FlytrapCare Carnivorous Plant Forums

Hey everyone,

I couldnt find a topic discussing what i wanted so here goes, i want to make my own LED grow light.

And i want to tailor it to vegetative growth, i can always make a full flowering light later, and i will try to find the following wavelengths:
Violet/blue: 410-435
Blue: 445-455
Red: 630-650
Red: 655-670
Orange: ~485

And maybe a few white leds to supplement the above, since it is not known whether there are specific wavelengths that are used for certain biochemical proceses.
I'm still unsure of which kind of leds to use, its one of the following:
5mm, 1w or 3w

I want the total to be less them 50 watt, since it will also be a test for the plants.
The 5mm leds:
Pros: are cheap and a lot (and i mean a freaking lot) of wavelengths are available, there cheap, no cooling needed
Cons: Not a high light output, so has to be placed close to the plants.

The 1w leds:
Pros: dont get as hot as 3w leds, so only a heatsink needed, higher light output.
Cons: do need a heatsink, less options available wavelength wise.

The 3w leds:
Pros: high light output
Cons: not really cheap, a lot of heat, so a big heatsink needed (maybe even a fan), limited wavelengths.

So, from my point of view, the 5mm leds are the best option, but the downside is they have to be placed close to the plant, i was planning on 3 inch.
For the ratio i was thinking: 1 white, 1 orange, 4 blue and 4 red.

Now, i might miss some stuff, and could only find info on led growlight by ppl that grow the stuff my country if famous for, and therefor geared towards flowering.

What would you choose, and why?
And did i miss anything?

If you haven't read...

Look up Patent US6921182 on google. Talks quite a bit about it.

A good place to look is aquarium forums. They're BIG on lights for plant growth and I know I've seen several build threads for LED grow light set-ups.

Hey Cyph3r, ya, ive browsed through it, but its a big amount of text (rather huge actually).

@parker679, do u have any links that u could pass? I only found canabis grower experiences.

My main forum for aquariums is plantedtank.net, I would poke around in the lighting section. I figure if it can grow aquatic plants through 12-24 inches of water it should be able to grow some CP's from a few inches away.

Just briefly poking around I found this link.

http://www.plantedtank.net/forums/showt ... p?t=195589

Im now in the process of seeying wether i can find the leds i want for a resonable price, looking at all options, so far Violet seems to be the hardest, or lets say, most expensive 410-435 nm)

Ok, done looking, decided to go for the 1w ones, found the following leds:
Colour Nm angle Forward voltage current(mA)
Red 660 120 2,5-2,7 350
Purple 420-435 120 3,0-3,6 350
Blue 450-455 120 3,0-3,4 350
White 6500(k) 120 3,2-3,4 350
Red 620-630 120 3,2-3,6 350

So i will be making the decision as soon as i can find a driver, its gonna be in the 40 watts range.

Ideally i would use this ratio:
1:5:3:3:5
White, purple, blue, red(630), red(660), for a 40 watt installation, it would be about 50 dollars excl. driver
Reason for this ratio is it covers nearly the complete PAR spectrum and a pretty high coverage at the PAR absorption peaks, but also some white leds to supplement the other colors that might be of importance in biochemical processes other then photosynthesis.

Now my question is, would people be interested in buying such a lamp? (since buying in bulk is way cheaper and while im at it, i could make a few extra)

After the hours of research i've put into this, i recon the leds i will use will be more efficient per watt then some of the more common grow light (ufo for example)

Honestly, if this works out, what would be your expected cost of one of these? And aprox. dimensions?
I would be interested pending the above.

I would also like to see some of the work that went into choosing those bands, and that ratio. (I also have been doing mad amounts of research about LED grow lights. There is so much misinformation out there, it's good to see someone actually use actual evidence than just wild claims.)

Hey nimrod, this depends on the wattage, im estimating 50 dollars for a 40w panel, them i will need a driver and maybe heatsinks.

For those bands and ratios i used 2 things (mostly):
PAR spectrum:


And individual spectra of the 3 main photosynthesizing molecules: Chl a, Chl B, and Cart


And from here i went looking for the best matching leds, and the par spectrum picture also shows why the argument about leds having not enough lumens, u need to look at it from a plants point of view.

Now for this i assumed 2 things: CP do need and use the same PAR as "regular" plants, and CP's have a normal distribution found in "normal" plants.
I choose this ratio because the dominant wavelengths for the 3 main photosynthetic molecules are well represented, and white to supplement wavelengths might be needed for biochemical proceses.


as i quote:
"In diethyl ether, chlorophyll a has approximate absorbance maxima of 430 nm and 662 nm, while chlorophyll b has approximate maxima of 453 nm and 642 nm.[19][specify]
The absorption peaks of chlorophyll a are at 665 nm and 465 nm. Chlorophyll a fluoresces at 673 nm (maximum) and 726 nm. The peak molar absorption coefficient of chlorophyll a exceeds 105 M−1 cm−1, which is among the highest for small-molecule organic compounds.[citation needed]
In 90% acetone-water, the peak absorption wavelengths of chlorophyll a are 430 nm and 664 nm; peaks for chlorophyll b are 460 nm and 647 nm; peaks for chlorophyll c1 are 442 nm and 630 nm; peaks for chlorophyll c2 are 444 nm and 630 nm; peaks for chlorophyll d are 401 nm, 455 nm and 696 nm.[20]"

That sounds good. But my only concern is that by the time you add drivers, and fans and whatnot to your unit, you will be pushing the same price for some of the UFO and UFO-like units already available. Yours might be slightly better, but is it really worth saving $10?

Well, for one, i can then experiment with ratios.

Well, if i can learn something new, and experiment while doing it and saving money, yeah its absolutely worth it for me =]

Im thinking about powering all leds by 2 12v drivers(one for the 660nm and one for the others, since the 660nm ones are 2.2v, others are 3,2)

http://led.linear1.org/led.wiz
Its a nice tool that im using.

Yeah, I should have guessed. You got that scientist gene in ya. However, I remember reading that you also have the curse of the student: No $$$.

I am also proposing an experiment regarding LED lighting and carnivorous plants... I think there might be some room for your LED tinkering in my experiment. Mostly regarding LED ratios.
I agree that using 1W LEDs are definitely the way to go, but my concern is how many of them are required to grow a VFT? Ratio is of a secondary concern to me, as I am not an electrician, and worry about my safety and the safety of those in my apartment complex

However, I will be attempting to determine what wattage is required for VFT vigor. And I will also be looking to make direct comparisons between red-blue "grow" lights using the same wattage as a comparable "day white" LED.
9W vs 3W
6500K white vs red-blue
If everything goes well, I would love to play with the wavelength ratio and number of bands in the "grow" light. But with only 9 spots for LEDs, I can't make your ratio fit. Do you think if I sent you a 9W ceiling fixture you could swap out the white LEDs with different colored ones? I think it might be easier to play with the ratios on a small scale before trying a 50W light fixture, but that's just me. Let me know if you plan to actually build one of these, as I would be very interested in the outcome of your tinkering.

Do you think it would be possible to replace the white 3W diodes with red/blue/orange/violet diodes (3W)? That would be ideal.

edit: Whoops! Didn't add the picture.

This will depend, i think u need to remove the heatsink first, and then de-solder the diodes, then u will need to find the same dimension right wavelength diodes to replace it with, i think its doable, but to be sure, u have to screw it open.

Well I am going to get one for my "pure white" trail jar, and when I do, I'll screw it open and show you the insides.
Where do you get individual diodes? If I wanted to have this done locally by someone, where would be my best bet?


/I really really want to make this thing a 15W, 5band, grow light... Think of how cool that would be.