AFCI's and bad sockets

[mark32]

I had the opportunity to work with some friends the other day that needed to finish a house asap. When the breakers were turned on, one 15a dp afci (Siemens) would trip as soon as any light on one of the circuits was turned on. After checking numerous splices we came to the great room that had six high hats, three on the sloped part of the ceiling and three on the flat, 18' high. All of the splices looked good, of course, I then suggested doing a continuity check between the neutral and ground of the cables between the lights. With the splice apart at the first high hat, on the flat, we had continuity from the ground to neutral down to the sloped high hat. My one buddy then took that splice apart and the continuity went away. Huh? He then grabbed the tester and said that the socket was bad. What had happened was this, on the bracket that the socket mounts to is a little tit that is suppose to keep the socket from spinning when a bulb is installed. (There is an indentation on the socket for this tit) The problem is that this tit was making contact with the same part of the socket that the grounded conductor terminates to. Luckily there is another indentation on the socket so he was able to spin the socket for that indentation to line up with the tit. This had to be done on all three Halo H745ICAT's! One this repair was done the breaker held. Just a heads up if you run into a similar problem. One question that remains, why would the circuit trip when any of the other lights on that circuit were turned on??

 

[Dennis Alwon]

Mark we have had basically the same problem with Halo H99 when you use the socket extender. Drove everyone crazy for a while. I rarely use those sloped ceiling cans but thanks for the heads up.

 

[S'mise]

Thanks mark. Interesting story. I'm sure others will run across this in the future and good to have the heads up.

The misterious tripping is because current flows back on the neutral (including where the bad socket is a ground fault) when any switch is closed.

Put any current on that branch circuit and it flows to ground. (remember all neutrals are tied together)

 

[cadpoint]

One question that remains, why would the circuit trip when any of the other lights on that circuit were turned on??

The fault of lines crossing/touching is still with-in the length of a parallel circuit!

 

[don_resqcapt19]

... One question that remains, why would the circuit trip when any of the other lights on that circuit were turned on??

There is a parallel path for the neutral current on that circuit via the incorrectly installed socket. This will trip the GFP part of the AFCI anytime there is a load on that circuit.

 

[mark32]

Thanks guys for the input.

 

[jmellc]

Thanks for sure. That could save any of us time and nerves trying to find a needle in a haystack. Speaking of socket wiring; also be watchful on appliance panel lights, control panel lights, etc. I've seen a lot of them over the years that have hot to the shell. Should always be neutral to shell and hot to center pin. I once had a range hood tripping breaker. I had wired exactly per label. Opened up hood and found several shells reversed. Once got a nasty shock from control panel light when changing a bulb. Same thing there, every shell in reverse. In every case, the equipment also carried a UL label. See my other posts on UL.

 

[jsmalone1]

I am having a similar issue. AFCI trips instantly when recessed can lights are turned on---take lamps out=OK. So what do I need to do with the sockets??

 

[ActionDave]

Isolate the cans from the rest of the house wiring, remove all the lamps and check for continuity shell to pin, shell to can, pin to can in the sockets.