LED grow lights

[Oznog]

I kinda wanna try an aeroponics setup for growing lettuce, tomatoes, spices, blah blah. Really. Not pot. I want tasty vegetables. I seldom buy lettuce because I make 2 sandwiches and throw the rest away in a week. If it's growing I'd just trim what I need and let the plant keep growing the rest of the leaves.

Just checking in on this. Is there any benefit to LEDs compared to CFL/HPS (high pressure sodium) lights? I saw something on being able to select the wavelengths the plants actually need. The installation could be versatile, I could mount the LEDs on flexible stems to arrange them as needed. Then again getting thousands of lumens in LEDs is pretty expensive. On the other hand I can get a LOT of Rebels off of Future Electronics for the hundreds (~$400) it costs for proper HPS growlights.

Does the photosynthesis/watt with LEDs actually go up above CFL/HPS results? I know LEDs are not remarkably better than CFL/HPS in terms of lumens/watt (and generally worse actually) but if photosynthesis only uses certain wavelengths out of the CFL/HPS output then I could see where color LEDs could produce more results per watt by producing only the wavelengths used. Less heat could be produced too, making a cooler growing environment which is better.

I tried to Google this and got a lot of pages of people selling shiat, whom I was not inclined to believe. Anyone have proper knowledge of this?

 

[LukeA]

Really. Not pot.

I have a friend who asked me to build a grow light for pot. I declined.

I don't have any proper knowledge, but it would seem to me that plants would like any light that isn't green.

 

[Oznog]

Yeah actually green is the wavelength the plants DON'T need. If they used it, it would be absorbed not reflected. It is unproductive to generate these wavelengths, it's wasted power and heat generation at the light and some of it will be turned into heat at the leaf. I know IR would just be dissipated as heat at the leaf and shortwave UV is surely damaging to the tissue not helping.

There are a bunch of charts showing there's two peak centers of usage, one in the blue range another in the red-orange. Apparently both are needed, there are descriptions of how the balance serves different biological functions and makes them grow vs flower or something but so far the pages sound like pseudoscience- trying to sound scientific in the lack of any actual data or scientific understanding. Basically speaking out of their asses.

Wikipedia says photosynthesis uses only 2% of the visible spectrum then "[citation needed]". Well, if one could give it only what it needed with LEDs that could in theory make LEDs a far superior choice. But this is Wikipedia.

I'm also unsure of whether a couple of tight bell curves at the red-orange and blue ranges will actually serve all the biochemical needs. The usage graph shows SOME absorption in other wavelengths... I wonder if they're essential biochemical steps of the photosynthesis process which would fail without small amounts of these wavelengths?

 

[Illum]

it would be a very expensive....and very tedious experiment...

I can say with confidence though....X1, Y0, YA, and X0 luxeons kill plants in a short time [all leaves turn white, a surefire sign that chlorophyll is dying out....] luxeon Cyan doesn't work either


NASA has been using LEDs to successfully grow plants in space...supposedly they use some UV lights in combination with the visible spectrum....but due to the widespread exploitation of using the LED for illegal operations, the information that can be valuable for agriculture is now a trade secret for drugsellers

 

[LukeA]

Yeah actually green is the wavelength the plants DON'T need. If they used it, it would be absorbed not reflected.

I know. I said so. Plants reflect green light, they don't absorb it. They absorb other wavelengths.

 

[Oznog]

http://en.wikipedia.org/wiki/Photosynthetically_active_radiation

Well I'm looking at this, it's getting interesting. It's no surprise then that white LEDs would kill plants, the wavelengths feed the mid-400's area ok but supply next to nothing for the peaks in the mid-600's range.

There are two peaks to service, one at around 450nm so a Royal Blue can supply that. Royal blue produces few visible lumens/watt but that's an eye sensitivity issue. Rebels' Royal Blue are rated in mw output which shows it's around 25% efficient which is "neat". Actually the top chart's breakdown shows that chloro-b's peak... ouch, Rebel-cyan almost hits it but no its tight line at 505nm is only "close". The Relative Power Distribution graph shows that it's at 0.6 relative power at 490nm and chloro-b's peak is down from 85% to 40% at the 505 peak emission of the cyan.

On the other side there's a peak need around 670nm which is difficult to meet with LEDs. Well Rebel-red would feed chloro-B's peak but not chloro-A. That 670nm peak is longer than red LEDs come in! I looked up Cree XR-C Red and it's like 635nm. The next step up is IR LEDs and that's way TOO long.

Reading up on photosynthesis, it appears this proceeds in 2 stages (the "Z-scheme"), and I assume each of those high/low wavelength peaks corresponds to a stage thus without both you have nothing. I don't know the difference the 3 photosynthesis types makes for each plant. Lettuce, tomatoes- do they need A,B, and carotenoids? How much?

The highly peaked nature of the graph IS promising. In theory if we could meet like 4 key spectral points, you could grow on a tiny fraction of the power since the process is so unresponsive to all that stuff in the middle.

Now I see why some pages suggested it as a supplemental since LEDs can't provide the longwave for chloro-a. This leaves unanswered questions. Specifically, I mean this shows absorption but not need to complete the 2-step Z-scheme. Could the process need a whole lot in the shortwave stage and just a little boost in the high end that a small CFL could fit? 'Cause that's possible. Do some plants have a lot of chloro-a vs chloro-b vs carotenoids?

 

[Sable]

I have two habanero peppers growing under a CreeBar hosting a RebelStrip from the Shoppe.

The RebelStrip has alternating red and royal blue Rebels from Future (Thanks, Wayne!) and is driven by a 750ma Xitanium.

It's working out almost supernaturally well.

 

[Oznog]

Huh. And that's the "red", not "red-orange"?
One red to one blue, alternating?
How many Rebels total?
Is there any additional light involved? Room light?

We had a discussion awhile back that running Rebel red or red-orange much over 350mA is almost pointless. The higher thermal impedance of this technology combined with the high efficiency loss at elevated die temps means there will not be a lot of additional light output even though at 700mA it's at 250% power (doubled current + increased forward voltage). It makes more sense to use more Rebel devices.

 

[Stillphoto]

Check these out
http://www.homegrownlights.com/

I actually found them while looking for some other indoor led lights. These seem to be what you need and are pretty cheap if you get the kits and solder it yourself.

 

[snarfer]

Looks like there are through-hole LEDs in 660nm. For example Lumex SSL-LX30448SRC, which is available from Digikey.

I remember reading a few years ago about a Dutch company that was using an LED system to grow produce underground. Seriously, not pot, it was in the newspaper. They had a system where the plants were put in little carts and sealed up in there. The LEDs were inside the carts, and the carts traveled slowly down a long track over a period of 4 to 6 weeks. At the end of the track the cart was opened and the produce harvested.

Of course the produce was probably completely tasteless like everything else in the Dutch supermarket.

 

[knabsol]

Hi!
This has been discussed here before so you can always pick up an old thread in this if you want some more information. I´m on the process building one for trying it out. I´m using the luxeon rebels. If you want more info I can allways post some pics and data later.
Good luck!

 

[Gryloc]

Oznog,

Go ahead and try it out. I found that royal blue emitters and a red-orange emitters work very well at keeping plants alive. …And it was not for pot.

I was in town a few months ago when this kid, who had a handful of picked four leaf clovers, handed me one of them. I thought that was pretty cool. So, when I got home from work, I decided to revive it because it was already limp and dead looking from the heat. So, I fashioned a mini plant holder, which consisted of a beer-cap and some looped wire soldered to it to support the tiny stem of the clover. Then I pulled out a spare royal blue rebel that I had (from a unique LED jewelry project) and a half-completed taillight project, which consisted of 3 red-orange rebels on a Cree star. I wired both in series to my CC bench power supply, and hung it about a foot over the picked clover. The red was overwhelming, so I only used one of the three r/o rebels in series with the royal blue.

So, with only 200mA of current going to both r/b and r/o Rebels at 0.3m, and a slight dribble of water in the beer cap, the clover perked right back within hours! I continued to give it water sparingly, and reduce the current down to about 20mA at night (200mA in the day). Surprisingly, this picked leaf lived for about a week before I noticed that the clover started to get "mature" (where the green darkened in the daylight, and the cells looked like they lived past their prime). Eventually, it died after a week (got limp and collapsed). I think that it lacked nutrients, or maybe the leaves just lived their longest. I am no expert on field plants, so I do not know how long to expect a few leaves to last.

I can imagine with a nice array of Rebels (r/b and r/o), your few lettuce plants will grow well. I also heard that the reds were used for the reproductive system, so I do not know if it is even required to grow plants strictly for consumption and not expect to collect seeds. It would be cool to actually be able to enjoy the tastes of food that got its power to produce food from LEDs. While you are at it, test most of your plants with the r/b and r/o LEDs, then maybe one or two with just royal blue. It would be neat to see what is required just for sakes of harvesting.

Do not forget that the royal blue and red orange LEDs make the leaves look black (normal) and lifeless. The light may be distracting if it is near where you work due to the odd colors (I got used to it). Have a white light source nearby to check daily how your plants are growing. You would need it to check to see if the tomatoes are becoming a healthy red, and if your lettuce is a nice green.

I plan to extend this experiment to grow more plants (venus fly traps and not pot). I found a light fixture at a PetsMart for reptiles that produce a large amount UV-B light. I got it but have not found the specs to see if the light peaks match that of photosynthesis. It uses two 9W 9" U shaped fluorescent bulbs. It is very small, but I hope it works as well as those coral grow lights (which I did not get because they are large linear fixtures). I may resort to making an array from what ever Rebels or Luxeon 1W or 3W emitters that I can find for cheap. The royal blue Rebels are more efficient than the old Luxeons, so I will just have to compare and buy what ever works best.

Good luck with the project. I hope you continue with the project since they Rebels are relatively cheap and easy to power (with low currents). I wonder if 12 blues and 12 reds will suffice… I guess you have to determine the actual light output of each LED color (depending on the array), and match it to the needed lux needed to grow plants. Then, you can adjust the emitter count and fixture distance to the plants accordingly. Maybe you can try using cheap optics to focus the light onto the plants (like the super cheap reflectors from KD/DX). Please post pictures with your updates. I want to see the fruits of LED light, and to see how the finished plants compare to store-bought. That would be awesome! Thanks.

-Tony



EDIT:

BTW, this neighboring thread in the LED section may help determine how well your red and blue LEDs fill the light gap in your spectrum for photosynthesis...
"Spectrographic analyses, pg. 3":
http://www.candlepowerforums.com/vb/showthread.php?t=200385
I remember seeing this at the LED museum long ago, but forgot about it. It is convenient how it sort of popped back up for you due to some updates.

 

[Oznog]

This actually started because of a thread elsewhere about silly as-seen-on-TV products. One was the "Aerogarden", a cute countertop garden.

Googling it shows that, except for the high priced seed kits, it *actually* works, and quite well! Grows damn fast. It is not hydroponics but aeroponics. Aeroponics leaves the roots suspended in air and mists them frequently. Aeroponics has been shown time and time again to be vastly superior to hydroponics. The initial $120-$200 price isn't all that bad but the "seed kits", well as my dad said, "that's how they git ya". Hahaha $20 for 7 seed pods of one type (except for the spices where you get several types). $20 would have bought a LOT of Romaine or tomatoes at the store! Ya people have already figured out how to make baskets and support medium so they can get it growing without that stuff. It's actually a bit funky because the pod needs to support the weight of the plant entirely since the roots are in air underneath. It would simply sink through a bit of starter soil and fall over or fall through.

Well naturally the next step of planning is hacking an existing one or hacking a whole new setup. There's some stuff on the internet about DIY for sure, making it into a big Rubbermaid container, the tough part being building a reliable mister that won't clog. Then weird changes like this LED thing come up!

 

[Oznog]

EDIT:

BTW, this neighboring thread in the LED section may help determine how well your red and blue LEDs fill the light gap in your spectrum for photosynthesis...
"Spectrographic analyses, pg. 3":
http://www.candlepowerforums.com/vb/showthread.php?t=200385
I remember seeing this at the LED museum long ago, but forgot about it. It is convenient how it sort of popped back up for you due to some updates.

Huh. It looks like the older red LEDs have a lot looser wavelength distribution, AND they're a longer wavelength, into the 675nm chloro-a longside peak. Interesting but it's hardly economical to produce light with hundreds or thousands of ancient low-output LEDs.

I'm still not sure how this works. Will it grow the same if we don't meet all 4 peaks of chloro-a and chloro-b? I saw where carotenoids are a different thing, responsible for some specific nutritional content. So leaving out the growing wavelengths for that would be bad, but it looks like the Royal Blue's gonna feed that nicely along with the chloro-a's shortside peak.

 

[Oznog]

Hmm. Well, Google shows widely varying recommendations of about 1000-10,000 lumens/sq ft, courtesy of the marijuana growers' dumping of info on the web. Let's throw a figure out there that the LEDs are 10x more productive per lumen and a middle-of-the-road 5,000 lm/sq ft x 2 sq ft growing area.

Btw, I would like to remind all that we ARE still talking about lettuce and tomatoes here. Really. Maybe strawberries. Surprisingly, there is no emoticon for "not-winking" as in "take this literally not wink-wink".

So, I need like 1000 color lumens. Well, I *can't* translate those lumens into royal blue mw. Ouch. Huh. I can make 1000 lumens with 25x 40 lumen 1W red Rebels. Again, I'd like to assert that reds should generally not be run above 350mA the returns are not really worthwhile. Well, I'm guesstimating the royal blues are more efficient and do more important work so it *may* require less but that's speculation, well, more speculative than these other speculations anyways. So I'm just gonna throw out the figure of 25x devices total, some red, some blue. Doing the Spidereye-6x boards really we're talking about 4x well-sunk Spidereyes here.

I did see some scientific evaluations showing photosynthesis has limits. That is, once you reach the maximum illumination needed per sq in of leaf, more will not help because the process is limited by the biological capabilities of the plant. However, there will be gains in ensuring that all the leaf area gets illumination and that's a trick since some leaves shade others, etc. I guess that's where reflectors come in.

The lambertian distribution troubles me, this is hard to deal with except to cover the surrounding area with reflective foil. The Spidereye concept isn't going to work with any OTC optic to make a focusable beam. If we place it 3" from a big plant then we'll exceed the photosynthesis limit locally on ONE part of the leaf, wasting the light power, whereas further out the leaves don't get enough light. Yet if it's place 2 ft away the leaves will get even light but there will be a significant amount of uncaptured light off to the side.

Running a 2sq ft garden off 25W, that's amazing. Even if I'm off badly in my estimations, 50W would still be amazing. This is hardly a lot of power. My refrigerator to keep the lettuce I bought at the store cool uses more than that. It won't heat up the growing area, which is good because I don't keep my house really cool in the summer and lettuce likes cool growing temps or the leaves won't be crispy. At least that's what I read, I've not tried to grow it even outside.

 

[tsask]

NASA has made some inadvertent strides here. Minor cuts heal faster in presence of some light frequencies, blue/red?. They have also made advances growing veggies.

For anybody reading this who has other intentions, the US Govt has placed enforcement at the top of the priority list. It is possible to get more prison time for an "illicit grow operation" than violent crimes like rape, murder etc. Stick to LED grown broccholi sprouts and life will be good.

 

[VanIsleDSM]

NASA has made some inadvertent strides here. Minor cuts heal faster in presence of some light frequencies, blue/red?. They have also made advances growing veggies.

For anybody reading this who has other intentions, the US Govt has placed enforcement at the top of the priority list. It is possible to get more prison time for an "illicit grow operation" than violent crimes like rape, murder etc. Stick to LED grown broccholi sprouts and life will be good.

Yeah.. that's because the US govt is more concerned with the tax dollars they're missing out on than people's actual safety. Everyone knows alcohol is a far more destructive substance. I imagine it would be much safer to grow a little bit though, than it would be to try and buy it. I've heard that here, 4/5 cultivation charges don't even make it to court.. but BC is known for it's leniency in that area.

I've seen a couple documented comparisons of LEDs to HPS, but none with proper high power LEDs. The HPS did better, same amount of watts, than the red/blue LEDs, but they were only 5mm guys. I think with some rebels, could be just as good as HPS or MH anyway.. it's just getting reds that are deep enough.. too bad there wasn't a high power LED between IR and RED.. that's what you'd want.

Of course the best way would be some mirror and lens system to focus daylight inside your veggie box.. and then only turn lights on if you need more daylight than you're getting at the time.

 

[red_robby]

I've looking into this for a little while myself, I found a pretty useful link

http://environment-center.blogspot.com/2008/07/growing-plants-with-leds.html

 

[Gryloc]

Really quick on light requirements:
Convert the lumen output of those figures from google to actual power output. Lumens are a solid unit (if they were measured correctly), but you would have to determine how many mW of energy of a certain color is required to create 1 lumen of humanly visible light of that color. There are lumen to radiometric power conversions somewhere I believe (I think).

So, find the spectrum of those similar to what those HID or fluorescent fixtures produce, and then find how many lumens (roughly) are produced from the certain reds and blues. Then, convert that to actual radiometric power for each, scale that number a bit for your sized veggie growing "operation" lol (serious face when laughing), and finally use those numbers to determine what is needed of the red and blue emitters.

Finally, the Royal Blue Rebels are efficient, as the Red/Oranges are about the same as they were a couple years ago, so why not use the Luxeon III R/O emitter? It is a beefier package that can handle more current, and produce around 190 lumens of red/orange from one emitter! You may need to spread that heat around a bit by using a thin copper heat spreader on the base of your existing heatsink (due to the higher W/cm^2 density), but it may save you a few bucks and make the red lighting less complex (with less parts). The Rebels may be your thorn because you may need many more of those. Did you check on the royal blue TFFC K2s? Maybe you can get more light density from one of those (like how you can by using the 3W R/O's). Just a suggestion.

Now I got to leave my workplace before they turn off the lights on me (atleast rush-hour will be less severe... slightly)

-Tony

 

[Oznog]

Neat. I looked around and once again the pot growers are a useful source of info. High Times magazine found this 80W "UFO" thing here did grow plants competitively well with a 400W HID. Well they were obviously growing pot but I'm assuming that if it can grow that functionally then it'll grow lettuce well too.

I'm curious what LEDs those are. I think there's supposed to be 80x 1W emitters here. They kinda look like they could be the sq 4-pin Superflux lights. They look far too close together to be 1w Luxeon emitters or anything but I'm not sure of the actual scale of the thing. Hmmm... might be questionable if this is actually an 80W fixture, and in any case we care about lumens out (in desirable wavelengths) not watts in. So it's really important what the device's lumens/watt actually is. Does this seem a bit small and enclosed for AlInGaP whose output drops off quickly with elevated die temps? 'Cause I don't see a clear intake-outlet differentiation on those vents which might suggest there's just a fan on the board or sink stirring things around rather than flowing.

They're unfocused lambertian devices by the look of it, unless they're actually just a bunch of 5mm T1-3/4 devices. They're claiming it's "directional" which does support the T1-3/4 hypothesis. Which would be... lame for $599. But it couldn't be 1W/device at T1-3/4, could it?

Actually based on the device's thickness and LED density I might suspect they've got a PCB carrying SMD LEDs. Maybe Cree XR-C or any of the small SMD LEDs out there- they come in 1W. But that would result in extremely poor dissipation using just board traces which would smother the output efficiency of AlInGaP-based devices.