Schedule a Meeting · Talk to a real woodworker: +1 937-890-5197 · 8:00AM – 4:30PM ET

GFCI protected circuits

GarthS

I have both a 500 that I have owned for 35 years and power pro 520 and never had an issue until I built a new house two years ago. With the new arc preventing breakers I have had to change out to conventional breakers. No problem since.

dusty

I am confused.

Does this mean that modern day electrical features are protecting us in a way that didn't exist before?

Is the new equipment (PowerPro) defective (non-compliant) in some way?

Is it comforting to know that you are not the only one with this issue?

DLB

dusty wrote I am confused.

Does this mean that modern day electrical features are protecting us in a way that didn't exist before?

Is the new equipment (PowerPro) defective (non-compliant) in some way?

Is it comfortng to know that you are not the only one with this issue?

I'm confused too. But for sure, depending on one's definition of modern, more of these features are protecting us. Ground Fault(GF) protection has been around, I believe since the 70's. Arc Fault (AF) detection is much newer first effective in 2002 I've read, and only in bedrooms. The National Electric Code (NEC) changes every three years and expands the circuits requiring GF protection, AF protection, or both, in every update. Complicating that, adoption of the version of the code varies by state and locality. IIUC we currently vary from NEC 2008 to NEC 2020. So, debatable whether we actually have a National Code in my view because the minimum requirements in AZ (NEC 2008) for example and TX (NEC 2020) are different. Neither of those is an isolated case, and all five versions from 2008 to 2020 apply somewhere. AF, GF, and combined solutions can be in a breaker, receptacle, or a disconnect. That might not be a complete list.

Ground Fault is well defined IMO, with a clear definition of a threshold and tolerances around the threshold. A PowerPro sensitivity is documented, but several of us report not experiencing it. My educated guess is that it has some ground current, not enough to be above the max trip threshold. But it may trip at the minimum trip current or it may break the threshold if other devices on the circuit have a ground current. For example if the PP has a 3mA ground current, that is not enough to trip. But two devices on the circuit with the same ground current, total 6mA, would be a mandatory trip where any protection device should trip.

Arc fault is not as well defined in anything I've seen so far. Some devices arc if they are working correctly. Brushed motors for example. Lots of variation there. My understanding is the devices work on a form of arc signature analysis with a goal of distinguishing between arcs and arc faults. Depending on what you did in USAF or elsewhere you may have some experience with signature analysis, not arcs but same principle. So, another educated guess here is that this is proprietary and one brand won't work the same as another on nuisance trips. If either a conventional or PP tripped an arc fault, I would want to know why, I don't thing either should arc as a normal function. (I have two houses, with zero arc fault detectors. I'm trying to buy a new house as an investment, it will have roughly 15 arc fault detection CBs, a few of those combined with GF.)

More recently, UL entered the fray and mandated a self test function for the interrupt devices. Adding more complexity, internal circuitry, and 'software.' Probably not software as someone with a military background would use the term, probably not even firmware. Programmed hardware. This is the case I wrote about in an earlier post in which the breaker can indicate seven different reasons for tripping.

- David

RFGuy

As I mentioned before I have no hands on experience with AFCI's like I do with GFCI's. However in reading online what I see mentioned many times is that AFCI's are known for nuisance tripping. So it really is no surprise, to me, that the PowerPro is temperamental on an AFCI breaker. AFCI's are intended to sense an arc which is essentially a leakage from the hot wire to the neutral wire. Any brushed motor is expected to have issues with AFCI nuisance tripping based on online anecdotes. In addition to this, the PowerPro has active power factor correction according to this (https://www.shopsmith.com/shopsmithpowerpro/pp_technology.htm). In my opinion this is also another solid candidate for why a PowerPro might trip an AFCI. Not to mention that the power supply is a switching power supply on the PowerPro which could result in switching transients on the line side that could trip the AFCI. Beefier input filtering would help this, but I don't think the PowerPro has sufficient line filtering inside it based on the multiple radio interference complaints for it over the years. Bottomline is the AFCI is a new technology and there are devices that are known to be incompatible with them...the PowerPro is probably one of those devices. There really shouldn't be any expectation that a PowerPro MUST work with an AFCI or even a GFCI (IMHO). However, if someone feels that way then I recommend they start lobbying Shopsmith to get them to fix the PowerPro to make it compatible because that is what would be required. Likely will require a huge bank of filter caps which won't fit inside the headstock but maybe you can hang them from the benchtubes underneath...

Hobbyman2

one thing that has been proven over and over is the false promise that you can not be shocked with GF protection with out tripping the GF circuit , this has been proven untrue over and over ,depending on the situation you body can act as resistance just as a lightbulb that has one side to the positive and one side to ground . unless it is a direct short the GF protection will not know the difference .If I recall this use to be called a indirect short . AF was usually found in special situations where there were combustible /toxic locations and were very expensive , hair spray was considered to be explosive , . not sure how the consuming device would ever know or detect the difference between a GF and AF circuit as the current is still just current , I have never read any where that it changed the wave form , in the situation described above wouldn't this be a amp issue with the old breaker or the load on the circuit ? after years of service some breakers wear out , springs may have over heated or maybe have broke after so many times of being reset or circuit load being maxed , the old Pushmatic breakers had a lot of issues with wearing out if the circuit was loaded on the high side of the breakers limits , and some were IMHO just had junk springs , but there are still a lot of them in use .JMO

JPG

Filtering? I believe line filters are the reason many devices trip GF devices. Filtering on an unbalanced line can be tricky(not totally effective). 120 in the US is unbalanced. 240 in the US is balanced. A filter that connects to the grounding conductor is asking for GF trips.

BTW one can get shocked with a GF, but that shock WILL BE short lived. After all it is the shock current that trips the GF. It keeps you from latching on and getting killed(or cooked).

DLB

I think the phrase "nuisance trip" may be used excessively, too. Any device that plugs in could have a latent arc fault until it gets plugged into a AFCI circuit which would detect it. Similar, probably way less likely, for a ground fault. Most users would not be qualified or equipped to distinguish between a nuisance trip and an actual fault. At the same time, I'm sure the nice people that came up with this did not do regression testing for the millions of designs and variations that plug in downstream of the protection device. But the whole point of ACFI is to detect faults that have been historically latent. Or to transfer money from one group of people that don't have a lobby to a group that does. ;)

Arc Fault is not required by NEC in the garage (yet) so I'd guess the number of users that have plugged one into a AFCI outlet are few. Seems that we'd need more data on that. Of course, that doesn't mean AFCI is not sometimes put in garages.

- David

edma194

The entire upper floor of my addition has combination AFCI/GFCI breakers with the fancy LEDs. The addition has it's own panel, the original house has those ancient regular breakers that don't cost $50 bucks a pop. I have to check what the basement breakers are in the addition, they aren't GFCI but AFCI may have been required. So far I've had trips on the basement breakers from likely simple overloading. I'm not overly concerned, if the breakers are AFCI and they get tripped occasionally by the PowerPro, it's not surprising from a DVR motor setup like that. I do wonder if some kind of line filtering could eliminate that problem, it would be helpful for all sorts of electronic devices.

I did note this bit from an article on AFCI Nuisance Tripping:

"Wiring Problems – when an AFCI circuit breaker is installed, the wiring for that circuit needs to be done a little bit differently. When AFCI devices first started being installed, it was common for electricians to have problems with nuisance tripping because the neutral conductors for different circuits were connected together at some point outside of the main panelboard. This has never been correct, but it would often happen with no ill effects, so nobody knew it was wrong. With AFCI circuit breakers, this improper wiring will cause the breaker to trip."

Should not be a problem in the addition but the old house is a log cabin with a mix of wiring installed over many decades and I'd have to check carefully what was done there if I ever upgrade, although I would likely rewire most of it even without a major upgrade just for piece of mind. Seriously, when I moved in there was knob and tube wiring in the attic with bare wires running inches away from cardboard boxes full of old clothes and magazines, not to mention plenty of other questionable wiring that needed replacement right away.

RFGuy

JPG wrote
Filtering? I believe line filters are the reason many devices trip GF devices. Filtering on an unbalanced line can be tricky(not totally effective). 120 in the US is unbalanced. 240 in the US is balanced. A filter that connects to the grounding conductor is asking for GF trips.

BTW one can get shocked with a GF, but that shock WILL BE short lived. After all it is the shock current that trips the GF. It keeps you from latching on and getting killed(or cooked).

I don't disagree with you on the unbalanced vs. balanced line differences in filter design and a leaking cap (in a filter) could theoretically trip a GFCI, i.e. a bad filter design could make things worse. Honestly I don't know the internals of an AFCI breaker so I am just speculating here. Any switched power supply is notoriously noisy and backfeeds noise and spurious harmonics into the supply main. Below is a graph from a published paper I previously pointed out here on the forum. In the graph, the top is the input line current & voltage going into the power supply for a switched reluctance motor design (should be similar to the PowerPro because it is also a switched reluctance motor), while the bottom graph shows the difference with a 1µF low pass filter added. As you can see in these graphs, the input supply is significantly cleaner on the filtered one. I suspect the spurious harmonics and noise getting back to an AFCI breaker from the PowerPro are causing the "nuisance tripping" and coincidentally this is also what is causing the radio interference experienced by some PowerPro owners. In a switched mode power supply, very large currents are being switched in/out of storage elements (L's and C's). Unfortunately the switches (FET) aren't ideal so there is a ramp as well as transit time to consider but also leakage currents due to parasitics in the storage elements. Because of this, it is possible that the hot and neutral current waveforms may not mirror each other over brief transients thus tripping the AFCI/GFCI. Filtering on the supply side of the PowerPro serves to not only significantly reduce noise & spurious harmonics, but also serves to provide a source of local charge to supply any short term transients due to the PowerPro - preventing the AFCI from even seeing them. At least, this is my theory and what I am going with at this time...

With regard to AFCI and GFCI, I do believe they save lives but there is always context to consider. I do believe that the great expansion of AFCI/GFCI requirements in the latest NEC is most likely the result of lobbying by breaker manufacturers. Personally I can see how an AFCI might save lives in a bedroom application, I am far from being convinced that AFCI adds benefit in a shop environment.

https://forum.shopsmith.com/viewtopic.php?p=265254#p265254

JPG

I am curious how a capacitor by itself performs a LOW pass function?

RFGuy

JPG wrote I am curious how a capacitor by itself performs a LOW pass function?

If I were trying to solve the EMI concerns for the PowerPro, I would buy/design an LC filter for the input, but regarding the published paper referencing a 1uF lowpass, I can't say for sure what their intent is since I don't think enough details of their setup is shared in their paper. However adapting it to discuss regarding the PowerPro, there could be enough R and L in the line between the breaker and the PowerPro that adding an input cap could create a sufficient lowpass filter (using the parasitics in the house wiring).

JPG

I consider a "C" across the input(Hot - Neutral) to be a High Pass Filter effectively snubbing the high Frequency stuff.

RFGuy

JPG wrote
I consider a "C" across the input(Hot - Neutral) to be a High Pass Filter effectively snubbing the high Frequency stuff.

??? A shunt capacitor will have low reactance to higher frequencies so it will shunt them from the hot to the neutral. This is a low pass function because low frequencies get through and high frequencies are shunted to netural in this case.

So, I dug up some more AFCI information that I think sheds more light on how they function and why a PowerPro might trip one. Below is the block diagram for what is happening in an AFCI. In the excerpt from this below you can see that the key point of how "bad" arcs are determined by an AFCI is by looking at whether the waveform is periodic or not. So, a well behaved brushed motor might be seen as a "good" arc, but asynchronous switching waveforms backfed into the supply line from a PowerPro could be perceived as a "bad" arc. Also spurious harmonics (from a PowerPro) being fed back could also trip the "bad" arc signature circuitry, I suspect.

Except from paper linked below:

A major design consideration was the ability of the AFCI to discriminate between “good” and “bad” arcs. “Good” arcs are characterized by being periodic or repetitive (occur each 60 Hz cycle) and can be non-sinusoidal. “Bad” arcs are characterized by non-periodic or non-repetitive waveforms.

AFCIs use electronic detection circuits to discriminate between good and bad arcing conditions. Once an unwanted arcing condition is detected, the control circuitry in the AFCI trips the internal contacts, thus de-energizing the circuit and reducing the potential for a fire to occur. Methods for the detection of “bad” arcs include looking at certain frequencies, discontinuities, and inconsistencies in the current waveform. For detection, both magnitude and duration of a particular half cycle are required. Some detection algorithms also look at rising or falling edges of an arcing current for their detection criteria. The industry voluntary standard for AFCIs requires a trip if 8 half cycles of arcing occur within a 0.5 second window.

AFCI.pdf

DLB

Combined breakers providing GFCI and AFCI are becoming popular for various reasons, even where only one type of protection is required. Hypothetical case: Garage receptacles require GFCI (true for decades) but has both AFCI and GFCI at the breaker, under warranty. PowerPro, under warranty, trips breaker. Who is paying for fault isolation, and corrective action? Automatic answer from these people is call electrician. Appears to me to require an oscilloscope with a current probe and ability to view per-trigger events (common now in digital scopes), knowledge of how the PowerPro should work, and knowledge of how the CB does work. Electricians, in my experience, might not be qualified to perform the relatively simple task they have. (Though they can occasionally pass a test on it.)

(Note - I have this exact case on the house I'm in the process of buying, except strike PowerPro and replace with Septic System. Septic, like a garage, requires GFCI but they don't have any GFCI-only breakers so it is on AFCI/GFCI. It is hard-wired so can't use GFCI outlet. Didn't trip the regular breaker they had it on when they were ignoring code. Didn't trip AFCI-only when they tried one. They are completely lost, and asking me as the prospective buyer for help. I think that's because they'd like me to tell them not to follow code. Nuisance fault? Maybe. Ground Fault? Seems likely. If no one was looking would the electrician follow code? Clearly not.)

- David

RFGuy

David,

It is a bit of a grey area, but I thought sump pump type applications are exempt from AFCI or GFCI requirements in the NEC. Presumably the septic application you mentioned falls under the same category as a sump pump. It should be reasonably safe to leave it on a regular CB if this is the case.

Having said this though, I don't know why a sump pump motor would trip it unless it has an internal problem. It was my understanding that these types of motors are almost always double insulated internally so in theory they should be next to impossible to trip a GFCI under normal circumstances. You didn't give specifics, but what kinds of tripping are they experiencing with the septic system and a GFCI or AFCI breaker?

paulrussell

An inline capacitor is high pass.

A capacitor to ground is low pass.

An inline inductor is low pass.

An inductor to ground is high pass.

DLB

RFGuy wrote David,

It is a bit of a grey area, but I thought sump pump type applications are exempt from AFCI or GFCI requirements in the NEC. Presumably the septic application you mentioned falls under the same category as a sump pump. It should be reasonably safe to leave it on a regular CB if this is the case.

Having said this though, I don't know why a sump pump motor would trip it unless it has an internal problem. It was my understanding that these types of motors are almost always double insulated internally so in theory they should be next to impossible to trip a GFCI under normal circumstances. You didn't give specifics, but what kinds of tripping are they experiencing with the septic system and a GFCI or AFCI breaker?

Sump pump, septic aerator, and septic pump all require GFCI (not AFCI YET), new features as of NEC 2020. The sump and septic requirements are in different sections, but both previously required a TM CB. Communications with the 'electrician' (meaning perhaps not licensed himself) on hand were difficult. We didn't speak the same language and the builder we were communicating through didn't really understand the technical language. So, I couldn't really determine if the breaker that tripped had a way to differentiate faults (like the flashing LED discussed earlier) or what it was attempting to communicate. Overly simplified logic: TM didn't trip, AFCI didn't trip, GFCI-only not available, AFCI/GFCI did trip. The only thing completely different is the one that tripped includes GFCI. They are trying to find a GFCI-only breaker and get the licensed electrician out there now. I explained the flashing LED thing and said there should be a manual there somewhere, but they couldn't find a manual or a breaker PN. I'm going to look for that on-line anyway, probably matches the key you posted earlier. If I was working on it the first thing I'd do is disconnect the septic system to isolate the problem between the electrical and septic systems just to get the right tech out there. That's easier in this case than in my PowerPro hypothetical.

- David

RFGuy

DLB wrote
Sump pump, septic aerator, and septic pump all require GFCI (not AFCI YET), new features as of NEC 2020. The sump and septic requirements are in different sections, but both previously required a TM CB. Communications with the 'electrician' (meaning perhaps not licensed himself) on hand were difficult. We didn't speak the same language and the builder we were communicating through didn't really understand the technical language. So, I couldn't really determine if the breaker that tripped had a way to differentiate faults (like the flashing LED discussed earlier) or what it was attempting to communicate. Overly simplified logic: TM didn't trip, AFCI didn't trip, GFCI-only not available, AFCI/GFCI did trip. The only thing completely different is the one that tripped includes GFCI. They are trying to find a GFCI-only breaker and get the licensed electrician out there now. I explained the flashing LED thing and said there should be a manual there somewhere, but they couldn't find a manual or a breaker PN. I'm going to look for that on-line anyway, probably matches the key you posted earlier. If I was working on it the first thing I'd do is disconnect the septic system to isolate the problem between the electrical and septic systems just to get the right tech out there. That's easier in this case than in my PowerPro hypothetical.

David

David,

Do you know what the AFCI/GFCI breaker is that tripped on the property? Was it the one you posted about earlier? I don't know the exact part #, but the CH (Cutler-Hammer) series from Eaton similar to what you posted before at Home Depot should follow what is in the instructions linked below. There are instructions on page 5 for how to get the trip code to display (not automatic) and the code resets 2 hours after it happened so you have to check it soon after it trips. This is the only Eaton document that I could find on this series AFCI/GFCI breaker from Eaton and it still only shows 6 trip codes, not 7, but perhaps there is a newer document coming eventually. Don't know if this helps you or not, but thought I'd give it a shot. Yeah, for something like this I think they have to fix it especially if you are looking to buy the property. If only GFCI is required per latest NEC in that area then that is all they need to do, but it sounds like there is a good possibility of current leakage in the septic equipment causing a discharge. More than likely they will need to replace the equipment and not just the breaker, but that is just my guess...

There was an interesting 2 star review for the breaker at Home Depot that you posted. Sounds like it still has issues, including that the fault LED looks like it is tripped when it is not (could this be what is happening there?) See below:

Definitely don't get any nuisance trips with the new Gen 6 rev, but about half of these have an error where the fault LED stays lit continuously. Eaton says this is a cosmetic defect in their breakers, but it does not affect their functionality. I have tested and confirmed this, but for a $60 breaker I expect better quality, especially for a CH (Cutler-Hammer) product. I would seriously hope they're already working on another revision to eliminate this. I would also like to know what in the current revision caused this error as it was not in previous generations, up to and including the 2019 release. The problem didn't start until the late 2020 revision.

eaton-resi-installation-instructions-il003022en-en-us.pdf

DLB

RFGuy wrote Do you know what the AFCI/GFCI breaker is that tripped on the property? Was it the one you posted about earlier? I don't know the exact part #, but the CH (Cutler-Hammer) series from Eaton similar to what you posted before at Home Depot should follow what is in the instructions linked below. There are instructions on page 5 for how to get the trip code to display (not automatic) and the code resets 2 hours after it happened so you have to check it soon after it trips. This is the only Eaton document that I could find on this series AFCI/GFCI breaker from Eaton and it still only shows 6 trip codes, not 7, but perhaps there is a newer document coming eventually. Don't know if this helps you or not, but thought I'd give it a shot. Yeah, for something like this I think they have to fix it especially if you are looking to buy the property. If only GFCI is required per latest NEC in that area then that is all they need to do, but it sounds like there is a good possibility of current leakage in the septic equipment causing a discharge. More than likely they will need to replace the equipment and not just the breaker, but that is just my guess...

They are "BR" series breakers and load center. I'm told that replacing the AFCI/GFCI with GFCI only worked. Which does not make obvious sense but, it worked and meets code. The BR AFCI and AFCI/GFCI units all flash codes, so there is at least some similarity to the document you sent earlier. The GFCI-only model just trips as far as I can tell, so you would not know if it was OC or GF that tripped it. I was not successful communicating the utility of reading the code to those working on the problem. They left the questionable AFCI/GFCI breaker behind, I may test it at some point. I will definitely get the code tables for the breakers I'll actually have. While trying to verify the some of their work I tripped some of the AFCI/GFCI breakers with my tester. They flashed a code upon reset. Five flashes, consistent with the table you posted earlier for a GF.

- David

RFGuy

DLB wrote
They are "BR" series breakers and load center. I'm told that replacing the AFCI/GFCI with GFCI only worked. Which does not make obvious sense but, it worked and meets code. The BR AFCI and AFCI/GFCI units all flash codes, so there is at least some similarity to the document you sent earlier. The GFCI-only model just trips as far as I can tell, so you would not know if it was OC or GF that tripped it. I was not successful communicating the utility of reading the code to those working on the problem. They left the questionable AFCI/GFCI breaker behind, I may test it at some point. I will definitely get the code tables for the breakers I'll actually have. While trying to verify the some of their work I tripped some of the AFCI/GFCI breakers with my tester. They flashed a code upon reset. Five flashes, consistent with the table you posted earlier for a GF.

David

David,

The previous document I shared here covered the Eaton BR breakers as well as the Eaton CH ones so the same table of LED codes should apply for each. As long as it holds (doesn't trip) on a GFCI breaker I would think you are reasonably safe. Presumably the GFCI breaker is good so there shouldn't be any ground fault condition. So, a GFCI has to trip on a leakage current as small as 6mA, but they are permitted several seconds to decide at that level. However, in the 20-30mA range of leakage current, GFCI's typically trip in 25ms or so. I mention this because I believe AFCI's are designed to trip in the 30mA range of leakage current (not lower). So, to a 1st order if the GFCI doesn't trip then most likely it isn't a leakage current causing it, but some other fault condition. I know AFCI's also have short circuit protection built in, I believe. This also applies to high inrush current which could trip an AFCI's short circuit protection. I don't know for sure, but presumably this is a "feature" that AFCI's have that GFCI's don't? So it is possible there is an arc-like event, e.g. high inrush current of motor turning on, etc. that is tripping the AFCI in the house you are interested in. Lastly, the AFCI is supposed to be able to detect a few different faults in house wiring. I thought the GFCI was similar in this regard and supposed to detect a neutral tied to ground at the load, but maybe there is some mis-wiring in the septic wiring causing the AFCI to detect it and trip where a GFCI is fine with it. If there are any other branch circuits sharing a neutral, I believe this will set off the AFCI as well. Below are a couple of articles that discuss troubleshooting an AFCI that trips when a GFCI doesn't that you might find useful. If you buy the house and move in, you could try installing the AFCI breaker and split up the loads tied to it in order to isolate where the problem is coming from (if you want it to be on an AFCI). Their electrician should have tried this.

AFCITripping.pdf

Sq-D-AFCI-Info.pdf

« Previous Page Next Page »