I've been reading some of your older messages about provision of AFCIs. This is quite fascinating as we don't have anything like these breakers over here (not yet anyway!).

Does anyone know exactly how these devices work?

http://www.zlan.com/arc.htm

Paul;
they sense an arc of (i believe) 75W or more

.......have you ever heard the expression 'tail waggin' the dog'???


[This message has been edited by sparky (edited 09-09-2001).]

Thanks for the link. I figured they'd obviously have to be more complex than the standard thermal/magnetic or even amplified GFI types.

The only way I can see for them to work is for the circuitry to look for short pulses of increased current but not trip on longer pulses. I can see how such a circuit could be developed with simple op-amps etc. in integrated form, though I wouldn't be surprised if they actually use a micro-controller to analyze the current pulses.

I just wondered if any more technical details on the actual arrangement were available.

Does the AFI rely on the control circuitry to trip on normal overcurrent, or does it incorporate a normal thermal/magnetic trip as a backup?


Most definitely!

http://www.zlan.com/afci_c4.htm

Paul,
i could'nt really answer any of your Q's , this (check pix in link) is definitly not my forte' .

i'd like to know how to detect a bad AFCI

Steve,

No "test button" as per GFCI?

I've yet to get a close look at one of these buggers...

Guess the "pair of wires and a really steady hand" won't do it...

Great Q... Just how do we test them?



I suppose one could run a spot (read arc) light on an AFCI circuit (If you could find one that runs on 120V 20A) and provide a known safe long-lasting arc...

[This message has been edited by sparky66wv (edited 09-10-2001).]

Virgil;
yes there's a test button on the breaker itself, just like a GFI. I just haven't found a tester to plug in thoughout the circuit like the one's used for receptacles with the GFI tester.
where would i get a calculated arc of say, 76watts's

Don't worry since this will be a problem sure the mahufacturers who got the code comitte to cram them down our throuts will soon have a expensive tester for us to buy.

Now there's another word that has a somewhat different meaning on this side of the Atlantic!

Thanks for that. The diagrams & notes on that site are very helpful. Looks as though as I was right about the microprocessor analyzing the current.

Did you follow the links there to the pages about testing an AFCI?

They seem to be using another micro-driven test unit which stores typical arc waveforms and then re-generates them to test the breaker, except it's done via the non-shunted neutral line only in order to keep the current down.

http://www.zlan.com/testing/testing3.htm

well it certainly looks $$$$

It sure does. I like the way the company mentions "moderate cost" and a few thousand dollars in the same sentence!

From the article in EC & M on AFCIs they are suppose to detect arc faults as low as an amp, these faults have a ccertain signature when viewed on an O-scope and this has been encorporated into the AFCI. The reasoning is that just a small arc will not trip a breaker but if allowed to continue could start a fire. Also in '99 they wanted to require all new construction to have the whole residence wired with these new devices but was shot down, so they settled for the bedrooms to start, then for 2002 code, they decided to include all outlets in the bedroom area and smoke detetectors. My question is if an arc is intermittant and trips the breaker it will be a real nightmare finding the offending outlet.

The facts collected to the point where we are with AFCI's are , at least in my point of view, questionable.

The reality will , in time, prove or disprove the manufacturing hype that many of us ( Like nesparky..) are assuming.

The diagnostics of AFCI activity will probably come from this forum before manufacturer's

Well had my forth and final trouble call today at a custom house we did for an EE I know. He asked me to put in the Arc Fault breakers in his new house. He and his lovely family moved in 3 weeks ago. Seems that almost every morning, when his wife drys her hair down goesw the master bedroom power. Three times the arc fault breaker tripped and would not reset. They told me they reset it several times before it would not reset. All wiring tests just fine as does the outlets. But let her run a hair dryer and good bye power.
Today those $28.00 arc faults are back at the supply house.(all 4 of them) Told them I want my money back. Regular 20 amp breakers are now in the panel. At the EEs request all outlet wiring in his house is #10-2G romex with all outlets pigtailed with #12.
Yes he paid for it - he is also not very happy with the new arc fault breakers.
Thankfully since he is an EE, he understands the problem-- His wife is madder than He**. She does not want me back.

geez...
try an' be a good doobie, an loose a customer !

Nesparky
I would take them out, since it doesnt go into effect until january '02, and by that time this house is pre-existing.

Newly-published book explains AFCIs

The new 2000 edition of Overcurrents and Undercurrents, published by Reptec, is the first book to explain the
technology behind arc-fault circuit-interrupters (AFCIs).

AFCIs protect against arcing faults whose characteristics that may not cause
conventional circuit breakers or ground-fault circuit-interrupters (GFCIs) to trip.

Overcurrents and Undercurrents also covers GFCIs and ground-fault protection
for equipment. Its author is Earl W. Roberts, long-time chairman of N.E. Code-making Panel 2.

The book costs $25 including shipping and handling, and can be ordered directly from Reptec at (860) 536-4496
or reptec1@aol.com.

Does Earl explain how to test them in the field Joe? We all could loose work the same way if we cannot provied Mama solid evidence that it's NOT the AFCI, it IS your hairdryer.

The arc fault c/b are gone from this house. That repair worked. Since i gave those pieces of junk back, the sales rep and tech support guys are sure trying to make excuses-sorry assist in problem resolution- but they do not want to give my money back.
Doubt if I'll buy them again. Even if it means not bidding housing work. Cannot afford numerous warrenty trouble calls and no back up by suppliers.

http://www.squared.com/us/products/circuitb.nsf/07a0210021262d45862564b5006e4f84/7d00b860907fc6a38525694600549bd6/$FILE/12303.pdf

Look here for testing information.

From the troubles you're having with these things I'm glad they haven't appeared over here. It's probably only a metter of time though.

When the manufacturers got their "current signatures" of typical fault arcing, I wonder if they bothered to compare these with those obtained from a motor which is sparking slightly at the brushes?

it might be interesting to compare the signatures and see how close they are. If sparking commutators is going to cause problems with AFCIs, I see you having a lot of calls if/when they start to be used on circuits feeding power tools etc.

I have a new custom home to wire in the next month. I plan on meeting the AFCI requirement for this house so I ordered 4 AFCI breakers a couple of weeks ago from the one and only supply house in this county. The only line they sell is ITE. I got my 4 Q120-AF breakers yesterday($30.63ea.)and after reading nesparky's post about the EE's house I decided to run some tests on one in my shop.Here's what I did: My shop is all in EMT so I dropped a foot of 3/8" flex out of the bottom of my panel to a handybox. I ran 3 #12 THHNs(black, white & green)to a Hubbell CR-20 duplex receptacle. I have plugged everything I could find into this receptacle trying to trip it. Blow dryers, circular saw, air compressor, old electric blanket(and ran over it with my truck while plugged in), Milwaukee Holehawg(badly in need of new brushes),and an old B&D drill that you have to shake to keep it running. I took an old 16ga. extension cord and cut the ungrounded conductor then twisted just 2 strands of copper conductor back together and plugged a Milwaukee 1800W heat gun into the cord. It ran until the strands got hot and burned in two but none of this tripped the AFCI breaker! The only thing I had that would trip it was my Ideal GFCI plug-in tester. It tripped it instantly. Does anyone have any other experiment suggestions? I would also like to know what brand nesparky had trouble with.

Good man Fred!
it is interesting that a GFI tester would trip an AFCI. I wonder if the 10ma?? is close to 75W ??