This may happen on the programmed start ballast, when the preheating voltage/current are not sufficient, so electrodes do not reach the required temperature and/or their emission layer is worn out or damaged in another way (so require higher temperature to emit electrons and/or the heat transfer from tungsten core onto the coat surface is poor). Same effect may happen on voltage mode heated electrode, when the filament is broken, but it's coat is still in good condition (e.g. result of mechanical shock; this happened to me while i was debugging my dimming ballast and the only visible problem was poor arc stability at low power setting - i thought the control loop dynamic stability was at fault, but i discovered the real cause later on)
Electrodes do not have to glow visibly, in fact the visible (incandescent yellow and/or pure thermionic emission white) glow indicate the temperature is too high (however this is considered of less problem then insufficient heating). At the generally recommended preheat temperature (so the filament resistance is 4x cold resistance) the incandescent glow is very dim red, so practically not visible (unless in dark room).
Do not forget, then good electrode coat is of white color, so it does emit the (incandescent) light only very poorly.
|
|
|
Logged
|