Note that if you want to compare different combinations of blue and
red, you'll have to compensate for different amounts of energy that
actually reaches the plants' surface, since most wavelengths can be
used to some extent, and some phosphors are more efficient than others
in converting UV to the desired wavelength. If the color is produced
by filtering through a colored coating, the difference will be even
more extreme. So a light meter, at least, will be important for
your experiments. You don't want to conclude that one color
combination is better than another when what's actually happening is
that the better combinations are just providing significantly more
total energy than the others.
Good luck, and let us know what you find out!
Many thanks for this in-depth answer.
My plan is to make my own PCB's with series of blue and red LED's and
eventually powering it up with a solar panel. Just for the fun of it. I'll
start with a normel power supply, though.
Chinese LED's are really cheap nowadays. Production of the PCB's costs me
far more than the LED's. Though I still have concerns about the wavelength.
Some suppliers are not able to give specific figures and they exaggerate the
amount of light they produce! If they say 14.000 mcd, you better divide that
by two. But still, at a price of about $0,07 a piece postage included it's
not bad. Suggestions for even cheaper suppliers are welcome!
What I would need, is a meter to measure the spectrum of the LED's. I saw
one for $199,- but will need to do some further research to be sure that
this is the right one for me.
http://www.specmeters.com/Light_Mete...ght_Meter.html
But I think that as long as I use blue/red LED's, I don't need to be
concerned about the effects of filtering of the light, because there is no
phosphor used in the pure colors.
I suppose I would need at least 1000-2000 LED's to start with, quite
acceptable if the experiments goes well, and I can use the panels for
50.000-100.000 hours.
Thanks again, and I think I should make a website for this experiment. If
so, I post a message on this list!
Ron