My Lighting Project - Will a Custom LED Panel Work?

[Cuttings2Colas]

A friend of mine recently got his state-approved license to grow, and I've suddenly found myself obsessed with learning how to grow MJ. I also used to live near a friend that had a garden, where I learned quite a bit. These are a few things I'm working on designing:

-- LED Lighting (home-made, very specific design)
-- Aeroponic/Hydroponic rubbermaid tub system (home-made, self-contained w/ sprinklers & reservoir, air-stone in res.)
-- High-speed ventilation, with pauses for CO2 (all through home-made carbon filters)
-- Aeroponic Cloning setup (15-site, RO Water, 100-200ppm, )
-- All of this built into a 6'Wx7'Hx30"D Locking Cabinet... that looks like a bedroom wardrobe cabinet. (really excited about this one)

This thread is dedicated to light & the potential of LEDs. I'm a bit of a geek when it comes to electronics, and I got drawn in by this project. I've learned incredible things about light in the past few weeks... light intensity for humans vs. light intensity for plants, which colors plants use most during different stages of their life, lux/lumens/mcd still doesn't make perfect sense (it will someday). I've looked at the Pros & Cons of different types of lighting systems - efficiency, light output, heat, light beam angle (more important than you think).

First, I'd like to say this: I'm not impressed, or even marginally satisfied with any of the LED grow lights on the market. There are a few main reasons why, and we'll get to those as you read on. For now, let's look at some of the positive aspects of LEDs.

Low Power Consumption. This is usually the first reason someone goes to LEDs, so let's get this topic right out in the open. The debate is this: To get LEDs to grow the same yield (quantity & quality) as a 600-1000W HID bulb, would you need so many LEDs that you were no longer saving power?

After exhaustive studying on converting candelas to lumens and how it sounds like it's not really possible, I found a chart that helped me out immensely. It has everything I needed to know.

----{See Image - Candles per Light.jpg}----

 

[Cuttings2Colas]

The LED setup I've been working on gives off about 9,000 candelas (cd). They run so cool, they can be touching the plants, though I'll probably give them a protective glass screen 1" away. Being one inch from the top of the plant means none of the light will be wasted. One very renowned grower says that every 12" your plant is from the bulb, you lose 50% of your light intensity, which is directly correlated to your yield. Move the light from 12" to 0", double the effectiveness of the light. This chart says that my LED panel will be very similar intensity to a 400W HID bulb at 14". You wouldn't want an HID bulb much closer unless you have a good light ventilation system. The test LED setup I'm looking at would use 200W. 200W Custom LEDs @ 1" is about equal to 400W HID @ 14". I might also add side-strips of LEDs to stimulate the growth of the lower foliage... stronger light, still under 300W.

Here's the bonus: These calculations are based on how the human eye sees light. Red or Blue LEDs that are putting out the same amount of light as "White" LEDs have a lower mcd (millicandela, 1/1000 cd) rating than the white. This is because the human eye sees Green and Yellow the BRIGHTEST, which are the two colors plants use the LEAST. With very few green/yellow LEDs, the room will appear darker to us (lower rating) and brighter to the plants... that's great!

----{See Image - Plant Light Curve.jpg}----

 

[Gmaccin6]

 

[Cuttings2Colas]

In this chart, we can compare the two. The solitary white line is the "Plant Sensitivity Curve," and the colored in area is what the human eye sees.

The current LED grow lights on the market talk about this, but it doesn't appear most of them have done any research. They say, "The plants only use the blue & red spectrum," or "Blue is used more during vegetative growth and Red is used more during flowering." Both of these statements are somewhat true, but severely misled.

Plants use all colors. Think of the light color as availability of nutrients. In their vegetative cycle, plants use more light from purple to blue to almost green. "Blue Spectrum" doesn't do it justice. In the flower cycle, plants use more light from the yellow-orange-red-deep red spectrum. "Red Spectrum" is quite an understatement. Notice I said the plants use "more" of specific colors during certain cycles. The plant still uses all colors in all cycles... it just uses some colors more during different stages. This chart shows what enzymes and plant organs use different colors of light. The left side is purple through blue, then green and yellow are the low center area, and orange-red-deep red are the right.

----{See Image - PAR action spectrum.gif}----

 

[Cuttings2Colas]

The sun emits its light throughout the whole spectrum. CFLs can be "cool white" or "warm white," which both have quite a bit of green/yellow (bright to us, dark to plants). Some people prefer to veg in Metal Halide (MH) and flower in High Pressure Sodium (HPS) because they are broad spectrum lights that have a bit more intensity in a particular color spectrum.

LEDs give us freedom from these limited bulb types by allowing us to specifically adjust the intensity of light in every specific color of the spectrum. If I get access to a color intensity meter, I can fine tune the light output to match any curve I want. That'll be incredibly useful. It'd take quite a bit of research, but I could even design separate veg & flower LED panels.

There are some pitfalls to LEDs, so let's not ignore them. With the right care in design, we can overcome the negatives and accentuate the positives. The light that is emitted from LEDs is very directional. If you look up LED Grows on YouTube, you'll see many "test setups" where the light looks like a straight beam, lighting up a small bright dot on the leaves of the plants. Also notice that most of these test grows exhibit signs of stretching - tall skinny plants caused by being too far away from the light. Not all LEDs are this narrow. All LEDs I'm looking at are at least a 30 degree beam, some much bigger.

Another concern with narrow beam angles is not enough cross-lighting in the plants. Plants in grow rooms with more than one light will benefit from both lights, not just the one directly overhead. If the LEDs are directional this could be greatly reduced. Less light on the under-foliage means light, airy buds (or none at all). OK for hash, but you'd like it to be better. This concern is why I mentioned the possibility of some LED light strips running up and down the walls to get more light down under the canopy.

The setup I'm looking at creating has 2,304 LEDs in a 1ftx1ft square. That's a ton of LEDs. Each one of them has to be individually pressed into a breadboard, I have to design an exact pattern to follow that gets the proper voltage to LEDs that run on different voltages, maximizes the space on the board, properly spreads the different colors evenly throughout the light, and I want it to look pretty decent. A step up from ghetto, at least. If they work as well as I hope they will, maybe I'll start having them manufactured.

More info as it comes... subscribe to the thread if you want the updates. Hopefully, there will be a test grow in the near future. The friend that just got his gardening license is interested, so I'm hoping to create a test cabinet. Exciting!

 

[Cuttings2Colas]

I just realized... if I make the lighting specific to what the plant actually wants, I'll have the closest thing out there to a PAR/PUR light. Freakin' badass.

What're everybody's thoughts on the setup?

 

[Cuttings2Colas]

So, I've been playing with the design more, and here's what I've come up with (see picture). It has three curves on it to compare: Plant Sensitivity Curve, Human Eye Sensitivity, and the light output by the current LED setup I've created.

This is my variation of the plant light curve, combining each of the previously mentioned charts, in a way. The human eye curve was taken from another picture. The LED light output comes from an Excel spreadsheet, where I can increase or decrease the number of LEDs for different wavelengths, and it automatically tells me the change in price, watts, amps, etc... and it updates the chart!

If the color measurements don't make sense, they're in nanometers (nm). 470=blue, 550=green, 600=orange, 640=red. The x-axis is the color spectrum from violet/blue to orange/red.

 

[Cuttings2Colas]

Since plants use Green/Yellow light the least, we're not wasting energy by producing those colors of light. However, we've broadened our color usage in the blue and red spectrums (both grossly under-stated), helping all of the cell organs and enzymes get the exact type of light they want.

You may notice that I didn't provide much light in the deep blue (430) and deep red (680) colors... simply put, those LEDs are much more expensive. I'm sure that after 2 or 3 harvests, you'd get your money back for having spent the extra money. The question is, would it be more beneficial to spend $130 for a few 680nm LEDs or $130 to double all of the other LEDs?

Either way, my friend's grow room is coming together, so I'll probably be placing orders soon... I'm excited to start building it!

 

[Treeth]

YAY for LED DIY

Good research.

I like the idea of exactly matching the generic plant par curve... very good optimization idea, but as you noted, not practical.

Nice profile-boob for the human eye curve

First: step up to a much more powerful one watt led or 5 watt led package, something from a reputable player like cree or phillips lumileds. They're products are worlds beyond the 5 mm leds that stick in breadboards. Think surface mount - These type actually need heatsinks, and not because they're less effecient.

If you insist, for ease, on sticking with a breadboard, you will still have to get a real heatsink on there... you know how many watts you're going to be pushing, Its going to make heat.


...you ideally want as much power as possible to drive sensimilla, for which a decent red is definitely the favored wavelength.

... so to answere your question MORE RED, and cfls for blue veg and filling in for the absent spectrums.

But really, I saw a guy literally fill up a bread board somewhere on this board with dees... and it got very, (ineffeciently) hot.

you really should rethink what your light source is,

...figure out whats current in the industry.

And damn this shit is expensive!

 

[Cuttings2Colas]

YAY for LED DIY


first off, which dees are you using?

For this setup, most are 3mm through-hole LEDs, one type is 5mm. The efficiency of each LED seems to be about equal, so I based light output on the input wattage (for the chart). This trial run will be built on breadboards... if it seems to be exactly what the plants want, I'd consider selling them in the future (probably after a little more research).

I've figured out the specific layout on the breadboards, but I don't really want to go into that detail here... this thread is more about light concept and if we can more efficiently provide the exact colors of light the plants need most. Let's hope we can.

 

[Cuttings2Colas]

YAY for LED DIY
I like the idea of exactly matching the generic plant par curve... very good optimization idea, but as you noted, not practical.

Thank you much... just to clarify, I think the resultant LED panel is very practical. Purchasing the super-deep red LEDs costs about 15 times the other LEDs, so I left the 680nm out of the equation.

First: step up to a much more powerful one watt led or 5 watt led package, something from a reputable player like cree or phillips lumileds.

You know, I did quite a bit of research on those, and I definitely need to do more. I had difficulty finding them in the colors I wanted. According to the math, this setup has more than enough strength to grow some very healthy plants. Perhaps my next test will be with the High-Powered LEDs... they're far beyond my budget right now.

If you insist, for ease, on sticking with a breadboard, you will still have to get a real heatsink on there... you know how many watts you're going to be pushing, Its going to make heat.

You hit the nail on the head with "ease." I'm no professional with a soldering iron. I can make it work, but it's not pretty. I'll definitely consider the heat-sink. I already plan to have fans blowing across the lights, so I'll have to see how hot my test panel gets. Any notable heat, and they'll get the heat-sink.

... so to answere your question MORE RED, and cfls for blue veg and filling in for the absent spectrums.

Considering what I've read about how red is more important during flower and blue during veg growth, and our goal is short plants (less growth) with fat buds (more flower)... I agree 85%. I want the test run to be as close to the spectrum as is financially feasible, then do experiments from there.

you really should rethink what your light source is,

...figure out whats current in the industry.

All in good time... these 3mm LEDs are impressing me already.

 

[Treeth]

Now that,

repvip,

is spamming.

Not intellectual property protection.

Alkemi, you coward, what is your source for these "models"?

 

[alkemi]

Treeth

if i answer that its like the cops askin me where my grow room is lol
but i will tell u specs on them to see if ur intrested in owning a few

 

[alkemi]

• Power Output: 50W
• Input Current: 0.46A/0.23A
• Input Voltage/Frequency: 110V-220V, 50/60Hz
• Lumens: 2000
• Power factor: >95%
• Total Harmonic Distortion: <15%
• Dimensions:
  - Diameter: 175mm
  - Height: 60mm
  - Power cord length: 1700mm
• Expected Lifespan: 20,000 hours

 

[Treeth]

What do I need THD for ??? hahaha...

I aint listening to the lights! <15% is terrible! Is it something from, idk, blose!?

 

[Cuttings2Colas]

• Power Output: 50W
• Input Current: 0.46A/0.23A
• Input Voltage/Frequency: 110V-220V, 50/60Hz
• Lumens: 2000
• Power factor: >95%
• Total Harmonic Distortion: <15%
• Dimensions:
  - Diameter: 175mm
  - Height: 60mm
  - Power cord length: 1700mm
• Expected Lifespan: 20,000 hours

I'd avoid those "Made in China" lights... they tend to be made with around 200-250 5mm LEDs. The 3mm LEDs I'm using have about 80% the output of their 5mm counterparts, and I'll be using over 1600 of them. That equates to about 6 times the amount of usable light from the unit I'm designing.

At the same time, there is only one light I've seen out there that's trying to do it right... http://www.hidhut.com/catalog/led-supernova-270w-led-array-enhanced-spectrum-p-1759.html

Eventually, I'd like to design a perfect light for veg and a perfect light for flower... that's probably a couple years out at this point. I'm really excited to get started. I think I'll be able to afford everything in June; at least enough to get my medical card and start some clones on a T5 or two.

The LED light tests will come soon after.

 

[Cuttings2Colas]

Also, here's an updated version of the chart I posted earlier... I got a more accurate representation of the Human eye curve. I also added a nice little signature-type thing to it.

 

[Shrike]

This is an interesting thread as many believe LED's will be the way to grow in the future...

What do you think of the LED's from this company?

http://www.prosourceworldwide.com/

At the same time, there is only one light I've seen out there that's trying to do it right... http://www.hidhut.com/catalog/led-supernova-270w-led-array-enhanced-spectrum-p-1759.html

Eventually, I'd like to design a perfect light for veg and a perfect light for flower... that's probably a couple years out at this point. I'm really excited to get started. I think I'll be able to afford everything in June; at least enough to get my medical card and start some clones on a T5 or two.

The LED light tests will come soon after.

 

[Cuttings2Colas]

This is an interesting thread as many believe LED's will be the way to grow in the future...

Thank you very much... without a grow card & place to grow yet (or money to get either), I'm just sittin' here researching everything. The cool part is that I'll go into my first grow with an incredible wealth of knowledge, plus the ability to find any info I might need in the future. This place is great!

As far as LEDs being the way of the future, I definitely agree it's the next step. Then we'll have something better after that, and something better after that until someday... we blow ourselves up.

What do you think of the LED's from this company?

http://www.prosourceworldwide.com/

They look like a decent product, comparable to most other LEDs on the market. I'm sure they'll work, and you can get some good yield, as long as you keep the lights close - one of the major benefits of LEDs (see pic 1).

All of the Red/Blue lights should work fine, and the fact that yours adds orange to the spectrum is a big bonus. Plants use all light... it may be possible to grow some sweet plants in just 2 or 3 colors. When you fill out the whole spectrum with at least some light, the plants do better. This is why it's recommended to have a CFL or two in LED grows, as it fills out the rest of the spectrum (also helps with cross-lighting).

My project is beginning with the whole plant sensitivity curve, then I'll eventually do side-by-side experiments that are stronger in one color or another. Actually, one of the first experiments I'd like to do (after a successful grow) is to create a rainbow light... purple one one end, moving through the spectrum to red on the other. I'll watch the plants during different stages of growth to see what lights they're gravitating toward, and how the plants that are directly under the different colors of light compare in size and yield.

I've also attached my layout design (just finished it last night), which shows the exact layout of all 1654 LEDs on the breadboards (see pic 2). It's a beautiful sight.

 

[Shrike]

Excellent design you have there for the LED's and looks like a heck of a lot of work...but hopefully well worth it. I'm kind of in the same boat as far as no money right now among other reasons to have to wait to start the grows. But having people like you and StinkBud, SoG, Earl and others on this board gives us all a tremendous amount of knowledge and experimentation that serves everyone well in the end. Like you, when I start, I'll be ready as hell.

Thank you very much... without a grow card & place to grow yet (or money to get either), I'm just sittin' here researching everything. The cool part is that I'll go into my first grow with an incredible wealth of knowledge, plus the ability to find any info I might need in the future. This place is great!

As far as LEDs being the way of the future, I definitely agree it's the next step. Then we'll have something better after that, and something better after that until someday... we blow ourselves up.



They look like a decent product, comparable to most other LEDs on the market. I'm sure they'll work, and you can get some good yield, as long as you keep the lights close - one of the major benefits of LEDs (see pic 1).

All of the Red/Blue lights should work fine, and the fact that yours adds orange to the spectrum is a big bonus. Plants use all light... it may be possible to grow some sweet plants in just 2 or 3 colors. When you fill out the whole spectrum with at least some light, the plants do better. This is why it's recommended to have a CFL or two in LED grows, as it fills out the rest of the spectrum (also helps with cross-lighting).

My project is beginning with the whole plant sensitivity curve, then I'll eventually do side-by-side experiments that are stronger in one color or another. Actually, one of the first experiments I'd like to do (after a successful grow) is to create a rainbow light... purple one one end, moving through the spectrum to red on the other. I'll watch the plants during different stages of growth to see what lights they're gravitating toward, and how the plants that are directly under the different colors of light compare in size and yield.

I've also attached my layout design (just finished it last night), which shows the exact layout of all 1654 LEDs on the breadboards (see pic 2). It's a beautiful sight.