In the horticulture research lab at the University of Florida. Would you like a tour?
But will it back up your claims ?
I saw your MCPCB you do not need metal core.
There is a better way.
How about a 4 foot by 4 foot PCB, with 10,000 µmol at near perfect uniformity across the 4x4 canopy, with a cost under $200?
Yes I know. But we need to have goals.
I have a philosophy where obstacles are all in the mind, and the impossible is only impossible until someone does it.
How about a virtual 4'x4', just cut out what is unnecessary?
How about three 16" PC Boards which end to end is 4 foot.
I found the most cost efficient PCB is 10.9" x 16.1". Add 0.1" to either dimension and the cost doubles. I can put 31 16" x 0.35" PCBs on one board with score lines.
0.35" wide and can be populated with 64 LEDs in 4 separate strings of 16, or 3 separate strings of 21. Both 16 blue 2.8V and 21 red 2.1V are around 43-45 Vf.
Will be powered with 48V.
A 48" heat sink with 6 boards attached, three on each side for a total of up to 384 LEDs.
Six of the circuit boards I can make for about $4. For six, $0.67 each.
I can buy 196 LEDs for under $150. Unless I can get another deal like this:
Now they need to be cooled. My heat sink is going to cost about $3 a foot in small quantity. So another $12.
Remember how you said you do not like the Olsen SSL 150? I love them. Why?
I can run them at inches over the canopy with near perfect uniformity.
Why 196 LEDs
Remember the Cree Horticulture Reference Design, the Gavita 1000W HPS killer? It had 196 LEDs.
So how much power is that?
SSL 150 flux at 350mA is 700mW Radiometric Flux for Blue
Flux at 350mA for Red is 500mW
A 100 each red blue would be 500 umol of photons.
But that means nothing. But that's the math, just the number of photons being emitted from 200 LEDs per second.
The Cree fixture was 400 umol at the canopy from 58 inches with 196 LEDs.
10x inverse square gain from 58" to 12"
So I'm looking at about 4000 umol for under $200.
And I want dimmers, that's another $12 for six of my new Buck Drivers.
I could use white too. (16 streaks from 16 LEDs)
I have a dual footprint design. Accepts Olsen SSL or any Cree XP including the efficient white XP-3G.
The blue square is SSL and the yellow Cree XP
notice how the thermal pad goes up to the heat sink screw hole.
The bottom blue row they are the 2.8v x16 x 4 spaced 6.35mm, the top red 2.1V x 21 x 3 are spaced 6.35mm. notice how the ones on the right is no longer even. 16" down the board they will line up again.
My current board is also dual footprint for Luxeon Rebel and Cree XP but wastes space. I love the Rebel Thermal design, it's perfect for my heat sink. But their Color C line is just too small for my eyes and it does not have the Rebel PCB footprint.
Max current is 1 Amp on the SSL 150 so I could boost the output about 2.5x. That would be 10,000 umol under $200.
If I spent another ten cents for the blue I can get 2 Amps. And still be under $200.
And the PCBs are spec'd with HR370 high temp rather then FR4 so I could cut the cost of the board but 6 for $4 is good enough.
With 18 years manufacturing experience, its under control.
Where can you buy them? Not likely going to sell them. Remember the solar test setup.
Do you know there are children that cannot study after school because their village has no electricity?
They have vitamin deficiencies. They need a solar powered vertical farm.
The Cree design was 550 watts. If I have 10,000 µmol I can dim them 10-20x saving electricity where it can easily be solar powered.
I want to give them away. Create a charity organization. Seriously.
Most likely I will make it open source. Put all the mechanical, electrical, and PCB drawings out there free.
Without goals nothing would get done.
Oh yeah, MCPCBs and thermal vias.
Would not a 100% copper path from the LED's thermal pad to the heat sink not be better than putting the heat sink on the opposite side of the board?
I found out that is true.
NOW THIS IS THE KEY See how the thermal pad goes to the screw hole?
See how the heatsink, the one attached to a copper water pipe, make a 100% copper path from pad to pipe?
These test results were the first prototype. Remember the red with the gap in my previous post? This is are the results.
That not bad I have made many improvements since. These results were from a dirty experiment, not a;ll the screws were in, just a shoddy setup.
I found it is important to screw them all down
And I do blue too.
I don't know what you guys have against magenta. This spider on the wall appears to like like it. But everything is not as it appears, see thumbnail.
Not as appears, it's on the ceiling not wall. I like magenta enough to use it in my office. I like spiders.
None proved to be better than common white LEDs.
Yeah, I think that has to do with thermal management. Red do not like any heat what so ever.
And I can do white if not.
.