An AF breaker may trip over several different things, which may be hard to distinguish. Here are a few :
- Arcing in parallel with the line (phase to neutral or phase to earth) through a place where the insulation is damaged and a conductive track is forming. The idea of arc fault breaker is, that the energy of the momentary shorts happening in the conductive track may be below what it takes to trip a normal breaker's electromagnetic release, unless they develop into a full high current arc (The odds of which get lower when the line voltage is lower, as in 120V vs 230V)
- Arcing in series with a load. This is one of the least reliable detection functions of AFCIs, as some series arcs dont have as clear signature (current peaks in the microsecond range and 10's of A that parallel arcs have). Some of them act more like a fairly stable resistance. There are some series arcs the breaker is meant to allow (like brushed motors), but there it helps that most such appliances have at least an X capacitor in them, which prevents most of the interference from the motor from coming out towards the breaker. (vs. a bad connection upstream, where all the interference is directly in series with the breaker). I imagine that the breakers will have some time-delay which is sufficient to ignore thermostats, as long as they click reasonably quickly
- Some devices plugged in still have parts under voltage even when they are off. For example, the AF breaker might (i guess ?) respond to self healing events of an X capacitor in an appliance line filter, even if it does have a mechanical on off switch and is switched off
- Interference that somehow match what the breaker is looking for. This might be induced both from within the circuit and from upstream (like from another circuit in the house, back through that circuit's breaker, panel busbar, and the AF breaker in question). I think some factors contributing to this sort of effects may be : Long circuits in general, especially circuits that extend outside of the house, or if the circuit wiring downstream is somewhere separated - as in phase and neutral wires taking different paths towards a point in the circuit. (This is not a proper way to install circuits, but had been found many times especially in lighting circuits that span between floors or across big areas of the house). The interference may come from many sources, including appliances etc, there have been cases of SW radio transmission in the area tripping AF breakers in houses...
As you mention your circuit, it seems to tick quite a few boxes of the latter....
How is the circuit laid out ? With what type of conductors ?
Disconnect all loads from it, remove any built in electronics (such as GFCIs, motion detectors etc) and perform an insulation test between all conductors (phase to phase, both phases to neutral, to earth, and neutral earth)
If you find abnormal readings, get to the place where the circuit branches to different locations, disconnect them to test the insulation separately and possibly locate the part of the circuit which might have bad insulation
Most AFCIs do have some low limit on the current they detect for parallel arcing. If the circuit is not in use (or you can make it not in use), you can watch it for lower currents - disconnect the phase end of it, and insert a low range ampermeter in series to monitor the actual current draw of the apparently open circuit. (Add an incandescent lamp in series for protection, in case some load appears which is higher than the ampermeter's fuse rating, or if the circuit indeed suddenly shorts)
An intact circuit will draw tiny (order of <1 mA), but most importantly constant current, coming from capacitance between the conductors and maybe distributed leakage currents for some insulation materials
A non intact circuit may (if you happen to catch it) draw higher current (order of few mA), and this current will sometimes tend to fluctate over the course of hours or days in a pattern - As a conductive track is slowly formed in the place of bad insulation, then at some stage it burns clear and starts over. (The addition of current limiting in series may affect the process as it limits the available current for the burn clear stage)
If the phase wire goes along with bare earth in a cable, or any similar setup where leakage is likely to be towards earth conductor or towards external conductive objects, then parallel arcing faults will also likely trigger GFCIs. If your AF breaker is not also a GFCI by itself, then adding a GFCI at the beginning of the circuit (after the arc fault breaker) will add an additional layer of safety, and might show whether there is indeed an earth leakage (the leakage may be intermittent)
You can try to swap AF breakers between this circuit and another, to see if there is some issue with the breaker. However, this test might be inconclusive. There might be a combination that's acting to trip it - There may be indeed something wrong with the circuit, but borderline, and the particular breaker might be with a little lower thresholds than another identical breaker. So if this test appears to solve the problem, you still dont know if it was interference from outside or there is really some very first stages of something happening in this circuit
If you can reconnect loads like groups of sockets permanently or semi permanently to other circuits, this will rule out parts of the circuit which may or may not be the problem (Just in case, do have an identical AF breaker on the circuit that you move the parts to). Also, it will reduce the amount of interference the circuit receives
|
