Arc Fault Current Interrupters - Operating Principles?

[Jeff Wisnia]

Would someone who has had a hand in the design or testing of AFCIs
please clue me in on their operating principles?

My curious mind wants to learn about the detection parameters used in
those AFCI "circuit breakers" now required by code for some new
construction locations.

Do they measure and count the current spikes (or the "noise") which
occurs when a "frying" short happens, but when the average current drawn
probably wouldn't be enough to trip a normal overcurrent breaker?

I've noticed those breakers have a pigtail lead which connects to the
ground buss in the panel they're mounted in. Does that mean they also
provide GFCI in addition to arc fault protection?

My question was prompted by an aquaintance who has an AFCI breaker in
his home which frequently pops when he turns off a tungsten lamp load
with a wall switch. I suspect the switch may be arcing enough when
opening to trip the AFCI and suggested he try replacing that switch with
one having more of a "snap action".

Thanks guys,

Jeff
--
Jeffry Wisnia

(W1BSV + Brass Rat '57 EE)

"Truth exists; only falsehood has to be invented."

 

[Charles Schuler]

Would someone who has had a hand in the design or testing of AFCIs please
clue me in on their operating principles?

That leaves me out but I'll post anyway as I have investigated this subject
and learned very little (proprietary information, I suppose). Seems to me
that the arcing might be modulated by 60 Hz and the sideband energy would be
spaced accordingly and the specturm would be somewhat predictable. Perhaps
several key sideband frequencies are sampled? Do they use DSP? ... possible
with modern ICs.

My curious mind wants to learn about the detection parameters used in
those AFCI "circuit breakers" now required by code for some new
construction locations.

Do they measure and count the current spikes (or the "noise") which occurs
when a "frying" short happens, but when the average current drawn probably
wouldn't be enough to trip a normal overcurrent breaker?

That's the idea ... the arc fault current is below the trip current of an
ordinary breaker and the detector is looking for the arc signature (see the
above about spectrum).

I've noticed those breakers have a pigtail lead which connects to the
ground buss in the panel they're mounted in. Does that mean they also
provide GFCI in addition to arc fault protection?

Some do and some don't.

My question was prompted by an aquaintance who has an AFCI breaker in his
home which frequently pops when he turns off a tungsten lamp load with a
wall switch. I suspect the switch may be arcing enough when opening to
trip the AFCI and suggested he try replacing that switch with one having
more of a "snap action".

Yes, switches often arc and old ones get worse in some cases. Also, switch
design and build quality are issues.

 

[Fred Bloggs]

Would someone who has had a hand in the design or testing of AFCIs
please clue me in on their operating principles?

My curious mind wants to learn about the detection parameters used in
those AFCI "circuit breakers" now required by code for some new
construction locations.

Do they measure and count the current spikes (or the "noise") which
occurs when a "frying" short happens, but when the average current drawn
probably wouldn't be enough to trip a normal overcurrent breaker?

I've noticed those breakers have a pigtail lead which connects to the
ground buss in the panel they're mounted in. Does that mean they also
provide GFCI in addition to arc fault protection?

My question was prompted by an aquaintance who has an AFCI breaker in
his home which frequently pops when he turns off a tungsten lamp load
with a wall switch. I suspect the switch may be arcing enough when
opening to trip the AFCI and suggested he try replacing that switch with
one having more of a "snap action".

The AFCI are the ultimate in fire protection breaker technology-
designed to look for the telltale arcing between adjacent wires with
damaged insulation. They are programmed to trip on a broad class of arc
signatures but if IIRC there is no standardization of the arc profile as
yet. That switch is not making a clean break of the tungsten lamp
current- try replacing it with a newer and higher rated switch.