AFCI service call, should I have done anything differently?

[Mr. Serious]

I did a search before posting this and noticed three other recent threads on the topic of AFCIs. I read some of the content and will go back and read the rest:

AFCI troubleshooting

Probably for all you service and punch list folks. Trying to find some thoughts and tactics to find root cause of problems with an arc fault circuit without having to open a bunch of back stabbed devices, fans, lights etc. This is of course when it's already verified that it's not the breaker.

forums.mikeholt.com

A story about dual function breakers, flickering lights, and galvanic corrosion

Hi Folks, Long time lurker, first time poster. I wanted to share my story for posterity here since this forum has helped me so much in the past. If this violates the rules in any way please feel free to delete, but I hope someone might find it useful so I'll give it a go. I'm the owner of a...

forums.mikeholt.com

https://forums.mikeholt.com/threads/i-have-been-on-the-fence-a-lot-about-afcis-part-of-me-thinks-its-corruption-between-manufacturers-and-code-makers.2563374/

The call came in this morning, husband and wife wanted someone out there right away to check 3 or 4 breakers that repeatedly kept tripping. On the phone he said, "It's weird."

I pulled up to a newer house, 2 or 3 years old, of the style popular nowadays here in the south-middle-west, white with natural wood trim. The house is newer, but the neighborhood is older, and there was probably a run-down house there before that got razed to make room for the new one.

The wife was there to let me in, she said now that I was there, the problem was not recurring anymore. She let me into the garage to inspect the panels. There were two panels right next to each other. The breakers were all on, but the ones they had been having trouble with were marked with masking tape. There were five of them, all double-pole 20s, with the panel label describing two circuits for each (MWBCs). There were at least 10 similar MWBCs in the two panels, and these five had been the ones giving them trouble. I looked at the circuit descriptions, and didn't see anything in particular that these 5 had in common. (But I took pictures, so I can re-look again if need be.) With the panel cover off I could verify the type of breaker, Homeline CAFCI, mainly double-pole 20s. A few other dedicated circuits were standard breakers, not AFCI. One other thing - each panel had a whole-house surge protector, and it was clear that they had been added later as an afterthought, but the homeowner said they had not added them, they were already there when they bought the house.

I verified all the connections were tight. Both panels were wired as a main panel with shared neutral/ground, and there was a 6 AWG ground wire connecting the two. Each panel had a 150A main breaker. Voltage 244: 121.1 on one leg and 122.9 on the other, checked in both panels. I looked outdoors, they had a 400A meter can with bypass lever, it was not sealed and therefore I was able to inspect all the terminations inside it as well, everything looked good, no problems. I had her start up one of the furnace blowers while I read the voltage. Voltage went from 121.1 to 121.8 on the leg I was measuring, but I couldn't be sure it was related to the furnace coming on. I explained I was looking for symptoms of a loose neutral, and not really finding them, and she said they hadn't noticed any lights dimming/brightening, so everything seemed OK there.

She said she had a science background, so I explained the theory of AFCIs, how they work and what they're supposed to do. And, I concluded that because there were multiple breakers tripping, not a specific problem with just one circuit, that whatever arc or radio-frequency noise was causing these to trip was originating outside of the house, and she should call the power company and see if they could find any problem. As I mentioned, it is an older neighborhood. The service was underground, but the power distribution lines are all overhead in the overgrown fence lines between rows of houses, and I know there is a lot of infrastructure there that is 50 to 70 years old.

I put all the covers back in place and charged a service-call fee. As I drove out of the neighborhood, I noticed there is a TV station there, only 2 blocks from the house. Not their main broadcasting tower, but a smaller tower with a doppler radar on it and a microwave link to their main tower. Could that cause noise that the AFCI breakers would pick up and misinterpret as an arc? I think it's more likely there actually was an arc on the power company lines. The homeowner said they had a similar problem about 6 months ago, but it went away and didn't happen again until last night.

 

[tom baker]

Normally an AFCI will only trip with about a 5 amp load.
You could substitute a GFCI breaker and see it it trips if it does not then its the AFCI
It could be the AFCI has a newer version out, but I seem to recall a thread on this site about QD/homeline no longer making 2 pole AFCIs...
So perhaps a call to homeline tech support and get the newer 2 poles while you can

 

[Mr. Serious]

Yes, I saw that thread also about the breakers no longer being manufactured starting this month. It sucks because I'm a new contractor and haven't saved up enough funds to be buying a bunch of materials I'm not going to use right away. I guess I could stock up on one box of each type (QO and Homeline).

 

[farmantenna]

just change them to dumb breakers and explain to HO then run away. If they don't want this change, explain the ramifications and get paid and then run away.

 

[K8MHZ]

And, I concluded that because there were multiple breakers tripping, not a specific problem with just one circuit, that whatever arc or radio-frequency noise was causing these to trip was originating outside of the house,

Look for big ham radio or CB radio antennas in the area.

 

[Mr. Serious]

Re antennas, would the doppler radar from the TV station 2 blocks away cause it? And if so, would a newer model of breaker help?

Looking back at the pictures I took, the double pole ones are HOM220CAFI. Issue No. DP-3640.

 

[farmantenna]

Look for big ham radio or CB radio antennas in the area.

then what? you have no case with ham radio. many devices must accept rf interference cited in fcc part 15(?). the disclaimer no one reads. when I was a teenager I was a ham radio operator with tower and trees and wires and was causing interference and got FCC complaints and I added mitigation techniques and I never stopped transmitting and never heard anything again

 

[retirede]

Is anyone else concerned that the voltage went up (admittedly slightly) when a load was added? A furnace blower isn’t much of a load.

 

[Mr. Serious]

I was thinking I should have had her put something in the microwave for 10 seconds on high power instead of turn on the furnace blower. As it was, whichever furnace she turned on was not located in the garage so I couldn't hear it turn on, and I wasn't at all sure if the voltage change was directly related to that, or just gradually going up anyway.

 

[romex jockey]

She said she had a science background, so I explained the theory of AFCIs, how they work and what they're supposed to do

go here Mr S>>>>

COMBINATION AFCIs: WHAT THEY WILL AND WILL NOT DO / combination-afcis-what-they-will-and-will-not-do.pdf / PDF4PRO

Manufacturer X provides a simple definition of a series arcing event [7]: “A series arc is an arcing incident across a break in a conductor. A common example is a cut across one of the two wires in a lamp cord, with a dangerous arc forming in

pdf4pro.com

~RJ~

 

[gar]

211004-2348 EDT

retirede:

No problem with a small line to neutral voltage rising when a load is placed on a panel. Think about it.

You have a power company transformer located some distance from the main panel, likely a center tapped transformer supplying 120-0-120 at the transformer, three wires from the transformer, each about the same resistance.

When these supply wires are loaded with a two terminal load (the blower motor) there is an an approximately equal voltage drop on the two loaded wires. In the transformer there is also a voltage drop on whatever winding is loaded. There is also a smaller drop on the transformer primary side. This primary drop reflects to both secondaries, but it is less than the secondary drop.

Now suppose that the primary drop common to both sides of the secondary is 1/4 the drop of half of the secondary drop, then if we load 1/2 of the secondary, and don't load the other half the loaded side sees a higher drop.

Next add in the voltage drop of the three distribution lines, and now a voltage drop at each half of the secondary wires at the main panel is probably greater than the voltage drop within the transformer.

Since there is no load change on one hot distribution wire there is no added voltage drop on that phase. Thus, if you measure the unloaded phase voltage at the main panel it only changes by the amount of voltage drop seen at the transformer, and the neutral drop. But if the loaded phase has sufficient current to produce a voltage drop on the neutral that is greater than the voltage drop seen on the transformer secondary, then you will see an increase in voltage of the unloaded phase relative to neutral in comparison with the loaded side to neutral at the main panel.

In my home I see about an 0.5 V rise of the unloaded side with a 10 A load on the loaded side, and about a 1.0 V drop on the loaded side.

.

 

[romex jockey]

Think about it.

so the more proximal to xformer , the more any inrush creates imbalance , the more possibility it affects toroidal ocpd's ?? ~RJ~

 

[K8MHZ]

then what? you have no case with ham radio. many devices must accept rf interference cited in fcc part 15(?). the disclaimer no one reads. when I was a teenager I was a ham radio operator with tower and trees and wires and was causing interference and got FCC complaints and I added mitigation techniques and I never stopped transmitting and never heard anything again

Some companies make special AFCI breakers that are immune to ham radio interference. IIRC, they say 'ham radio' on the breaker.

 

[gar]

211005-1057 EDT

romex jockey:

Nothing to do with inrush. It is simply a result of ohms law, and relative phase relationships.

.

 

[retirede]

211004-2348 EDT

retirede:

No problem with a small line to neutral voltage rising when a load is placed on a panel. Think about it.

You have a power company transformer located some distance from the main panel, likely a center tapped transformer supplying 120-0-120 at the transformer, three wires from the transformer, each about the same resistance.

When these supply wires are loaded with a two terminal load (the blower motor) there is an an approximately equal voltage drop on the two loaded wires. In the transformer there is also a voltage drop on whatever winding is loaded. There is also a smaller drop on the transformer primary side. This primary drop reflects to both secondaries, but it is less than the secondary drop.

Now suppose that the primary drop common to both sides of the secondary is 1/4 the drop of half of the secondary drop, then if we load 1/2 of the secondary, and don't load the other half the loaded side sees a higher drop.

Next add in the voltage drop of the three distribution lines, and now a voltage drop at each half of the secondary wires at the main panel is probably greater than the voltage drop within the transformer.

Since there is no load change on one hot distribution wire there is no added voltage drop on that phase. Thus, if you measure the unloaded phase voltage at the main panel it only changes by the amount of voltage drop seen at the transformer, and the neutral drop. But if the loaded phase has sufficient current to produce a voltage drop on the neutral that is greater than the voltage drop seen on the transformer secondary, then you will see an increase in voltage of the unloaded phase relative to neutral in comparison with the loaded side to neutral at the main panel.

In my home I see about an 0.5 V rise of the unloaded side with a 10 A load on the loaded side, and about a 1.0 V drop on the loaded side.

.

Yes, sir. I get that.

You get a 0.5V rise with a 10A load.
He has 0.8 rise with maybe a 4A load.
Of course it all depends on the impedance of the circuits.

I guess my point is that some further investigation may have been warranted.

 

[Mr. Serious]

Yes, sir. I get that.

You get a 0.5V rise with a 10A load.
He has 0.8 rise with maybe a 4A load.
Of course it all depends on the impedance of the circuits.

I guess my point is that some further investigation may have been warranted.

So, there may have been a problem with a loose neutral after all. Either that or higher than normal impedance somewhere in the utility wiring. Either way, it's on the utility side so it's really their problem, but I shouldn't have discounted the 0.8 volt difference as inconsequential. Perhaps I should have taken the time to inspect the upstream power source.

Thinking back on it now, I realize I really should have moved my meter lead from one hot leg to the other repeatedly, so I could verify if the other leg fell in voltage when that one rose. Since I don't have that information, there's a chance it still could have been just a general rise in distribution voltage throughout the neighborhood. Perhaps a regulator somewhere stepped up while I was making the measurement.

 

[oldsparky52]

So, there may have been a problem with a loose neutral after all. Either that or higher than normal impedance somewhere in the utility wiring. Either way, it's on the utility side so it's really their problem, but I shouldn't have discounted the 0.8 volt difference as inconsequential. Perhaps I should have taken the time to inspect the upstream power source.

Thinking back on it now, I realize I really should have moved my meter lead from one hot leg to the other repeatedly, so I could verify if the other leg fell in voltage when that one rose. Since I don't have that information, there's a chance it still could have been just a general rise in distribution voltage throughout the neighborhood. Perhaps a regulator somewhere stepped up while I was making the measurement.

Monday morning QB'ing, yea probably not.

I thing I might have used a hair dryer on its highest setting as a load and test both phases while it's off and then on. But again, MM QB'ing.

 

[RRJ]

I had a similar case where there had been a surge from utility company and I ended up replacing the breakers because their chips were acting out. Of course this building did not had a surge suppressor.


Sent from my iPhone using Tapatalk

 

[tthh]

I've seen nuisance tripping on AFCI's that I attributed to compact fluorescent bulbs. New breaker, same problem. Maybe the CF bulbs were crappy, but changing them all out for Philips LED bulbs fixed the nuisance tripping.