[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[APD] Idle Thoughts on LED Illumination

Everyone is doing, while I just sit and think. :-)

White LEDs are somewhat more expensive and probably not as efficient or versatile as RGB colored LEDs.

Quick e-bay survey shows bright whites at $0.23 ea in 1000 quantity plus shipping, while red, green and blues (high intensity) run at or below $0.20 ea *including* shipping at only 50 quantities from Hong Kong.

Only apparent advantage of whites is that you can pack more in a smaller space than three colors. Disadvantage is the lack of ability to control the exact "whiteness" (aka color temp.). [May be OK if you really like to look at 8600K (bluish) lamps.

The suppliers all expect you to waste power by converting 12Vdc to 20mA with a resistor. Since 9V will be dropped in the resistor and only 3 in the diode, that seems to be wasting about 3/4 of your input power. [Starting with 3X as much heat as you really need!] You possibly cannot operate with diodes in series to get around that, as their Voltage drop at a given current isn't predictable enough to keep current within safe bounds.

The safer solution may be to use a constant-current supply (aka battery charger) if the diodes are pretty well matched. You can then use series/parallel combinations without risking overdriving them.

Since the band gap, hence forward Voltage drop isn't the same for red, green and blue diodes, you may need three independent power sources for the three colors. At least that lets you tune for a pleasing color. :-)

Otherwise, I'd start by trying to put groups of three in series of RGB and parallel those up to see if the constant current through each set is close enough to not burn some out. A standard car battery charger might be perfect, as you can often set them at 3A or 6A as needed.

Wish they would spec LEDs with mW of output. Candelas are utterly useless for our purposes and impossible to easily convert to PAR. Note that some blue LEDs are around 470nm, which is bright in candelas and photosynthetically useful, while some are relatively dim, in candelas, at 400-440 nm and plant-useless. Be careful, here.

Red LEDs are readily available at a useful 670-630 nm. All LEDs put out a pretty wide spectrum (compared to a laser) so the photosynthesis isn't a problem if you just get close. That is, they aren't such a narrow line source that they won't activate the chlorophyl a and b spectra until you get below about 440 nm or well above 700 nm. [Those parts of the spectrum aren't visible enough to be useful, either.]

Grinding off and polishing or immersing have been suggested as ways to get the narrow beams spread out more. A third, fairly easy, fix would be to pot the lenses completely in a layer of clear GE silicone potting compound. You would lose about 5% of the light at the silicone air surface, due to reflection and again at the air-water boundary. Final efficiency would be about 90% (95% of 95%) of what immersion would give. Not too bad.

Enough idle thoughts for now. Asbestos britches going <ON> in case I got some of the above wrong. ;-)


Wright Huntley -- 760 872-3995 -- Rt. 001 Box K36, Bishop CA 93514

    "...there are only a limited number of things that government
 can do more effectively than individuals or other organizations
 can do."
     -- T. Sowell

_______________________________________________ Aquatic-Plants mailing list Aquatic-Plants at actwin_com http://www.actwin.com/mailman/listinfo.cgi/aquatic-plants