Im thinking these are "ideal" ratios with these "ideal" wavelengths:
Vegetation Flowering
455-25%---455-10%
465-35%---465-10%
505-5%----530-5%
590-5%----590-5%
625-5%----625-5%
635-5%----635-20%
645-5%----645-20%
660-10%---660-20%
735-5%----735-5%
2:1 B:R---7:2 R:B
Ok so 135 PAR watts/m^2, which is equal to 600µmol/s/m^2 or 36000 lux (lm/m^2)(this is based off a HPS light though), is very sufficient for lighting a grow area. My preliminary test unit is going to be a hollowed out computer shell. The dimensions are .35m long x .177m wide x .35m tall. my total grow area=.06m^2. I want to have at least 400 µmol/s/m^2 for 1 plant (for now). I am planning on getting lowryder seeds since my grow height is so short. Since I am not looking for yield so much, just differences
in plant growth with different variables, the lowryder will be perfect to experiment with. My first grow will be a 6'x4' area. I was thinking about 12 or so blueberry and white widow plants will be nice. Before I spend $4000 on that grow I want to test
and find the best rations and colors to use and such. That is what this is. So 1 plant, lowryder, wont need a crazy amount of light since its so small....but who cares im gonna blast it with LEDs!!!! Here is a forum with a guy who used 100W of 5mm LEDs to grow a lowryder plant:
50w LED grow - Lowryder 2 - Page 3 - Grasscity.com Forums.
If he got 7g from 100W of 5mm LEDs with only 2 waveleengths and all the problems he had during the grow...I cant imagine how high-power LEDs and CO2 and hydro are going to effect the growth (well hopefully in a good way!!!)
And since I am using lowryder this grow I skip the vegetative phase- so I am only going to use the flowering wavelengths (which will make this alot cheaper)(also can ditch the 735nm
since there is no need for the change in light time [12:12] since it needs 18-24 hrs a day):
455-10%
465-10%
505-5%
590-5%
625-10%
635-20%
645-20%
660-20%
7:2 R:B
I want to have 135 PAR Watts/m^2 (or 600umol/s/m^2) for my growing area. But since this is a single plant and my grow area is not 1 m^2, but .06m^2, I cannot simply say 135x.06 for calculating the light needed as this would be an underestimate (since the plant would normally need more than .06m^2 for grow). I say for 1 lowryder that 50 PAR Watts (or 222 umol/s) is more than enough to provide optimum lighting for 1 lowryder plant (good-bad?? dont know for sure about this amount).
Here are the LEDs broken down to PAR (I hope this is correct):
-color---manuf.---wattage-output-PAR--umol/s/m^2
455nm-LuxeonRebel--2.5W---13lm--.4W-----1.5
465nm-LedEngin-----15W----210lm--4W-----15
505nm-LuxeonRebel--2.5W---80lm--.25W----1
590nm-LuxeonRebel--2.5W---65lm--.125W---.7
625nm-LedEngin-----10W----320lm--1.5W---7.8
626nm-LuxeonRebel--2.5W---85lm--.45W----2.4
635nm-LedEngin-----10W----320lm--2.13W--11.3
645nm-Epitex-------4.5W---12lm--.12W----.7
660nm-LedEngin-----5W-----17lm--.4W-----2.2
735nm-Epitex-------7.8W---11lm--?4W?----???(not to sure about this because 735nm is outside the PAR range[well from most sources at least, i did see one with 350nm-750nm for PAR], the normal PAR range is 400nm-700nm).
These are the LEDs I will buy for my flowering array:
color--part#--price-quantity-total lumens-total PAR-total umol/s
455nm-(2)LXML-PR01-0225--$3.24----31-------403----------12.4-------46.5
465nm-(1)897-LZ400B215---$35.65---2--------420----------8----------30
505nm-(2)LXML-PE01-0040--$3.60----3--------240----------.75--------3
590nm-(2)LXML-PL01-0030--$2.67----3--------195----------.375-------2.1
625nm-(1)897-LZ420R110---$27.20---1--------320----------1.5--------7.8
626nm-(2)LMXL-PD01-0040--$2.53----15-------1275---------6.75-------36
635nm-(1)897-LZ400R110---$24.00---4--------1280---------8.5--------45.2
645nm-(3)L645-66-60------$3.00----14-------168----------1.68-------9.8
660nm-(1)897-LZ110R205---$11.05---10-------170----------4----------22
subtotal:$476.92--83-------4471---------44-------202.4
Actual ratio:
455-9%
465-9%
505-5%
590-5%
625-35%
635-28%
645-4%
660-5%
7.8:2 R:B
The ratios are a little off of the "ideal" ratio, but since I do not want to spend $600 to get 1000lm of 660nm light, I have adjusted them to make them more cost-efficeint. Eventually
I will gett more 660 and 645 to make the ratios more red (more like 30% 660 and 20% 645nm). Its jsut the 660nm and 645nm have lower outputs (much lower in fact) so I would have to get
20 of them to just get the same amount of lumens as 1 625nm LED.
The total input power needed for the LEDs is 321 watts. And hopefully pulsing LEDs will prove successful so then I will only use 32.1 watts at a 10% duty cycle. If that works...only 33W
for 1 plant that is receiving 1/3 the PAR that a 600W HPS emits, sounds like a dream..all I can do now is pray that works.
Some nice websites for converting photometric units:
Designing A Simple Lighting Layout - Venture Lighting - Lighting A Better World http://www.kitchenculturekit.com/lighttable.PDF
Ok I looked for a good 4 hours for 440nm or somewhere close. Epitex offers 420, 430, 450, and 470nm wavelengths. The only problem is that the 420 and 430nm have only 14mW of power
at 8W, way way way way too inefficient for us to use (I refuse to waste 8W on only 14mW of radiated power). The 450 and 470nm have higher outputs (40mW and 120mW), but still that is too innefficient for 8W power consumption. Luxeon has a 455nm rebel LED. I am going to use a few of these to help with chloryphyll a absorption since it has higher ansorption at 455nm than 465nm (chloryphyll a has about 7% absorption at 465nm, but 40% at 455nm). But the 10W 465nm LED from LedEngin will be the main powerhouse for blue light production. As soon as a 440nm (or even 420nm if there is enough output) or something close to that is availabe, I will buy some. I feel like I am cheating everybody here by leaving out the some blue wavelengths
(like 420nm and 440nm), but I cannot continue waiting for those colors to start my testing. I will continue searching and as soon as I can I will buy them and put them into testing. I believe
the reason there are not any 440nm is because that is the peak wavelength, and most efficient, to cause damage to eyes. I read this somewhere off wikipedia, for got which article though.
I changed the green wavelength from 530nm to 505nm. Cannibis has overall absorption of 30% at 530nm, and about
50% at 505nm. The 505nm is more of a cyan than a green, but will still serve as the "green" light (by that I mean light that is reflected by the plant since there is less absorption in
that spectrum so the light contiues to reflect and recycle itself). On the cannabis absorption graph I did notice that the colors are shifted to the left. By this I mean the graph shows green being right about 500nm (should be about 530nm), which also makes the red area look less red than it should be (explains your question you had a while ago SnS, about why your lights dont look orange but 635nm in the graph looks orangish). No big deal I just wanted to point that out.
Opie what do you think about using cyan (505nm) instead of a more green (530nm) color? There is still alot of reflection at 505nm so it will still continued to be reflected and absorbed
again and reflected again.......recycling itself basically (just like 530nm does, but not as much).
I am still trying to decide between the 625 and 626nm LEDs. The 625nm is the 10W from LedEngin, and the 626nm is from the 2.5W from Luxeon Rebel. 4 of the luxeon rebels will produce 1800mW and 340lm at 10W. 1 of the LedEngin will produce 350lm and 1500mW at 10W. So the luxeon rebel is only slightly more efficient (only 1 lm/W to be exact), so I do not compare between these two from that. One advantage of the luxeon rebel is they are alot cheaper than teh LedEngin. It takes 4 of the rebels to have MORE output that the LedEngin. It would take $10.68 for 1800mW (340lm) from the luxeon rebels, but $27.20 for 1500mW (350lm). So for $30 I could have 5400mW from luxeon rebels, or 1500mW from LedEngin. That is a big factor, however there is also a con to the rebels. Since they produce less light, I will have to have more of them to make the same light, meaning less heat-sink space available (or it could mean making a larger array). I think I will buy both, and decide which one to use later.
The luxeon rebel are definately alot better for the money, but the LedEngin makes it easier to hook-up and wire (and also drive!!)
You may wonder why I have 625,635,645,and 660nm for the red wavelengths. Well this is because peak absorption for chloryphll a is 667 and 642 for chloryphyll b. Overall the peaks are 640nm and 665nm. But since I can't find and 640nm LEDs, I hope that 635 and 645 will cover that absorption area enough. Overall cannabis has about 85% absorption at 640 and about 93% at 665nm. At 635nm there is about 77% absorption, 83% at 645nm, and about 94% absorption at 660nm. So I hope that by using 635,645,and 660nm there will be about the same absorption as 640nm and 665nm alone. The 625nm is for phycocyanin, a pigment that recycles "wrong" colored light into the more efficient 660nm light (not neccesary but everything will help). The 590nm wavelength is for phycoerythrin, which has the same function as phycocyanin.
DRIVING LED's:
This was my plan for driving the LEDs....Using Picaxe microcontrollers to control the current going to the LEDs. Each Picaxe microcontroller can control soooooooo many things that I plan to use one to control my grow (take input readings from humidity,temperature,CO2,ect, and turn on fans, whatever I hook up with it to control the temp/humidity/CO2). Anyways a Picaxe 28X2 can have up to 16 output channels. Hooked up to the output channels are the strings of LEDs. I am going to connect a computer power supply to the controller to use that power to power the LEDs. Then with the Picaxe I can control PWM for dimming, I can control pulsing, and also put the lights on timers and calenders to turn on and off when I want them to (turn off at night, ect) I can run about 3-4 LEDs per series depending on the forward voltage, and since there are 16
channels I could run about 64 LEDs from one controller (but of course I am not going to work that microcontroller that much, I will run say 20 from it and just have multiple microcontrollers.
Ok about testing....here is what I had planned for experimenting with LEDs(later though..after this first test):
Plant #1: control plant. HPS light (100-250W).
Plant #2: 465nm and 660nm LEDs
Plant #3: 455,465,645,and 660nm LEDs
Plant #4: 455,465,505,590,625,635,645,660,and 735nm LEDs
Plant #5: 455,465,505,590,625,635,645,660,and 735nm LEDs-pulsed at 1% duty cycle (100us width)
*when I say Plant #, I mean plant group since there will be 2-3 plants per group.
*all plants are planted at same time
I think this should be enough experiment groups as this will let us know:
1) difference between HPS and LED growth
2) effect of additional peak wavelengths (455 and 645, hopefully 440 and 640 later)
3) effect of green (cyan,505nm) and amber (590nm) wavelengths
4) effects of pulsing LEDs (could prove to be VERY worthwile to pulse LEDs)
5) time differences (maybe LEDs grow plants faster...)
6) overall yield-HPS vs. LED
7) g/kW/hr yield-HPS vs. LED