AFCI Theory of Operation
- Thread starter be_shane
- Start date
- Status
charlie
Senior Member
- Location
- Indianapolis
Re: AFCI Theory of Operation
Part of the circuitry works like a GFCI in that it measures the current going out to the load and compares it to the current coming back from the load and if it is over its set range, usually 30 to 50 mA, it will trip. The GFCI range is 4 to 6 mA.
The rest of the circuitry will look at the arc signature and if it matches whatever is set up for that type of AFCI, it will trip. This part does not work unless the current is over 75 amperes.
Part of the circuitry works like a GFCI in that it measures the current going out to the load and compares it to the current coming back from the load and if it is over its set range, usually 30 to 50 mA, it will trip. The GFCI range is 4 to 6 mA.
The rest of the circuitry will look at the arc signature and if it matches whatever is set up for that type of AFCI, it will trip. This part does not work unless the current is over 75 amperes.
al hildenbrand
Senior Member
- Location
- Minnesota
- Occupation
- Electrical Contractor, Electrical Consultant, Electrical Engineer
Re: AFCI Theory of Operation
be_shane,
I'll second that! The usual reference for this question is to link to Zlan.com.
What I personally find disturbing is the thunderous silence from the big four manufacturers of the AFCIs I can purchase to install in my client's equipment.
Sure, there is a lot of verbiage about how arcing starts fires, and a lot of advertising jingoism, but there is a paucity of facts about the detection circuitry and the algorithms employed in any one physical AFCI to determine friendly arc signature from foe.
In my mind, the detection circuitry of an AFCI is a signal processor. It will only be able to process signals above a minimum detection threshhold. At a bare minimum, I can improve the performance of the "package" (my branch circuit and the OCPD/AFCI) if I learn about signal degradation in various branch circuit structures.
Once a single AFCI receives the "signal", how does that specific processor operate on the information contained therein?
Broad brush information is also available at UL, Inc. that attests to the regulations the AFCI must satisfy. But, again, this is off to the side of the question. . .
Another source of links is at Arc Fault.org.
be_shane,
I'll second that! The usual reference for this question is to link to Zlan.com.
What I personally find disturbing is the thunderous silence from the big four manufacturers of the AFCIs I can purchase to install in my client's equipment.
Sure, there is a lot of verbiage about how arcing starts fires, and a lot of advertising jingoism, but there is a paucity of facts about the detection circuitry and the algorithms employed in any one physical AFCI to determine friendly arc signature from foe.
In my mind, the detection circuitry of an AFCI is a signal processor. It will only be able to process signals above a minimum detection threshhold. At a bare minimum, I can improve the performance of the "package" (my branch circuit and the OCPD/AFCI) if I learn about signal degradation in various branch circuit structures.
Once a single AFCI receives the "signal", how does that specific processor operate on the information contained therein?
Broad brush information is also available at UL, Inc. that attests to the regulations the AFCI must satisfy. But, again, this is off to the side of the question. . .
Another source of links is at Arc Fault.org.
Re: AFCI Theory of Operation
Posted by Charlie:
1) A short to ground or neutral? If so, what causes the arc? Is a short that is repeatedly made and broken required? And wouldn't a simple breaker detect this?
2) A bad connection or intermittent open that causes the inductance in the circuit to produce arcs through an inductive kick?
3) Both of the above?
Steve
Posted by Charlie:
OK, Charlies post makes me realize I don't know exactly what an an "arc fault" is. Every definition I see says the same thing: "a fault that causes arcing". More specifically, I would like to know what type of fault? I'll make it a multiple choice. Is an arc fault:
1) A short to ground or neutral? If so, what causes the arc? Is a short that is repeatedly made and broken required? And wouldn't a simple breaker detect this?
2) A bad connection or intermittent open that causes the inductance in the circuit to produce arcs through an inductive kick?
3) Both of the above?
Steve
Re: AFCI Theory of Operation
I have learned that there are two types of arcs. The Load arc that occurs any time a switch opens or closes. These are, if you will, "good" arcs. Then there are Parallel arcs. These are "bad" arcs and are caused by to power to ground or neutral partial shorts. This is the type of fault you would find if someone pinched romex too tight with a staple or set a heavy metal object on a cord, like a lamp cord or something. Arc Fault Breakers have some sort of differntiation circuitry that allows them to detect the difference between those two and trip the breaker if the fault current exceeds the limitations of the device. How that works is still somewhat of a mystery to me.
I have learned that there are two types of arcs. The Load arc that occurs any time a switch opens or closes. These are, if you will, "good" arcs. Then there are Parallel arcs. These are "bad" arcs and are caused by to power to ground or neutral partial shorts. This is the type of fault you would find if someone pinched romex too tight with a staple or set a heavy metal object on a cord, like a lamp cord or something. Arc Fault Breakers have some sort of differntiation circuitry that allows them to detect the difference between those two and trip the breaker if the fault current exceeds the limitations of the device. How that works is still somewhat of a mystery to me.
charlie
Senior Member
- Location
- Indianapolis
Re: AFCI Theory of Operation
1) A short to ground or neutral? Yes. If so, what causes the arc? A breakdown of the insulation through a damaged conductor. Is a short that is repeatedly made and broken required? No. And wouldn't a simple breaker detect this? Yes, most of the time.
2) A bad connection or intermittent open that causes the inductance in the circuit to produce arcs through an inductive kick? The type AFCIs that are on the market now will not detect a series arc.
1) A short to ground or neutral? Yes. If so, what causes the arc? A breakdown of the insulation through a damaged conductor. Is a short that is repeatedly made and broken required? No. And wouldn't a simple breaker detect this? Yes, most of the time.
2) A bad connection or intermittent open that causes the inductance in the circuit to produce arcs through an inductive kick? The type AFCIs that are on the market now will not detect a series arc.
Re: AFCI Theory of Operation
Thanks for the background info. Charlie. From the usual explinations I see (like in the NEC handbook), I'm not sure the author really understands the source of the arc.
As for a simple breaker detecting the arc, I can see a possible situation where an intermittent arc would not last long enough to trip the breaker. For example an arc that was on for a millisecond, and off for a millisecond, over and over. So I think I follow your answer of "Yes, most of the time".
But I still don't understand your answer to:
Steve
Thanks for the background info. Charlie. From the usual explinations I see (like in the NEC handbook), I'm not sure the author really understands the source of the arc.
As for a simple breaker detecting the arc, I can see a possible situation where an intermittent arc would not last long enough to trip the breaker. For example an arc that was on for a millisecond, and off for a millisecond, over and over. So I think I follow your answer of "Yes, most of the time".
But I still don't understand your answer to:
What process produces the arc if there isn't any switching? Can we agree that a "bolted" fault (where the line and neutral are bolted together) would not produce an arc?
Steve
charlie
Senior Member
- Location
- Indianapolis
Re: AFCI Theory of Operation
Is a short that is repeatedly made and broken required?
No, it is not required to make and break repeatedly for the AFCI to operate. It can be a normal dynamic arc.
What process produces the arc if there isn't any switching?
Damaged insulation from a staple, a table leg across an extension cord, or something similar.
Can we agree that a "bolted" fault(where the line and neutral are bolted together) would not produce an arc?
I agree. However, most faults are arcing faults, it is seldom that you encounter a bolted fault.
Is a short that is repeatedly made and broken required?
No, it is not required to make and break repeatedly for the AFCI to operate. It can be a normal dynamic arc.
What process produces the arc if there isn't any switching?
Damaged insulation from a staple, a table leg across an extension cord, or something similar.
Can we agree that a "bolted" fault(where the line and neutral are bolted together) would not produce an arc?
I agree. However, most faults are arcing faults, it is seldom that you encounter a bolted fault.
Re: AFCI Theory of Operation
This is interesting. Charlie says, "The type AFCIs that are on the market now will not detect a series arc." However any GFCI will detect and trip on a parallel arc. So what is the advantage of AFCIs if they cost more and don't do anything more than GFCIs?
Opinion: the extra money should be better spent on installing more outlets so as to reduce the use of extension cords.
~Peter
This is interesting. Charlie says, "The type AFCIs that are on the market now will not detect a series arc." However any GFCI will detect and trip on a parallel arc. So what is the advantage of AFCIs if they cost more and don't do anything more than GFCIs?
Opinion: the extra money should be better spent on installing more outlets so as to reduce the use of extension cords.
~Peter
charlie
Senior Member
- Location
- Indianapolis
Re: AFCI Theory of Operation
An AFCI will trip on a parallel arc. The GFI portion of the AFCI will only trip on a ground fault. A parallel arc is from the phase wire to the neutral wire and the GFI portion will only see it as load and not operate. For the AFCI to operate on the parallel arc, it has to reach the 75 ampere threshold.
An AFCI will trip on a parallel arc. The GFI portion of the AFCI will only trip on a ground fault. A parallel arc is from the phase wire to the neutral wire and the GFI portion will only see it as load and not operate. For the AFCI to operate on the parallel arc, it has to reach the 75 ampere threshold.
iggy2
Senior Member
- Location
- NEw England
Re: AFCI Theory of Operation
Arcing can start and stop spontaneously. With AC, the voltage waveform passes through zero, so the arc can cease on it's own. (With DC, the arc is much more difficult to extinguish since there are no zero crossings. You never want to use AC equipment that is intended for circuit interruption on a DC circuit unless approved for the purpose. AC fuses, for example, rely on zero crossings to extinguish the arcing when the elements fuse. Inside fuses, some arcing is good, since the rising arc voltage reduces current flow.)
But once an arc starts, it tends to be self-promulgating. Heated gas/plasma is conductive, and can result in more arcing. The carbon given off as a result of the hot plasma charring nearby products (conductor insulation, rug fabric, wood framing) is conductive, and can lead to more arcing. The burning/burnt insulation is no longer insulating, and can result in more arcing.
What was the question?....
Arcing can start and stop spontaneously. With AC, the voltage waveform passes through zero, so the arc can cease on it's own. (With DC, the arc is much more difficult to extinguish since there are no zero crossings. You never want to use AC equipment that is intended for circuit interruption on a DC circuit unless approved for the purpose. AC fuses, for example, rely on zero crossings to extinguish the arcing when the elements fuse. Inside fuses, some arcing is good, since the rising arc voltage reduces current flow.)
But once an arc starts, it tends to be self-promulgating. Heated gas/plasma is conductive, and can result in more arcing. The carbon given off as a result of the hot plasma charring nearby products (conductor insulation, rug fabric, wood framing) is conductive, and can lead to more arcing. The burning/burnt insulation is no longer insulating, and can result in more arcing.
What was the question?....
Re: AFCI Theory of Operation
Charlie,
Here's the catch:
"For the AFCI to operate on the parallel arc, it has to reach the 75 ampere threshold." Since the AFCI and the GFCI and even an ordinary breaker all will trip at 75 anps [we're talking 15 and 20 amp circuits here], then the AFCI doesn't offer any advantage.
~Peter
Charlie,
Here's the catch:
"For the AFCI to operate on the parallel arc, it has to reach the 75 ampere threshold." Since the AFCI and the GFCI and even an ordinary breaker all will trip at 75 anps [we're talking 15 and 20 amp circuits here], then the AFCI doesn't offer any advantage.
~Peter
hurk27
Senior Member
- Location
- Portage, Indiana NEC: 2008
Re: AFCI Theory of Operation
An AFCI will trip on a parallel arc that has an arc that only averages 20 amps RMS but reaches a peek to peek of 75 amps, a regular breaker would only see the RMS value and never trip. A 20 amp RMS would have a peek to peek value of 169.7 amps.
[ December 24, 2004, 12:56 AM: Message edited by: hurk27 ]
An AFCI will trip on a parallel arc that has an arc that only averages 20 amps RMS but reaches a peek to peek of 75 amps, a regular breaker would only see the RMS value and never trip.
A 20 amp RMS would have a peek to peek value of 169.7 amps.[ December 24, 2004, 12:56 AM: Message edited by: hurk27 ]
charlie
Senior Member
- Location
- Indianapolis
Re: AFCI Theory of Operation
Peter, I have been saying that for a couple of years.
Wayne, I believe you are talking about the combination type because the information that I have gotten has never indicated it was the peak to peak values.
It is my opinion that the Code Making Panels were sold a bill of goods about the operation of AFCIs. They do not do much better than a standard class B GFCI. However, it would look foolish to remove the requirements for the AFCIs only to reinsert them after the combination type were produced. This also gives the manufacturers experience with the AFCI technology so the combination type will work better when they finally hit the market.
Peter, I have been saying that for a couple of years.
Wayne, I believe you are talking about the combination type because the information that I have gotten has never indicated it was the peak to peak values.
It is my opinion that the Code Making Panels were sold a bill of goods about the operation of AFCIs. They do not do much better than a standard class B GFCI. However, it would look foolish to remove the requirements for the AFCIs only to reinsert them after the combination type were produced. This also gives the manufacturers experience with the AFCI technology so the combination type will work better when they finally hit the market.
Re: AFCI Theory of Operation
You guys are saying the AFCI threshold is 75 Amps? That is arc welder current! Don't you mean 75 milliAmps? Or am I missing something?
As for the theory of operation, I would speculate that the AFCI contains circuitry which detects the RF currents associated with the arc.
You guys are saying the AFCI threshold is 75 Amps? That is arc welder current! Don't you mean 75 milliAmps? Or am I missing something?
As for the theory of operation, I would speculate that the AFCI contains circuitry which detects the RF currents associated with the arc.
- Location
- Massachusetts
Re: AFCI Theory of Operation
Yes.
- Location
- Bremerton, Washington
Re: AFCI Theory of Operation
75 amperes was determined to be the most common fault current in residential applications. Setting the threshold lower would result in nusiance trips. Keep in mind a 20 CB may take 220 amps to trip.
75 amperes was determined to be the most common fault current in residential applications. Setting the threshold lower would result in nusiance trips. Keep in mind a 20 CB may take 220 amps to trip.
pierre
Senior Member
- Location
- Westchester County, New York
Re: AFCI Theory of Operation
As current flows through a conductor, there is a sinusoidal wave shape which is typical for current that is "normal".
As an "arc" in the current path is caused, the signature of that wave shape changes (the arc's signature). The different type of wave shapes for the different type of arcs have been tested and stored in computer programs.
Now these Arc Fault detectors utilize this data and the microprocessor in each unit will determine if the particular arc it senses is a "good Arc" or "bad Arc" (based on the signature produced by the arc).
The current type of Arc Fault Circuit Breakers only detect the arc in the wiring to the outlet device. There will be available soon a "Combination" type detector which detects the wiring and cord connected appliance, lamp, etc...
As technology improves, so will this breaker.
Merry Xmas.
Pierre
As current flows through a conductor, there is a sinusoidal wave shape which is typical for current that is "normal".
As an "arc" in the current path is caused, the signature of that wave shape changes (the arc's signature). The different type of wave shapes for the different type of arcs have been tested and stored in computer programs.
Now these Arc Fault detectors utilize this data and the microprocessor in each unit will determine if the particular arc it senses is a "good Arc" or "bad Arc" (based on the signature produced by the arc).
The current type of Arc Fault Circuit Breakers only detect the arc in the wiring to the outlet device. There will be available soon a "Combination" type detector which detects the wiring and cord connected appliance, lamp, etc...
As technology improves, so will this breaker.
Merry Xmas.
Pierre
al hildenbrand
Senior Member
- Location
- Minnesota
- Occupation
- Electrical Contractor, Electrical Consultant, Electrical Engineer
Re: AFCI Theory of Operation
Pierre,
Are you aware of any studies, research or detailed product technical info that explains what GE, SqD, CH & Siemens have done?
I mean, each of these manufacturers has a solution that reduces a world of arc signatures and other high speed, high energy transients in to a binary state. . .Good arc do nothing. . .Bad arc operate disconnect.
How have they done that?
edited for punctuation
[ December 24, 2004, 05:07 PM: Message edited by: al hildenbrand ]
Pierre,
Are you aware of any studies, research or detailed product technical info that explains what GE, SqD, CH & Siemens have done?
I mean, each of these manufacturers has a solution that reduces a world of arc signatures and other high speed, high energy transients in to a binary state. . .Good arc do nothing. . .Bad arc operate disconnect.
How have they done that?
edited for punctuation
[ December 24, 2004, 05:07 PM: Message edited by: al hildenbrand ]
- Status