What's new
TerraForums Venus Flytrap, Nepenthes, Drosera and more talk

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

DIY LED plant light project

so, i have had a bit of an interest in reefkeeping for quite a while now. i dont actually have my own tank, but i think about it all the time, and look at corals, clams, fis, etc. my brother is the one who originally got me interested in it, as he has a 40 gal soft coral tank (which i am currently taking care of while he is at college). it is becoming somewhat common for people to build DIY high output LED setups for their reef tanks. there are several sites which sell supplies for building these kinds of lighting systems, which i have been looking at from time to time. my brother has said that someday, he might want to upgrade to led, but he isnt very good with technical stuff, like soldering and putting stuff together, etc, where as i tend to have a knack for that kind of thing. i have also heard about LED plant lights from time to time on here. i have been thinking a lot lately about making some sort of grow rack, or indoor greenhouse setup in my basement. when i started thinking about lighting, my mind immediately went to LEDs. i figured that i could make something suitable for plants with the LEDs that these reef-lighting websites are selling, if i got the right spectrum. the most common colors used in DIY reef light setups are the blue and cool white LEDs. but, as i searched through them, i figured out that these sites sell more way colors than just those, to balance out those other two . green, cyan, red, amber, even UV/purple LEDS are available from them. such colors are used in fancy DIY setups in small amounts, often with each color/group of colors connected to its own timer/dimmer, in order to simulate sunrise/sunset, moonlight, etc. for the corals. some of these light colors would be great for plants, others, not so much. i selected LEDS with the best colors for plant growth, a heatsink, driver, etc on RapidLed.com.http://www.rapidled.com/ i decided that i want to make this with RapidLed's solderless LEDs, even though i know how to solder, just 'cause it would be easier(that, and i cant find that *bleep*in' solder iron! where could it have gotten off to?:crazy:) anyways, im not going to be ordering these anytime soon, as i am kindof broke at the moment, and this is about $110 total, but my b-day is just around the corner (nov. 26), and i also will work my but off doing chores around the house, mowing the lawn, etc., so i think that i can manage to get enough money!

once i have the cash, its LED time!:banana2:

as for the colors i picked, i am planning to get 12 LEDs, 8 warm white, and 4 red, on a heatsink predrilled for exactly 12 LEDs

heres how i plan on arranging them:

0 = 1 warm white LED

0 = 1 red LED


0 0 0 0 0 0 0 0 0 0 0 0



and here's all the parts (minus the solderless connectors and thermal grease)

the warm white LEDs
http://www.rapidled.com/solderless-cree-xp-g-warm-white-led/

the red LEDs
http://www.rapidled.com/solderless-cree-xp-e-red-led/

the heatsink
http://www.rapidled.com/drilled-tapped-1-4-x-24-aluminum-heatsink/

the driver
http://www.rapidled.com/mean-well-lpc-35-700-constant-current-driver/


any advice, suggestions for a better choice of color for the diodes, etc will be very much apriciated!:-O



and by the way, i know that there are other, cheaper (and to some, more effective) ways of getting light for your plants. i am aware that some of you consider LEDs to be a bad choice, due to their price. i am not planning on getting LEDs because they are the cheapest way to provide light for my plants, i'm am choosing them because of their efficiency (even though i am not the one paying the electric bill!), and because i think they will be the "next big" thing in horticulture, and are awesomely high tech. yes, i am choosing these lights cause i think they they are awesome. some of you will probably think that that is kind of dumb. well, its true, they are awesome! if i can get the right colors of diode, then they should provide some great lighting for my plants! so please no comments trying to convince me not to go LED, if i can get the money, i will most likely be getting these lights.
 
They are a lot of fun to play with and since you asked...
Just a quick look at your component selections....

1. The 700ma driver is less than half the maxium driver current for the LED's you've selected (Think of it as them being on half power)
2. The spectral combination is very weak in the blue wavelengths.... might want to bump up the white to more like 5000k and a higher CRI
3. Increasing the driver current will also likely require the use of active cooling (e.g., fan/s) and greater heat sink area. (probably the reason the driver is only half power)
4. You didn't mention what optic angle you are going with... and the distance between LED's (give some idea of blending)
5. What type of plants are you wanting to grow using these?
LED's provide a very concentrated light source and shadowing is a concern with high canopy type plants. They are much better suited to low canopy plants (Drosera, Cephalotus...etc)

Caveat, fun to play with.... but not yet a competitive option in our application IMHO
Be sure to read the factory tech sheets on LED cooling and follow their recommendations carefully.
 
Last edited:
Hey Sarracenia,
For the last month and a half I have been running 4 10w (6700K) LED's over my lowland Neps and have noticed that they like it more than the T5 HO's (4-39w). On my Highlands I have 4 15w Par38's (3700K) which replaced 6 54w T5 HO's, they don't seem to like it as much as I had them too close at the start which made them all sulk

Cheers,
Kase
 
should i go with the cool white or neuteral instead, or mabey a combination of colors? will runing these LEDs at half power still be enough light for the plants? if would not mind getting a fan or two if i amp up the power on the lights. oh, thats right, you need to get some sort of lens to get a better optical angle... weel, ill look into that, and let you know what i choose. any reccomendations? and low growing plants are exactly the sort of thing i was planning to grow under this, i was thinking pygmy/rosetted dews, cephs, pings, etc.

thanks for the advice Av8or!
 
i think you would be best with mostly cool white with some blue and red mixed in, go to a site that sells grow lights and see the spectrum used with metal halide and hps lights , there designed to grow the most herbs per watt so the picked the light colors used most by plants may colors just reflect off the plant
 
should i go with the cool white or neuteral instead, or mabey a combination of colors? will runing these LEDs at half power still be enough light for the plants? if would not mind getting a fan or two if i amp up the power on the lights. oh, thats right, you need to get some sort of lens to get a better optical angle... weel, ill look into that, and let you know what i choose. any reccomendations? and low growing plants are exactly the sort of thing i was planning to grow under this, i was thinking pygmy/rosetted dews, cephs, pings, etc.
thanks for the advice Av8or!

I like the neutral white if you are going to add more red.... the exact ratio of blue/red etc depends greatly on the specific plant. Some plants can grow just fine with no red or no blue, but most need some blend of the two (in addtion to other wavelengths)

As a rule of thumb 20% blue seems to be a good standard to shoot for... however, this is looking at the spectrum only though the eyes of photosynthesis performance. There is much more to it than that... and this is where the knowledge base becomes clouded.

Green for example, it was long though that green was pretty much useless but now it has been shown that green is important for canopy penetration. Other wavelengths prevent the sporation of botrytis (why sunlight seems to kill it) while another wavelength is needed for the sporation of beneficial trichoderma.

To cover your bases going with high CRI values is the safest bet. That way you closely mimic the performace of natural sunlight.... Not only will you supply the light that is more "natural" but the plants will "look" good too..... and are more easily diagnosed when a problem occurs

There are so many variable involved that no one can truly give you black and white answers.... IMHO mimic natural light as much as possible for long term performance.



What optics you use will depend on spacing between the LED's and the distance to plant.... you want the light to be "blended" when it arrives at the plant.

As far as "powerful" enough.... again it depends on distance to plant.... I prefer the drivers that are dimmable, that way you can crank them up to maximum output or turn them down as the need arises. I dont think the driver you selected is adjustable.... maybe it is, I didnt read that deep about it. The one's I used had an external potentiomenter connection.... it was the cats meow, you could have several LED's all with variable outputs depending on the needs of the plant or to compensate for less than ideal thermal mitigation.

Plant lighting is a very ford vs. chevy type subject.... ask 20 ppl and get 20 conflicting answers

My recommendations are based on my personal experience and research, nothing more.

I would recommend you start off with one or two trial setups before you invest much money. At least it will give you some ideal of best practices on spacing, power, spectrum etc.

and MAKE THEM RUN COOL!

HTH's
Av
 
I dont think the driver you selected is adjustable.... maybe it is, I didnt read that deep about it.

its not. ill look into getting an adjustable one, you make a good point, that it would be easier to compensate for the lights being closer, or further away from the plants

which one do you recommend?
http://www.rapidled.com/dimmable-drivers/


I would recommend you start off with one or two trial setups before you invest much money. At least it will give you some ideal of best practices on spacing, power, spectrum etc.

that might be a good idea, i was thinking that this would be a trial setup, and upgrade it later by making copies of it, and then putting them one behind the other, to make a larger grow area, or perhaps just sell the old heatsink to another grower/reefer, and buy a bigger one, like one of these:

http://www.rapidled.com/drilled-tapped-6-x-20-black-anodized-aluminum-heat-sink/

http://www.rapidled.com/drilled-and-tapped-4-25-x-23-heat-sink-for-36-leds/

do you think i should go smaller with my trial setup, like mabey using one of these (six LEDs):

http://www.rapidled.com/drilled-tapped-1-4-x-12-aluminum-heatsink/
 
SX,

The LED's you selected have a maximum current flow of 1500ma... I would look for an adjustable constant current driver that could at least provide 1000ma.
Im having a hard time finding the term "constand current" in any of the descriptions.... you might want to contact their tech question email and talk to them. I just dont know.

I would recommend that you also give Luxeon a look..
http://www.luxeonstar.com/

Av
 
  • #10
I use a professional LED grow light, which has a combination of 660nm red, 615nm orange, 455nm blue and 3700K warm white, in a ratio of 3:1:1:1.

You might want to reconsider the wavelength of the red LED, as 630nm is a little low on the chlorophyll absorption band: http://www.biologie.uni-hamburg.de/b-online/e24/3.htm. 660nm might be a better choice.

Whilst I can see certain benefits in choosing a high CRI white light, in that it should give a spectral distribution, one must consider that some of the light output will be of little or no use to the plant, as it is simply reflected, thus is wasted energy: http://www.botany.uwc.ac.za/ecotree/photosynthesis/spectrum.htm.
 
Last edited:
  • #11
True, but what part is not needed?.... and aesthetically, the plant will not look "normal" to the human eye otherwise.

If we start looking at the needs of accessory pigments the spectral distribution becomes much more complex
pigment.gif


In addition to the need of preventing botrytis sporation or facilitating trichoderma sporation...

and we are back to ford vs. chevy, harley vs. honda, M16 vs. AK47, excellent arguments from both sides.

no one ring to rule them all :)
 
Last edited:
  • #12
I've tried several white LED lamps and to be honest I have not had good growth results from any of them. This is not to say that certain white LEDs won't work, but I've yet to find which ones. As as example, I had my U. campbelliana under a Philips (Cree) LED lamp and growth rate was glacial. I later moved it under the aforementioned hybrid 3:1:1:1 lamp and the growth rate is very good. If you want white light then unless you are trying to achieve high wattage density you might be better off using fluorescent lamps, which use phosphors to produce white light in much the same way as white LEDs do, but using cheaper technology.

The 3:1:1:1 LED lamp I use has white light in it, so that helps with the aesthetics. Though there is still an obvious purple hue it is much less noticeable than if using traditional red:blue only grow lamps.
 
  • #13
If I was going to build one using today's technology....

I would use High CRI 4100-5000k Whites with supplimental red and blue as needed. I would also use dimmable driver circuits.
We can grow great looking plants with white fluorescent bulbs, I see no reason why LED's cant do same all things being equal...

I just dont think their price is competitive yet compared to T5's...
 
Last edited:
  • #14
One of the big advantages of LEDs over fluorescents is they give the grower more opportunity to pick particular wavelengths attuned to the plants requirements. Fluorescents indeed do produce good results, but there can be quite a lot of waste energy in terms of plants requirements. One of the 'tricks' that light manufacturers can use to make lamps look brighter is to boost the green output, which the human eye is most sensitive to - but not too useful for plants.
 
  • #15
True.... but you have to be careful in culling a certain wavelength. Research done by NASA for the International Space Station showed that green is important for canopy penetration. It also prevented the plant from dropping their lower leaves...

Their research also showed that what one plant needs can be very different than what another plant needs.... sometimes these differences were rather dramatic.
 
  • #17
Could debate this all night :-O

I'm not disagreeing that accessory wavelength are of use, but one must be careful that they do not become the primary ones when selecting a grow light. I will do a little more research regards the green wavelength, but the fact that most plants are green indicates that they reflect that colour.

A good compromise, which be both appear to agree on, is to use blue and red LEDs, with accessory wavelengths provided by white. I have a personal preference to 660nm reds over 630nm, as most of the chlorophyll absorption charts I have seen peak at 660/662nm.
 
  • #18
some of these look good. can i just use white instead of having to add red and blue, if i were to use a combination of cool,, neutral, and warm whites?

also, would i be able to attatch them using screws to the predrilled heatsinks from RapidLED, or are they a different size? the stars look the same as the ones on RapidLED

http://www.luxeonstar.com/ANSI-White-CoolBase-Rebel-Stars-s/348.htm

http://www.luxeonstar.com/ANSI-White-10mm-Rebel-Square-CoolBase-LEDs-s/394.htm

what about getting multiple colors of LED on a single star like these, for beter color blendng? are these the same size as a normal star, or are these bigger, to fit more diodes on them? i would have to get a lot of fans and cooling stuff for these, right? i am assuming that they need more cooling due to the fact that there are three diodes in close proximity to each other, so there is twice the heat, correct? i cold get a couple extra fans if needed.

http://www.luxeonstar.com/Predefined-CoolBase-LED-Assemblies-s/446.htm

http://www.luxeonstar.com/Any-3-Rebel-LEDs-Mounted-On-A-Tri-Star-CoolBase-p/sr-03-custom.htm
 
  • #19
Could debate this all night :-O

I'm not disagreeing that accessory wavelength are of use, but one must be careful that they do not become the primary ones when selecting a grow light. I will do a little more research regards the green wavelength, but the fact that most plants are green indicates that they reflect that colour.

A good compromise, which be both appear to agree on, is to use blue and red LEDs, with accessory wavelengths provided by white. I have a personal preference to 660nm reds over 630nm, as most of the chlorophyll absorption charts I have seen peak at 660/662nm.

I agree Carl, on both points......... and honestly, I doubt the green would matter much in sarra's application. However, for larger nep type applications where canopy penetration might be more critical.. :)

some of these look good. can i just use white instead of having to add red and blue, if i were to use a combination of cool,, neutral, and warm whites?

also, would i be able to attatch them using screws to the predrilled heatsinks from RapidLED, or are they a different size? the stars look the same as the ones on RapidLED

http://www.luxeonstar.com/ANSI-White-CoolBase-Rebel-Stars-s/348.htm

http://www.luxeonstar.com/ANSI-White-10mm-Rebel-Square-CoolBase-LEDs-s/394.htm

what about getting multiple colors of LED on a single star like these, for beter color blendng? are these the same size as a normal star, or are these bigger, to fit more diodes on them? i would have to get a lot of fans and cooling stuff for these, right? i am assuming that they need more cooling due to the fact that there are three diodes in close proximity to each other, so there is twice the heat, correct? i cold get a couple extra fans if needed.

http://www.luxeonstar.com/Predefined-CoolBase-LED-Assemblies-s/446.htm

http://www.luxeonstar.com/Any-3-Rebel-LEDs-Mounted-On-A-Tri-Star-CoolBase-p/sr-03-custom.htm

Sarra...

Can white along grow plants.... yes,
Can they be enhanced with addtional blue and red, again most definitely yes,
(Even when I select fluorescent bulbs, I still look at their spectral curves for red and blue performance)

Will more LED's on a single star produce more heat... oh yes, absolutely

Will multiple colors on one star improve color blending.... yes

Twice the heat?.... for some reason Im thinking it increases exponentally, but I cant remember why.

I used a heatsink paste that was also an adhesive.... but that is probably old school technology by now.

You're on your own now mate :)
 
  • #20
Couldn't resist.... dang

Custom Tristar:
SR-03-CUSTOM Your Choice Of Any 3 Rebel LEDs Soldered To A 20mm Tri-Star CoolBase
[LED 1: 5000K ANSI White - 200 lm (LXW8-PW50)] (85 Cri)
[LED 2: Deep Red - 580 mW (LXM3-PD01-0260)] (655nm)
[LED 3: Royal-Blue - 840 mW (LXML-PR01-0425)] (447nm)

Already have an adjustable driver/heatsink/fan from my previous trials...
Well at least you know what "I" would order... LOL


:p
 
Back
Top