My test WindowFarm with DIY LED fixtures
3:24 pm in electronic components, energy consumption, environmental impact, Materials and Resources by Christopher McCool
http://cgrantmccool.com
Using the following website with a quick, easy DIY solderless LED fixture as a guide, I set-up and made my own lights! http://www.dabblings.net/subpage2.html
I’m no expert with electricity, but I understand enough and had enough research skills to pull off a modified version of the DIY set-up given in the site. My LEDs were TopBright LEDs
I have since reworked this set-up to add a Red bulb to each fixture. This was a bit tricky because the Red LEDs I was using required a separate circuit from the White/Blues which had the same electrical draws. While each fixture’s Blue and White circuits are all self-contained in their own project enclosure, the red bulbs are also in series with each other from enclosure to enclosure.
Hmmm . . . I don’t remember posting this . . . I was still writing it, in fact!
In any case, It should be noted that TopBright LEDs are not the best match for photosynthesis in terms of LED efficiency, but they are close enough to see great benefits after just one week. Also, as these are supplemental lights, it’s not as big an issue as it would be if they were only lit artificially. I’m counting on the window light to fill in the spectrums I’m missing.
http://3.bp.blogspot.com/_pFQ0wrHWd1k/St4V0NN2CuI/AAAAAAAAAFg/FGKwRvOB9g0/s320/action_spectrum_en.jpg (Link to Photosynthesis light chart)
I will be making a better, cleaner set-up in January, perhaps a few different designs (a single, large DIY UFO LED light, one per column, instead of individual fixtures for each planter).
That looks great! Did you happen to experiment with any other color combinations? When you say you added a red bulb, does that mean you’re running 2 red, 2 blue, and 1 white in each array or is it the same 1 red, 2 blue, 1 white fixture you posted on your blog?
I’d like to see a cost/time/productivity comparison with your fixture and conventional full spectrum CFLs. Neither are the best option for the only light source but can work supplementally.
Also, what was the final cost of each fixture?
Cost per fixture was about $10, I think. The Perf Boards were about $1 each fixture and the enclosure about $2.50 Resistors cost me $1 for the radioshack 10ohm resistors (set of 5) and about $5 for the 47 ohm (50 units) off of ebay.
I paid about $25 for 15 LEDs ( 5 red, 5 white, 5 blue ) . . . about $1.60 – $1.70 per LED. I’m thinking about windowfarming up my living room window, which would take 20 – 40 of these fixtures. I imagine that it would drive the cost down a fair amount.
I’ll be doing a better documented (on all levels) version of this in January after I get back from a month of travels
When I say added a red LED, I meant added it to the White and Blue circuits. The first version and first pictures didn’t have the red in it. The picture of the circuit board with 4 LEDs and 2 resistors was the first one to have red added, but the circuitry started getting messy/tricky.
I haven’t tested any other color combinations, but one of my reasons for using White, instead of blue was simple aesthetics (although White LEDs are mostly blue spectrum anyways). I have no doubt that 3 Blues would be better than any combination that includes White.
Past experience has informed me that the yellow/orange spectrums that most LED growing lights omit are helpful for flowering plants but not overly useful. My next version will likely be a 6 LED fixture of Blue, White, Red and UV
B WH
UV Red
WH B
I’m hoping this will give me more even distribution of the light/spectrums across each plant, although the Red will still require its own circuit and adding more LEDs probably means having to start soldering.
All 3 of the plants have basically doubled their growth rate since transplanting last week, or significantly improved, health-wise. The basil in the top planter (in the video you can see that they couldn’t even stay standing up because of weak stems/top-heavy leaves) are all standing on their own, now.
I’m convinced that this will be better than CFLs overall. That’s considering growing efficiency, aesthetics/light pollution (aesthetics was the only reason I wasn’t using CFLs before this. Even with reflectors, etc. it’s a lot of light pollution, a lot of wasted electricity, and ugly), and cost to build. I think cost will probably end up being 1.5 – 2X as much as a CFL light string.
Other LED alternatives would be to wait for LED Christmas lights to go on sale and use the Blues, Whites and Reds in your windows this christmas!
I found these on the net:
http://www.dealextreme.com/details.dx/sku.5256
It’s strips of ‘white’ LEDs which give off a slightly bluish light.
I’m not certain how well-suited thee light is, but…
These strips works with a normal 12V supply, and can be ‘hacked’.
Internally, there’s a resistor for every 3 LEDs, so it’s possible to cut it in chunks of 3/6/9/12 and so on, and just connect 12V to the ‘bus bars’ at the end of the chunks.
I have 3 of these strips, but haven’t reanny experimented with them, yet.
There are other strips and LEDs, driver circuits(for 3V LEDs) and everything on the site, too.
because I lack adequate window space, and! my windows are rather cold, since i am so far north! I researched and set up with lights. this is what I found… LED lights are one of the most expensive for initial set up, and lumens are still too low for anything but complentary light. Compact Florescents are still the most cost effective + efficient for close up lighting. cost $10 for 4 CFL 6500K temp 92 watts total 6600 lumens. Metal Halide tend towards the blue spectrum great for vegetative growth, be careful of the heat! the new 70 watt MH will fit into regular sockets and do not require ballasts, and are aprox $20. per bulb with 6250 lumens.
Blue light (5000K+ light temperature) for vegatative cycle, good for things like lettuce, basil, anything that does not need to flower, and the growth stage for anything that does.
Red light (3000K-) for flowering/fruiting cycle, combined with reduction of light cycle times.
Additionally, it is recommended to use aproximate 2,000 lumens per plant (or sq. foot) for maxium growth.
LED is a wave of the future… let me know how your experiment works out….
update… LED for the blue/white light is running about maximum 30 lumens per watt,
white light LED about maximum150 lumens per watt.
compared to CFL, at 23 watts for 1650 lumens, 100 watts for 1500 lumens LED, and watts = $$, CFL with balanced grow spectrum lighting still seem the better buy…
Information gained from:
http://members.misty.com/don/led.html
@mccoolcg That is a cool little set-up! Aesthetically it is a lot more appealing than a CFL string. Definitely useful for tight spaces, commercial LEDs are ludicrously expensive. To help your little LEDs be more efficient you may want to pick up some mylar or other reflective material, and line the back of the bottles with it, so any overshot light is reflected back on the plants.
@deidre Where did you see the 70W MHs that don’t need ballasts?! I would like to get my hands on some of them. Also, 2000 lumens is the minimum light for most plants I believe, the sun outputs ~10000 lumens/foot in growing seasons, so the closer to that, the better. Surpassing it is even better, if you have enough airflow/ humidity control.
The efficiency run-down would be HPS > MH > CFL > LED > Incandescent (including halogens). But HPS is best served to flowering due to yellow/orange hues, MH to vegetative due to bluish hues, and both traditionally have the extra cost of a ballast. And like Deidre said, they get hottt. But they both have very high lumen/watt values, and do not drastically lose luminosity with distance. The CFLs do lose luminosity fairly quickly, and need to be kept with 6″ of the plants, ideally within 2″. And LEDs are the same I believe. Without reflectors they are all going to waste about 2/3s of their light.
You can’t do a lumen comparison between CFL or other grow bulbs. LEDs are very spectrum specific. So even if less lumens per watt, they are more efficient per lumen than other bulbs that emit full spectrum light, as well as UV and infrared.
To compare to a blue LED at 435-460nm you would have to figure out what percentage of energy/output is going to that wavelength specifically, for CFLs or other growing bulbs).
Why is that? Because of the plants’ inability to utilize mid-range wavelengths? So the lumens that would be allocated at those wavelengths is wasted on the plant therefore wasting energy?
Electricity to light, LEDs are definitely most efficient. My efficiency run-down is just based on what I have read about the effectivity of the assorted types specifically in plant growth. While LEDs are great, and will likely be the future, right now I haven’t heard many great things from anything other than the $300+ per lamp uses. My only critique is the lack of intensity, which is paradoxically why they are so efficient, because the light relates to photosynthesis, but the intensity relates to transpiration (which needs to go up equivalently with a plant’s photosynthetic reaction to be effective).
Totally loving this. Great system @mccoolcg. Sorry if we posted this for you. There’s a bug that makes us have to approve some posts and sometimes we accidentally end up publishing posts that were not yet ready to publish by accident as they’re all in the same queue. I’m excited to hear how your plants fare and to follow this how-to myself. I want to see how I can set up a test to compare it with the LEDs the Finnish team made. I’ll send them your link. I think they are mostly talking on a Facebook group in Finnish but I’ll see if I can get Nikko over here.
This is some interesting stuff @mccoolcg – I’m just stumbling onto this project almost a year after your original post, but some thoughts from an engineering perspective:
1. LEDs vary tremendously in efficiency per watt consumed. Right now, Cree is probably making the best stuff (check out their XLamp series). Those LEDs you’re using are really not optimized for light output, they’re just meant to be used as indicators.
2. related, contributing to the overall efficiency of light delivered from the silicon is the lens on the LED – again that’s where someone like Cree or similar has spent a lot of R&D $ to ensure best radiation pattern. By default, LEDs emit a Lambertian spread, which just means that there isn’t much focusing at all. Look for 90-120 degree lenses.
3. A big point in overall efficiency is how to provide the appropriate current and voltage to the LED. Typically this is about 3V and current will vary based on LED size. In your design, you are fixing the current delivered using a resistor – that certainly works, but the resistor is actually consuming a lot of energy. Instead, there are integrated circuits that can operate at 90%+ efficiency from a number of places – you’re going to want to use something that is using a “switching” driver.
Does anyone have info on how much time to have LEDs on? My green beans seem to love it, but the cabbage is getting very old and has not yet produced the head.
Okay i am trying to build a LED lamp myself and the best i have found so far would be the OSRAM Golden Dragon plus series.
With an optical efficiency of up 79 lm/W (red), 35lm/W (blue) and 107 lm/W (white) they seem to be quite nice LEDs.
My question is:
What would be the best light mixture? Are the white ones required?
According to Wikipedia a good mix may be 8 to 1 for red to blue.
But if you look at a Chlorophyll spectrum the blue wavelenghts should be more dominant.
I read that NASA is experimenting with LEDs to grow veggies on the ISS.
They didn’t provide any details but the LEDs looked something like violet. So more blue than red.
What is your experience with LED grow lamps?