Which starting method gives the highest life at fluorescent lamps?

On the smaller wattage, i.e. up to 20w, I've noticed that lamps with trigger start (aka rapid start for smaller wattage) ballasts seem to last longer in most cases with the fixtures I've used. Of course, it will depend how often they are switched on/off, but they seem to last twice as long as preheat...for the 120v reactors. The F40 seem about the same as I've seen on the fixtures at work.

From what I have read by ballast manufacturers and in other articles, programmed start (electrodes heated before start & then shut off) gives the highest lamp life, followed by rapid start, preheat, and then instant-start. However, if a lamp is lit once and then lit continuously until it fails, I would imagine instant start ballasts would give a longer life because the high OCV would keep the lamp going until the electrodes melt and the lamp loses vaccuum, whereas on other types of ballasts, the lamp would go out when the ballast can no longer supply the voltage needed to keep a steady discharge.

Thanks for the infos!

actually i think programed start would make the lamp last the longest instant start would not provide sufficient cathode heat causing it to slowly sputter away and fail after 5 years. ive witnesed it happen.

I seem to remember that General Electric used to claim a longer average life for their F40T12 when used on rapid start circuits as opposed to preheat circuits. IIRC this came from an 1980's lamp catalog.

Again, depends on the use, amount of times switched on and off and temperature.  Preheat is very reliable at starting at least down to freezing (0c/32f), whereas many less than full power rapid start ballasts will not start well if cold.  If switched on once a day and run for 8-10 hours, I would imagine preheat and rapid start would compare close in longevity.  I have many F40 preheat fixtures I use daily and I rarely ever change a lamp, even though many of the lamps I have installed were used when installed!  Frequent switching of either preheat or rapid start will shorten their life.

Preheat is not all identical, i would guess that blink-happy starters will kill the lamp earlier cause some of the flashes come after too little of preheating

I think one of the problems with (smaller wattage) PH lamps in the US is that most of the choke ballasts are for 14w-22w, meaning that the 20w are seriously underdriven in many cases...I've measured currents as low as 250 ma, and the low current possibly causes low electrode heat which in turn looses efficient electron emission causing premature wear.

Is there reason why higher lamps in the US are not preheat, while they used to be so before ?

Is there reason why higher lamps in the US are not preheat, while they used to be so before ?

Higher wattage lamps need a transformer ballast anyway (120V is too low as an OCV), so it was no more complication to add two 3.6V windings and a bit higher OCV to get rid of the starter.
Moreover if that mean faster and smoother startup...

I think I remember reading ( GE maybe ) years ago = 12,000 hours preheat and 20,000 rapid start.


Powell

@Powel: These numbers were meant for the "home use" (2.7hours per start) cycle.

 YES  indeed !!!     The 12,000 vs 20,000 was rated at 3 hours per start in the literature !


Powell