LED lights with pics

[liplayer101]

it totals 2032 leds mixed

[jason5710]

i built a 180 bulb LED fixture for about $30 and it works GREAT!

took a look on ebay and purchased the following:

100 x 660nm 5mm red leds = $2 shipped
100 x 470nm 5mm blue leds = $7 shipped
1 x 30w (from an HP printer) with a 500ma capacity = $10 shipped
100 x 1 ohm resistors = $5 shipped

i went to the dollar store and picked up a rubbermade box with a white plastic top.

i layed out a grid to evenly space out the leds (about 1-2cm apart). I used the following link to figure out how many leds i could use per row with the 30w power supply i was using:

http://led.linear1.org/led.wiz

you will need to know the forward voltage, and forward current (usualy in mA) of the led in order to use the calculator.

leds have such an advantage over hid lamps for many reasons;

leds produce no heat
leds allow for specific light wavelength choice
leds use way less power
leds are safer to work with
leds cannot burn your plants

im sure there are more reasons, but im too busy trying to get over how well it works!

DONT PAY $400 for something you can build on the cheap!

[SmknVTEC]

Originally Posted by liplayer101

The led lights are the same as the panels n such. u think that ebay one is bad???

Not necessarily I just can't vouch for it.

Good work jason5710, let us know how it works.

Here is some basil that I cloned with the aeroponic cloner and grew under the LED's.

[aaaaaa]

Anybody else been following this thread (link: http://boards.cannabis.com/grow-log/...ow-log-36.html)? You have to register to view pics but its defiantly worth it. That guy has some beautiful plants growing under 3 Procyon's with huge amounts of resin!

High times did a grow using the UFO led lamp and also got increased resin production giving a marked increase in potency. And now this grow that makes 2 led grows with increased resin production under LEDs. LEDs are the future and the future is bright.

[wtfk]

I'm a total newbie to growing anything, and a newbie to this system. In fact, I registered last night so I could see your pictures and understand the topic before making any big, expensive mistakes. I've literally been reading this thread straight through the night, because I really want LED light to work for growing my medicine, as it works for lighting my desktop (and only my desktop, not my office.)

In short, I want to say I hope I'm not stepping out of line by jumping right into your discussion. I want to first say your work so far has been a tremendous help to those of us trying to learn. I particularly appreciate the wealth of information from VTEC, Rye_led, Green Thumb and Budster (not to mention whoever posted the Plant Response graph, in case it wasn't one of you.)

The reason I'm chipping in here is that I couldn't help but think back to studying calculus in engineering school (the only place I ever used calculus.) One of the first things we learned about was the Riemann Sum, and how useful it is for approximating the area under curves (and you can bet, though it had limited use to Newton, it's pretty useful in digital approximations--sound familiar?) Long story short, one way to approximate the area under a curve is to fill it up with rectangles whose area can be easily calculated. As the rectangles get smaller and smaller, you can fill up more of the area (think "more, tinier rays of light, of progressively higher wavelengths.")

I see the emitters as little rectangles to fill up your area, and each photon from a "similar" emitter as a longer rectangle under that area. The application of this concept is that you need an array of many different kinds of emitters, in the correct quantities, to best approximate the light source the plant wants. As has been posited, we're best off configuring a unit exactly for cannabis, not just a generic "grow light" if we hope to perfect LED lighting for our medicine--that configuration is the area under the curve, which cannot be filled using just one or two narrow wavelengths (rectangles.) Notice how the union of the two wavelengths intersected with the Plant Response curve represents a minor portion of the potential light the plant could utilize (notice also that the area under the curve corresponding to 627nm is approximately 6-10 times the area corresponding to 470nm, and you'll see we're onto something.)

What does this mean? It means, for starters, I suggest that VTEC's system is (or, for that matter, the ones we've been seeing on ebay are) a good starting point, but it's missing an important greater variety of LEDs. May I suggest some Royal Blue emitters (455nm) and some Blue emitters (470nm) and some 627nm (Red) emitters, as well as 660nm (Red) emitters all together in the appropriate quantities to form a slightly more perfect Riemann Sum of light? (for the next iteration of projects, anyway.)

I think the above configuration would be an improvement over everything we've seen, and any other "missing" portions of the spectrum the plant responds to would be an even greater improvement, if/when available (different wavelengths of red/orange and violet.) Further, making a "cannabis unit" of light for a cluster of emitters that could be individually positioned could allow the positioning of, say, a half-dozen clusters per plant, in the ideal growing positions, to maximize the portions of the plant's receptors contacted. I don't know much about the frightful electronic circuit I'm implicitly specifying--I'll leave that to you gurus.

In summary, throwing more of the same wavelength or throwing white light at the holy herb won't utilize the unique properties of LED light, but choosing the right variety may. I believe the key is to give the girl more of the variety (in the proportions dictated by area under the Plant Response curve) that she wants and needs.

Thank you for your time. I hope those of you who already know how to get something to grow find the idea stimulating and can use it to make your design more nearly perfect so a cowardly clod such as I can copy your work after you've tested it out, and possibly even grow my own medicine without inviting into my home all the alphabet soup agencies known to man!

(FYI, I almost bought one of the "as seen on ebay" designs until I read this thread and decided it's best not to. If I had not read your personal accounts, I may have wasted that money.)

Addendum: Wow! 500th post in this thread. That's almost as cool as the 420th!

[SmknVTEC]

wtfk, welcome to the city and thanks for posting. The info on the Riemann Sum is interesting. I will count on you to plug in the numbers to the equasion because I plan to take calculus and analytical geometry in the future. However it makes sence that using different types of emitters can widen the spectrum. Instead of doing that I know that however my friend Guru has added some royal blues into the mix and he says they are growing nice. Remember we use thwe Luxeon III Emitter. I hope by this June I get up there to him and I can see what he has going on up there and how to assemble the systems etc. Until then good luck with your LED grow and keep us posted.

[wtfk]

Thanks for the warm welcome, VTEC.

Actually, I don't know what the calcs are because I don't know much about the nature of light or the electronics of LEDs, and I don't think too much serious calculation could be inferred from the diagram earlier in the thread--just some ballpark figures. Looks like you have 12 627nm red and 11 470nm blue at the moment. Let's try generating some new numbers mathematically using my interpretation of the graph.

The idea, of course, is to get the sums of the intensities of the LEDs to fit under the curve. You could do what I did by holding a scale (ruler) up to a printout of the diagram and marking it up to derive raw numbers (say, in inches.) If I do this, I get the following values (I divided through by the sum of the original numbers, so they all add up to 1--approximately, and it was just a statistical fluke that the raw data added up to almost exactly 17):

435nm: 	3.000	0.176
470nm: 	1.3125	0.0775
590nm: 	0.875	0.0515
617nm: 	3.000	0.176
627nm: 	3.875	0.228
660nm: 	4.937	0.290
		16.999	0.999

Now, how many total emitters do you want? Let's try 23 (just because it's the number you have now.) Multiply the unitized values by the number of emitters you want (23) and round to a whole number:
 

• 4 435nm Royal Blue Luxeon III
• 2 470nm Blue Luxeon III
• 1 590nm Amber Luxeon III
• 4 617nm Red-Orange Luxeon III
• 5 627nm Red Luxeon III
• 7 660nm Red (Cree?) of the same intensity as Luxeon III

Note that this gives you approximately three times as many emitters in red orange and amber hues as blue hues--that's about what the red-and-blue-only people are doing, only this is refined to populate more of the spectrum.

The larger the total number of emitters you choose, the better the approximation. That would be in the ballpark of the power draw and emitter sites you currently have, and should perform significantly better. Obviously, the red-orange, and even more so the amber, will only provide a marginal advantage over the configuration that does not have them, but I believe replacing the blue with even more of the royal blue will provide a great benefit, and adding 660nm red (notice that this is in addition to more 627nm red) will provide a big boost. On the other hand, I could be all wrong, but it feels right to me. This exercise is obviously almost as much art as science. You may need a fudge factor for lumens-to-einsteins. I'm not sure.

You might want to share the chart to your buddy, so he can see where all this crazy conjecture is coming from. Come to think of it, seems I may have seen a photograph of such a bulb, but I doubt it had high-power emitters. I suppose if you could do something like I described, you might have a shot at a high-performance LED unit. My main thrust is that I think it's kinda fruitless (pardon the pun!) trying to overdrive just two wavelengths of the spectrum (one each of red and blue) to make such a lamp work. I would expect the above design to perform much better.

Even then, I can tell by looking at the marked-up graph that there's still a moderate hole to the left of 435nm and a significant hole to each side of 660nm. I'd characterize the hole to the left of 435nm as perhaps 20% of the potential left-end (perhaps 15% of the whole potential) and the holes around 660nm as approximately 30% of the potential right-end (perhaps 20% of the whole potential), meaning if we're lucky, we might be getting around 65% of the potential spectrum that your girls can benefit from.

Again, I could be totally wrong, which is why (if I didn't scare you with the numbers or my cavalier WAGs) you'll consider testing the idea.

Addendum:

You'll probably be tempted to just use 12 627nm Red Luxeon III emitters, instead of 5 of them plus 7 660nm emitters. One way to test my theory, if you had an objective way of measuring the results, would be to try with just 5 of the red Luxeons, then with 12 of them. I'd bet you'll see pretty much the same results, meaning the other 7 red emitters are wasted energy, unless they're in a different part of spectrum (I.E. 660nm.)

[Genie]

here is another LED light i have come across,
http://www.ultraledlights.com/full_s...grow_light.htm
any thoughts on this product, also has anyone tried/had success with flowing using LEDs? or are they only comparable to conventional lighting for growing?

[Beck Bristow]

my question with that light, Genie, is why do they have those UV leds on them? their explanation doesn't make sense to me... maybe it has something to do with that kicked bowl in front of me?

[wtfk]

Genie,

I see some problems with the page you referenced. My first concern is they're charging $500 for 25 watts. Even if they have the perfect spectrum, 25 watts probably isn't enough--probably more like 50. They seem to be using low-power emitters, which don't penetrate/travel far.

Their graph is so small, you can't tell what it's supposed to be, but it's totally different from the graph someone else posted earlier in this thread depicting the spectrum response of cannabis. Cannabis doesn't care much about orange light, and even less about yellow light, if the previous graph is correct. Plus, there should be more blue LEDs in at least two wavelengths, including violet (not just ultraviolet), etc. Some people say UV doesn't even stimulate a growth response (the web page doesn't even seem to claim it does, if you read carefully.) In fact, they don't seem to be designing for cannabis, which, for LED, would be mandatory, so far as I can tell. Perhaps this is optimized for tomatoes, or something. The point of LED is not wasting power on the portion of the bandwidth you're not interested in.

In short, the purpose of LED is not to reproduce the full spectrum--it's to try to reproduce the spectrum a specific plant is interested in. I don't think this light appears to provide it in quality or quantity--at least for cannabis.

It's a pretty pattern, though. ;-)

I'm still waiting for someone to get big buds from LED light, and so far, many people have put in an honest effort but come up short. I salute all of you for your valiant attempts. Keep posting your results!

[aaaaaa]

wtfk- have you seen this link? http://www.ledsmagazine.com/features/4/11/2
its about led color mixing.

[wtfk]

Originally Posted by aaaaaa

wtfk- have you seen this link? http://www.ledsmagazine.com/features/4/11/2
its about led color mixing.

Looks interesting, thanks! I'm a big fan of LED light, and I'll definitely check it out. At least up until gas plasma lighting systems mature, LED lighting appears to be "the future."

[Genie]

Originally Posted by wtfk

Genie,

I see some problems with the page you referenced. My first concern is they're charging $500 for 25 watts. Even if they have the perfect spectrum, 25 watts probably isn't enough--probably more like 50. They seem to be using low-power emitters, which don't penetrate/travel far.

Their graph is so small, you can't tell what it's supposed to be, but it's totally different from the graph someone else posted earlier in this thread depicting the spectrum response of cannabis. Cannabis doesn't care much about orange light, and even less about yellow light, if the previous graph is correct. Plus, there should be more blue LEDs in at least two wavelengths, including violet (not just ultraviolet), etc. Some people say UV doesn't even stimulate a growth response (the web page doesn't even seem to claim it does, if you read carefully.) In fact, they don't seem to be designing for cannabis, which, for LED, would be mandatory, so far as I can tell. Perhaps this is optimized for tomatoes, or something. The point of LED is not wasting power on the portion of the bandwidth you're not interested in.

In short, the purpose of LED is not to reproduce the full spectrum--it's to try to reproduce the spectrum a specific plant is interested in. I don't think this light appears to provide it in quality or quantity--at least for cannabis.

It's a pretty pattern, though. ;-)

I'm still waiting for someone to get big buds from LED light, and so far, many people have put in an honest effort but come up short. I salute all of you for your valiant attempts. Keep posting your results!

Many thanks for the info, buying this light would have been once costly way to find out :-)

[SmknVTEC]

wtfk great info man, thanks for replying to the thread. We can use some smart cats like yourself here at the city. I hope you stick around and continue to participate in the threads. Plus rep brother.

[Genie]

ok, so is the genarel gist of it , LEDs so far work good for growing, but for flowing best to switch back to the Convental lighting systems? even if this is correct we still save electricty by using the LEDs thru the long lighting periods of veg growth and reduce the tell tail heat sig, or am i reading things wrong? forgive me if i got this wrong , like i say, still quite new to this with my first grow.