Tubes vary a lot in output. T4/5 triphosphors on hf can produce over 100 lpw, low efficiency well aged halophosphate T12s a dim fraction of that.
NT
Tubes vary a lot in output. T4/5 triphosphors on hf can produce over 100 lpw, low efficiency well aged halophosphate T12s a dim fraction of that.
NT
This one is kind of nice. Got one to replace the 4 13W CFL bulbs in our aquarium. It was on sale for about $40 about a year ago.
Michael
Life Years: 45.7
I'll be long dead by then. lol
r aquarium. It was on sale for about $40 about a year ago.
.product.100223617.html
Hahaha. Bequeath it to the kids or grandkids then!
I took this set apart, and the electronics seem really simple. I *think* t he LEDs are wired in series, and then the mains voltage just powers them. Maybe half of the LEDs light up on one cycle, and the other half light up o n the other cycle... I don't remember if I saw diodes on the circuit board. Tubes were just plastic cylinders frosted white, and the LEDs were spaced an inch or so apart, pointing down.
Michael
Wow, what junk! The (yes, they are expensive) commerical LED lighting power supplies I am using are isolated form the mains, and a true DC current source. When you shut them off, you can see the capacitors drain for a moment. I really like the DC lighting - no flicker! I will be doing more of these as I get around to it.
Jon
Maybe I was mistaken, or I didn't take it apart thoroughly enough. I found a teardown video of the same Feit LED fixture, and this guy found some more circuit boards inside than I did.
Around the 8:40 mark, he seems to have uncovered some sort of switchers... I don't remember those when I took mine apart.
Thanks,
Michael
Life Years: 45.7
I expect the plastic tube will yellow with time, but that depends on the material. I know a *lot* of the light covers on fluorescent fixtures turn yellow after a few years. That costs a *lot* of illumination.
I don't care if they are run from DC or not. I can't see lamp flicker at 60 Hz although I could see it on a CRT, but that had motion involved. Actually, these will flicker at 120 Hz, not 60, so I can't imagine anyone can see that.
The down side of the power supply will be the change in intensity with power line voltage which *does* vary, sometimes rapidly.
He found two controllers for each tube. Looks like the first board converts to high voltage DC (likely unregulated) and the second board regulates the current. Interesting. 38 watts though, so much more than half the power of a similar T8 fixture.
OK, good to hear there's more to these than originally reported.
Well, this first batch of retrofits is in the kitchen, but I also have some CFLs on my milling machine, and there the strobe effect is very strong. Although I've never been fooled, it would be possible for an untrained person to think the cutting tool was not turning when it was, when the machine happens to be at certain speeds. That is dangerous.
I will eventually get around to retrofitting those positions, not for energy efficiency but for safety.
But, I am aware of flicker. I don't think I see that much 120 Hz flicker, but if the emissivity of the filaments on the tubes is not matched, then the tube could conduct better one way than the other, and that would cause some
60 Hz component to the flicker. I can definitely see that. It isn't really pronounced, but I can tell it is there, when moving my eyes rapidly or something. Some people complain that fluorescent flicker gives them migraines.Jon
I thought CFLs upped the frequency to many kiloHertz? A friend got one many years ago and their TV remote stopped working. I figured it was because they chopped the AC to a higher rate for a smaller inductor.
Filaments? You mean LED strings? I see the dots from auto taillights when I move my head. Most people I tell about this think I'm nuts. I've never seen anything flicker with LED Lights, I guess they are DC.
Most modern CFLs with electronic ballasts do, and they have a cap that at least reduces the 120 Hz component. Although the cap is so small to optimize power factor that they still flicker.
But, many early CFLs had magnetic ballasts. I have a number of these around here. Some were made using the 9" circular fluorescents, and the tubes could be replaced, which I have done.
These custom fixtures on my milling machine used "stick" style CFLs, just a
3/8" glass tube folded in half, and socketed into the ballast. I separated the ballast from the socket and enclosed the CFL lamp in a plastic tube so it would be protected from impact, water and chips. I built this at least 15 years ago, and they still work fine. With the magnetic ballast, they do flicker, and the strobe effect is very visible on spinning tools.No, old-style fluorescents have filaments on each end. If the filament's emmissivity is not equal, you could get a 60 Hz component to the flicker as well as a 120 Hz component.
Yes, I see a LOT of cars use PWM on tail lights. I think they use the same LED for turn, brake and night illumination, and use the PWM for the dimmer illumination function.
I have seen some really CHEAP LED lights that have a capacitor, bridge rectifier and the LED string. These would flicker at 120 Hz. I hope all the newer stuff has some current regulator components. Well, I got a few LED bulbs at the store recently, when the first one fails, I'll have to see what's inside. (Might be waiting a while...)
Jon
but only close to end of life, and it's still swamped by the 120Hz. So not a risk. And no-one has been installing mains frequency iron ballasts for decades.
NT
Just for fun I aimed a fast photodiode amplifier circuit at the fluorescent lamp on the bench (with electronic ballast) and it "flickers" at about
19 KHz. Ordinary magnetic ballast fluorescents flickers at 50 Hz (they almost estinguish at each zero crossing, probably there's some phosphor persistence because light never goes to really zero level). Old tungsten lamps also flickers at 50 Hz but thermal histeresis of the filament gives much less light level ripple than fluorescent lamp with magnetic ballast. I've never checked an ordinary CFL but I guess it's powered with some KHz inverter too and I expect a similar light ripple as the electronic ballasts.Frank IZ8DWF
Just for fun I aimed a fast photodiode amplifier circuit at the fluorescent lamp on the bench (with electronic ballast) and it "flickers" at about
19 KHz. Ordinary magnetic ballast fluorescents flickers at 100 Hz (they almost estinguish at each zero crossing, probably there's some phosphor persistence because light never goes to really zero level). Old tungsten lamps also flickers at 100 Hz but thermal histeresis of the filament gives much less light level ripple than fluorescent lamp with magnetic ballast. I've never checked an ordinary CFL but I guess it's powered with some KHz inverter too and I expect a similar light ripple as the electronic ballasts.Frank IZ8DWF
I'm a bit late to the thread... James, how hot did the instantfit led get?
George H.
Oh, and no electrostatic "crud" from the high frequency ballasts.
George H.
The ones in my office "go" at ~25kHz, I know because I can see the electrostatic pick-up on high input impedance circuits.
George H.
Yup. most fluorescents produce a lot of wideband RF, maybe one strong peak, too. We work with sensitive stuff all the time, and if we pull a cover off the housing, we can see it.
We have an AM/FM broadcast site about 5 blocks away, too, and they come through quite nicely.
Jon
I was picking up Bollywood stuff on the spectrum analyzer- apparently there's a Hindi AM station a km or two away.
--sp
I don't remember. Not very, to the touch, but they could be cooking deep inside that plastic tube.
Maybe I'll set one up and shoot it with the IR thermometer. But not soon-- too busy.
Cheers, James Arthur
Have something to add? Share your thoughts — no account required.
Ask the community — no account required