Phycoerythrin - strong absorption peak around 560nm. Phycobilins: not
in higher plants. Phycoerythrin's a phycobilin, and not found in
higher plants.

Iron, Manganese, Copper, Magnesium
Phosphorus, Potassium, Nitrogen

Carbon Dioxide
Water

Really considering the bubbling for the roots, to let soak, but that is
complex, and involves moving parts and so on, and still need to drain
that. Then again, the laser reflector/detector and motorized trusses
on the canopy with variable LED/canopy spacing is also having a variety
of moving parts.

It seems some of the enzymes are key, the idea is kind of a concurrency
situation. The photosystem II is releasing the oxygen, hrm...

"chlorophyll b efficiently absorbs light in 450 to 500nm"?

"The 26-kd subunit of light-harvesting complex II (LHC-II) is the most
abundant membrane protein in chloroplasts. This subunit binds seven
chlorophyll a molecules, six chlorophyll b molecules, and two
carotenoid molecules."

chl a 7
chl b 6
carotenoid 2

plastoquinone, plastocyanin, plastoquinone: a highly lipid soluble
electron carrier

The idea with the concurrency is to avoid backlogs, and to get the
right light wavelengths and concentrations so that the reaction centers
operate in greatest efficiency and harmony. The idea is to get the
Calvin cycle really running, dark or light-independent reactions.
Maybe then with the pulsing, the idea would be to find the right duty
cycle of pulsing for the various photosystems I and II, so that they
are activated and so on, and then to pump them.

Photosystem II --via-- plastoquinone -- electron to cytochrome b_6/f
complex
photosystem I --via-- plastocyanin -gets- electron from cytochrome
b_6/f --with-- ferrodoxin -- reduces NADP+ to NADPH.

noncyclic photophosphorylation

Then, Calvin cycle.

So, definitely focus on the chlorophyll a inputs. That is around 667
and 435. So, what, well, hmm... That means just drive the red leds and
run the actinics, but I'm worried a little bit about UV radiation and
promoting the carotenoids to help protect against UV damage.

Chlorophyll b.

Chl b is an accessory found in peripheral light-harvesting complexes,
LHC's. These complexes usually contain three xanthophyll, two luteins,
and one neoxanthin, and nearly equal amounts of chl a and chl b, 7/5 or
8/6 for the major LHC-II, with na a/b ratio of 1.4, bound to proteins
(LHCP's) that are encoded in the nuclear genome and imported into the
plastid after synthesis in the cytosol.

http://www.biomedcentral.com/1471-2229/1/2

So, it seems chl b has more to do with chloroplast assembly than
photosynthesis. So, it is still in much smaller quantities than chl a.

Now, emitting directly at 680 and 700 might be not worth it, even
though P680 and P700 absorb at those, because, the antennae will be
plenty busy feeding them.

Hydroponics

media

airstone vs. ebb/flow

basically the media is going to be soil type thing, in a seedling tray.
Then, those are transplanted into perlite or vermiculite or so, with
gravel. Then, daily, with the turn on of the lights, water with some
nutrients floods the pots, and then is allowed to quite drain.

pump circuit: have when power is sent that it latches and runs for
some time, then resets when the power is off.

275nm UV-B?

660nm LEDs

make a bunch of little pucks or panels with them. It's nice to
consider an enclosure, for example the plastic friction-fit boxes with
a hole drilled for the lead out and then sealed tight. There are heat
dissipation problems with that. One notion is to use thermal
conductive, not electrically conductive paste over the board under the
LEDs and around a heat sink plug that goes through the box. Then that
is sealed, basically with the idea of keeping mositure off of the box
and making them easy to clean, and the metal post that supports the
board is then outside the box as a heat sink. Something like a largish
rivet. So anyways then a 660nm panel needs a constant current driver,
to begin, towards later using pumping pulsing laser diode type inputs
or the rotating wheel type that distributes a pulse among the panels
lessening each's duty cycle but washing around back-EMF.

In that way as much as possible the efficiency and protective circuitry
would be off-loaded to the power supply. So, the panel would have a
square array of LEDs, series in parallel. Each series should have a
small resistor just to ballast a small bit. The maxim current
regulators are only around a dollar.

That involves getting a bunch of perfboard. Damn, that stuff's
expensive, need to find a radio shop.

Then, there are design questions about the voltage of the power supply.
Basically, around 12v would be good, but maybe 24 would be better. It
would be nice to find a high-efficiency power supply, instead of
bridge-waving with a filter network and then current regulator. To
maintain the current is one thing, then another. Another notion is to
have little current regulator for each panel, which is important
instead of having a monolithic power supply, because then, say, the
rectifier can be off a high frequency with a high frequency power
rectifier. In that way, the efficiency of the rectifier is increased,
as inductive coupling is better. What is that, with capacitors or
something? Jeez.

The idea then is to have the panel drivers go in parallel off of the
power supply, which provides, say, regulated voltage DC, er, damn.
What is a "switched mode power supply."

12VDC PFC universal, PFC is "power factor correction"

Wha... how many amps go through? Ohm's. The LED's are conductors.
They have forward bias. They dissipate. Damn. Probably a good idea
to look at the first lines of the spec sheet for the important
information as electrical components.

Ooh... this MAX5033/MAX5035 looks really cool, with current regulation
to .5 and 1.0A at various output voltages, including various output
voltages... and stock inductors, running off a variety of input
voltages... and valued! Hmm... 500mA / 20mA = 25 20mA LEDs... Those
seem great. If they run at 12 V, and then the LED has a typical 1.8
voltage, then, erm..., Ohm's law? What is voltage drop there? Does it
just go by P_d power dissipated? Power dissipated is 60mW, 12V x 1A =
12VA = 12000 mW? Divided by 60 is 200? That doesn't seem right
either. Running at 3V x 1A, 3000mW? Divided by 60 is 50? Maybe that's
better, but then there's only 3 volts? Damn that requires some more
learning.

Again the idea is to have color coded jack plugs that interconnect the
dumb LED panel with the power supply and controller, err... so then the
controller modules plug into the main power supply and then they each
run a line to an individual panel... unless the panels could be
connected in various ways and run into a hub of sorts, ie series in
parallel, ..., for various placement of the panels or nodes, with screw
attachments, double-sided tape, and so forth, in UL approved packages
for sale to the public.