Building LED lights from facts, no theories

[knna]

It is better to appear stupid than to open one's mouth and remove all doubt. --Mark Twain

[glutek]

f*uck :/ knna, I'm sorry. actually I think it's not that way you wrote - think some ppl found it too clever and/or need some time to 'metabolize' the infos. me personally just have no time at the moment and temporarily stopped writing in all threads.
I know it's bitter and I feel bitter often when trying to talk about leds with ppl I know. esp marketers cause a frustration
and I do appreciate your work! and the par energy calculator was smth that impressed me
anyway, while thankful for the thread, my focus is not on the photosynthesis rate but on the photoreceptors driving specific responses, esp. flowering. I promised much to the ppl in the other thread long ago and still play with the papers, but want to gather as much I can and present it fully, then I reveal another and another and it goes deeper, dont know when to stop at certain level to systematize and present it, so I read read and don't write,
plz don't take it personally at least me is interested and appreciating! also have to admit I really learned from you. If you expect polemics, for the moment I say imo photosynthesis rates are much less important than ppl think, as leds usually do good veg and little buds. and in the photoreceptors realm there are many things already and widely not considered theories, and web is rich in it. secondly, that is rather a question than a polemic: do you have any prove that plant needs only a certain amount of blue and not a proportion to total light? not talking about pores or stomatas opening and stetch blocking, but 'bout all other responses?
ps. nice to see an European here!

[knna]

Hey, Glutek, im glad my Bulb Analyzer tool is useful for you.

Im too interested on qualitative and no quantitative effects of light quality. But the problem is this topic is way more specie dependent than light productivity. And effect are very often not linear, but modulated, and it mean lots of experimentation to be able to get any valid conclusion. Very similar experiments have different results

For example, about blue requeriments for healthy growing, it strongly depends of the plant specie considered. There are some than requires a relatively high percentage of blue light, and there is some than not require blue at all, as wheat.

About cannabis, there is little scientific data about. Most i know is for my personal experience, and is very difficult any not controlled experience proves anything. We know that cannabis requires little blue because it has been grown sucessfully using HPSs emiting very little of it. I want to try to grow cannabis only under red light and see what happens. I really suspect that cannabis not strictly require blue to grow healthy. Ive read of cannabis grown under LPS lamps, so under a absolute lack of blue, sucessfully. But ive never seen it.

What ive noticed is cannabis is a little demanding specie in terms of light quality. It thrive under almost any type of light, and its basically a photon's whore as wheat. At to what point it need a given amount of each waveband, i still dont know, but for sure is little.

Of course, not requiring not mean cannabis wont benefit for using more complete spectrums.

[cture]

Hello Kanna!
Tell me please if I have understood you right: you can find out most energy efficient led for a given wave length and assemble the prototype?

[knna]

Hello Kanna!
Tell me please if I have understood you right: you can find out most energy efficient led for a given wave length and assemble the prototype?

Not exactly. I can calculate accurately the efficiency of a given led, as anybody else interested on doing it.

And according to it, choose the better models for each given wavelengh you wish to use. Or if not precise wavelengh tuning required, i can say what model avalaible should give you better results for plant lighting.

LEDs wavelenght is a consecuence of chip molecular structure and composition. Manufacturers choose to produce chips wich emit the most usable wavelenght for many different applications and at what thay can get the best photometric output.

In the practice, it mean we can use blue and white leds very well, because their spectrums and peak wavelenghts are well suited to be used with plants. But on the red side, we have a problem. The highest photometric perfomance (lm) is produced with AlInGap chips emitting around 632-633 nm. At that peak wavelenght, the compromise between radiometric efficiency and photometric conversion is max, so manufacturers arnt really interested on producing other red wavelenghts in high power chips.

That mean that is very difficult to obtain chips emiting on longer wavelenghts, specially close to 660nm, wich would be the better for us. There is no a technical problem to do them, its just lack of interest of LED chips manufacturers. Indeed, AlInGap chips emiting at 650nm have the highest radiometric efficiency, that is what interest us.

But that wavelenght have only one main application (ours), and our demand is not enough to any manufaturer makes a production line of it.

So unfortunatelly, we cant get red leds of those wavelenghts at a cost effective price. We have avalaible a narrow range of wavelenghts between to choose what we prefer.

What i can do is helping in that task, by calculating of how many photons per burned watt emits each LED, so you can choose if using a 660nm peaked led emitting 1 uE/w or a 635nm LED emiting 1.3 uE/w, for example.

This should change on some years. Currently, all LED manufacturers are concentrated on developing high efficients white leds, and especially, warm white leds. This mean all the reseach funds are dedicated to InGan chips. Along the last 5 years, InGan (blue, green, white) leds have strongly improved its permomance (about 3x), while red leds are almost sttoped in their development.

But when the 150 lm/w has been reached on commercial LEDs (about 2 years, 3-4 for warm whites), then the competition is going to translate to improve the light quality. And this requires high efficiency red leds emiting on longer wavelenghts for color compensation. After years sttoped, now some companies are adquiring machines for AlInGap leds, preparing that time. Then we should have very high efficiency 650nm red leds easily avalaible.

But currently, its almost impossible to get them. The longer you can get on high power and efficiencies high enough are 640-645nm peaked red leds. And getting them is a headache (im already had spent last month discussing with distributors about it).

[cture]

That is great that you are doing this. Cause at the moment I had carried out biological research on the wave lengths and PAR wattage required (I just can't really convert enrgy of light in to photon flux). Apparently we do need leds higher than 640, especially wee must have 680nm and 700nm as theese wave lengths affect photosythetic rate and photomorphogenesis of the plant.
So these are the wave lengths I have discovered are preferable to have.

420
430
440
450
460

620
630
640
650
660
670
680
700

720 (optional)
We need to maintain approximately the same light energy power output on each wavelength.
Total light energy power should be around 350 Watts per meter sq (500 PAR (photosynthetic active radiation) watts for eqatorial sun).
Also you may have a look at my link in signature. There you may find spectras for chlorophyll a and b and carotenoids.

[knna]

You must notice that led arnt lasers (wich emits on very narrow wavebands), and emits along a relatively long waveband:

[attachment=o206959]

The nm rating of a led refers only to where it put more energy, but it still emits light of a lot more wavelenghts. So you dont need to use so much differents wavelenghts close to the next. For example, a Royal Blue led is going to cover very well the from 440 to 460nm. And still decently 10nm more at each side.

So you can simplify that distribution of leds. For the blue range, with two different colors, Violet and Royal Blue (but Violets on the 425-430nm range arnt very efficients, compared to those peaked 20nm shorter). And for the red, with a typical red peaked at 625nm and a deed red peaked at 670nm.

Most of the peak nm you mention are rare to find. If you still want to use so many different SPDs of each color, for fine researching on the propierties of each one, you will need to buy them at a very specialized company, as Roithner Laser. They have all, but you will need a good budget to afford it.

[knna]

We need to maintain approximately the same light energy power output on each wavelength.
Total light energy power should be around 350 Watts per meter sq (500 PAR (photosynthetic active radiation) watts for eqatorial sun).
Also you may have a look at my link in signature. There you may find spectras for chlorophyll a and b and carotenoids.

Dont guide too much by absortion spectra of photosynthetic pigments. One thing is their absortion in lab and other very different how they perform on live plants.

Photosynthetic response dont follow at all the pattern of chlorophills absortion. Many LED experiments have failed to be guided by chlorophills absortion instead by photosynthetic response, that are very differnt.

On the other hand, controling accuratelly the energy emission of such narrow wavebands is very complex. You are going to need a science rocket equipment to do that.

If you want to check reactions to very narrow wavebands, then probably laser diodes are a better choice than LEDs.

350W/m2 of irradiance is pretty high. Its about 1500uE/m2. Think that max irradiances at noon at 40ºN rarely goes over 1800uE/m2. And the daily average is way lower. You are going to need to use CO2 supplementation in order to use those average irradiances sucessfully.

[cture]

What if we go another way: you could provide me with spectrum graphs of light emitted by leds you find that are most efficient and I will pick the needed.

[cture]

Or another one.
Match the spectrum to
420-490 30%
490-620 10%
620-700 60%
What kind of soft do you use to build spectra graphs, may I have a link?