AFCI issues

Status
Not open for further replies.
sparky_magoo

sparky_magoo

Senior Member
Location
Reno
I went on a service call today. It was a new house. HO complains two bedrooms quit working. I found, at the breaker, the hot wire was touching the screw on the AFCI breaker. It was not "landed" under the screw. The insulation, on the hot wire, was burnt about three inches past the AFCI breaker. The wire melted in two pieces. The AFCI never tripped. Obviously, the loose connection was arcing at the AFCI breaker. The AFCI never tripped. I cut the wire back & landed it on a new AFCI breaker. This is the second time I have repaired fire damage, from an arcing fault, on an AFCI ckt., which never tripped.

My question is, what good are AFCI's?
 
Good at making money for the company that makes them. I pay about $37. for a single pole.
 
I'm primarily a semiconductor engineer. It's obvious, to me at least, that a microcontroller undoubtedly with some added circuitry, AtoD's and stuff, is used to run through a program that senses voltage/current irregularities. Does anyone know where the program used is published or defined?

Regards,
 
I agree, its all about the money! These large manufacturers such as Square D, GE, Siemens, Cutler Hammer "contribute" so much money into the code making panels, that they can dictate what goes into the codebook so that it is revenue generating machine for their companies by requiring homeowners to have this stuff installed in their homes. I know many of us grumbled when GFCI's came out and were mandated, but I for one think GFCI's are a great idea and certainly have a place, although maybe not to the extreme that the NEC has been requiring their increased use every cycle. This whole deal with the AFCI's is something that I still cannot seem to grasp. I think they are just a bad idea and if the 2008 code ups the requirements and makes us install them on every circuit in a home, I think it will be a nightmare!
 
RDR said:
I'm primarily a semiconductor engineer. It's obvious, to me at least, that a microcontroller undoubtedly with some added circuitry, AtoD's and stuff, is used to run through a program that senses voltage/current irregularities. Does anyone know where the program used is published or defined?

Regards,
Click to expand...

I've suggested this before -- I'd look over the product literature, or perhaps molded into the body of the breaker, for patent numbers. Then go look the patents up on the USPTO website. The inventor must submit enough details for someone "skilled in the art" to make one of their inventions or else the patent is invalid.

I've read about half a dozen AFCI patents, and they are definitely not to be read while operating heavy machinery, and what I've read should do what it's supposed to do.

The basic idea is that they have circuits which sense recurrent transient current spikes. After some number occur in a given period of time, the breaker is opened.

Have any of y'all, especially the EE's in the crowd, ever found out why these things fail to operate in situations where they so clearly SHOULD be operating? Is it that the logic is wrong, or the sensors aren't sensing, or they things just break and don't fail-safe?
 
The insulation, on the hot wire, was burnt about three inches past the AFCI breaker. The wire melted in two pieces. The AFCI never tripped. Obviously, the loose connection was arcing at the AFCI breaker.
Click to expand...
That was not a parllel arcing fault. The only type of arcing fault that the AFCI can directly detect is a parallel arcing fault. That is a fault between the grounded and ungrounded conductors. A loose connection is just that...a high resistance point where I^2R heating is produced. Some have called these series arcing faults, but as series arc is not self sustaining at dwelling unit voltages...it self extinguishes the next zero crossing. The high impedance connection just produces heat and the only way that the AFCI can detect this is when the heat melts enough insulation to create either a parellel arcing fault or a gound fault (all AFCIs have 30 to 50 mA GFP) or in this case, since the AFCI is also a standard thermal magnetic breaker, if the heat at the thermal element would reach the trip point.
Don
 
Julie,
There are any number of reasons why the AFCI might not detect a fault. The first is the fact that they are only looking for parallel arcing faults, which in my opinion are rare compared to the most common electrical failure...a poor connection. Also the AFCI arc fault sensor does not even start to look at the arc unless the current is in excess of 75 amps.
Don
 
I've seen some electric motors that have a little red button. If the motor is overloaded and overheats, the unit trips and the button pops out and the current is cut out.
Perhaps if these simple devices were used instead of a fancy electronic gizmo, the result would be a more reliable method of detecting the arc fault and shutting off the current? These devices could be installed not only in the breakers but also in the receptacles and shouldn't cost much. Where else besides a breaker or a receptacle would an arc fault occur?
~Peter J. Michael
 
don_resqcapt19 said:
That was not a parllel arcing fault. The only type of arcing fault that the AFCI can directly detect is a parallel arcing fault. That is a fault between the grounded and ungrounded conductors. A loose connection is just that...a high resistance point where I^2R heating is produced. Some have called these series arcing faults, but as series arc is not self sustaining at dwelling unit voltages...it self extinguishes the next zero crossing. The high impedance connection just produces heat and the only way that the AFCI can detect this is when the heat melts enough insulation to create either a parellel arcing fault or a gound fault (all AFCIs have 30 to 50 mA GFP) or in this case, since the AFCI is also a standard thermal magnetic breaker, if the heat at the thermal element would reach the trip point.
Don
Click to expand...
Most of the arc fault problems I have seen are series, from loose connections. They may not be self sustaining but they certainly do melt the device, plates, and char the wood around the metal box. I have started to use them on older wiring that is not updated on service changes but this makes me think they are almost a waste of customers money.

Nebraska has not required AFCIs and reports of this type will not likely get them required the next Code cycle. I would like to know how many of you are satisfied with the perfomance of AFCIs
 
Most of the arc fault problems I have seen are series, from loose connections. They may not be self sustaining but they certainly do melt the device, plates, and char the wood around the metal box.
Click to expand...
They are not really arcs...just a poor connection producing high heat. Under the right conditions a few watts of load can result in a "glowing connection" (one that is actually so hot that it is glowing red) without tripping either a standard breaker or AFCI. The AFCI should trip once there is enough damage to the insulation so that the fault becomes a parallel arcing fault or a ground fault...the question is, is this before the fire starts?
Don
 
Status
Not open for further replies.
Top