AFCI Tripping- Anyone have any suggestions?

Status
Not open for further replies.

BMacky

Senior Member
Location
Foster City, CA
I recently remodeled a home where we converted a bedroom into an office, but by definition, the closet, which now holds a printer, copier and a bunch of networking gear, still makes this space considered a bedroom. So I ran a dedicated circuit to all outlets in this space, including the smoke, which is interconnected to the one outside in the hallway. Nothing else in this space is fed by any other circuit. During the remodel phase I was sure to be certain nothing existing had been tied into this new circuit.

The day after I installed the AFCI (Murray/Siemens single-pole 20A combo) they called me to say the "office" breaker had tripped while shredding paper. I went to the home that day and did a little research. The shredder that was used was on another circuit in the house that fed an outlet in a room several feet away. I traced the circuit to a breaker that happened to be directly below the AFCI in the panel. The breaker was an old Westinghouse tandem 15A, and was pretty loose. I pulled it and inspected the tab on the bussing, thinking maybe it was causing some arcing in the panel that might be somehow backfeeding its way into the arc fault circuitry, perhaps via the neutral (but how?) since this was stabbed down on the opposite phase as far as the bussing goes. I changed the breaker to a new one with a much tighter fit. I also did a thorough check of all other terminals in the panel, since I had the panel open. All seemed fine.

I buttoned things up and asked the cutomer to monitor what was going on at the time it trips if it did trip again. A week went by, maybe 10 days.

Well, it did trip, and they were using the shredder when it happened. They emailed me a list of all loads on at the time of the trip, so I have some data to work with (or some red herrings to scratch off what hair is left on my head).

The question is if there is any possibility that this shredder had caused the problem, even on a different circuit. I posted here to get some feedback from the field, and will do some more research as well. I have found that reading other posts related to afci's is good to do, but haven't stumbled upon any new data that helps. I will be contacting the mfr as well, but will be happy to hear from any of you that have suggestions.
 

Dennis Alwon

Moderator
Staff member
Location
Chapel Hill, NC
Occupation
Retired Electrical Contractor
I don't know how but it may be possible the shredder is causing the problem. I would try another AFCI first and then I would move that CB away from the shredder CB-- I have no idea why I would do that but I would try anything.

I had a similar experience with 2 GFCI's years ago that fed a large pond area outside. I could not get the GFCI breakers to not trip- moved them all over the panel. I final installed two GFCI recep. just outside the panel and never had a problem again. I cannot explain it but I am certain that nothing else changed but the GFCI Breaker to GFCI recep.:-?
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110102-1728 EST

I think that it is entirely possible for a voltage transient on the input side of the AFCI to cause a false trigger of the AFCI. This might not even occur thru the arc fault detector in the AFCI but might capacitively couple to the SCR gate that drives the coil that trips the breaker. In other words further into the system than where the signal processing is done to detect and arc.

If this is the case, then a low pass filter at the input to the shredder might reduce (eliminate) the problem.

With great difficulty, in others words many tries, I have been able to create a voltage transient on the input side of a GFCI that causes the device to trip, and I believe it is via the SCR gate because of the circuit board layout.

.
 

BMacky

Senior Member
Location
Foster City, CA
More info on this issue

More info on this issue

OK, so I did a little more troubleshooting (actually just ran a few different scenarios) on the circuit in question. I had the homeowner repeat the existing trip conditions and of course it did trip, but it took quite a bit of shredding to do so. I then moved the shredder to another circuit (recall the original trip was NOT on the AFCI-protected circuit) and we were able to repeat the trip there as well. The tripping occured much quicker (immediately in most cases) when the shredder was run in reverse. SO, obviously we're headed for the conclusion that it definitely has something to do with the shredder.

As Dennis mentioned regarding the movement of the breaker within the panel, I tried a couple of other positions but still came up with the trip condition.

The sub-panel that the AFCI is in feeds another sub-panel in the house. I tried plugging into an outlet fed from that (2nd) sub and it also tripped the AFCI as well.

The next attempt to resolve will be the TVSS device on the shredder. If that doesn't fly I will swap the breaker.

Thanks for the replies and advice. I will keep updating with results.

What's gonna happen now that we have to install AFCIs practically everywhere in the house now?????:confused:
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110103-0833 EST

BMacky:

An answer to your PM question.

I use a Corcom 5VR1 filter in 8' Slimlines to reduce RFI (Radio Frequency Interference) back into the power line, and in a power supply for our gaging equipment.

This filter consists of a two stage LC Pi filter (LC means Inductance and Capacitance), and capacitors from each line to enclosure, and this means connection to the EGC and ground.

The rating is 5 A 120/250 V. Corcom, http://www.cor.com , http://www.cor.com/Series/PowerLine/R/ , is now part of Tyco.

Corcom has many other models. You did not mention the shredder current. This will be a determining factor in your choice. Also you could add an MOV across the line between the filter and the shredder. The filter should be close to the shredder.

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
The day after I installed the AFCI (Murray/Siemens single-pole 20A combo) they called me to say the "office" breaker had tripped while shredding paper. .

Is this the Siemens brand of breaker with indicator LEDS? If so is it indicating the trip is due to an Arc vs Ground Fault?

Out of curiosity have you tried running the shredder on the AFCI protected circuit itself?

Does removing any of the other loads -located on the AFCI circuit make any difference?
 

beanland

Senior Member
Location
Vancouver, WA
Algorithm

Algorithm

AFCIs use a complex algorithm to decide whether an arc fault is occurring. Universal motors, with brushes, can cause arcing that can look to some AFCIs as a fault. Noise suppression at the shredder may help. Very possibly the shredder has a noise suppressing capacitor that has failed. A different brand of AFCI may solve the problem.

There is no testing or rating program for AFCIs to detect whether they falsely detect arcs that do not exist. UL testing just tests the AFCI for a very limited series of faults that do not represent appliance loads that may be connected. This is why power tools used during construction often cause AFCIs to trip. I would love for Consumers Reports to run a full set of AFCI tests to determine how reliable AFCIs really are.
 

jcase

New member
Somewhat related information, perhaps it will be useful to others in the future

Somewhat related information, perhaps it will be useful to others in the future

Just adding my own personal experience with a similar problem, in case this information is of any use to others in the future. I'm not claiming to be any expert, just a homeowner who has spent a lot of time researching this issue online.

I've had a Sole F80 treadmill for over 3 years - works great, never had any problems. I lived in a 1992-built house, and the treadmill recepticle was on a normal 15-amp circuit. Then I moved in 2010 to a new house. The house was built in 2008, and the original builder also finished off a 3rd floor bonus room in 2010 (using the same electrician). All the bedroom and bonus room circuits have 15 amp AFCI circuit breakers (everything is Eaton Cutler Hammer), per the code since it's a new house. My treadmill trips those AFCI circuits (15 amps, AFCI) every time - I have it on the 3rd floor, but I also ran an extension cord to various other recepticles on the 3rd and 2nd floors with the same result. It worked fine only if I ran an extension cord downstairs to a non-AFCI circuit, but that's not a realistic solution as I have small children and a wife who doesn't want to see extension cords strung all over the house. My builder's electrician told me I was out of luck, that he's seen motor-based equipment like treadmills trip AFCI circuits, and there's nothing he can do since the code prevents him from putting a non-AFCI circuit in. Fair enough, I can't ask him to risk his license...but I'm not about to write off a $1500 treadmill.

So I did a lot of reading online (including mikeholt.com) about others who had similar problems. I found several instances where people had put SOMETHING in between the treadmill and the recepticle that solved the problem. In some cases, that SOMETHING was just an extension cord; in other cases, it was a portable GFCI (don't understand that one, as treadmills should also trip a GFCI); and in other cases it was a surge supressor. The working theory is that whatever that SOMETHING is, it somehow masks or filters the treadmill's arcing signal (arcing because that's how a treadmill motor works) enough that the AFCI doesn't trip over it. I finally had luck by putting a specific surge supressor in between the treadmill and the AFCI recepticle. Run-of-the-mill power strips didn't work. Next a surge supressor with a 'noise filter' didn't work - that 'filter' was EMI/RFI noise reduction of up to 42 decibels, according to the packaging. Being stubborn and house-poor, I tried one last filter before resorting to expensive electrical re-work. The last filter was designed for home theater, with a stated EMI/RFI filter of up to 75 decibels, from a frequency range of 150kHz to 1MHz. It was a Belkin brand. I've also seen a Tripp-Lite brand (ISOBAR models) that has similar EMI/RFI filtering, but I couldn't find that in stock locally. So if you've got a similar problem, pay close attention to the level of EMI filtering in a surge supressor, and look for something better than 40 decibels.
 

cgj

New member
electronic transformer on separate branch circuit - seperate neutrals

electronic transformer on separate branch circuit - seperate neutrals

i was called to a home i had wired two years ago with a afci in a bedroom upstairs tripping. every time a light was turned on, did not matter which one, would trip the afci breaker. could not find any problems with connections or nm cable in room, so started turning off breakers in subpanel until afci no longer tripped. it was a low volt transformer plugged into separate branch circuit downstairs. plugged in new low volt xfmr and no more problems. brought low volt xfmr home and put on a test bench with two different branch circuits and hooked up a picoscope to capture waveforms. have jpeg of captured waveforms
the low volt xfmr's filtering circuit no longer worked properly and injected electrical noise onto the neutral bussbar in subpanel, so it did not seem to matter if it was not plugged into afci circuit. have had this happened two times before in subpanels with 4/0 al ser feeders but not yet in mainbreaker panel with bonding jumper in same panel as afci.
tried different manufactures afci and only two of them did not have problem unless xfmr plugged into thier circuit. plugged 1500w heater into afci circuit and loosened wirenut, waveforms from loose wirenut and bad low volt xfrmr nearly identical when reading neutral
 

wolfeman

New member
Afci

Afci

I think we are going to run into a whole lot of problems with afci's on all recepts and lights. I have run into issues with afci's tripping because a 3 wire was stapled to tightly and when staple loosened it would hold? My concern is how do you turn on lights without some arc on that circuit, I know the breakers are supposed to realize the difference but seriously?
 

ELA

Senior Member
Occupation
Electrical Test Engineer
brought low volt xfmr home and put on a test bench with two different branch circuits and hooked up a picoscope to capture waveforms. have jpeg of captured waveforms

plugged 1500w heater into afci circuit and loosened wirenut, waveforms from loose wirenut and bad low volt xfrmr nearly identical when reading neutral

Where's the beef?
 

mutualrepect

New member
Location
dayton,oh,usa
Electrical Technician

Electrical Technician

I recently remodeled a home where we converted a bedroom into an office, but by definition, the closet, which now holds a printer, copier and a bunch of networking gear, still makes this space considered a bedroom. So I ran a dedicated circuit to all outlets in this space, including the smoke, which is interconnected to the one outside in the hallway. Nothing else in this space is fed by any other circuit. During the remodel phase I was sure to be certain nothing existing had been tied into this new circuit.

The day after I installed the AFCI (Murray/Siemens single-pole 20A combo) they called me to say the "office" breaker had tripped while shredding paper. I went to the home that day and did a little research. The shredder that was used was on another circuit in the house that fed an outlet in a room several feet away. I traced the circuit to a breaker that happened to be directly below the AFCI in the panel. The breaker was an old Westinghouse tandem 15A, and was pretty loose. I pulled it and inspected the tab on the bussing, thinking maybe it was causing some arcing in the panel that might be somehow backfeeding its way into the arc fault circuitry, perhaps via the neutral (but how?) since this was stabbed down on the opposite phase as far as the bussing goes. I changed the breaker to a new one with a much tighter fit. I also did a thorough check of all other terminals in the panel, since I had the panel open. All seemed fine.

I buttoned things up and asked the cutomer to monitor what was going on at the time it trips if it did trip again. A week went by, maybe 10 days.

Well, it did trip, and they were using the shredder when it happened. They emailed me a list of all loads on at the time of the trip, so I have some data to work with (or some red herrings to scratch off what hair is left on my head).

The question is if there is any possibility that this shredder had caused the problem, even on a different circuit. I posted here to get some feedback from the field, and will do some more research as well. I have found that reading other posts related to afci's is good to do, but haven't stumbled upon any new data that helps. I will be contacting the mfr as well, but will be happy to hear from any of you that have suggestions.[/Q. Failure to terminate the ungrounded (hot) conductors to different phases can result in overloading the neutral conductor, I believe if the shredder is on same phase or if their neutrals are shared then bam problem with overload.
 

tallgirl

Senior Member
Location
Great White North
Occupation
Controls Systems firmware engineer
Are you absolutely, positively, 100,000% sure the shredder is really on that other circuit?

The only other thing I can think of, weird as it may be, is that they have some kind of power line filter gadget on that AFCI that's sending the result of the filtering back to the AFCI at the wrong time for its tastes.
 

djd

Senior Member
djd

djd

110102-1728 EST

I think that it is entirely possible for a voltage transient on the input side of the AFCI to cause a false trigger of the AFCI. This might not even occur thru the arc fault detector in the AFCI but might capacitively couple to the SCR gate that drives the coil that trips the breaker. In other words further into the system than where the signal processing is done to detect and arc.

If this is the case, then a low pass filter at the input to the shredder might reduce (eliminate) the problem.

With great difficulty, in others words many tries, I have been able to create a voltage transient on the input side of a GFCI that causes the device to trip, and I believe it is via the SCR gate because of the circuit board layout.

.
This could be simply transient voltage from the shredder, and or are you saying it is the afci, I think the shredder but I might swap the breaker position with the garbage disposal.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110710-2026 EDT

djd:

What I am saying is that an AFCI or GFCI should not trip resulting from abnormal voltage on the AFCI or GFCI input (the bus). Abnormal includes large voltage transients. The AFCI or GFCI should only trip as a result of something on its output side.

I do suspect the shredder is the source of the noise, and that filtering its input may solve the problem. But still I would describe the breaker as being defective in design if it trips from input noise with moderately high probability.

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
110102-1728 EST
I think that it is entirely possible for a voltage transient on the input side of the AFCI to cause a false trigger of the AFCI.
.

110710-2026 EDT

djd:

What I am saying is that an AFCI or GFCI should not trip resulting from abnormal voltage on the AFCI or GFCI input (the bus). Abnormal includes large voltage transients. The AFCI or GFCI should only trip as a result of something on its output side.
.

Gar,
Your responses might seem a little inconsistent at first glance but I think I understand what you are saying.
I would qualify that "under normal conditions" an AFCI or GFCI should not trip due to transients or "noise events". However we know that abnormal electrical transients and noise events do occur and can cause electronics of all sorts to perform erratically or abnormally.

110710-2026 EDT
I do suspect the shredder is the source of the noise, and that filtering its input may solve the problem. But still I would describe the breaker as being defective in design if it trips from input noise with moderately high probability.
.
I do not agree that a breaker is necessarily defective in design if it trips due to "noise".
A person needs to qualify what they mean by "noise".

Any good quality electronic device is designed to withstand a certain level of noise and hopefully tested to a standard at a particular level of interference.
Any electronic device can be purposefully caused to act erratically if a high enough level of "noise" is introduced (depending upon what level of noise immunity it was designed to withstand ... )


If the upset is due to a high voltage transient then I can see either a GFCI or AFCI false tripping ... whether the transient were on the input or output side.

If the upset were due to normal leakage or arcs then one would expect that only load side events would cause a trip.

There have been a number of reports of AFCI's tripping due to "noise" on adjacent circuits (not loads on the AFCI in question).
The ability of the AFCI to discriminate between downstream and upstream arcs may be compromised by the level of noise interference introduced.

I suspect (but have no direct test data to support) that an AFCI that is loaded to the minimum arc detection current of 5A (assuming a "non-noisy load"),
may be upset by an upstream noise source more easily than if it were loaded at less than 5A.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110711-1253 EDT

ELA:

My comments derive from experiments on the Leviton GFCI that we have previously discussed.

With great difficulty I have been able to create false tripping of the GFCI from voltage transients on the input side. I believe this would be greatly reduced by a redesign of the board layout to the SCR. There is an unnecessarily long path from the LM1851 to trigger the SCR without protection from transient coupling. At the National chip is a capacitor, C2, that is effectively 1/2 to 3/4 inch away from the SCR gate-cathode terminals, but is electrically across those terminals. Thus, a long path length to couple into the gate. This is what I am describing as a defective design.

I would judge that the capacitor should be directly across the gate-cathode pins, and also possibly a small resistor, maybe a 47 to 470 ohms at the gate in series with the trace from the LM1851M.

I have not modified one of these to see if this would reduce the probability of a false trigger.

The possibility of dv/dt at the SCR anode is another possibility, but I think there is enough junk on the anode side that a high dv/dt there is unlikely.

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
Gar,
Yes I recall, and as I pointed out in our previous discussion you can easily reproduce false tripping in a GFCI when you use "EFT- electric fast transient" test equipment.

It is total speculation on your part to point at the SCR gate. For all we know it could be the LM1851 chip producing a false output when it latches up due to a transient event.

Unless a person is testing with "real" test equipment that can readily reproduce false tripping it is all speculation. I will grant you that it may be an educated guess but still a guess ... until testing proves what is actually occurring.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110711-1631 EDT

ELA:

I agree what I am suggesting is speculation, but it is something that pops out visually.

It could be the LM1851, but like the SCR anode supply the LM1851 supply has a fair amount of filtering.

Testing with well defined test equipment is obviously the best way, but one can still get some useful information from a crude setup. If you have an installation where moderately consistent tripping occurs with a GFCI or AFCI and you change the problem device with a different model or manufacturer, and the tripping is eliminated or greatly reduced in probability, then you have some valid information to determine the next step.

.
 
Status
Not open for further replies.
Top