afci breakers

[roger]

Since when, in the name of safety ,would you rely on a device to test itself?

When it is the only recognized way to perform the test.

BTW, I don't ("in the name of safety") have much confidence in the AFCI manufactures claims either.

You do realize that GFCI's and so called GFCI remote testers share the same issues as the AFCI remote testers don't you?

Roger

 

[ELA]

The GFCI is in a totally different catagory from the arc recognition portion of a AFCI and I would not compare them.

For a GFCI it is simply a matter of creating a differential current greater than 5 ma and can be done using a resistor.

Arc creation and characterization are way more complicated matters.

 

[76nemo]

I have read so many replies on this. If an AFCI tester can simulate so many arc's per time period, than why does the test button deter from a good tester????? I know they can't simulate EVERY arc fault, but they can simulate a fault that SHOULD open the breaker. I emphasize on SHOULD.

I have read soooooo many debates on this that it has become a headache. A good tester can simulate a fault that should normally open the circuit. Obviously the EXACT fault may NOT be able to be duplicated but should be able to perform a test that simulates the AFCI breaker to open, not genuinely, but well enough to have the breaker recognize a certain fault, one that which in real life would cause a problem to set the breaker off in the first place.

Ughhhh, I have NO real answer. How can one say the test button is superior over any one particular tester?????????

 

[crossman]

No, they create an actual arc...something a tester cannot do.

Why can't a tester create an actual arc?

 

[brantmacga]

according to the picture on the front of the new square D combo afci's, if you set the house on fire, the afci will protect you from shock. try that.

maybe you could plug in an extension cord and run a drill bit through it. that supposedly makes them trip too. if i were an inspector, i would require method #1 for an approval.

 

[roger]

The GFCI is in a totally different catagory from the arc recognition portion of a AFCI and I would not compare them.

For a GFCI it is simply a matter of creating a differential current greater than 5 ma and can be done using a resistor.

Arc creation and characterization are way more complicated matters.

You missed the point, the fact is that the only true way to test either device is with their on-board test buttons.

Roger

 

[ELA]

You missed the point, the fact is that the only true way to test either device is with their on-board test buttons.

Roger

I think I disagree. Who's point?
When you say the only "true" way -can you qualify what you mean?

I heard you when you said that a true AFCI tester may not yet be available to the general public. I accepted that as true.

I maintain that there are other "true" ways to test each device. They may not be accepted by some but none the less true in my minds eye.

Perhaps the definition of "true" is weighted by the level of expertise involved in performing the testing. The test button is the most accepted method for the general public. There are other high level lab type testing that I would consider "true" ways of testing. With interest in doing so, and enough money invested, there could also be a hand held device that could be considered a "true" test.

 

[wptski]

How can anyone state that the "only" approved way to test a AFCI is the test button when it also has been stated that nobody knows how they work and each manufacturer seems to be doing it a different way?

I know that the answer will probably be because it's the "only" NEC approved way!

 

[crossman]

Seems the "most accurate" method to see if an AFCI works is to mimick an arcing fault which could cause a fire on the circuit downstream of the breaker.

Maybe the only "approved" method is the test button, but that does not make it philosophically the proper way to do it.

At a trade show awhile back, SquareD had a booth and had an external apparatus connected downstream of an AFCI breaker. The external apparatus looked somewhat like a micrometer with a hot wire connected to the movable portion and a neutral connected to the stationary side. They would dial the hot closer and closer to the neutral until it sparked and the breaker would trip.

So, there is certainly an external method to test an AFCI. Whether it is approved or not is not the point. Fact is, there SHOULD be a generic tester which can test any brand of AFCI.

Again I ask: Why can't an external tester create an arc that can test whether an AFCI breaker is working?

 

[roger]

How can anyone state that the "only" approved way to test a AFCI is the test button when it also has been stated that nobody knows how they work and each manufacturer seems to be doing it a different way?

I know that the answer will probably be because it's the "only" NEC approved way!

Ok, take it up with UL.

http://www.ul.com/regulators/AFCI_Indicators_Regulators_Page.pdf

Notice in the answer to the question in the link below states that AFCI indicators are not testers

http://www.ul.com/regulators/mayjune2005.pdf

and take particular notice of where is stated that the only accepted method of testing is the "test" button on the device in the following link.

http://www.ul.com/regulators/julyaug2003.pdf

If anyone knows of a more recent develoment please post it.

Roger

 

[al hildenbrand]

So, there is certainly an external method to test an AFCI. Whether it is approved or not is not the point. Fact is, there SHOULD be a generic tester which can test any brand of AFCI.

Again I ask: Why can't an external tester create an arc that can test whether an AFCI breaker is working?

I imagine the Combo AFCI as a wide band signal processor. The processor must first detect, and then discriminate ALL of the possible signal types of "bad arcs" from all of the signals it observes.

There will be a nearly infinite number of specific "bad" arc signatures that must cause the AFCI to trip.

No one test procedure that can be executed in the field can test a customer's AFCI an infinite number of times to verify that the AFCI does, in fact, SEE each and every possible "bad" arc signature.

Some time in the future, I expect that our regulatory agencies will compromise on a testing paradigm that has three or five different arc types, arc types carefully chosen to be in the very different, yet more common, areas of all the possible arc signatures. And then we, as the field technicians, will be told, "Do this, this and this, and that is all that is needed for the truth that is out there."

A more restricted procedure is what, in fact, is happening right now. We are told by UL and the manufacturer's that all the truth we can handle, or will ever need, in a court of law, is the truth derived by pressing the AFCI breaker test button and having the breaker shut off.

 

[crossman]

Thanks for the reply, Al. Your commentary is concise and makes sense. Maybe someday all the "waveshape signature" stuff will be wrapped up in a few practical types of arcs with a tester that can be used externally.

 

[ELA]

I have always been curious just what the test button on the AFCI does.
Up until now I have never seen a schematic of an AFCI.

I found page 5 of this doc interesting. I cannot vouch for its accuracy, I just found it on the net :smile:

http://www.pfeiffereng.com/The Arc Fault Circuit Interrupter.pdf



If it is accurate then see how the test button is just a digital input to a microcontroller. Very interesting ....

 

[don_resqcapt19]

If it is accurate then see how the test button is just a digital input to a microcontroller. Very interesting ....

That would not suprise me as that is exactly how a lot of the test buttons and magnetic switches on smoke alarms work.

 

[crossman]

That would not suprise me as that is exactly how a lot of the test buttons and magnetic switches on smoke alarms work.

I don't know much about fire alarms, but when I have seen certification testing done on a system, they run around with little cans of spray smoke and actually test if the heads can function under real-world conditions.

I would think, since the children's safety is at issue here, that a real world functionality test should be required by the NEC with a "arc-fault box" tester.

 

[ELA]

That would not suprise me as that is exactly how a lot of the test buttons and magnetic switches on smoke alarms work.

It does not surprise me that it does not surprise you.

Does it make you curious as to what the microcontroller does to test the unit? Perhaps all it does is energize the output to trip the breaker.
It makes me very curious to know more.

 

[wptski]

Ok, take it up with UL.

Notice in the answer to the question in the link below states that AFCI indicators are not testers

and take particular notice of where is stated that the only accepted method of testing is the "test" button on the device in the following link.

If anyone knows of a more recent develoment please post it.

Roger

Who first refered to or changed the name of a AFCI Tester to AFCI Indicator? Define a indicator? Is a Wiggy a tester or indicator? Would a tool be called an indicator becasue it only has lights? It still tests. Testing is a function.

What indication do you get when a AFCI breaker is tripped?

 

[roger]

Who first refered to or changed the name of a AFCI Tester to AFCI Indicator? Define a indicator? Is a Wiggy a tester or indicator? Would a tool be called an indicator becasue it only has lights? It still tests. Testing is a function.

Did you take your questions to UL as I suggested?

What indication do you get when a AFCI breaker is tripped?

What I would expect to see is no power but, if using an external device (indicator) to try to cause this condition, and it did not end with the AFCI breaker tripped, it still wouldn't mean the AFCI is not working properly since the external device is not a recognized tester.

With that said, if an inspector used this as reason to red tag an installation, he/she would be in the wrong.



Roger

 

[iwire]

I don't know much about fire alarms, but when I have seen certification testing done on a system, they run around with little cans of spray smoke and actually test if the heads can function under real-world conditions.

That (at least around here) is entirely up to each city or town. Some want a source of real combustion (we use a fancy tool that holds a punk) or some areas want the canned smoke while others think those previous two methods harm the smoke detectors so this group has use use the magnets.

 

[Lxnxjxhx]

the test buttons

the test buttons

BITE (built-in-test-equipment) by definition, cannot include all device components when it does its test.
And, as to a button that gooses some input on a microprocessor as a substitute of injecting real arcs into wiring, this borders on the ridiculous. It assumes the design is sound. Unless there is a gross failure, all you're doing is confirming the designer's wishes.
The problem is, is the device he/she designed capable of dealing with what the real world throws at it? All manus are invested in the "yes" answer. I would probably be, too, until I come up with the improved version.
There is a pressure to be first in the marketplace, at the expense of merchantibility.