Arc Fault breaker operation

[rkrieger]

OK, I think I understand how a GFI breaker works by sensing the difference in current between the neutral and hot conductors, but is an ARC fault breaker the same concept?? How does it sense arcing and sparking some distance from the breaker? Probably similar to a GFI but does anyone know for sure??

Thanks from the "newbie" again.

Dick Krieger

 

[cadpoint]

Where all dealing with it!

Where all dealing with it!

In the new breakers its just sitting there ready to release, just like any service breaker. An Arc Fault Current Interurpt, yes it senses an inbalance of the current through the use of applied electronic therory,
IE which is within the breaker & yes it works.

 

[winnie]

The approximate hand-waving description:

The AFCI circuit has a sensor which measures the actual pattern of current flow. Think about what an oscilloscope does, but rather than creating an image for a human to look at, the measurement is loaded directly into a small computer.

This computer is supposed to look for specific patterns of current flow associated with arcs. The better the job of pattern recognition, the better the AFCI will differentiate between arcing faults and normal operation.

Different loads have different operating characteristics and different current flow patterns. An arcing fault will look quite different than a resistive load or an induction motor. Unfortunately, some loads have totally normal current flow patterns that look quite a bit like arcing faults. These loads make the job of differentiating between faults and normal operation quite difficult.

If your computer has too broad a definition of the current flow pattern that looks like an arcing fault, then you will have lots of nuisance tripping. If the definition is too narrow, then your breaker will miss real arcing faults.

Until recently, this has been so much of a problem that only current flow that looks like an arc _and_ is well in excess of the breaker trip rating will be treated as a fault. This has the benefit of speeding up the response to an arcing short circuit, but this approach cannot respond to 'series arcs', meaning faults where the current passes through an arcing connection and then through the balance of the load.

As of January 1, 2008, so called 'combination' AFCIs will be required. These have to properly detect arcing down to 5A, and should be able to detect arcing faults.

It remains to be seen if manufacturers can actually deliver a product which can both detect series arcs _and_ not have excessive problems with nusiance tripping.

-Jon

 

[Physis 3]

What AFCI's do is "compare" the waveform, or what, as Winnie says, you would see on an oscilloscope, to a preprogrammed or stored set of waveform profiles.

It doesn't care about the distance from the event. Any changes caused from distance are beyond negligible.

Probably similar to a GFI but does anyone know for sure??

The only similarity is that they both detects things that a typical circuit breaker wouldn't. Other than that they are completely different animals.

A GFCI is very simple in how it tests for a fault and works very well with respect to what it's asked to do.

An AFCI on the other hand is still very much in the development stage and it's performance is at best very iffy.

Edit: Error B , I think it was?

 

[BVD]

The approximate hand-waving description:

The AFCI circuit has a sensor which measures the actual pattern of current flow. Think about what an oscilloscope does, but rather than creating an image for a human to look at, the measurement is loaded directly into a small computer.

This computer is supposed to look for specific patterns of current flow associated with arcs. The better the job of pattern recognition, the better the AFCI will differentiate between arcing faults and normal operation.

Different loads have different operating characteristics and different current flow patterns. An arcing fault will look quite different than a resistive load or an induction motor. Unfortunately, some loads have totally normal current flow patterns that look quite a bit like arcing faults. These loads make the job of differentiating between faults and normal operation quite difficult.

If your computer has too broad a definition of the current flow pattern that looks like an arcing fault, then you will have lots of nuisance tripping. If the definition is too narrow, then your breaker will miss real arcing faults.

Until recently, this has been so much of a problem that only current flow that looks like an arc _and_ is well in excess of the breaker trip rating will be treated as a fault. This has the benefit of speeding up the response to an arcing short circuit, but this approach cannot respond to 'series arcs', meaning faults where the current passes through an arcing connection and then through the balance of the load.

As of January 1, 2008, so called 'combination' AFCIs will be required. These have to properly detect arcing down to 5A, and should be able to detect arcing faults.

It remains to be seen if manufacturers can actually deliver a product which can both detect series arcs _and_ not have excessive problems with nusiance tripping.

-Jon

This has the benefit of speeding up the response to an arcing short circuit, but this approach cannot respond to 'series arcs', meaning faults where the current passes through an arcing connection and then through the balance of the load.
Just for clarification, the series arcing would be considered a load type? Not real familiar w/ what the combo types are supposed to do. The branch circuit type just protect the circuit upt to the outlet-right? Then the combo type will also pick up arcing in the load itself?? I know I'm centuries behind,but deal mostly in commercial work. Thanks

 

[George Stolz]

Take a look at the FAQ on AFCIs.

As insensitive as the branch/feeder type AFCIs are, they can only theoretically detect arcing between an ungrounded conductor and another conductor. The more sensitive combination types should detect arcing at a point along a single conductor, if they work as billed.