We are confusing arcs with sparks. Anyone who has ever seen a cigarette lighter knows sparks can start a fire if the fuel is easy enough to light. You can also see pretty big sparks from low voltage supplies.
When Harvey Johnson from CH was explaining the AFCI to me in the 90s (before it was really known outside their labs) the example was a pinched lamp cord behind the bed, buried in cotton dust bunnies. That broken or shorted wire could easily create a spark big enough to light the cotton without tripping the O/C device.
Their mission was to design a circuit that would detect that situation and open the circuit.
Whether the ones they came up with actually work as advertised will be borne out by history. We are on something like Version 3.2 tho. I would certainly not have much confidence in the Ver. 1.0 device we were forced to use in 2002.
Greg, do us a favor the next BBQ you have with your neighboring manufacturing reps, and ask them
specifically what changes in the detection algorithm(s) have had UL1699 amended
~S~
