Then, with the controller logic removed, there are questions about
pulsed supplies and so on, but moreso initially about component values
in the panels and nodes, where a node is basically a single large
high-output LED in a case of some form with perhaps a resistor to
ballast it to the node lines, or something.

"DO NOT PLUG THE GROW-TRON POWER SUPPLY JACK PLUGS INTO THE MEDIA CARD
PORTS OF A DEVICE OR ANYWHERE ELSE EXCEPT AS DESIGNATED. INSERT
GROW-TRON POWER SUPPLY JACK PLUGS ONLY INTO COMPATIBLE JACKS ON
GROW-TRON PRODUCTS. GROW-TRON POWER SUPPLY PLUGS CONTAIN DANGEROUS
VOLTAGES AND AMPERAGES. WARNING: CLASS 1 LASER DEVICE, DO NOT STARE
INTO LED EMITTERS. USE EYE PROTECTION AROUND LED EMITTERS. WARNING:
FLUORESCENT BULBS EMIT ULTRAVIOLET RADIATION. USE EYE AND SKIN
PROTECTION AROUND FLUORESCENT BULBS. INTERNATIONAL SPACE STATION NOT
INCLUDED. WATER IS TOXIC IN LARGE QUANTITIES. MADE in USA. ALL
COMPONENTS MADE in USA, except some nice pieces from JAPAN."

So, that levity aside as I'm still an idiot about the LEDs, hmm... Try
and keep the ballast with the power supply for centralized, or
concentrated, cooling.

With the form factor of the power modules, they should basically stack,
to form a compact block. Then, there's the problem of wires coming out
of them. It might be possible to consider having a hub of some sorts,
with having precut and identified jacks or simply fitted connecting
plugs, and then they multiplex of a sort to some hub with thick wire,
and they then go to a distributor, to reduce the number of wires
running from the grow chamber to the power supply cabinet or rackmount.
That's all complicated. I'm still over at cavemen poke with stick.

While that may be so, it's good to consider how this could work very
well, and even having synchronizing control circuitry among the power
bricks so the controller can pulse them all how it wants, even a bus
and the SDK.

With these MAX5035, with the 24Vin and 12V out at 0.5A it say 94%
efficiency. That's good. They can drive the LED panels. A problem:
they're only a buck or two, but that's in 1000 lot quantities. Then
other parts of the high-efficiency LED driver as well cost money.

For powering the pump, consider a 555, that when it receives power goes
to on, and then after the time period shuts it off. That might be
usable with a potentiometer of sorts.

The idea is to have a simple, self-contained timer circuit. It seems I
can get some relays that will allow driving a power strip off a simple
timer, and then have the circuit for the pump only run for a brief
amount of time, compared to the 20/4 and 12/12 lights on.

Basically it is going to be the CF's first and then as it is
constructed the LED array(s). One 120VAC plug for the 3 CF ballasts,
and hopefully one 120VAC plug for the LED driver, and one 120VAC plug
for the pump timer relay. Then there are the fans, they might be on a
different type circuit. I want damn quiet fans that don't move much
air, damnit. Then for cooling they can blow over wicks for evaporative
cooling.

Then, hmm..., there is temperature to control. Hmm... one notion is
the waterbath to the aquarium heater, ... for heating, and then fan
over wick for cooling.

So, the CFLs, should I get warm spectrum or not? Hmm... Well, why
not. I want to to have symmetry of sorts. Hmm... another question
is: 50/50's or Actinic 03's?

In terms of cost, well this is quite a bit, but then again it's very
high-tech and also should be low-maintenance on the plant, with a long
lifetime, and it will be interesting to build the design. The parts
list is as above. As well, it will just be an experimental chamber for
a long time. Hmm... Basically the lights and reflection for chamber
preparation are to begin, with a rack and so on and so forth. Hmm...
again the adjustable shelves thing is good.

If the rack is raised and lowered, then the reservoir sits on the
bottom at some point, then the pump would need a couple feet of head.
Again the fill and drain is unclear, but just having a smaller diameter
drain should be OK, then I want the pump to stop if the reservoir is
empty, to maintain prime and not waste electricity.

So, preparation would involve a wash-down of the tank, maybe vinegar or
something stronger to really clear it out, all-natural and so on, and
nothing that would mess up the seals. Hmm... then get the reflective
coating and make a shower pan fo the bottom and some black plastic
underneath it, the lights should already be there and then test the
mylar for emissions through and so forth. Also test out the lights for
heating and so on. Then the mylar or other reflective coating is set
up, inside the thing. I don't want to spraypaint or permanently coat
the glass, it's a fishtank. Also, I don't want humidity in the
plastic. Hmm... dessicant? Nah. How to fasten the plastic? Glue
bits? That might work pretty good. I could find some water-soluble
glue, ie Elmer's, the plastic is light and so forth and not subject to
much strain, and they should tear out leaving removable pieces for
cleaning them up.

That should be easy to set up, i.e., without designing circuits and so
on and so forth. That's just some mechanicals. Try and keep the
mechanicals/structurals neat, simple, clean. Materials are under
consideration in terms of use, price, availability, primarily
availability and price, machinability, yet something along the lines of
generic metal shelf raiils might work very well for most purposes. ..

The roofdecks and various shelving and so on, ....

Well, maybe in a couple months I'll get into this. For now, that is
all. Well, it's interesting.

Driving the cooling and ventilation fans with variable speed seems to
be the same problem as the constant current for the LEDs, in that when
the fan's DC motor is started it initially has no resistant, so it
starts hard. Then, to current control the fan would be a good thing.
Again, efficiency is a major goal. Having a constant current fan
controller off of the same regulated voltage that the other constant
current devices are receiving would be a good thing, as the fan would
be operating in a similar range. As well, designing the constant
current with the constant voltage type things with the dimmer leads to
the notion of pulsing control, and sharing the pulsing around all the
various panel and node drivers that are running at the same frequency,
and that each would have their own little hardware component to save
implementing thatmonolithically within the power voltage regular, from
basically wall current, but conceivably from 12V or 24V power, but
basically universally as the point is that there will be a voltage
regulated 24V supply of some kind for as many amps as all the module
circuits would draw and so on and so forth, towards keeping it under
hundreds of watts definitely, and preferably with no output modules on
approaching a very high power factor. The constant current circuits
seem to be drawing some few percent of the amperage they regulate.

So, the LED panel controller. The idea of getting an LED die and
diffusing it, because there are positive temperatore control notions,
is a good one. Boy the miracle technology.

The optically regulated LED, the frequency goes into a photodetector on
the same die for analog regulation discs. That's optically small form
factor optical junctions to discs, and discretized analog data into the
pixels, simple as digitizing. The wires, they have separate optical
fibers for the feedback over the device. Why optical? No electronic
interference. What have the LED and photodiode on the same IC, in
channels on top of each other. The frequency of the wavelength changes
with the current. Feed the current right back into the photodiode.
Then, it's a light wavelength. Band gap.

Then, the high precision diode device can have a feedback based on the
light curve, i.e, 3-D chromaticity? Also it is about polar signals.
The connector isn't two wires, it's polished and high precision, they
stick together, for as far as I care they can be magnets. That's the
device, the mirrors and distributed over the surface. The bandgap,
though, it would seem to be in-crystalline, glueing the ICs together
with clear plastics.

The plastics, they are insulators, here they are not dielectric, it's a
plastic hybrid semiconductor.

Photonic band gap Wave Guides on a chip. Light, bouncing off of
something, in reflection. Why? want I want is a high frequency timing
source. The diode, i actually converts an electrical signal to light,
of a wide freqeucny I guess. It bends, around the mathematical field,
it reflects it's light, it reflects off mass, it must do something
about electromagnetic radiation in light emitting diodes.

The crystal, it seems to be less thermally conductive that the
amorphous material. Yet, what about metallic ceramic? Well that is
semiconductor itself. The idea for a robust power laser diode is to
use very gross methods. For example, analysis of obsidian, mica,
similar crystals/amorphous materials, eg the buckyball, graphite and
carbon diamond.

Doped graphite.

Building in the thermocouples to the integrated circuit. It's a layer,
it 's generally hoped to be insulator, electrically, but not thermally?
Wait, whuh? Oh, they go out to thermo-pins in the DIP socket, with
different materials for the heat pipes. Yeah, heat pipes. A problem
might be the heat ones obviously radiate, but the cold ones gets
electrical deal there. The hot ones get more resistance, and the cold
ones less. Wait, vice versa.

uk.rec.gardening
rec.gardens

Plug it in! Plug it in!

Grow SILICON. An open book test: carbon, silicon, ___.

That represents about five days of research.

Ross