There was a thread a long time ago, either here or at brickboard.com, on the cussedness of these valves. The consensus was that they tended to stick over in one position once activated, through obvious lack of use. There was a warning about unscrewing the switch, that it was difficult to stop it leaking, and very difficult to re-engage the threads. I recall that as an experiment I tried unscrewing my spare unit, and confirmed that this was indeed so. The switch is made of hard plastic, which is difficult to start in the threads, and very tight. It was suggested that having ascertained that there really was no fault in the brakes, the easiest solution was to disconnect the wire or take the bulb out. The initial triggering is caused by an imbalance in the pressures in the two brake circuits, in this case presumably only a momentary difference, not indicating an actual brake fault. I did suggest that a way of re-centring the sliding piston would be to lever back a brake caliper piston on the appropriate circuit, and then stamp on the pedal. This would pressurise one circuit immediately, but not the other because the caliper piston would have to take up the clearance first. The shock, if applied on the correct side, MIGHT re-set the switch. Just an idea - I don't know whether anyone has actually tried it.