Startup coss of CFLs

[Adam Funk]

Interesting, thanks (to both of you).

[Grimly Curmudgeon]

Oh, I can't recall exactly - New Scientist or Practical Wireless, or similar. The original article did mention the life-shortening effect of start-ups compared to the actual cost of running the tube. There was no mention at all of the start-up cost of power - that's a total irrelevance. That was back then - things have moved on and the goalposts have shifted. Modern electronic starters are, afaik, much kinder to the tubes than the old trigger starters and ballasts and tubes last much longer with them. Tubes still lose emmissivity, to the point where they need to be replaced after x-thousand hours, but I've noticed that old tubes which previously would have simply refused to start, will carry on working, even at much reduced levels.

[Andrew Gabriel]

Actually, switchstart on 240V supplies for tubes from 18W to

80W works extremely well, and almost no worse than an electronic ballast with a microcontroller controlled starting sequence. On the lower power tubes, there is barely enough power to operate the starter (certainly not optimally), and with 8' tubes, the mains voltage and tube voltage are a bit too close together, particularly if the tube is cold, and it may not start.

Switchstart on 120V supplies is really bad though, and only works at all on short tubes, but seriously reduces tube life.

The electronic retrofit starters usually try to start the tubes without preheating by generating a stream of HV pulses. This ignites the tube faster, but at the cost tube life. There are different types with different timing options though.

Most electronic ballasts preheat filaments for least wear, but they can be designed to perform instant-start instead. I have a number of the instant start type installed, but they've got very hard to find in the last 5-10 years. (Unfortunately, the term Instant Start is now also used to an mean electronic ballast which might have a 0.5-2second preheat time, so it's very hard to tell which ballasts will really be instant start without getting one and trying it. Some have even changed from instant start to preheat without changing the ballast model numbers!)

Instant start fluorescents have always been much more common in the US than anywhere else, and they even have tubes with single pin connection at each end which can only be instant started.

Loss of emissivity causes a fairly instant death, when the last bit wears off one electrode. Only one tiny bit of it is used at any one time (the hot spot), so the tube will still work fine when 99% of it has gone, but not when the last bit goes. Some crappy control gear will carry on running a tube after the emission material has gone as a cold cathode tube, but that will only last an hour or two, before the tube melts or cracks or the control gear burns out due to higher power and voltage output.

Dimming of tubes is due to the phosphor coating wearing out. With more recent lamps, when it's also accompanied by a longer run-up time, particularly if there's a dim pink glow when it gets really bad, it's due to the mercury running out. Tubes are now only dosed with just enough mercury to last the expected life, whereas they used to have an excess so it never ran out before something else died first. The mercury is slowly lost into many of the tube components (glass, phosphor, electrodes) during operation, where it no longer works.

[Adam Funk]

...

Interesting, thanks.