loading

What type of LEDs are used for grow LED lights?

I've been googleing about the advantage of LED grow lights and how it saves energy and how they last longer. After looking to buy them I realized they were very expensive. However, I also found a few do-it-yourself guides which I thought were pretty neat and straight forward. The only thing I didn't see was if the LEDs they used were special or just any red/blue led lights can be used. Hope to get an answer and save a few bucks but most importantly save energy, Thanks in advance !!

sort by: active | newest | oldest
RikJ13 months ago

how can you calculate what the best mix of led diodes is than? because the spectra is found every where but no info about the mix of red bleu green uv and infra red

Short answer: Plants evolved under sunlight. The best thing to do is to duplicate sunlight, which means providing full spectrum lighting.

Less short answer: Chlorophylls turn light into energy in plants. Chlorophylls are greenish pigments, and each type is sensitive to particular wavelengths of light. Land plants use chlorophyll a and b. Both types have two small spikes--small frequency ranges that they're most sensitive to. Type a absorbs UV (below 400 nanometers,) up to around 470nm, and also at the red end in the ~560nm to ~700nm range. Type b absorbs more weakly in the UV--let's say ~420-490nm, and orange/red light at around 600-680nm. It's easiest just to lump those sets of ranges together and say 400-500nm, and 600-700nm. (These aren't hard numbers; just me glancing at a graph and estimating.)

Most commercial LED grow lights put out 5-8 times as much red light than blue. Plants don't use 8x as much red light as blue, though. It may be partly because red LEDs have been around longer, and are cheaper and more available than blue. Violet, "hyperviolet" (near UV,) and ultraviolet (outside the visible spectrum,) LEDs are a relatively new thing. Bright blue LEDs have only been available since the early 90s. A photosynthetic response graph of chlorophll a & b shows that light from below 400nm to over 700nm is absorbed. We probably don't know everything about how plants use light, but you can provide a only couple of frequency ranges and grow nice, healthy plants. That response graph has lumps centered at ~440nm and ~640nm, but those are peaks, not the whole mountain.

Fancy commercial LED grow lights might have 10 emitter types at different frequencies, while the cheapest have 2 or 3. Some even have customized emitters with specially formulated phosphorus to create special light magic. Apparently, they're working very hard to cover the visible spectrum, while imbuing some colors with mystical importance. Seems to me that it's easier to just use full spectrum lights to begin with, possibly supplementing with some deep blue and far red. The grow setups I've seen in recent years have used plain, old CFL bulbs; 2700K and 5000k. Maybe not ideal, but it does tell me that unenhanced white light can grow big, productive plants.

In building LED grow lights there seem to be two general approaches: 1) Use a small number of high-power (5 or 10 watt) white/broad spectrum Chip-On-Board (COB) LEDs; mostly warm white (2700K,) and some daylight (5000 Kelvin,) or 2) Use a lot of individual LED emitters at 5, 6, 8... maybe even 10 specific frequencies. My thumbnail is that you'd want to cover two ranges, say, 400-500nm, and also around 600-700nm, but that isn't a scientific answer, it's just an easy answer. Yes, you could focus on 650nm, but that's right in the valley, between the absorption peaks of types a and b chlorophyll; you want both sides of that.

I've found all sorts of good and bad data & dis/information, but this theory+practical write-up a good one: http://goo.gl/F8IVc6. This guy uses four 2700K (warm/red,) COB LEDs, and one 5000K for a 2'x2' space. His build and gear are kind of high-end/overkill, I think, and components he used are less expensive now. He supplements with two CFL bulbs, which I would also do. Since that was written, cheaper, more efficient LED COB modules have hit the market. There are good inexpensive drivers, and easier approaches to cooling. (like aluminum square stock w/ a fan at each end.) He doesn't bother salvaging obvious stuff like old computer power supplies, which are good, cheap, and easy to find. I suspect now you could make something equivalent for under $100. I'd use old PC power supplies, cheaper drivers, aluminum square stock w/ airflow as heat sinks... I truly don't think that many exotic LED colors are necessary.

In nature, the plant goes through two main phases: vegetative and flowering. Vegetation starts at the beginning of summer, as days are getting longer and the sun gets high overhead. Toward the end of summer, days get shorter, and the light gets redder as the sun sinks toward the horizon. (Sunlight coming at more oblique angle passes through more atmosphere, and is more yellow/red than when straight overhead.) Shorter days induce the plants to enter their flowering phase.

When growing under lights, some people provide more blue/daylight light in the veg phase, and more red in the flowering stage. The duration of the light is key, but the color change may have an effect. I don't know. I do know that you can induce flowering by going from 18-20 hours of light per day to 12 hours without adding red light. I like to try and mimic nature, but I can't say whether the plants appreciate it or not. It makes sense that the flowers might be attuned to the redder light of late summer.

I know you can't use normal red and blue LED's for growing plants as they emit a very narrow range of wavelengths. I don't believe there are any specifically "UV" type LED's, though I've seen many infrared. I do know that certain normal diodes of the non light-emitting variety actually do emit UV radiation, though at what wavelength, and what intensity, I don't know.

I know that your post has been here for some time. There are uv leds that can be purchased but I don't know the wavelength they emit. Thought you would like to know.

There's the LUXEON Z UV package. Those go down into the ultraviolet, and up above 400nm into what's being called "hyperviolet." Those span the 380-420nm wavelengths. They cost ~$15-20/ea. stevesleds.com has SemiLeds Hyperviolet stars that emit at 415-425nm. Those are currently $3.90/ea.

If you're doing something like curing dental epoxy you might need a 365nm light source, but I'm not sure whether that would help with photosynthesis or not. It might! Great way to burn your retinas! (Seriously. Wear UV goggles when working with intense UV/HV sources. Retinas don't grow back.)

For LED grows you want to know the ratio between spectrums first of all. That prereq depends on what sort of crop you are growing.

And the better your aim, the better result.

The majority of the emitters in a LED-array should be in the 660nm spectrum for general use.general. The next largest would be (depending on who you ask) 740nm or 450nm. Personally I'd focus on 740nm after 660nm for a good result and make the third largest the 450nm. The border spectrums to green from both directions are active, but in general, there is little point in adding emitters for these. A spectrum between the 450nm and 395nm could however be usable, but from what I've seen, those are fairly hard to come by.

Keep in mind LEDs usually vary in their spectrum +-20nm. So with a middle spectrum you should cover the overlapping spectrums as well as is the case with 740nm, the spectrum you really aim for is 730nm.

Be advised 740nm is visible to humans even if it's faint which IR isn't, but it is to plants. Some use IR as well as supplementary spectrum.

The general multiband tailormade horticulture LED spectrums are:

~400nm (UV) >1%

~420 (Purple)

~450nm (Blue)

~480nm (Turquoise)

~590nm (Yellow)

~630nm (Red)

~660nm (Deep Red)

~740nm (Far red)

~850-950nm (IR) >1%

I cannot stress enough how important the ratio between the spectrums are. You are better off buying 100W warmwhite+coldwhite emitters, they provide enough light to give a good result. If needed you can always add supplementary spectrums in the UV and Deep/Far Red and IR spectrum.

Mogger6 years ago
Plants use the Photosynthetically Active Spectrum, also called PAR. But it is more involved: Chlorophyll photosynthesis has two bands with peaks, one in the blue range (maximum at 430- 440 nm) and the other in the red range at 670 nm (info from an old GRO LUX TL grow lamp datasheet). Now, Phototrophism has another band which is responsible to let the plants grow low and broad or high and slender. Here it is is a curve with two peaks, starting around 410 nm with a peak at 430 nm, and a second one at 475 nm, sloping down finally at 500 nm. If we want to to grow plants with LEDS, we must match the maximum curves with a mix of LEDS consulting the LED datasheets- Spectrum curves. Until now, I did not find yet LEDS in the precise spectral range that the plants need- so we must again mix here and find LEDS which are the closest to the maximum in the curves. I was selecting Blue, Royal blue and RED Luxeon Leds. Also important is the lumen ( light output ) of the LEDS and their distance over the plants, with RED LEDS beeing a lot weaker than the blue ones. So in any design, the RED leds must outnumber well the blue. There are a lot of LED grow lights out there now- but the PRICES! CREE LEDS seem to be the best ones at the moment for this use and I just checked out Luxeon today. Chinese exporters offer also a lot of grow lights and there are even aquarium lights now. (LED GARDENER e-journal is one nice source of information- google it up). In any case, plants may differ in spectral requirements and the best is setting up experiments and test run the LED rigs. (For example: we have an in vitro Lab running during 12 h a day 80 pieces of 120 cm white TL lamps of 40 Watts with old reactors which consume also 40 watts, wasting energy in heat. This is a load of 6.4 KW every hour. Therefore I am looking to stop this nonsense and work with LEDS.
Red and blue...Cree's are the best, you can buy 1W LED's that way. Only problem is having to buy them in bulk (last time I looked I had to buy 1K at a time @ $3-4 each). I would put intensity controls on the red and blue circuits and tune until the plants appear black or at least very dark colored. Basically if the plant appears black then it is not reflecting any light, therefore there is no wasted light (plants reflect the green therefore they are not using it).
the intensity controls would be something to see, actually something on the whole setup you've got would be fantastic
noahw7 years ago
monchito7 (author)  noahw7 years ago
So from what I understood, yes you can use regular LEDs to grow (would probably need lots of them to make a decent amount of light though) bujt they suggest high power LEDs which gives better results. I guess now my research will be more on high power LEDs lol
110100101107 years ago
the way i understand it : each plant needs various ranges of light - each for another need. the light source should match as much as possible the needs of the plant - supply enough light in each range here are some images that show this (the light sources are taken from lamptech.co.uk) some plants need dawn-like light to grow and day-like light to flower
plant.jpeghalide.jpeghelium.jpgpl.jpgsodium.jpgtungsten.jpg
monchito7 (author)  110100101107 years ago
awesome graphs really gives a visual for some question i had, thank you !
rachel7 years ago
LEDs emit light on only one wavelength. The sun emits a huge range of wavelengths. LEDs are not made in all of the wavelengths found in sunlight. If LED grow lights work, it must mean that plants only need certain wavelengths, or are able to use a more limited mix. This makes sense, but I would expect it to vary somewhat by plant species. This doesn't answer the question but hopefully gives some context.
monchito7 (author)  rachel7 years ago
From what i've been reading so far you need red and blue LED lights the blue would help the seedlings and the red promotes flowering, there are many website out there that sell them already arrange in a certain matter but they range from 150$-600$ it's very new and doens't create much heat or use much energy.
monchito7 (author) 7 years ago
Thanks for the replies everyone much appreciated.