"bud--" wrote in message
eb.com...
On 3/2/2014 8:55 AM, Robert Green wrote:
Clearly figuring out which device is actually more spacious is something
I'll leave to someone with the actual engineering specs of both devices
and
LOTS of time on their hands. I was just surprised that they were so
equal
in my rough measurements.
The size of the AFCI receptacle can make it harder to install if there
are a lot of wires in the box. The AFCI is larger than a simple
receptacle. (Same is true for a GFCI receptacle.)
Not only is it larger, one end is 1/4" thicker than the other, so you have
some space to tuck wires but not an awful lot. I had no troubles with
new(ish) 12/2 Romex using them as pass throughs but this is new work and I
always try to use roomy metal boxes for RTE - room to expand. I've never
had to take stuff OUT of boxes. The only time I gained any space in a
wiring box was replacing the failing 40 year old light switches with newer
ones that were remarkably small in comparison. But soon afterward I was
cramming large X-10 devices into smallish boxes. Fortunately they used
pigtails which meant wire nuts but a lot of flexibility in wire cramming.
Unfortunately wire nutting stranded and solid wire is an artform and skill
at it evolves slowly. )-:
For the really critical stuff I once chiseled out the whole thing but that
turned out to be a project that I immensely underestimated, time-wise and
since then I tend to choose any other method. Even installing a new box in
a plaster and lathe wall is easier than working around an old box. The damn
Reader's Digest book made it look a lot simpler than it turned out to be. I
can still see the faint outlines of the canal I dug to string new wire after
the old one broke right at the entrance to the switchbox. The front porch
light switch is where I learned to be very gentle with old wiring and the
toilet shutoff valve was where I learned to be very gentle with old
plumbing.
For determining how many wires can connect to a box, a receptacle or
switch volume is factored in. I am surprised the larger volume of a
GFCI/AFCI is not accounted for.
Really? I assume you're talking NEC.
I know of several problem outlet boxes in My Very Old House that those units
would not fit into because X-10 devices of similar dimensions did not fit.
Twisting and fussing with those old, cloth covered wires is very dicey
business if you hate chiseling plaster as much as I do.
The images at the top all have pigtails. Mine had a pigtail. I assume
it's
looking for some difference in something but I was unable to find out
why
they required the pigtail the last time I looked. I should have
researched
it further but that was a dealbreaker and once I discovered the outlets
that
were much cheaper I didn't care much *why* the pigtails were there.
AFCIs require the neutral be wired though the device for the same reason
GFCIs do. For the ground fault feature the AFCI runs the hot and neutral
through a current transformer and looks at the difference in the current
between them. If it is 30 mA or more the AFCI trips.
"Some difference in something" I was right! (-: So the effect is that the
neutral of the load isn't mixed into the pool of shared neutrals directly -
it has to pass through the breaker so the GFCI component can determine if
the neutral from the load is or is not at ground potential. I assume that
is what the white pigtail is always at because it's connected to the neutral
buss bar which is bonded to ground at the panel.
The latest thing I read about AFCI operation is that they captured the
signatures of over 100 different arc sources and try to pattern match any
detected arc to their internal arc sig library. That's how they exclude the
arcs produced by motors, switch contacts, vacuum cleaners and other "good
arc" creators.
Rather than my inept paraphrasing, I'll just post the URL and the abstract.
A Method for Parallel Arc Fault Detection and Identification
http://www.atlantis-press.com/php/do...er.php?id=4448
Abstract-Every year, fires bring great loss and damage, and
arc faults is one of main causes. But it is identified that the
conventional circuit breakers fail to interrupt the circuit when
an arc fault occurs. It is necessary to find new protecting
techniques to improve the safety of circuits and reduce the
probability of fires. This paper shows a simple method of
detecting parallel arc faults. In this thesis, two significant
characters of parallel arcs, the "shoulders" and the high
amplitude, are obtained by experiments and then a circuit for
discriminating the parallel arcs depending on them is designed.
Other normal working household appliances, such as
computers, air-conditions, are also have the similar characters,
not easily distinguished from arc faults. The waveforms of
those appliances are studied and compared with arcs. Results
show that they still have different features in "shoulders" or
amplitudes. At last, the correctness and reliability of this
method are tested by an AFCI-tester.
Some interesting diagrams . . .
stuff snipped
I assume the conduit (or is that word misapplied - is "metal-walled"
more
inclusive of other types of wiring?)
Also may include metal sheathed cable - MC and AC.
Was it ever called armored cable or BX? That's what I seem to remember from
NYC wiring codes. That and my poor Dad regularly cutting himself on that
coiled sheath.
is required to protect against other
types of arcs that can occur?
It adds safety for parallel arcs (L-N, L-G). With romex you can get a
parallel arc from a nail or screw driven through the cable. A lot harder
to do with metal walled, and the metal provides some fire protection if
you manage to do it. There is much higher current available for parallel
arcs.
Is that because L and N are directly connected with no intervening load?
--
Bobby G.