On 2 Feb 2004 09:11:08 -0800 bob peterson wrote: | I wonder how many lives, if any, will be saved by this particular code | change. That's supposed to be the reason for the code, but economic | interests seem to be what really drive things.
If arc fault detectors were real arc fault detectors, I'd want to have them in many places (and bedrooms first). But based on the designs I have seen, they are detecting arcs that specifically draw higher levels of currents, though not as high as a short would produce that a regular breaker would pop on. That's an arc between hot and either neutral or ground, but not a physical contact between conductors. It's a genuine fire hazard that normal breakers don't catch fast enough, so they are a good thing in that respect. But they are being marketed as doing what their designs won't do, and that's catch other forms of arcing that can also be dangerous, but have relatively low current (well under the whole circuit rating). Loose plugs, for example, will arc and/or produce hot spots, and can cause a fire. These AFCI devices can't detect that. If they could, it might be worth it.
I don't have an actual AFCI right now to test with, but it looks like they won't be as susceptable to RFI as GFCIs are. Well, I hope not. RFI will put a common mode current on the wires, which won't cancel out on the sensor. They will need a common mode RF filter to prevent the problem. AFCIs have a higher trip level, so maybe they won't be affected as much. Also, AFCIs will go in the breaker panel, mostly, and there is better opportunity to put some RF filtering on the wires.