Not so much if the relatively new "all electronic" ballasts are used. They step up the frequency into the kHz range generally. I'd have to look it up to remember the details, but higher frequency operation has lots of advantages, both to lamp life and ease of starting.

ETA: Not lamp life, but *efficiency* (how much light you get per input watt of power, while fluorescents, and the HIDs are much more efficient that incandescent, in terms of the actual radiated power of visible light vs total input power, this figure is vanishingly small), actually. High frequency can actually reduce life at some point. It's funny. If you were to operate a lamp with a DC current, the anode side will wear out more quickly. The anode is what gets the brunt of the heating, with electrons slamming into it. With AC, you spread that out over both ends. But at some point, the higher frequency will start reducing life.

-Richard

And for more than you probably wanted to know about electronic ballasts. I just read a little paper on finding the optimum driving waveform for a flourescent tube. Optimum was defined as something that minimized current crest factor (ratio of peak to RMS), and hopefully minimized harmonic distortion. These help extend lamp life.

A sine wave was not optimum, nor was a square wave. It wasn't a fully analytic solution. The had a range that represented two competing characteristics of driving waveforms, with in between waveforms, then looked at where the optimum seemed to be.

Interestingly the optimum turned out to be an ellipitical waveform. That is, take half an ellipse and periodically extend it in glide symmetric fashion. With modern electronics, that is possible to do easily enough.

If any are interested (and I hope you are -- tell me and I'll quit rambling about all this odd stuff that peels my bananna so to speak), here's what's afoot. Arcs are non-linear. The resistance is not constant with current (resistance decreases with current, which gives a runaway effect). And so if you drive an arc with a sine wave voltage, you don't get a true sine wave current response. If the non-linearity is small, the harmonics are small, but the effect is there. The math of arcs gets complicated -- and that's just a taste of full plasma physics.

The idea here was to work with the non-linearity and find the driving waveform best suited to it.

-Richard

I have a story on this topic from my misspent youth.

I was taking the girl I was dating home, driving late at night through a bad neighborhood, when someone shot out the front passenger window. I was turning toward her at that instant, and saw a cone of shattered glass in the wake of the bullet. That was over 20 years ago and I can still remember it like it was yesterday.

__________________
"Scientific progress goes 'boink'?"