Hi all,

Consider slimline and other instant start-only lamps. Does anyone suppose that running such a lamp on a proper electronic ballast might be advantageous over running it on the proper magnetic ballast? Now, of course the magnetic ballast will statistically last longer and the electronic ballast will use less power... But what about lamp life? Might a given lamp last longer on an electronic ballast than a magnetic one? I figure the high frequency AC would definitely be beneficial, but perhaps the modified waveform of electronic ballasts isn't such.

I suppose this exact same question could be asked regarding a magnetic preheat ballast with an electronic starter, vs. an electronic programmed start ballast, as well.

What do you think?

My Sylvania F96T12/CWX/SS lamps are beginning to show wear from use on a Keystone ballast. I don't even use the fixture that much. I've decided to only run 60 watt lamps so I can preserve 75 watt lamps for the future.

Not sure if it affects life or not but I see t12s used on electronic ballasts get bands where on magnetic ballasts they don't.  I had a pair of post-EPACT, pre-2003 GE Watt-Miser 34w cool white lamps running on an Advance 2XF32T8 electronic instant-start ballast in a frequently switched location (namely,  a bathroom)  and they did about a year but had brownish bands before they failed.

TheHF electronic ballasts give off only really a sinewave (the "modified sinewave" is only on the output of the UPS or so).
And other aspect, with the high frequency the ionization level in the lamp, so it's resistance, can not follow the waveform anymore, so even sharp current/voltage edges donot yield any high current spikes. (normally the resistance decreases with the current, hence the resulting constant voltage even during the mains frequency sinewave)

The unability of the ionization level to follow the high frequency waveform has yet another advantage: The arc does not extinguish during the current zero crossings, so the arc does not exhibitthe reignition spikes. These spikes are then responsible for the majority of the runtime wear (not the ignition cycle wear) of the mains frequency ballasts, so the lamp ages way slower.

As the starting can not be influenced, the high frequency ballasts would definitely do yield longer lamp life...


@ace100w120v: TheF34T12 are lamps, which could be started with heated electrodes, there the missing heating is, what makes the instant start ballasts wearingthem off faster. But that is not because they are high frequency, but because they start the lamp with cold electrodes and let only the cathode fall dissipation as the only heat source there.
And on top of that the F32T8 ballast feedthe lamp by 0.25A, while the F34T12 is rated for 0.43A, so it is operated at just about half of the rated curent (so at about 20W), what makes the heating phase (when the cold cathode operation damages the electrodes) way longer.

I see t12s used on electronic ballasts get bands where on magnetic ballasts they don't.

I've read this statement on here many times, and it confuses the heck out of me. Don't ALL lamps eventually develop bands as they wear? I've never seen one that hasn't, and I've certainly never heard of it being independent of the ballast used.

It depends on the operating mode and pattern, as well as (to big extend) the lamp design.
And ther are two types of banding: One is just mercury, which have condensed on the electrode assembly (mercury is more attracted to metals) and the ignition sputtered it onto the glass walls around. It tend to evaporate as the glass warms up, but frequently it does not evaporate completely. It could be recognized by rather sharp edge it forms around hottest points of the glass wall during lamp warmup, as it evaporates from there. This type of banding does not show anything about real lamp wear. It appear only after few operating cycles of a new lamp, after the mercury slowly moves onto the electrode assembly (because initially the mercury is dropped just on the glass, so resist on the lowest point; on the electrodes ends up only that part, which had evaporated during operation to form the discharge atmosphere).

And the other type is the sputtered metal frrom the electrode assembly. Normally this happens only after all the emission mix is either consumed, or otherwise damaged. The emission layer is white, so even as it gets sputtered around, it won't be visible. Only after it is consumed completelky, so the bare tungsten start to act as the cathode, only then it get spouttered around and form the blackening. This blackenning type then appear on the very end of the lamp life, as the "tungsten cathode" lasts only few hours. But it is really heavy black sputtering...

Other form happens when the lamp is cold started and the shape of the electrode assembly is so, the first cold electrode discharge goes from other structures than the filament. Then the material of those other structures get sputtered around. As this des not affect the cathode coat, this tyupe of blackening does not affect the life either (usually there is far eough material, so the sputtered amount does not affect the construction), but it does block some light and form not so nice banding. This type is then connected with not so optimum electropde shaping and could be prevented only by proper preheating (so the emission coat takes over all the current immediately since the ignition)

All lamps blacken/age somehow eventually but these had brownish bands, not blackening (though the EOL one got that too).  It makes sense though that starting was rougher...it certainly didn't last the rated life; granted, it was used but little/no where when I got that lamp).  A year of frequent/short switching cycles wasn't that long in terms of hours so yeah they died sooner for sure.

@Medved: And ther are two types of banding: One is just mercury, which have condensed on the electrode assembly (mercury is more attracted to metals) and the ignition sputtered it onto the glass walls around...

The recent ecological ("green tree" etc.) lamps contain much less mercury than old "full mercury" lamps. I wonder if this fact decreases the level of this type of banding?
Btw. I've read lots of complaints on those low mercury lamps here at LG. In what ways are they worse than full mercury lamps?

One of the degrading mechanism is the mercury irreversibly reacting with other components of the lamp, mainly oxygen from the phosphor. This then consumes the mercury, so the amount of mercury available for lamp operation decreases.
Traditionally this was solved by dosing sufficient amount of mercury, so even when part gets consumed, there is still plenty left for the lamp operation. But on the low mercury dose types the mercury may get consumed before anything else fails and so become the lamp life limitting factor.

Other aspect is the tolerance of the mercury dosing: With old tube types few droplets of mercury were just poured into the tube before sealing it. But the mercury tend to stick in droplets of certain size, so it means the standard deviation of the dose tolerance is about the size of such drop.
So if you just lower the overall dose, the tolerance "gaussian curve" start to reach significant values at the points of insufficient dose, so such lamps with shorter lifetime become more frequent.
The problem with the quality is, the end of line testing could reveal just when the mercury is missing completely. But when there is some mercury, just not enough for long life, it is not possible to recognize by the end of line test (the lamp still lights, as at that time there is enough mercury for the discharge). With "full dose" there were incidents, when the mercury was not put there at all (glitches on the machine; these defects are spot by the test, so such lamp rejected), but the low dose cases were practically not happening at all.
General rechnique to suppress this tolerance is to dilute the mercury into some other material (so fom an amalgam alloy), so then for the dosing you have again reasonable size chunks. Often this amalgam is made as a pressure regulating mesure as well. But this means another problem: Such amalgam tend to block the mercury by itself, so it has to be heated during the lamp operation. That means it has to be attached to some position with exact temperature, usually electrode assembly. And there come other problem: It may get loose. And that means it won't be sufficiently heated, so mercury starvation again.

Other aspect is, people frequently mix up cases of mercury starvation with other defect appearing in the recent time: Air leak, it look in very similar way (weak, purplish glow).
Well, in the first place the users don't have to care, in both cases a poor manufacturing quality caused the lamp to die too early, period.
This come from the extreme pressure on the manufacturing cost, at the same time the need to use some more expensive methods (more expensive phosphors, low tolerance mercury dosing,...). One way to lower the unit cost is to run the machinery at higher throughput rate, so you do not need so many of the machines. But that means, there is less time left for each of the operations, include sealing off the exhaust tube.
Normally it was rather slowly heated, so it collapsed and sealed when it was of very thick consistency, so it just kept pretty thick walls on the complete seal. Now if you want to speed this up, the only thing you may do is to increase the flame power, so it heats the glass faster. But that means the glass is not heated evenly, so some section become really liquid. This liquid is then sucked in by the vacuum, but as it is so easily flowing, instead of gradual collapse of the tube, it forms a kind of bubble on the end of the exhaust tube. And this bubble then has rather thin wall, what sometimes become so thin, it lets the air pass through slowly, poisonning the inner atmosphere. Again, this lamp works perfectly on the end of line test, so pass through to the customer...