meltmanbob
Active Member
Hey guys this is my first post! Keep in mind I'm very new to a lot of this stuff but I've always wanted to try my hand at growing. I'm not at a point to start growing but in the mean time I've been wanting to work on the lighting. I'm sure a lot of this has been gone over in some detail before but I haven't found much that is very detailed.
So here's what I did and how I thought this through. First I need a reference which most logically I thought would be the sun, also later comparing to MH and HPS lights. I looked up the energy from the sun that reaches Earth using the term "irradiance." Found out that the other term that is used in "insolation." Then I looked up how much of that actually makes it to the ground and what wavelengths but this is where the information became less clear. The general idea was that about 1000w/m^2 hit the ground. Some sites said 43% of that was in the visible spectrum (400nm-700nm, as they classify it), some said 44%.
Next up were the insolation spectrum distribution charts that showed how much of the suns energy was at what spectrum and the chlorophyll absorption spectrum distribution charts. When looking up the chlorophyll stuff I learned about P.A.R. which is then what I looked for. I've attached the 2 charts I found and used.
Once I had the charts I used Googles Sketchup and drew 2 large rectangles, painted the faces of them images and re-sized to fit 1 image and not tiled. Then I drew a 5" line and repeated it 60 times end to end to make 300" which for my scale equals the 300nm spread over the visible spectrum. The vertical axis was drawn 100" and guide lines drawn horizontally to match the scale according to the image. Once this was done, I took each rectangle with an image and scaled it up and re-positioned it to get it to match up with the axis in Sketchup as best I could. Next I drew vertical lines at every 5" interval from the horizontal axis up to the curve. Then I pretty much played connect the dots to create a new smoother curve. A few connecting lines were adjusted to more accurately show the average rate of change in light intensity over that 10nm spectrum.
For those of you who've taken calculus this should be familiar for the rudimentary way of finding area under a curve but the point is that the area under that sunlight distribution is the value of the energy in that spectrum. There is still plenty of room for improvement but then again, ideally I would have the actual data points that I could plot on the computer and find a regression equation to represent that set of data and then integrate it to find the area. This would have been the preferred method since it probably would have taken less time and would have been the most accurate but I couldn't find the data.
This is an area I could use a lot help on and that is the light emission output curves on LEDs. I can't really find much info regarding how far up and down the spectrum the light is emitted. I saw one specific LED datasheet that showed it's peak around 624nm and had a spread (for lack of a better term) of about 25nm so that 95%+ of it's light was emitted between 600-650nm. I don't know if a 25nm spread is common etc but that info would help.
Since I don't have a whole lot to go on with response curves of the LEDs, I'll move on.
So I'll break this down the way I understand it so bear with me. Basically plants have things such as chlorophyll that absorb light and they each do different things. Some are more primary, some are accessory or secondary such as if I remember correctly, chlorophyll b functions to support and promote chlorophyll a's duty. All of their combined absorptions gives us P.A.R.
Ideally we would know exactly what light absorbing parts there are, how many, what they do, when they do it, etc. This way we could custom tailor the lighting to the specific and unique plant and at the right times or under the right conditions. This is where I think a lot of LEDs are lacking but hey what do I know.
I still need to find more info on the concentrations of the major light absorbing parts.
So the first big design question I think would be what to model this after. I think it's pretty obvious that using only the main peaks in the plants absorption is not enough because that's only part of the picture. As you can see on the P.A.R. graph, the red end has a lot more than just the chlorophylls which suggests that there are a lot of auxiliary, accessory, or secondary light absorbers that work in that spectrum.
I'm wondering if the plants absorption is pretty consistent throughout it's different stages or not. I'm also assuming that blue is for veg because the blue triggers certain growth patterns or qualities that are ideal at that point and then red for bloom. What I want to know is what the other colors trigger. Or is it just the light that determines veg or flower?
So what I've been getting at is modeling the lights output to match the plants absorption or input. The other option is to model after the sun but here is the dilemma; PAR shows absorption is the green,yellow,orange spectrum but it's less efficient so if we model the lights output to the plants input, the actual energy absorbed by the plant will be even less. If we model after the sun then we would be wasting more energy but the plant would be getting a more complete spectrum.
Now that I think about it, LEDs compared to MH and HPS etc is the bottom up vs top down approach. I was actually thinking in terms of food that LEDs are like trying to build a meal out of the basic components of proteins, fats and sugars. No wonder getting it right is difficult.
Anyway, knowing what the spectrum's other than red and blue do to the plant would help because then I could make a judgement call as to how much I should add in.
The thing is I took the data I got from the images and factored in 93lumen/w for sunlight to figure out how many lumens there were in each 10nm spectrum segment which allowed me to figure out how many lumens were actually being absorbed.
Something I found a bit odd is that a lot of people have used the 1000w/m^2 to justify the 100w/ft^2 thing but what I don't get is do they realize that more than half of that energy the plant doesn't use. It's more like 45w on a bright clear day in the visible spectrum.
Sorry for the mess I'll try to come back later and clean it up but I've been going through a lot of info and just wanted to get some of what I've learned and I'm thinking, out there to get some help and more ideas.
So here's what I did and how I thought this through. First I need a reference which most logically I thought would be the sun, also later comparing to MH and HPS lights. I looked up the energy from the sun that reaches Earth using the term "irradiance." Found out that the other term that is used in "insolation." Then I looked up how much of that actually makes it to the ground and what wavelengths but this is where the information became less clear. The general idea was that about 1000w/m^2 hit the ground. Some sites said 43% of that was in the visible spectrum (400nm-700nm, as they classify it), some said 44%.
Next up were the insolation spectrum distribution charts that showed how much of the suns energy was at what spectrum and the chlorophyll absorption spectrum distribution charts. When looking up the chlorophyll stuff I learned about P.A.R. which is then what I looked for. I've attached the 2 charts I found and used.
Once I had the charts I used Googles Sketchup and drew 2 large rectangles, painted the faces of them images and re-sized to fit 1 image and not tiled. Then I drew a 5" line and repeated it 60 times end to end to make 300" which for my scale equals the 300nm spread over the visible spectrum. The vertical axis was drawn 100" and guide lines drawn horizontally to match the scale according to the image. Once this was done, I took each rectangle with an image and scaled it up and re-positioned it to get it to match up with the axis in Sketchup as best I could. Next I drew vertical lines at every 5" interval from the horizontal axis up to the curve. Then I pretty much played connect the dots to create a new smoother curve. A few connecting lines were adjusted to more accurately show the average rate of change in light intensity over that 10nm spectrum.
For those of you who've taken calculus this should be familiar for the rudimentary way of finding area under a curve but the point is that the area under that sunlight distribution is the value of the energy in that spectrum. There is still plenty of room for improvement but then again, ideally I would have the actual data points that I could plot on the computer and find a regression equation to represent that set of data and then integrate it to find the area. This would have been the preferred method since it probably would have taken less time and would have been the most accurate but I couldn't find the data.
This is an area I could use a lot help on and that is the light emission output curves on LEDs. I can't really find much info regarding how far up and down the spectrum the light is emitted. I saw one specific LED datasheet that showed it's peak around 624nm and had a spread (for lack of a better term) of about 25nm so that 95%+ of it's light was emitted between 600-650nm. I don't know if a 25nm spread is common etc but that info would help.
Since I don't have a whole lot to go on with response curves of the LEDs, I'll move on.
So I'll break this down the way I understand it so bear with me. Basically plants have things such as chlorophyll that absorb light and they each do different things. Some are more primary, some are accessory or secondary such as if I remember correctly, chlorophyll b functions to support and promote chlorophyll a's duty. All of their combined absorptions gives us P.A.R.
Ideally we would know exactly what light absorbing parts there are, how many, what they do, when they do it, etc. This way we could custom tailor the lighting to the specific and unique plant and at the right times or under the right conditions. This is where I think a lot of LEDs are lacking but hey what do I know.
I still need to find more info on the concentrations of the major light absorbing parts.
So the first big design question I think would be what to model this after. I think it's pretty obvious that using only the main peaks in the plants absorption is not enough because that's only part of the picture. As you can see on the P.A.R. graph, the red end has a lot more than just the chlorophylls which suggests that there are a lot of auxiliary, accessory, or secondary light absorbers that work in that spectrum.
I'm wondering if the plants absorption is pretty consistent throughout it's different stages or not. I'm also assuming that blue is for veg because the blue triggers certain growth patterns or qualities that are ideal at that point and then red for bloom. What I want to know is what the other colors trigger. Or is it just the light that determines veg or flower?
So what I've been getting at is modeling the lights output to match the plants absorption or input. The other option is to model after the sun but here is the dilemma; PAR shows absorption is the green,yellow,orange spectrum but it's less efficient so if we model the lights output to the plants input, the actual energy absorbed by the plant will be even less. If we model after the sun then we would be wasting more energy but the plant would be getting a more complete spectrum.
Now that I think about it, LEDs compared to MH and HPS etc is the bottom up vs top down approach. I was actually thinking in terms of food that LEDs are like trying to build a meal out of the basic components of proteins, fats and sugars. No wonder getting it right is difficult.
Anyway, knowing what the spectrum's other than red and blue do to the plant would help because then I could make a judgement call as to how much I should add in.
The thing is I took the data I got from the images and factored in 93lumen/w for sunlight to figure out how many lumens there were in each 10nm spectrum segment which allowed me to figure out how many lumens were actually being absorbed.
Something I found a bit odd is that a lot of people have used the 1000w/m^2 to justify the 100w/ft^2 thing but what I don't get is do they realize that more than half of that energy the plant doesn't use. It's more like 45w on a bright clear day in the visible spectrum.
Sorry for the mess I'll try to come back later and clean it up but I've been going through a lot of info and just wanted to get some of what I've learned and I'm thinking, out there to get some help and more ideas.
