White LED theory question

[jpeg]

All the white LED's used in the flashlights I see commercially available today are all blue with the phosphor coating, which significantly reduces the brightness of the LED.

It would make sense to me to instead have the three primary colors on a die, and ocillate through them above 60Hz, which would (I think) give the illusion of white light without the intensity loss. Are there any heads doing this now? If not, why? Just curious...
-J

[Alaric Darconville]

The oscillating circuit would have a loss of its own-- and the diodes would have to be very closely matched so they produce as close to white as possible.

Now, all three diodes were lit at once, and the intensity was balanced just right, it would seem that that sort of white LED would be more efficient -- just very expensive to make and do quality control on...

[The_LED_Museum]

There is a very good dissertation on this subject at http://members.misty.com/don/ledrgb2w.html if you're interested.

[WildRice]

its a catch 22. If you want to use the LED to throw the light then you will have 3 different focal spots, overlaping in some places.

if you re-focus (grind down) the LED using a reflector, the 3 dice are still in different locations. you get different color intensities at different parts of the reflector.

Now, for general light, it can be done ie milky white LED, or using a neutral density filter, but at the cost of intensity.
Jeff

[PhotonWrangler]

I have a couple of surface-mount RGB LEDs with a milky white encapsulant. Even with the milky diffused coating, I can clearly see the individual chips when they're lit. Thus they would still produce some amount of spillover of the individual colors if used for general illumination.

[jpeg]

Oh, good Lord, Telephony, not ANOTHER name change??? [img]/ubbthreads/images/graemlins/yellowlaugh.gif[/img]
Thanks for the article... It looks like he's mostly comparing the balance of power to the red, and swapping blue LED's to see what works best... I don't think he ocillated the colors though. I heard the ARC4 pulsed like that, and if you waved it fast enough you could see the flicker. (I can't see it though) If the ARC4 really is using a microchip to maximize battery life by pulsing the LED, I would figure it wouldn't be that hard to pulse the three colors in sequence...
I assume of course, there would be some minimal color separation on the edges of things, but I thought it would make a neat experiment for a super bright/efficient white, even if it looked like a bad acid trip if you waved it around too fast. [img]/ubbthreads/images/graemlins/tongue.gif[/img]

Alaric, yeah QC would be a bitch unless one could hook each die to a separate adjuster, and then manually tweak the brightness...
Tell me THAT wouldn't be a cool flashlight...
Three sliders instead of an on/off switch... I want one... [img]/ubbthreads/images/graemlins/smile.gif[/img]

[kakster]

Dont the different colour LED dies degrade at a different rate too? That would mean the RGB white LED would colourshift after a while.

[PhotonWrangler]

[ QUOTE ]
kakster said:
Dont the different colour LED dies degrade at a different rate too? That would mean the RGB white LED would colourshift after a while.

[/ QUOTE ]
Very true. The LED-based signs correct for this with a built-in aging curve that's stored in flash memory, so it varies the PWM duty cycle and/or the current drive to the individual colors to keep the color temperature constant over it's life. A little messy, but an example of what an
RGB illumination source would need. Or maybe an RGB color sensing photocell triad that can sample a portion of the light output and adjust accordingly.

Or three sliders. [img]/ubbthreads/images/graemlins/grin.gif[/img]