I'm also glad I did not install an AFCI breaker because it would constantly trip (I'm assuming). Do note however, nothing particular ever happened anywhere in this circuit as far as shorts, breaker tripping, lights dimming, sparks, or anything else. This is all happening today because I'm on a mission to AFCI/GFCI protect all of our outlets/lights

Please help!
 
  #2  
Old 02-05-18, 07:07 AM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
The more I think about this, the more confused I am getting:

- Downstream of that outlet is 3 occupancy sensors (1 of them is a 3 way) which all use ground as EGC - no neutral in those switches.
- Furthermore, unlike other old cables, this particular cable is fully cloth (gray) from the breaker box to until it reaches to this outlet. From that point on it is in bx. Now throughout this circuit there are ground connections to metal boxes and I always checked using a multimeter ground to hot and it was 120v. And indeed if ground werent working, those sensors wouldnt function.

Could it be possible that somewhere in this cable because its not in BX originating from the box that somehow the neutral is used as ground?

My theory is, someone somehow connected a neutral to the BX armor which is then used as a ground everywhere. When i mean everywhere I mean several light fixtures, switches, and outlets and some of this grounding was done by licensed electrician btw couple of years back when we had moved in. By grounding I mean replacing 2 brong outlets with 3 probg regular outlets and wvery one has its ground attached to the metal box with ground screw.

Can someone explain to me how these metal boxes are acting like a ground when the original line cable from the breaker box is full cloth? And could this grounding downstream (to possibly neutral) be why this GFCI keeps tripping?

Its not tripping in downstream installations because where its installed before its line was bx cable whereasnow the line is cloth and load is bx.

I need a way to diagnose this because I m very worried now about all these grounded boxes and where they are grounded.
 
  #3  
Old 02-05-18, 07:30 AM
P
Member
Join Date: Feb 2016
Location: United States, Virginia
Posts: 1,513
Received 125 Votes on 99 Posts
Is it tripping on a ground fault or arc fault? I assume GFCI is tripping, so if the sensors are using ground as a neutral then that will trip a GFCI.

I thought you originally had a GFCI outlet protecting this circuit with no problems?
 
  #4  
Old 02-05-18, 08:17 AM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by pattenp View Post
Is it tripping on a ground fault or arc fault? I assume GFCI is tripping, so if the sensors are using ground as a neutral then that will trip a GFCI.

I thought you originally had a GFCI outlet protecting this circuit with no problems?
Its tripping as soon as you hit reset, literally instantly. I m guessing that is gfci not afci. I tried 2 new unopened outlets.

Part of this circuir downstairs, lets call it he subcircuit, works fine. However what works fine is downstream of all the lights and sensors.

That being said, I do have another case where there is an AFCI/GFCI outlet upstream of 2 occupancy sensors and bunch of other lights and I dont have this problem.

The three notable things that are different in this circuit vs the other circuit that has sensors and works with gfci are:

- In a one of the light switch boxes, there is a small light bulb that is powered by the hot wire. Imagine a tiny bulb that is the size of a light switch. It takes up as much space and is right next to the switch it is connected to. The whole setup looks very old. Imagine a jumper wire that connects to the two ends of a small very tiny light fixture (there is no neutral). This is so that we know the lights in laundry room are on. So when the light is on the power travels through the jumper on the small bulb back on to the hot wire.

- Two of the 4 occupancy sensors in this circuit are three way sensors and are made by different company than the ones I use elsewhere which doesnt have the issue with GFCI.

- There are a total of 4 occupancy sensors on this circuit vs 2 in the other circuit that works with GFCI.
 
  #5  
Old 02-05-18, 09:42 AM
P
Member
Join Date: Feb 2016
Location: United States, Virginia
Posts: 1,513
Received 125 Votes on 99 Posts
If the little light blub is on the circuit that is protected by the GFCI and the return path from that light bulb is using a ground and not a neutral it will trip the GFCI instantly. The GFCI monitors power going out on the hot and returning on the neutral, if it is not balanced it will trip. Some of the return power over the ground causes the imbalance.

The indicator light should show which type of fault tripped the outlet.
 
  #6  
Old 02-05-18, 10:13 AM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by pattenp View Post
If the little light blub is on the circuit that is protected by the GFCI and the return path from that light bulb is using a ground and not a neutral it will trip the GFCI instantly. The GFCI monitors power going out on the hot and returning on the neutral, if it is not balanced it will trip. Some of the return power over the ground causes the imbalance.

The indicator light should show which type of fault tripped the outlet.
Could you please explain what you mean by indicator light? As soon as I press reset the red light comes on for a split second and then no light is on gfci.

That little light bulb has 1 incoming wire and 1 outgoing. There is no neutral and there is no ground (the box is grounded with BX cable’s armor assuming).

Its wired like this: The outgoing wire from the switch to the light fixture is connected to the left hand side of this little bulb thing, imagine like a metal rod with a screw on left and screw on right and in the middle is a threaded area that bulb goes in. The incoming wire is on left screw.. Then the bulb is there in the middle. On the right hand side is the wire that goes to the actual light fixture.

So imagine the hot wire to the fixture has this little bulb in the middle with no neutral wire. The power goes through this bulb. There is no return path for current. I dont even understand how it would work now that I m thinking about it.
 
  #7  
Old 02-05-18, 12:56 PM
P
Member
Join Date: Feb 2016
Location: United States, Virginia
Posts: 1,513
Received 125 Votes on 99 Posts
I think I understand the indicator light wiring on the switch but not knowing all the internals of the setup of the switch sensor it's hard to provide any guidance on the tripping of the outlet or if that is the cause. If you look at the instructions for the outlet it should state what the indicator light does when the outlet trips on a GFCI fault vs. an ARC fault. I believe the indicator just goes out on a GFCI fault but may change color or flash when it is an ARC fault trip.
 
  #8  
Old 02-05-18, 02:07 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by pattenp View Post
I think I understand the indicator light wiring on the switch but not knowing all the internals of the setup of the switch sensor it's hard to provide any guidance on the tripping of the outlet or if that is the cause. If you look at the instructions for the outlet it should state what the indicator light does when the outlet trips on a GFCI fault vs. an ARC fault. I believe the indicator just goes out on a GFCI fault but may change color or flash when it is an ARC fault trip.
This indicator light has nothing to do with the sensors. It was probably put in 70 years ago looking at how t was wired. I will take a picture when I go home and post.

I fail to understand a how light bulb can glow by just hot wire. This bulb is on the hot wire leaving the switch (the switch in question is not sensor its a regular 3 way switch).

How can I precisely point out if any current is flowing to that light switch box from this indicator bulb using my multimeter without electrocuting myself or causing fire?

I m thinking maybe the box is acting like a neutral/ground for the little light bulb because of the way its mounted.
 
  #9  
Old 02-05-18, 07:22 PM
Tolyn Ironhand's Avatar
Group Moderator
Join Date: Nov 2007
Location: United States
Posts: 13,802
Received 252 Votes on 220 Posts
Can someone explain to me how these metal boxes are acting like a ground when the original line cable from the breaker box is full cloth?
If the cable from the panel is only cloth wires, and there is no ground wire or other metallic path in the cable, then it is an ungrounded wiring method. That being said, it is possible that the circuit is somehow getting a grounding path someplace in the house. How I can only guess.

I think I understand the indicator light wiring on the switch
Some lighted switches will draw a very small amount of current through an incandescent bulb. It is not enough to make the lamp glow, but it is enough to light the switch. The return path is the neutral wire connected to the lamps in the fixture.

You have already determined that the GFCI/AFCI device is good by disconnecting the load wires and it resets and holds. If I was troubleshooting it I would start by removing switches, receptacles, and light fixtures and search for anything that is wrong. Make sure there are no neutrals touching anything grounded and that there are no hots sharing one neutral like what is done in a multiwire circuit.
 
  #10  
Old 02-06-18, 07:17 AM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by Tolyn Ironhand View Post
If the cable from the panel is only cloth wires, and there is no ground wire or other metallic path in the cable, then it is an ungrounded wiring method. That being said, it is possible that the circuit is somehow getting a grounding path someplace in the house. How I can only guess.


Some lighted switches will draw a very small amount of current through an incandescent bulb. It is not enough to make the lamp glow, but it is enough to light the switch. The return path is the neutral wire connected to the lamps in the fixture.

You have already determined that the GFCI/AFCI device is good by disconnecting the load wires and it resets and holds. If I was troubleshooting it I would start by removing switches, receptacles, and light fixtures and search for anything that is wrong. Make sure there are no neutrals touching anything grounded and that there are no hots sharing one neutral like what is done in a multiwire circuit.
How can I check to make sure how much current is flowing through the ground of that circuit?

I have done some basic multimeter troubleshooting and I can see neutral to ground is reading 0.283 volts when lights are not on and 0.49 when the lights are on. Now this is not a scientific test because there are low voltage items hooked up on this circuit like a water dispenser ir small night lights.

I want to be able to point out how much current is in the ground and for the record its still beyond my ability to compherend where that ground is hooked up I have 6 circuits with pure cloth wiring and they all show grounded boxes.

Using an extension cord I also checked ground to ground and I get about 0.09 volts between two different circuit’s ground.

This is probably the most difficult circuit to troubleshoot at my house as it is connected to:

- Spot lights in kitchen
- Entry way lights
- 1st floor bathroom room lights and sensor
- 1st floor bathroom gfci
- Laundry room sensor and light
- Garage sensor and light (only 2 LED fixtures)
- Basement entrance sensor and light (only 2 LED fixtures)
- 2 outlets in living room
- Flood light outside (which works)
- Outlet in kitchen that is hooked up to water dispenser

You see what I m talking about?

Can you tell me how exactly would one check each of these boxes? Do I simply check the box vs a known ground (extension cord from another circuit) and see if DMM picks up anything?
-
 
  #11  
Old 02-06-18, 07:49 AM
P
Member
Join Date: Feb 2016
Location: United States, Virginia
Posts: 1,513
Received 125 Votes on 99 Posts
You will need to use an analog meter. A digital meter will give false readings even when there is no voltage. Readings of 0.49 and 0.09 volts on a 120V system using a digital meter are most likely what is known as phantom voltage readings.
 
  #12  
Old 02-06-18, 12:04 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
I have feeling there is a shared beutral somewhere or my floodlight from 20 years ago is the issue
 
  #13  
Old 02-06-18, 07:33 PM
Tolyn Ironhand's Avatar
Group Moderator
Join Date: Nov 2007
Location: United States
Posts: 13,802
Received 252 Votes on 220 Posts
Measuring the way you are is a waste of time. A GFCI will trip on 4-6 miliamps of current which is too small for a standard clamp on meter.

I would start with the main suspects on the circuit: The sensors, the flood light, etc. Remove them from the circuit and see if the GFCI/AFCI will hold. If not, continue troubleshooting. You can also disconnect the circuit in the middle, that way you can tell which half the fault is on.

Note: If the flood light still works with the circuit off, that would be a big red flag to me.
 
  #14  
Old 02-07-18, 06:42 AM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by Tolyn Ironhand View Post
Measuring the way you are is a waste of time. A GFCI will trip on 4-6 miliamps of current which is too small for a standard clamp on meter.

I would start with the main suspects on the circuit: The sensors, the flood light, etc. Remove them from the circuit and see if the GFCI/AFCI will hold. If not, continue troubleshooting. You can also disconnect the circuit in the middle, that way you can tell which half the fault is on.

Note: If the flood light still works with the circuit off, that would be a big red flag to me.
Is there a way to troubleshoot for a "major" possible red flag without dismantling everything?
 
  #15  
Old 02-07-18, 02:00 PM
T
Member
Join Date: Oct 2005
Location: USA
Posts: 1,693
Received 47 Votes on 46 Posts
This may require a bit of more advance troubleshooting. Do you have a DMM, or VOM? Please give the manf and part no. and we can see if you can measure low AC currents.

I'm headed into measurements of the Hot side and Neutral side leakage to ground.
 
  #16  
Old 02-07-18, 02:35 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by telecom guy View Post
This may require a bit of more advance troubleshooting. Do you have a DMM, or VOM? Please give the manf and part no. and we can see if you can measure low AC currents.

I'm headed into measurements of the Hot side and Neutral side leakage to ground.
Thanks, I have this DMM

https://www.amazon.com/dp/B00NWGZ4XC...&pd_rd_w=1jukB

As of today, I have now installed AFCI/GFCI outlets throughout the entire house, they are on every single circuit with the exception of the circuit that has the microwave/fridge/oven (which I did it on purpose) AND this one circuit that gave me the problem.

In addition to installing AFCI/GFCI on every circuit, I have also converted every single other 2 prong or 3 prong non TR outlet to TR outlet and of all this work, I have not had a single issue with AFCI or GFCI tripping (and we are talking about probably 20 circuits). So I m pretty sure this isn't some stupid mistake on my side for this particular circuit (and I did check all my personal work in this circuit).

To sum up, I would like to find out what is wrong with this. Initially I was thinking its the flood light but today I installed another AFCI/GFCI on another circuit that has another flood light and it works fine. The flood light on the problem circuit also works fine.

The three things I am not able to cross out as not problems are:

- This is the only circuit which has 4 occupancy sensors, other ones that GFCI doesnt trip have either 1 or 2.
- TWO of the FOUR occupancy sensors are made by LUTRON vs the other ones that dont upset GFCI is LEVITON. The Lutron ones are the ones I had to use for 3 way switches.
- As I mentioned, this circuit has a switch (just a regular 3 way switch) where via a small jumper cable, a small light bulb is being lighted up inside the switch box. I do not see a neutral wire on this light bulb. The hot wire going to the actual light fixture (which does have a neutral) does a little "detour". I dont understand how can a light bulb glow without a neutral wire.
 
  #17  
Old 02-07-18, 02:50 PM
T
Member
Join Date: Oct 2005
Location: USA
Posts: 1,693
Received 47 Votes on 46 Posts
So, your clamp on has a 2A current range and 3 1/2 digits. You can read down to a milliamp or so.

If I was faced with this problem, I'd pull the fuse/breaker and the problematic GFI... recept out, and remove the four leads. So, you are looking at 4 wires. a pair towards the fusebox and a pair going to all the loads. All loads removed or OFF.

I would connect JUST the two black together, and loop the current probe around the black wires (only). Fuse back in, and then on the 2A scale, what do you read? IF over 10 mA, then that is why the device trips. If not over 10mA, then we need to look at the neutral side, but using the Ohmeter. Fuse back out, measure neutral LOAD Volts to ground. what do you read? If low, then try Ohms to ground, fuse still OUT.

NOte that there is one fault type that is hazardous, a shared neutral with another circuit. You MIGHT have 120V on that neutral, even if you pulled the fuse to that BLACK wire. carefull!
 
  #18  
Old 02-07-18, 03:02 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by telecom guy View Post
So, your clamp on has a 2A current range and 3 1/2 digits. You can read down to a milliamp or so.

If I was faced with this problem, I'd pull the fuse/breaker and the problematic GFI... recept out, and remove the four leads. So, you are looking at 4 wires. a pair towards the fusebox and a pair going to all the loads. All loads removed or OFF.

I would connect JUST the two black together, and loop the current probe around the black wires (only). Fuse back in, and then on the 2A scale, what do you read? IF over 10 mA, then that is why the device trips. If not over 10mA, then we need to look at the neutral side, but using the Ohmeter. Fuse back out, measure neutral LOAD Volts to ground. what do you read? If low, then try Ohms to ground, fuse still OUT.

NOte that there is one fault type that is hazardous, a shared neutral with another circuit. You MIGHT have 120V on that neutral, even if you pulled the fuse to that BLACK wire. carefull!
Thank you so much for your response

I have some follow up questions:

- When you say fuse out, you mean breaker off? I dont have a fuse
- When you say current probe, you mean the hook thing that opens and closes on the DMM not the red/black probes correct? So I just open the hook and hang it on the 2 black wires that I connected
- In this setup (where I manually connect 2 black wires possibly with a wing nut), what do I do with the neutrals? They stay on the receptacle or do I also wire nut them or just leave them disconnected?
- Also can't I do this neutral to ground reading even now with that breaker off and possibly without disconnecting the receptacle/manually connecting the black wires? Maybe I should do that first as the initial test?
 
  #19  
Old 02-07-18, 03:08 PM
T
Member
Join Date: Oct 2005
Location: USA
Posts: 1,693
Received 47 Votes on 46 Posts
yes, fuse or breaker
yes, clamp around the BLK wires that are connected only by a wirenut.
neutrals floating in space.
The new recept is in another room.
Lets save the Neut test for last.

the blk/red test leads are for Volts or Ohms only. Its in the book!!

Also, describe where you are finding "ground".

SAFE; rubber shoes, insulate wire ends wtih wingnuts, don't kneel on knees, touch only one wire at at time, even if you think they are cold, etc.

And, verify with a non-contact voltage tester or other reliable method.
 
  #20  
Old 02-07-18, 06:33 PM
D
Member
Thread Starter
Join Date: Jun 2017
Posts: 438
Received 0 Votes on 0 Posts
Originally Posted by telecom guy View Post
yes, fuse or breaker
yes, clamp around the BLK wires that are connected only by a wirenut.
neutrals floating in space.
The new recept is in another room.
Lets save the Neut test for last.

the blk/red test leads are for Volts or Ohms only. Its in the book!!

Also, describe where you are finding "ground".

SAFE; rubber shoes, insulate wire ends wtih wingnuts, don't kneel on knees, touch only one wire at at time, even if you think they are cold, etc.

And, verify with a non-contact voltage tester or other reliable method.
Sorry, what do you mean by new receptacle in another room? Am I not doing this on the problem receptacle - which is the first in series?

The ground is the metal box OR i use an extension cord from another circuit and use its ground hole.

Also after doing several of these outlets today I now know that there indeed is a ground wire in these cloth cables. BUT strangely the ground wire is connected to a ground screw that is OUTSIDE of the box, on top of it. In a lot of boxes, you basically have the line/load cables goijg into the box, held by clamps, then the bare ground wire EXITS from the same hole that cable comes in and then a screw is screwed on top of the box and these ground wires connect to that. So there is ground and boxes are grounded which explains why I was able to have ground connections qhen I also screwed in to the box with my own ground wires.

Anyways.....

I just realized something very strange.... when this receptacle, the first one in the circuit (or so I thought) is fully disconnected, nothing in the circuit works EXCEPT one thing - the recessed kitchen lights. How is that possible? How can the line first go to kitchen lights and then come back to this receptacle? Because there would be no neutral unless they ran 3 conductor wires and these are not that and our kitchen was remodeled 10 years ago vs this 75 year old cloth cable.

So I m thinking maybe someone decided to “steal” a neutral from somewhere else to complete the circuit. Obviously no easy way to know but if this is the case then its pita, no way to fix it.
 
  #21  
Old 02-07-18, 06:43 PM
T
Member
Join Date: Oct 2005
Location: USA
Posts: 1,693
Received 47 Votes on 46 Posts
Sorry, what do you mean by new receptacle in another room?
My lame attempt at describing removing the recept from the debug exercise altogether.

the first one in the circuit (or so I thought) is fully disconnected, nothing in the circuit works EXCEPT one thing - the recessed kitchen lights. How is that possible?
So, then the lights are the first, and the recept second? I don't follow on why this is odd or a defect. The lights have a switch on the BLK side, and the lights pick up the Neutral, all before they head for your first recept.

In any case, the GFI trip ckt is only looking at the load on the LOAD output.
Verify you meant to say a few posts ago, GFI recept and not GFI breakers.
 
 

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