Trying to understand AFCI breakers for troubleshooting.

[Wayne Collins]

Why does the AFCI breaker trip when you have a grounded to grounding short. For instance: metal romex clamp screw in metal light box, (box is grounded through grounding) penetrates and touches grounded conductor. There is no potential difference between them. Same place. Is it because at the sub panel they seperate and come back together at the main panel? Also on a non arc fault branch circuit would there be a thermal hot spot at clamp screw? Thanks. Wayne Collins



--------------------------------------------------------------------------------

 

[76nemo]

No potential between them? Does your DMM say that, or are you guessing that?

 

[Chamuit]

Nevermind .

 

[Dennis Alwon]

I believe the op is talking about a grounded to grounding short. Thus if the neutral touches the egc it will trip the arc fault. I believe the arc fault will do as the GFCI in this case, and sense the difference in current between the hot and the return path (the neutral). If the grounding conductor comes in contact with the neutral then some of the current will travel back on both conductors. This differnce will be sensed by the arc fault and it will trip. I think that's right.:smile:

 

[don_resqcapt19]

The grounded to grounding short creates a parallel path for the grounded current. Some of the current will flow on the grounding conductor and if this current exceeds ~30mA the GFP part of the AFCI will open the circuit. On a non-AFCI circuit, this is unlikely to create a hot spot as long as there are no problems with the grounded conductor between the panel and the grounded to grounding short.

 

[480sparky]

It is sensing a ground fault, something it is designed to do. That ground fault will trip it.

 

[ELA]

Both new GFCIs and AFCIs have a circuit added specifically to check for a neutral to ground short on the load side. They incorporate a second coil that "injects" a signal onto the load side wiring so that it can sense the neutral to ground short.

 

[don_resqcapt19]

Both new GFCIs and AFCIs have a circuit added specifically to check for a neutral to ground short on the load side. They incorporate a second coil that "injects" a signal onto the load side wiring so that it can sense the neutral to ground short.

Wouldn't that mean that the AFCI should trip on a grounded to grounding short without a load? As far as I know the AFCI does not do that.

 

[ELA]

Don,
I tested one here and it trips on a neu-gnd (load side) short with no load present.

http://forums.mikeholt.com/showthread.php?t=104129&page=3

 

[ELA]

Don,
I have to qualify that I do not directly know that the AFCI was "designed to" detect the neu-gnd short without load power as the GFCI is.


While it did trip on the unit I tested it was unusual to me that it required 0.1 ohms or less resistance to trip. I had tested a combination GFCI a while back it it would trip on 30 ohms or less.

So it could be that my test breaker or conditions were an anomaly. While this unit I tested did trip on a direct short (neu-gnd) it does make me wonder why it took so little resistance to make it happen.

I would be currious to hear of others experience. In a quick search I did not find anything directly in writing saying that the AFCI must detect this type of fault.

 

[ELA]

Don,

Thanks for questioning my comment about the AFCI having similar internal sensing specifically for sensing a neu-gnd short (like the newer GFCI does).

After further testing I now feel that was incorrect for me to say.
I had just started trying to learn more about AFCIs via testing one I purchased -since I do not normally encounter them in my daily work.

I never felt quite right about the test result that I pointed to in the other tread and after review I see what was happening and it makes more sense now. The AFCI is sensing in the "normal differential current" mode and not via any special injection core like is in the GFCI.

I reviewed my test setup - that was done in my garage :grin: . The circuit I was using I had thought had no upstream loads but I was incorrect. The test was not valid due to the conditions shown in the diagram below.
(The AFCI was not mounted in the load center but setup for a bench test)

I moved the extension cord to another outlet that I was sure did not have an upstream load and the AFCI no longer tripped on a neu-gnd short.


So this now poses another question for me.

Can AFCIs be used in a sub-panel ?

If so wouldn't that create the same type of tripping I experienced in my test - due to the neutral voltage drop buildup above the main cabinet neu-gnd bound point- due to any loads feed by the sub panel?

 

[jerkteacher]

Wouldn't that mean that the AFCI should trip on a grounded to grounding short without a load? As far as I know the AFCI does not do that.

I have cut across a neu and ground with no load for a light (ie: the switch turned off, no light in place yet either) and have had the arc fault trip nearly every time. I have also had apprentice who couldn't device push a recep. in the box and have the neutral and ground touch, will also cause a trip.:smile: