AFCI (yes again)

[ActionDave]

..... Opinions being what they are I never really expected everyone to agree on this topic but i did hope for one or two open minds on the subject.:angel:

Open or sympathetic? Seems like you are trying to direct traffic down a one way street.

I don't doubt that you can put some wires and stuff under a glass bowl, hook it up to your fancy gizmo and make it work a la UL Nine Five Six Two Nine Double Seven stroke A in front of friends and family. Big deal. Take your act on the road and and see how it plays in front of a live audience on your own instead of forcing the audience to buy tickets.

If safety was the primary concern we would be miles ahead by ditching AFCI's and moving to more ground fault protection like other developed nations.

 

[MasterTheNEC]

Open or sympathetic? Seems like you are trying to direct traffic down a one way street.

I don't doubt that you can put some wires and stuff under a glass bowl, hook it up to your fancy gizmo and make it work a la UL Nine Five Six Two Nine Double Seven stroke A in front of friends and family. Big deal. Take your act on the road and and see how it plays in front of a live audience on your own instead of forcing the audience to buy tickets.

If safety was the primary concern we would be miles ahead by ditching AFCI's and moving to more ground fault protection like other developed nations.

Glass Bowl.......hmmm....anyway opinions are like ********, everyones entitled to have one.:jawdrop:

 

[mbrooke]

I am very familiar with those documents...

There design was a lot more user friendly, and it did open the door for the ability to allow very complex logic and self diagnostic abilities through its interface. Its a shame it was not refined upon.

I get the manufactures are part of NEMA, but I don't believe that the issue. The part that concerns me is CMP allowing AFCI requirements into the code.

Look I am not blind to some potential issues...

Its good that you have no had problems, but others are having many. The technology is imperfect for the same reason as others have stated earlier: it is to premature. The used to detect arcs is logic is to primitive.

I have no doubt that in testing AFCIs perform as indeed. I know they do so since no UL label would exist on them. But the real world is not a laboratory, and as far as I know there is no testing outside of manufacturing facilities that actively make sure AFCI will not nuisance trip.


Sure the AFCI gets a UL label because it passes tests, but what verification sticker exists to say these devices will not trip on a UL listed plasma TV or vacuum cleaner? None.

In terms of the GFI (GFP), it is assumed that because it has a protective theory of it's intended use, to protect the volitive circuitry in the devices themselves that it does add to the overall fire protective safety of the device.

Explain, that is simply so untrue. No offence but that discredits a lot of what you have said. How does differential logic protect the circuitry within the device?



Deferential logic plays a profound role in fire safety within a branch circuit. Literally half of an AFCIs protective features are obtained via 30ma protection. A hot to ground fault, a neutral to ground fault and crossed circuits; from wiring errors, compromised insulation, over driven staples, ect produce a current differential. It does not even matter if this current difference has an arcing signature or not. Its a fire hazard. If current is leaking to the equipment grounds then a hazard along with a code violation is taking place. I can not even think of one instance where 30 milliamps and over of current leakage is allowed or normal on a branch circuit. Consider the fact that about 90% faults are hot to ground, thus the brunt of fire protection is done via differential. Same for the majority of NEC wring errors like crossed circuits: detected by GFP.

In fact, the 30ma logic allows for passing of about half the UL tests and I have heard that is actually the reason why AFCI makers put it in there to start with. As soon as the test places current on the EGC the device trips, and no arc logic is necessary to accomplish that. This guarantees a 100% chance of passing without needing to design logic specifically for that part of the test. Heck if one gave a GFCI breaker to be run through those tests it would pass about half of them.



Arc logic is only placed in AFCIs where GFP reaches its limitations. Those are parallel hot to neutral faults and low level arcing in series with the load. Of note, if one set the instantaneous magnetic trip low enough on a breaker, you could probably pass a series arc test. But in case, the goal of arc logic is to catch these.


It is true that some manufactures like GE have taken GFP out of AFCIs, but that is only to facilitate their use in MWBC and reduce customer complaints of tripping caused by electrician wiring errors. Which I do find ironic CMP is not concerned about those, ie they aren't requiring 30ma protection for AFCIs. An AFCI without GFP will let a standing neutral to ground fault continue as well as crossed circuits which are hazards in themselves. For starters, the inductive heating in ferromagnetic conduits (as well as any lose conduit connections that may spark) caused by this objectionable current is a fire hazard.

Well i will buy into the fact and even remove the assumed portion; yet UL is a paid association now who is charged with evaluating products for their intended use and currently under UL 1699 the GFI component is not required because it plays no role in the intended function of the product. While some manufacturers did (and some still do possibly) rely on this feature to pass the complete testing procedure for UL 1699 compliance others have obviously developed a method that passes without the GFI (GFP) function. What you are presenting is that experts believe that the GFI (GFP) function is a critical component that must be included and that when manufacturers leave that out they are some how immoral or pandering to the cost of production so I need to address that.

You are correct, UL does not require it and manufactures do indeed use GFP as a way to pass some tests worry free. I get that.

In purely detecting arcs GFP is not essential, however I can think of a slew of code violations some even fire hazards that arc detection will not pick up on where GFP will catch it immediately.

The issue is not that manufactures are taking out something that's not required but rather the fact the CMP is not mandating it.

The NEC clearly states that an electrical system shall be free of defects and errors, but how do we know that? We don't unless we test the installation or even better have an OCPD that picks up on it. How many North American electricians test circuits? None. How many AFCIs can catch them? Only the ones that have GFP.


And to say such errors are to rare to justify GFP is an understatement. Since the inception of AFCIs countless wiring errors have been detected. In fact many AFCI reps will tell you close to 90% of AFCI nuisance tripping is the result GFP catching ground faults/wiring errors.

That is an issue I need to address.


Surely you don't believe the CMP should ignore this issue?

Firstly I represent a manufacturer (not of AFCI's) and our products are produced in accordance with UL standards. We do not shoot to exceed the standards, we shoot to meet the standards...

I do not blame manufactures, I blame the CMP. In the end it is them that have all the power.

AFCI technology speaking...on the most basic level of ARC detection their are specific characteristics that are replicable, modeled and perfectly acceptable to be used as foundation points in arc detection. The research that takes place are indeed now days done on finished product. However, continued research is constantly being done by the manufacturers to enhance the AFCI Technology because the manufacturing business is also a game of who gets it to market first and who enhances it to over shadow the competition, trust me (or not, thats your choice) every day manufacturers are trying to enhance and one-up the competition if only to say our product beats your product and NEMA, the collective body, can't restrict that competition.

I trust you. But that isn't stopping nuisance tripping from UL listed appliances.

When looking at the merits of a "premature" ...

And that's a half truth escape.


When one introduces an unreliable technology it back fires. There are and have been wonderful technologies with wonderful promises, but relying on them full scale is/was not economical even dangerous. Do you want a nuclear reactors built without all the safety bugs ironed out because said reactor will save some money on coal? I think not.


Same goes for AFCIs. When they are tripping on UL listed devices they get taken out. They force electricians to loose faith in the CMP, it forces an underground market of violating code.



I can see you making the argument "well, that is their decision not ours" And that may be true, but that simple defense does not changing a complex reality.

I believe the problem originally was that AFCI's were touted as the Holy Grail and when random issues came to light (even if installer error) it tainted that statement made by various manufacturers or invested advocates but it does not diminish the technology and advancements in terms of arc detection at least in my less than intelligent mind.

AFCI technology when done right is a layer of safety, but it needs improving.


But ultimately, the biggest contributor of electrical fires is not arcs, its glowing connections. Why is it that in 2015 no device or code mandate exists for every connection to have a device capable of shutting down the circuit when a glowing connection arises?

I personally believe that the largest hurdle...

And that is why CMP should not have mandated AFCIs until this was fixed. At this point a customer is nothing more than a research lab. AFCI will be swapped until the right on is found or they will all be failed, removed, and replaced with a regular breaker. How is that different than R&D?

Again, if the logic was amped up, it would get easier to differentiate between everything.

 

[kwired]

And that is why CMP should not have mandated AFCIs until this was fixed. At this point a customer is nothing more than a research lab. AFCI will be swapped until the right on is found or they will all be failed, removed, and replaced with a regular breaker. How is that different than R&D?

Not only is the customer a research lab, they are paying for service calls, and a majority of those calls are not even producing data for the people that are trying to make improvements to the product. Not that well of a controlled research environment, but IMO there was too much $$ on the line to not push these to be in code sooner instead of later, and that is all on the manufacturers and whatever lobbying they did to convince the CMP it was needed now instead of later.

Since they did that why should we ever trust that they did the right thing later on?

 

[mbrooke]

Not only is the customer a research lab, they are paying for service calls, and a majority of those calls are not even producing data for the people that are trying to make improvements to the product. Not that well of a controlled research environment, but IMO there was too much $$ on the line to not push these to be in code sooner instead of later, and that is all on the manufacturers and whatever lobbying they did to convince the CMP it was needed now instead of later.

Since they did that why should we ever trust that they did the right thing later on?

Tell me about :rant: On top of that the electrician isn't reimbursed for all the call backs with the customer thinking the electrician screwed up rather than the device.


But hey, every home built now built spends on average a literal 18 times more on OCPDs. Same for service upgrades and circuit extensions. No economic boom could do that, not even if new home construction increased 8 times the present day amount. Talk about profit gain.

 

[mbrooke]

Just to correct an error in post #103, where I said

if one set the instantaneous magnetic trip low enough on a breaker, you could probably pass a series arc test

I meant to type in parallel arc test.


A parallel arc producing 75amps would trip a breaker with a magnetic trip of only 70 amps.

 

[romex jockey]

Gentlemen, may i take the opportunity to post this jem>

COMBINATION AFCIs:
WHAT THEY WILL AND WILL NOT DO

thx

~RJ~

 

[edlee]

...............At this point a customer is nothing more than a research lab. AFCI will be swapped until the right one is found or they will all be failed, removed, and replaced with a regular breaker. How is that different than R&D?

.............

That is exactly the truth. It is apparently how the industry works. The manufacturers can't do extensive real-world testing , the NFPA doesn't care what anything costs as long as it adds to fire safety, and UL makes it's money from every approval sticker it puts on something.

So they pressure the code-making authorities to force us to use these things and figure that over time and many iterations of the technology they will eventually get them reliably useful.

In the meantime we all have to put up with crazy false trips, call-backs and hard decisions about whether to do code-compliant installations. I've been in the trade long enough to remember some of the same problems with GFCI's. But at least with GFI's the Code moved at a slower pace in terms of how extensively they were required in homes.

 

[Dennis Alwon]

Gentlemen, may i take the opportunity to post this jem>

COMBINATION AFCIs:
WHAT THEY WILL AND WILL NOT DO

thx

~RJ~

I found a few errors in the article - stopped reading after the experiment he did. The afci will not trip on arc if it doesn't see a load of ~4 amps. If the author created an arc with a small load then it was an invalid experiment. I have seen afci trip on arc ( or at least that is what I surmised)- I use GE and they do not have any gfp in them so it must have been the afci that tripped.

 

[mbrooke]

That is exactly the truth. It is apparently how the industry works. The manufacturers can't do extensive real-world testing , the NFPA doesn't care what anything costs as long as it adds to fire safety, and UL makes it's money from every approval sticker it puts on something.

So they pressure the code-making authorities to force us to use these things and figure that over time and many iterations of the technology they will eventually get them reliably useful.

In the meantime we all have to put up with crazy false trips, call-backs and hard decisions about whether to do code-compliant installations. I've been in the trade long enough to remember some of the same problems with GFCI's. But at least with GFI's the Code moved at a slower pace in terms of how extensively they were required in homes.

And why should they do extensive testing? Code will always require them. And when all brands are mediocre, not only does it save money, but no one manufacturer will look bad.


GFCIs have more valid bases, were moved in slower as you said, and nuisance tripping was a LOT less of an issue.

 

[don_resqcapt19]

I found a few errors in the article - stopped reading after the experiment he did. The afci will not trip on arc if it doesn't see a load of ~4 amps. If the author created an arc with a small load then it was an invalid experiment. I have seen afci trip on arc ( or at least that is what I surmised)- I use GE and they do not have any gfp in them so it must have been the afci that tripped.

I would expect that Dr. Engle fully understands the limitations of AFCIs...he was one of the chief design engineers for Eaton's AFCI.

 

[kwired]

FIFY

And why should they do extensive testing? Code will always require them. And when all brands are mediocre, not only does it save money, but all manufacturers will equally look bad.


GFCIs have more valid bases, were moved in slower as you said, and nuisance tripping was a LOT less of an issue.

 

[MasterTheNEC]

Explain, that is simply so untrue. No offence but that discredits a lot of what you have said. How does differential logic protect the circuitry within the device

It discredits nothing....only the notion that limits your perspective of what the intent of UL 1699 is. The device is designed to detect arc conditions as defined in the UL 1699 Standard and nothing more. The GFP function was added to pass a specific portion of the test and was indeed in those devices. If the manufactures (who have removed GFP) have determined a way to meet the test then they can openly remove the GFP component....they choose a different method to protect their internal components.

You talk about a lot of technical substantiation but lack the posting of it. How about posting a technical explanation of what you BELIEVE the AFCI Device is supposed to do (in your mind) versus what UL 1699 says...and then cross reference the two.

No offense taken my friend....we just have a different of opinions is all.

 

[mbrooke]

It discredits nothing....only the notion that limits your perspective of what the intent of UL 1699 is. The device is designed to detect arc conditions as defined in the UL 1699 Standard and nothing more. The GFP function was added to pass a specific portion of the test and was indeed in those devices. If the manufactures (who have removed GFP) have determined a way to meet the test then they can openly remove the GFP component....they choose a different method to protect their internal components.

You talk about a lot of technical substantiation but lack the posting of it. How about posting a technical explanation of what you BELIEVE the AFCI Device is supposed to do (in your mind) versus what UL 1699 says...and then cross reference the two.

No offense taken my friend....we just have a different of opinions is all.

Answer the question first before wanting me to move forward. How does GFP protect the circuitry of an AFCI? By that you mean the internal electronics of the breaker itself?

And I am not grinding an axe with UL. I don't think the standards are an issue. I think the issue is AFCI nuisance tripping on so many devices. Again, that has nothing to do with UL testing. And again, UL testing doesn't test against nuisance tripping.

I know what the intention is of an AFCI: to trip on a dangerous arc fault. I want it to do that according to the UL test methods IN ADDITON to not nuisance tripping.

 

[MasterTheNEC]

Answer the question first before wanting me to move forward. How does GFP protect the circuitry of an AFCI? By that you mean the internal electronics of the breaker itself?

And I am not grinding an axe with UL. I don't think the standards are an issue. I think the issue is AFCI nuisance tripping on so many devices. Again, that has nothing to do with UL testing. And again, UL testing doesn't test against nuisance tripping.

I know what the intention is of an AFCI: to trip on a dangerous arc fault. I want it to do that according to the UL test methods IN ADDITON to not nuisance tripping.

I had a long discussion today with an engineer who i would consider is expert on the AFCI subject and he explained a lot of things to me that I wish I could explain here in terms of debunking Mr. Engle and his claims. One of which that he attended a lot of the original meetings with the CPSC which is incorrect....but thats a different story and I wont get into that issue as it is none of my business.

In terms of the GFP, it did add a level of protection to internal circuits no different that it does in the GFCI device as well with the raised element of personal protection as well. However, at the GFP level in an AFCI it only served to provide some limited protection to the circuitry but the main function ( in some of the manufacturers) was to aid in detecting arcs in nonmetallic sheathed cable where the bare EGC was present...it helps in that area and still does in a few manufacturers that still have the GFP function.

I think the issue is this, you want a HOLY GRAIL and you just will not get one. You can't test the algorithms in a basement, you can't replicate the arc condition by doing a test in your basement and touching two conductors in an effort to create an arc that will some how be detected as a BAD arc...it is just not possible and this was confirmed today again with the engineer.

Also...DO NOT try to twist a statement that may get misstated in a debate...the issue is still what it is. The GFP is inherent to offer some level of protection to the internal components of the device but it was not the actual intent of it being added to the device , the intent was to pass the original UL 1699 cotton ball test (ignite test) and due to the nature of the test the GFP helped that test pass and some manufacturers still have that component in their products ( like Square D).

But what I found interesting is that you wish to have a perfect product and only have that product go to market when you or someone else perceive it meets all possible arc conditions and that is just not possible....as stated today by the engineers who make the devices, it is not possible as the detection of an arc is a complicated thing and while microprocessors can detect cross shoulders and cross over points in the arc as it works through the waveform, nothing will always be absolute.

Now I am sure that you will say if it can't be absolute then it should not go in the code.....I just happen to disagree.

I can tell you that their are a large number of IEEE professionals who disagree with Mr. Engles statements in his now outdated whitepaper. The document is almost like a cult classic because the statements are in many cases fictional and can't be substantiated.

I can tell you this, the cost of development and continued development of AFCI's by only one company ( I wont mention) is well over 8 figures. Thats at the least $10,000,000.00 on a single product. Thats testing, development and so on. Yes, don't feel sorry for them because I am sure they will make that back but then again lets not forget who mandated the technology...it was not the manufacturers. The manufacturers were just smart enough to develop a product that meets the needs but those needs continue to change and so does the technology.

Look I am not going to sit here at debate AFCI's with you based on a selected few individuals who claim to have issues with AFCI's.


You also stated "GFCIs have more valid bases, were moved in slower as you said, and nuisance tripping was a LOT less of an issue."...can you substantiate that statement. The GFCI was introduced in the 1960's and it was met with alot of resistance....how on earth can you make that statement......in fact you make alot of statements that ask ME to substantiate yet I have not seen a single one from your perspective.

So....tell me about the testing you have witnessed.....I will GLADLY tell you about mine, I will gladly tell you in depth the things that AFCI's look for but I will also tell you that their are limitations that still need to be overcome...one being a frequency issue that is still a lingering concern but not wide spread enough to call for mass removal of these potentially life and properly saving products.

Also for those that call for AFCI Branch/Feeder devices that some (including the author of that whitepaper) offer a BETTER solution of protection from false tripping issues then try again. Most all of the manufacturers abandoned the advancement testing of AFCI B/F devices when they (Manufacturers) moved to the combination devices....why....because their was no longer a need for testing the B/F models anymore.

Now with the 2014 NEC they (AFCI B/F) have resurfaced but as I am told today....they are not being produced and have not been tested for the same wave characteristics and frequency issues as the combination devices have since they stopped worrying about them. So bringing them back would result in an increased level of false trip issues in the view of the manufacturer.

Anyway...look....I am not sure if you are just looking to get a rise out of me or really do care about the conversation. The fact is unless you can substantiate tests other than what the manufacturers do and comply with then the argument falls on deaf ears. I clearly know how these devices look to function, what characteristics that are looking for, what shoulders and half-cycle key points that are building into an algorithm but again each manufacturer is different in how they approach it and I am quite frankly tired of arguing over it...call me wrong, call me crazy...call me what ever you like but I speak directly from the mouth of the manufacturer and so thats all I can say on that.

For those that say the AFCI's are worthless.....So Be It......for those who believe....So Be It.....at the end of the day..accept it unless you can change it. Also in terms of UL 1699...you could indeed make it a UL issue in terms of the STP and changing the standard to require GFP in the standard if you can have them remove the statement about "glowing connections" and convince the STP that GFP is vital for the AFCI function..until then you will never force the manufacturers to do anything other than comply with the current standard.

 

[don_resqcapt19]

I had a long discussion today with an engineer who i would consider is expert on the AFCI subject and he explained a lot of things to me that I wish I could explain here in terms of debunking Mr. Engle and his claims. ...

Dr. Engle was one of the lead AFCI development engineers for Eaton. I will take his statements over an unnamed engineer any day.

 

[MasterTheNEC]

Dr. Engle was one of the lead AFCI development engineers for Eaton. I will take his statements over an unnamed engineer any day.

Well.....I won't go into details as it serves no purpose but you can believe whom you wish. I can tell you the belief in the IEEE community is not as agreeable with his statements.

Also let me clarify......I would NEVER take anything away from Mr. Engle...he is without a doubt a pioneer in the AFCI technology industry. The only problem is that other leading (and it does not serve them well to disclose them, they did not write the paper) AFCI individuals who have worked to advance the technology disagree with much of the statements and assumptions in the 2012 IEEE white paper...that is all I am saying.

Is Mr. Engle way smarter than me...YEP......never once wish to discredit him at all. Just that his opinions are only that in terms of his white papers theories and since that time many in the field who do this for a living have stated that his assumptions are incorrect, which is all I am saying.

At the end of the day......I can't convince anyone of anything much less individuals on this forum and I don't want to ( yes, iwire I don't want too...lol)...just felt it was a good debate and at the end of the day.....believe in what you want......yes I drank the kool-aid so I believe they provide a value worth paying for.

 

[readydave8]

I don't like AFCI's.

Then again, I didn't like GFCI's in the late '70's-early 80's, now I do.

 

[George Stolz]

Well.....I won't go into details as it serves no purpose but you can believe whom you wish.

Well, that's kind of a problem because 95% of this thread consists of you hinting and winking and generally avoiding any specifics while stating you'd love to go into specifics. You'd love to discuss details of a technical nature as to why AFCI is the best thing since the first coming, but you don't want to discuss details. You'd love to name-drop whoever toured you through the factory, but you don't want to. You seem to have a conflict here which I am remedying now.

Also let me clarify......I would NEVER take anything away from Mr. Engle...

Again, with your internal struggle. You don't want to run him down, yet you believe everything he has claimed is hog-wash. But, alas, you don't want to go into details as to why, aside from some unnamed source "in the know" told you, and that statement is supposed to carry weight.

At the end of the day......I can't convince anyone of anything much less individuals on this forum and I don't want to ( yes, iwire I don't want too...lol)...just felt it was a good debate and at the end of the day.....believe in what you want......yes I drank the kool-aid so I believe they provide a value worth paying for.

That is one point you've made abundantly clear. Over, and over, and over, and over...

You've made your case, the opposing argument has made it's case, and this forum will likely never settle the issue, if there even is one to settle. This is a vapid thread that has run it's course. In closing, I wish to thank all participants for keeping it civil.

-George

 

[mbrooke]

AFCI Response

AFCI Response

The other thread is broken...

...in terms of debunking Mr. Engle and his claims. One of which that he attended a lot of the original meetings with the CPSC which is incorrect....

I would be happy to hear more, you can PM me if you feel comfortable.

In terms of the GFP, it did add a level of protection to internal circuits no different that it does in the GFCI device as well with the raised element of personal protection as well. However, at the GFP level in an AFCI it only served to provide some limited protection to the circuitry...

Can you please explain further how a GFP circuit within the breaker protects the AFCI circuit within the breaker?

I think the issue is this, you want a HOLY GRAIL and you just will not get one. You can't test the algorithms in a basement, you can't replicate the arc condition by doing a test in your basement and touching two conductors in an effort to create an arc that will some how be detected as a BAD arc...it is just not possible and this was confirmed today again with the engineer.

And that's just it, if nuisance tripping can not be solved then they should not be required. GFCI are already a HOLY GRAIL as well as standard thermal magnetic breakers: nuisance tripping has been resolved.

I can agree here, a home workshop test is not documented nor can claimed to have the same if any level of control. In other words it holds no merit in the sake of an argument because the claim can not be proven. You are correct here.

But, FWIW, it helps more in my own opinion and in my own world to reach a conclusion about them.

Also...DO NOT try to twist a statement that may get misstated in a debate...the issue is still what it is. The GFP is inherent to offer some level of protection to the internal components of the device but it was not the actual intent of it being added to the device , the intent was to pass the original UL 1699 cotton ball test (ignite test) and due to the nature of the test the GFP helped that test pass and some manufacturers still have that component in their products ( like Square D).

Please in detail elaborate how GFP offers protection within the device itself.

GFP may help pass a cotton ball test and any arc the is phase to ground, but again: GFP catches a boat load of wiring errors in addition to a phase to ground fault. Again, I know its not mandated, but for the sake of wring errors CMP should mandate it. Without a doubt GFP is another level of fire protection.

FWIW RCDs are profusely used in Europe where phase to ground voltages are much higher. They do catch a lot of issues when they happen to crop up.

But what I found interesting is that you wish to have a perfect product and only have that product go to market when you or someone else perceive it meets all possible arc conditions and that is just not possible....as stated today by the engineers who make the devices, it is not possible as the detection of an arc is a complicated thing and while microprocessors can detect cross shoulders and cross over points in the arc as it works through the waveform, nothing will always be absolute.

Im not talking about every arc condition, rather nuisance tripping. But if that requires an AFCI to pre-know 99.9999% of every single current ripple on earth not to nuisance trip... well that's what it takes then.

And he is correct, arc detection is complicated... so complicated that to get something reliable far more computing power is needed than what current AFCIs have. Looking for where current spikes and drops as an arc restrikes/extinguishes in a sine wave is not enough. Logic must look for further discrimination of you will nuisance trip.

An example is adding a logic circuit that looks for current draw patterns. Ie if current is going up and down in a sine wave with multiples of say 4, 6, 8 12, ect ect (or how ever many coils in the rotor) that would indicate a universal motor. Such motors have this finger print since the carbon brushes bridge (short circuit) one or two commuter contacts before going on to a new as the leave the prior one while the rotor is spinning. When this circuit senses this, it can restrain (override) the arc logic circuit that is sensing the brushes arcing. Thus a nuisance trip on a vacuum cleaner is avoided.

This is just one example of arc discrimination. An arc can further be discriminated by the pitch, angle, RMS and peak restrike values, patterns, linearity, ect ect then comparing it in relation to something... but all that takes logic, research and modeling. Even if the R&M are perfected, you still need logic.

Now I am sure that you will say if it can't be absolute then it should not go in the code.....I just happen to disagree...

Im sure people are speaking out against it, denying his claims ect. And I know he is not perfect, but neither are his opponents.


But IMO he was on to something with his central processor. Maybe he knew AFCI needed more computing power, and perhaps that the logic would need to be changed over time as our knowledge/experience of arc detection changed?

What do you think about this?

I can tell you this, the cost of development and continued development of AFCI's by only one company...

Oh, they will make that money back over and over and over. Investment is wise.

But who mandated them? Who got them into the code? This is the real issue IMO, as with any mandatory law its own fall out must be understood before it is passed.

Look I am not going to sit here at debate AFCI's with you based on a selected few individuals who claim to have issues with AFCI's.

But those individuals are not opinionated posters on You Tube or some News Forum. They are the people first hand installing them in addition to getting call backs.

You also stated "GFCIs have more valid bases, were moved in slower as you said, and nuisance tripping was a LOT less of an issue."...can you substantiate that statement. The GFCI was introduced in the 1960's and it was met with alot of resistance....how on earth can you make that statement......in fact you make alot of statements that ask ME to substantiate yet I have not seen a single one from your perspective.

Ok, I will agree, GFCI were not met with open arms by everyone. I can however prove GFCIs were introduced over time with NEC archives.

As for substantiations you, (for example) have not substantiated how GFP protects the electronics of an AFCI even though Ive asked multiple times. How does it? Im clueless. You know something I do not?

So....tell me about the testing you have witnessed.....

I wouldn't feel comfortable going into detail in public on all of it, mostly what was done at verification lab.

But, you can defiantly tell me what you feel comfortable in sharing.

Also for those that call for AFCI Branch/Feeder devices that some (including the author of that whitepaper) offer a BETTER solution of protection from false tripping issues then try again...

I offered one with the vacuum cleaner scenario. And it is one of many that utility AFCI protection uses. Pattern analysis outside of what an arc might look like are key to mitigating false tripping. #2, all the keys to the problem take logic, far more than a resi breaker has.

Now with the 2014 NEC they (AFCI B/F) have resurfaced but as I am told today....they are not being produced and have not been tested for the same wave characteristics and frequency issues as the combination devices have since they stopped worrying about them. So bringing them back would result in an increased level of false trip issues in the view of the manufacturer.

And that's just it? WHY? WHY? WHY? Why do customers have to put up with this? Why is joe public R&D? Why tarnish reputation? Why open the door to DIY modification?

Anyway...look....

I am not trying to get a rise out of you. I care about this. Something that adds hundreds of dollars and more to a new home or panel change that is giving customers headaches demands honest answers. It demands serious reevaluation.

Also, saying it again, analyzing an arc itself is not enough. You need logic to accurately discern that data accurately.

For those that say the AFCI's are worthless.....

I have a better solution: withhold AFCI requirement until they can have thorough thoroughly R&D.


GFP is not necessary to an arc itself, but to fire protection.


I would have done what Europe did. I would have required GFP on most 120 volt and latter 240 volt circuits.


This would give protection to most fires in addition to catch wiring errors by poor electricians.


I would then get the ball rolling on Glowing connection research, in addition to AFCI research. I would ONLY mandate those once all bugs were ironed out.


That would have been a real game plan.