AFCI troubleshooting

[hillbilly1]

Found another issue with AFCI’s. I have previously had them trip within seconds of each other. POCO pole mount transformer connections arcing. Found today, beyond a reasonable doubt, AFCI’s do detect arc’s on the line side. Sitting in the kitchen today, hear a bunch of them trip. Reset them a couple seconds later, they all trip again. Reset again, trips again. Checked voltage at the panel, all good. Go out to the pole, I can hear arcing while the A/C is running, stops when it is off. Called the poco, they came out, sure enough, another bad connection on their side. Breakers now all reset and hold. Transformer is about 30 years old, and this makes the third time the connections have given problems in the last two.

 

[hbiss]

Found today, beyond a reasonable doubt, AFCI’s do detect arc’s on the line side.

Put that in a sticky somewhere here!

-Hal

 

[drcampbell]

That seems inevitable, since the line side is connected to the load side by a pair of dry relay contacts when it's not tripped.

 

[kwired]

Found today, beyond a reasonable doubt, AFCI’s do detect arc’s on the line side. Sitting in the kitchen today, hear a bunch of them trip.

kind of always figured that "arc signature" it is looking for should be present anywhere in the circuit - the "circuit" that current takes includes the source as well as feeders and service conductors.

Transformer is about 30 years old, and this makes the third time the connections have given problems in the last two.

Are they possibly "remaking" connections but still have metallic components that were overheated in previous failures but not replaced?

Re-terminating a conductor into same connector might be asking for trouble again in relatively short time, need to cut back to "good conductor" and replace terminal that has been overheated to get back to having a good connection.

 

[synchro]

Found another issue with AFCI’s. I have previously had them trip within seconds of each other. POCO pole mount transformer connections arcing. Found today, beyond a reasonable doubt, AFCI’s do detect arc’s on the line side. Sitting in the kitchen today, hear a bunch of them trip. Reset them a couple seconds later, they all trip again. Reset again, trips again. Checked voltage at the panel, all good. Go out to the pole, I can hear arcing while the A/C is running, stops when it is off. Called the poco, they came out, sure enough, another bad connection on their side. Breakers now all reset and hold.

Were there AFCI's that didn't trip? The reason I ask is it would be interesting to know if the AFCI's would trip with just a noisy bus voltage, but without a noisy current going through them if they have no load. Since the bad connection is now fixed, that may be a moot point unless you noted such details.

 

[hillbilly1]

Were there AFCI's that didn't trip? The reason I ask is it would be interesting to know if the AFCI's would trip with just a noisy bus voltage, but without a noisy current going through them if they have no load. Since the bad connection is now fixed, that may be a moot point unless you noted such details.

All seven would trip. You could cut any one on, in a couple of seconds it would trip. None would hold regardless of order of reset.

 

[hillbilly1]

kind of always figured that "arc signature" it is looking for should be present anywhere in the circuit - the "circuit" that current takes includes the source as well as feeders and service conductors.

Are they possibly "remaking" connections but still have metallic components that were overheated in previous failures but not replaced?

Re-terminating a conductor into same connector might be asking for trouble again in relatively short time, need to cut back to "good conductor" and replace terminal that has been overheated to get back to having a good connection.

They replaced the lugs the last time, I could tell because they used a multi tap lugs that were bright and shiny. I wasn’t there this time they fixed it, so hopefully they cut out the bad wire.

 

[Ravenvalor]

My takeaway from this is that the next time a customer calls with AFCI tripping issues I should tell them to call the POCO first.

 

[kwired]

Were there AFCI's that didn't trip? The reason I ask is it would be interesting to know if the AFCI's would trip with just a noisy bus voltage, but without a noisy current going through them if they have no load. Since the bad connection is now fixed, that may be a moot point unless you noted such details.

Back when AFCI's were fairly new thing here (we had them amended out of code until adoption of 2008 NEC) I remember we were still working in a newly constructed home, had many permanent circuits energized, most the load was lighting. Out in the country and not really all that unusual for a power "blink" to happen occasionally in rural systems - basically a recloser likely has been called on to operate. I recall several AFCI's tripped but not all of them. Kind of also remember figuring out that the ones that tripped likely had load at the time and the ones that did not trip were likely not loaded at all.