LED test on SM100

[SantaMonica]

9 days of growth after cleaning. Fluorescent on left, LED on right. LED is the 50 watt Grow Light from EshineSystems in China. Actually uses 41 watts per the KillOwatt. Was raised up 3/4 inch to fit in middle of scrubber window. Camera is unfortunately an old one, since new one is being repaired. Growth was very similar on both sides; too similar to tell them apart. Growth on the LED side was floating a bit higher since it had no light near the bottom. LED was about $140 including shipping, and I asked for the black case:

http://www.radio-media.com/fish/SM100%20LED%20Test%201.jpg

http://www.radio-media.com/fish/SM100%20LED%20Test%202.jpg

http://www.radio-media.com/fish/SM100%20LED%20Test%203.jpg


Video:
http://www.youtube.com/watch?v=LKcib4YaoTc
LED:
http://www.eshinesystems.com/growlight/ ... ght-3g.htm (http://www.eshinesystems.com/growlight/50W-led-grow-light-3g.htm)

[Ace25]

Awesome! Thanks for stepping up and experimenting with LEDs yourself. Looking forward to updates over the next couple months to hear what your thoughts are after trying them. Really looking forward to hearing your thoughts at the 30 day mark on color, growth, and density (weight) of your harvest in comparison to each other.

It does seem already, that after only 9 days, that the LED light is looking very promising. Wattage is about the same, price is almost comparable (LEDs costing slightly more due to shipping costs now), but the big saving will come from not having to replace bulbs if the LED light can survive long term.

[Rumpy Pumpy]

Well done SM - Looks like LED scrubbers will be the future then.

[Marksfish]

I spoke to these people today and delivery to the UK is around £100 landed, using DHL. If you manage to avoid the VAT (which I doubt as DHL will automatically add it), you will be looking at about £80.

Mark

[SantaMonica]

12 day cleaning:

http://www.radio-media.com/fish/SM100LEDTest4.jpg

http://www.radio-media.com/fish/SM100LEDTest5.jpg

[Floyd R Turbo]

It looks like the LEDs power through and grow more green. This supports my theory that the T5HO has a lot of 'wasted' light that the algae doesn't know what to do with, so you get the yellow/brown growth. I can't wait to get my fixtures and run my experiment...

[Aeros]

Brilliant work SM! would love to hear your take on the outcome. And any anecdotes you have about the test.

[maglofster]

Very nice!

Have you thought about using the slightly larger (three leds wide) 100W version? I think it would fit better? Or would that be overkill?

[SantaMonica]

I don't know if it would fit. As it is, these need a lot of testing. And they are not waterproofed.

[Floyd R Turbo]

Yes the round ones from Gotham / Lighthouse have been tried and one of them fried from the moisture. So if you run them remote from the sump or fed via a dry air source, then you're much better off. But they definitely cannot be ran in an enclosed sump cabinet with an open sump, IMO. They work on SM's setup because his run on shelves next to the tank.

[kbb0118]

Nice! I'm looking forward to seeing how your fixture holds up to the humidity! My LED fixtures did horribly.

[Ace25]

I don't think SM has to worry about humidity with his setup. ;)

[kbb0118]

I meant for my SM100 (which is in my stand).

[nickq]

How audible are the fans in the LED unit?

[Floyd R Turbo]

I meant for my SM100 (which is in my stand).
You didn't have an SM100 - didn't you have an open screen with the two round fixtures, or am I totally losing my mind?

[nickq]

As I am now getting an SM100 built in the UK I've been looking into lighting. I really like the idea of this LED as the bulb cost of replacement fluorescents is not great, not to mention the environmental impact.
So I was chatting with eshinesystems today who are watching this thread also, they have recommended the following configuration of LEDs in the 50W unit for the best algae growth:
60% blue 455-460nm
25% red 655-660nm
10% daylight 5500k
3% violet 420-430nm
2% UV 380nm

I guess it makes some sense as the chlorophyll b peak of absorption is higher in the blue than the chlorophyll a in the red. I am assuming that this is related to SMs previous explanations of photons absorbed directly correlating to nutrient absorption. However I have read that there are more photons per watt towards red than blue, but more energy in a photon in blue, so probably the overall energy is equivalent or something? So are the number of photons more important to the energy per photon?

I only want to buy one set of lights, my wife will not be impressed if I spend loads more cash on experimenting :|

Cheers,
Nick.

[Floyd R Turbo]

Those guys at eshine don't know algae. There's no way they should have suggested that much blue. That's a vegetative growth spectrum. Also there's no need for the 5500K or the UV.

All attempts at LED so far have show a major lean towards red.

You would be better off with a 7:1 or 8:1 ratio of red to blue and do 50% of each. That's 425 and 460 for blue and 630 and 660 for red, nothing else.

[Floyd R Turbo]

Actually from all the research I have done, we're the only ones who really have any kind of data whatsoever on what grows algae. We being those that have built scrubbers (myself not included at this point). So don't go by what someone else 'thinks' at this point, because they're purely guessing. I have had long conversations with a very helpful gentleman who has a horticulture degree and 15 years in the industry, and he has only referenced 2 studies that even come close, one from NASA and one from I believe Cornell University (or another one that starts with a C)

[nickq]

The growth in SMs pic looks plentiful anyway, so I guess the off the shelf unit he's got there is a good bet.

[kbb0118]

I meant for my SM100 (which is in my stand).
You didn't have an SM100 - didn't you have an open screen with the two round fixtures, or am I totally losing my mind?

I built a SM100 too. :D I just haven't had time to make a thread about it.

[Floyd R Turbo]

so did you put LEDs on that one too? Which fixtures and which ones failed on you (the round ones or the unknown ones)?

[kbb0118]

so did you put LEDs on that one too? Which fixtures and which ones failed on you (the round ones or the unknown ones)?

Sorry for the confusion. The round hydroponic lights did not hold up against the humidity. I have not put LEDs on the SM100 (yet). For now I am using Nova fixtures.

[SantaMonica]

How audible are the fans in the LED unit?

Fans are quiet; you only hear them if you turn the flow off.

[SantaMonica]

T5HO Side, 7 days:
http://www.radio-media.com/fish/SM100LEDTest6.jpg

LED Side, 7 days (T5HO used for pics):
http://www.radio-media.com/fish/SM100LEDTest7.jpg

T5HO Side, 7 days
http://www.youtube.com/watch?v=p_zEupsHHmk

LED Side, 7 days (T5HO used for pics):
http://www.youtube.com/watch?v=bYjeiweuW24

Top Side, 7 days (left side is T5, right side is LED)...
http://www.youtube.com/watch?v=BH7-tXN5LB8

[Ace25]

So what are your thoughts so far? I know it is still early and not all that conclusive. From the videos it seems I can make some initial conclusions (especially the Top Side video in the above post), but don't really want to say mine because I want to hear first hand thoughts from you first. You are going to be able to tell us better than anyone else watching videos because you can see it in person. ;)

[inkidu]

A dawn of a new age.

[SantaMonica]

I'm going to have to swap sides first, to make sure it's an even test.

[lahon]

any futher comment?

I agree to Inkidu------NEW AGE-----

IT make the world better ? T5 is a danger to my home (earth) and my wallet

[SantaMonica]

I'll post more pics after the 3rd cleaning. So far both sides are the same.

[kerry]

I just finished my 4 LED light bar with 3W LED's. Right now I am 2" away and do not have any lens's just straight LED's. Any suggestions if 2" away is OK?

[SantaMonica]

I have no idea.

[kerry]

Well, 2" is what we will go with for the next couple days. When I had the lens's on the LED's at 2" away it turned the area of focus on the screen back to white, so I guess the 3Watt LED's are very strong. I will ride out 2" away with no lens and see what becomes of it.

[kerry]

SM, is your LED array the 1Watt LED's? How far away are those from the screen?

[Floyd R Turbo]

That is a 1W LED array and it is 2" from the screen. bare 3W leds are too much 2" from the screen. You will need a diffuser. search LEDs builds on this site and you'll see.

[SantaMonica]

As the growth hits the window, it's about .5 inches away. I've not noticed any spots or holes, so apparently the water refracts the beams enough for even growth.

[SantaMonica]

Updates on the LED test. 3 pics from 2nd cleaning, and 1 vid from 3rd cleaning:


2nd cleaning, 7 days, top:
http://www.radio-media.com/fish/SM100LEDTest8.jpg


2nd cleaning, 7 days, T5:
http://www.radio-media.com/fish/SM100LEDTest9.jpg


2nd cleaning, 7 days, LED:
http://www.radio-media.com/fish/SM100LEDTest10.jpg




3rd cleaning, 11 days:
http://www.youtube.com/watch?v=nBEkVdEQ_oQ
.
.
.
.

[lahon]

the led side seem fresh green than T5

[Aaarrrggg]

If the LED experiment proves successful, will the current Santa Monica 100s be upgradable to LEDs?

[kotlec]

Do you extend cleaning from 7 days to 11 days ?

[SantaMonica]

I clean when I can.

It will not be available unless you buy the lights yourself from China and take your chances. The lights were designed for hanging over plants; not only are they un-sealed, but they have an open frame, and they suck air/salt/moisture in via the fan.

[kerry]

I clean as needed to. I notice my feeding levels dictate my cleaning times. When I feed heavy its 4-5 days, lighter weeks I can go 7-9 days. My algae is more brownish so I try to clean it before it gets really thick because it can fall into the water and is hard to remove due to the height of the scrubber tank.

[Aaarrrggg]

It will not be available unless you buy the lights yourself from China and take your chances. The lights were designed for hanging over plants; not only are they un-sealed, but they have an open frame, and they suck air/salt/moisture in via the fan.
Ah I see... I guess what I mean is. One day if you decide to make the SM scrubber LED based, will the existing SM100s be backwards compatible?

I'm seriously thinking about investing in one of your SM100s, but before I do I'd like the assurance of knowing that I could upgrade to LEDs later. I'm not nearly handy enough to do it myself. I of course would be willing to pay extra for the new components; I just don't want to have to buy a completely new scrubber down the road.

[SantaMonica]

I can't say for sure.

[Aaarrrggg]

Fair enough... Sorry I know it is a tricky question, I'm just trying to plan ahead! :)

[Floyd R Turbo]

any packaged LED system will generally have built-in fans. The ones I just got are similar in fashion but different in form factor. Fans on the back that blow into the fixture and vent out the edges. No seal between the back side and front side of the PCB board so moisture getting to all parts of the fixture is a definite concern. This makes designing an LED scrubber system that much more complex and that less cost effective...

[kotlec]

Thats why I go DIY route. Oversized heatsink eliminates fans. After fixing LEDs to heatsink wiring them I spray several layers of acrylic lackuer over LED side. That keeps moisture away from sensitive parts and makes whole design moistureproof of some level.

[SantaMonica]

Getting ready for the 5th cleaning. All growth has been about equal to the T5 side so far. Two noticeable things:

1. There are no hot growth spots, even with the LED's so close to the algae; growth is evenly spread.

2. Since the unit runs cooler, the scrubber window does not get warm, and thus there is some condensation on the inside (whereas the T5 side is warm, without condensation). This condensation may or may not help with the diffusion of light from the LED's.

[jaczal]

Hi,
My scrubber is already working 9 months gallery https://picasaweb.google.com/111713615096669370036/Solniczka?authuser=0&feat=directlink. I used LED (2500K - 3500K), but a bit too little power and works on contact. Two screens at 300cm2, with each scraped about 30 grams of algae every week (tank 140 liters).
That completes the necessary elements to perform the test lamp on blue and red LEDs.
Yours

[SantaMonica]

I don't know if it's increased, or it I'm imagining, but the sound from the fans on the LED are detectable from across the room, even with two scrubbers emptying into the sump and the sump door open.

[jnad]

I don't know if it's increased, or it I'm imagining, but the sound from the fans on the LED are detectable from across the room, even with two scrubbers emptying into the sump and the sump door open.

Heiio!

Sorry to hear the fans are starting to be noisy, if you think you can hear a increased sound you probably not are imagining. I am an electical ingenieer and i have some experience with fans from work. My experience with cooling fans is that they has to have very very good quality for long life, especcaly in though envierments. When a dealer advertise with buildt in quality fans there is no garantee for good quality.

I hope your new scrubber designs for 2012 is constructed for leds with no need for cooling fans, fans are no good in the livingroom with the wife sitting 2m from the aquarium looking at the television :D

jnad

[Floyd R Turbo]

Could be that it's a cheap Chinese product. Just sayin. I'll have to pay attention to the fixtures I have once I get them up and running.

[Ace25]

newegg.com? $10 for a fan high quality fan.. they are a consumable part even on a $750 Ecotech Radion LED light... just sayin... I am sure they were cheap fans to start with but replacing fans is something that has to be done on a regular basis anyway (with a better quality fan they will last longer).. heck.. I have to replace them yearly on computers all the time. I have yet to have to replace fans on my LED lights on my display because I went with 140mm fans that run at 10v and only spin at 800 RPM, super quiet, actually no noise at all, but they are big so they still push a lot of air for moving so slow.

[Floyd R Turbo]

Yeah, replacing the fan with some nice ball bearing quality fans will probably go a long way towards long-term use of the fixture. Stock fans I'm sure are junk.

[nickq]

I have the same led fixtures as in this test. The fans have been noisy since day 1 for me, they seem to run very fast.
It's not too bad though as they are on mostly at night.
Now if I could make my scrubber quiet I would be very happy :? It thunders along at the recommended flow rates.

[Doompie]

Also got the same LED units, but on two sides..

It's now 4 days after cleaning, and still the center where the LED's shine directly, there is real yellow ´slimey´ growth.
Around the LED area the growth looks as normal..

Already about doubled the distance to the screen from day 3, and yesterday decided shut off half of the bars on each side. One side only the top ray is on, other side it's the bottom ray.

now it seems like there is some green starting to show..but nothing like before.

http://www.doompie.net/plaatjes/IMG_5543.JPG


So should I reduce hours? or increase feeding? or increase flow? or increase iron? or..?
Thanks.

[Ace25]

I would suggest reducing the hours. My screen looked like yours on my 42w CFL scrubber until I cut the time in 1/2, now it only runs 9 hours instead of 18 and the screen is now solid green (although not much algae growth, at least the entire screen is green instead of yellow slime).

[Floyd R Turbo]

Yellow growth, if you're talking about the yellow rubbery type, but I think yellow slimy is probably in the same category, is a result of not enough nutrients being delivered to the algae compared to the amount of light it is getting.

What that means is either you do not have enough flow, or enough nutrients in the system to support that much algae growth, or both. I would double-check your flow rate first.

If you read up on the new lighting suggestions, which involves doubling the light output and halving the photoperiod (mainly for CFLs and T5HOs so you can use the lamps longer) this is brought up. Ace is right here, cut the photoperiod back until you no longer get yellow growth.

But IMO this is only if the flow is not the problem first. Your pic shows quite a bit of water in the bottom of the box, so either you have a lot of flow, or a very slow bottom drain, or both.

[Doompie]

I think you got it right Floyd..incidentally cut the flow in half, now LEDs are still both half on, but flow is at 100% again..will see how The algae react tomorrow and otherwise switch on all the LEDs..

Thnx allot!

[SantaMonica]

After about 10 or so cleanings, without a doubt the LED side is better. Grows sooner, thicker, and one that I did not expect... the light penetrates deeper to keep the roots alive longer, which makes sense because since green absorbs blue and yellow, the red continues through it.

[Floyd R Turbo]

Awesome. Now if there was only a water-tight and rust proof fixture that was quiet.

At least now I won't have to bother with doing an LED vs T5HO experiment.

Thanks for posting!!!

[Gigaahxl]

The increase in penetration to the roots could also be partially attributed to the source point nature of the LED?
All light from the LED goes directly from the diode to the target with little to no reflecting off other objects in a much more focused manor.

While its been awhile since I've read up on these things.If I recall Metal halide lights are prized in part for their depth penetration(in water) over T5 due to being more point source light in nature.

Just figured I'd throw that out there as a possible contributing factor along with the benefit of being able to tailor the spectrum to your needs..once those have been identified.

[Gigaahxl]

Another thing after looking at the eshinelight used.Due to the fact that the light is used at such close proximity to the target. The 5 Blue LED used in the array could easily be a waste of time seeing as the areas furthest away from the blue(and probably not getting any blue light at 1" distance) seem to grow fine.

Santa Monica, was there ANY noticeable growth difference in relation to the blue LED in the array? It may be that its not a huge deal, but I've used LED's at distances like that and the light pattern I witnessed leads me to believe the blue is not even required.

I'm sure at the 5+ inches that the light was intended to be used at the blue would mix semi-evenly with the red. However at the distance we are talking here if you want to mix colors the diodes I would think need to be nearly butted up against each other. I've seen PCB/Heatsinks that have 2 up to 12 LED spots on them in tight groups.

[SantaMonica]

I don't think the blues are doing anything.

[Floyd R Turbo]

One of the ideas I was going to try on my benchmark testing was to cover up the blues completely and see how much of a difference it makes. I have a feeling it is miniscule though...

[SantaMonica]

In addition to the blues probably not doing anything, there is also the two fans which look to be about 5w each. This would be 10 w of of the 41 real watts that is not producing growth. So less than 31 real watts, maybe 25 if you remove the blues, is doing more growth than the 47 real watts of the T5. So it might be safe to say that half of the wattage of 660's can be used to get the same growth as fluorescents.

[jnad]

Hmmmm, this is both good and bad news. I have orderd two units of 9W 660nm leds for my small nano scrubber, i probably will burn the algae of with this units then?

jnad

[RkyRickstr]

Imma try six 3w leds on each side.. all red.. think will be enough?.. screen is 10 x 6

[Gigaahxl]

IMO 1w LED are best. You'll have to back the 3W LEDs a ways away from the screen to get the right distribution and intensity

[SantaMonica]

You can always run less hours, and if it still burns, run less current.

[Gigaahxl]

Correct me if someone finds better information. From what I can gather the epistar 1w LED are rated at (roughly) 35lm. one place said 35. one said 30 to 50. we can just call it 35 best bet.

Check the rated lumens of the LED's being used. "3w" LED's don't put out 3 times the lumens. 3w LED's are usually actually closer to 2w and are usually less than double the lumens of a 1w LED. Companies that make high quality "3w" LED like cree you can actually drive them to 3w and above get more output but your efficiency per watt takes a dive due to increased junction temperature.

I've done a lot of work on tank lighting and spent a good deal of time over at nanoreef talking to and messaging a couple people there like Evilc66, blasterman, and a few others about LEDs spectrum, light spread,spot lighting effects from LEDs etc.

Since we've just started to hammer out loose guidelines thanks to Santimonica and others. I know for myself I'm extremely grateful to santamonic for confirming what he has in spectrum and wattage thus far. Its good to keep these things in mind because One 3w LED and Three 1w LED's can be considerably different when it comes to light output.This isn't even getting in to variations in lumen output from mfg to mfg when comparing similar LED's.

SANTAMONICA: Thats very interesting the wattage measurements you took. 41 watts -10w for the two fans. -5 for the blues roughly. This is all before transformer AC/DC conversion loss and the loss in the drivers for the LED's as well? or is this after the step down and conversion to DC?

[SantaMonica]

It's power from the wall. And actually I measured again and it's 47 watts, not 41.

[Gigaahxl]

I just got back from talking to the very very helpful folks at eshine about the 50w grow light. They pointed out to me and it had escaped my attention that 1w red LED's run lower voltage than regular white LED and most RED led of that nm run at 2.4v at .33 amps. This makes "1w" 660nm red LED's electrically about .8 watts each. This accounts for a lot of the wattage in vs wattage we expected difference.

They specifically gave me the 2.4v .33amp figures that are used for the red in their system. which is awesome. Most companies support couldn't or wouldn't provide that information. Props to the E.shine Systems guys.

[Floyd R Turbo]

In having my fixture designed, I had much discussion regarding 1W vs 2W vs 3W. The fact is that a 3W chip does not pull 3W anymore. It used to, but then the chips became more efficient, and instead of calling a 1W chip a 0.6W chip (which it is) and a 2W chip a 1.1W and a 3W a 1.7W, they leave it alone for marketing. This may change very soon as it is getting pretty silly and out of whack.

But the fact remains that 3 'standard' 1W chips will provide more lumens than 1 3W chip. This is also why doing a 1W tight array is likely more effective (for scrubbers AND for DT lighting) at achieving even light coverage than a 3W chip array. So people have to get over the idea that the 1W chip isn't as good as a 3W chip, because on a bang-for-the-buck basis, it very much is a player.

[Gigaahxl]

In having my fixture designed, I had much discussion regarding 1W vs 2W vs 3W. The fact is that a 3W chip does not pull 3W anymore. It used to, but then the chips became more efficient, and instead of calling a 1W chip a 0.6W chip (which it is) and a 2W chip a 1.1W and a 3W a 1.7W, they leave it alone for marketing. This may change very soon as it is getting pretty silly and out of whack.

But the fact remains that 3 'standard' 1W chips will provide more lumens than 1 3W chip. This is also why doing a 1W tight array is likely more effective (for scrubbers AND for DT lighting) at achieving even light coverage than a 3W chip array. So people have to get over the idea that the 1W chip isn't as good as a 3W chip, because on a bang-for-the-buck basis, it very much is a player.

It is very nice to have other people around that understand that. This is the reason that last LED DT lighting systen I did I used 1W LED's. It almost makes MORE sense to me. especially when you take in to account the large intensity drop off of LED's the further you get away from center and spotlighting effects many 3w users get.

[SantaMonica]

I hope they keep making the 50w bar light, since I'm using it as the model LED light in all the upcoming 3D drawings.

[RkyRickstr]

So, the question is.. can a driver that was designed for 12 3w leds run 36 1w leds?

[kotlec]

Yes and no.
LED driver is designed to provide particular amperage , not wattage. 1w leds use different amperage than say 3w.
One must always choose driver acording to amperage output required. Roughly it is 330-350mA for 1W led, 700mA for 2w led and 1A for 3W led. Deppending on how many leds you want to drive you must look what voltage boundaries driver has. Also you must know what voltage drop your led have.

Still you can risk to run 3 parallel strings of 1w leds. This way you will still use 1A divided in to three strings. But if any one led blows open, then 2 remaining strings will receive aditional current as driver tryes to maintain same 1A. And all we know that too high current is what kills leds fastest.

[Floyd R Turbo]

There are also constant-voltage drivers so the above is not always true. Running parallel strings can be safe if quick-blow fuses are used also.

[kotlec]

Constatnt voltage drivers are not for directly connecting LEDs. They mostly are for string leds. Strings have resistors incorporated . And resistors are to limit current. Use CV drivers with 1w or 3w leds at your own (great) risk ! :evil:

[Floyd R Turbo]

actually there is a large thread on RC and a board developed by TheFishMan65 that uses CV meanwell drivers. Yes, you have to know what you're doing, but it's a heck of a lot cheaper. You can run 8 strings of 6 LEDs off one driver board and if you configure the board right, 6 channels of PWM dimming.

[kotlec]

Unless he invited somthing new recently. I was following that thread , and even done diy drivers for DT lighting. But those were CC Cat4101 PWM dimable drivers. Using them till now.

Will be checking for news on RC.

[Floyd R Turbo]

Hmm, maybe I'm wrong. I could have sworn those boards used CV drivers.

[Floyd R Turbo]

Nope, I was right. They're using 24V power supplies on the DIY driver board thread here:

http://www.reefcentral.com/forums/showt ... ?t=1759758 (http://www.reefcentral.com/forums/showthread.php?t=1759758)

look at post 969 for one example.

It is true that for the average hobbyist, using a standard meanwell CC driver is easier, because you just direct connect and solder and away you go, for a setup needing a ton of LEDs (like over 100) then you can run into issues with having many, many of those meanwells. The DIY driver allows you to run many more LEDs off a single driver.

Post # 859 shows his driver board. you can use one or two PSs (the blue connections on the end).

But, this is way off topic.

[kotlec]

Not that I want to disappoint you , but ...
This is CC driver. Exactly one I did myself. It uses 24vdc suply. But driver itself is buck type CC . You can change output current via exchanging resistor R3. Driver chip CAT4101 actually is CC driver itself.

First picture shows driver itself and second shows ordinary PSU. Next is Arduino board which sends PWM signal to the driver.

[Floyd R Turbo]

Ok I guess I'm sanding off a little rust, been a while since I went through the motions trying to figure this out. What I was stuck on is that the PS is CV, but thank you for pointing out that the board itself transforms this to a CC source. I guess that clears up a confusion point for me. The meanwell ELN drivers are one piece that supplies CC and the PS on the DIY is CV so that's where I got lost. Most people just use the MWs and when your build gets large, you end up needing quite a few and the $ goes up...

[RkyRickstr]

Pooooffff!!!.. guess what that was.

Im gonna take a pic of my driver and have you guys help me pick the right leds.

[Gigaahxl]

If I recall right making a cheap LED driver isn't that hard. I think I made some Current limiting circuits awhile back that where adjustable current. With a cheap and easy to get LM317 IC a resistor and a POT. Just add X volt DC supply depending on how many LED's you need to run...as long as you don't exceed the power dissipation limits of the IC. But custom matching the input DC voltage to the number of LED in series can minimize that.

[Ace25]

I have seen the DIY drivers like Gigaahxl mentioned, seem ok and cheap enough, but when you can get meanwell drivers for just a few $ more it just isn't worth my time to make a driver. Meanwell makes large drivers that can handle lots of LEDs for under $100ea. I have a friend making a 240 LED light and using 3 Meanwell HLG-185 drivers but he could get by with using 2 HLG-240 drivers more than likely. Both are dimmable via all methods (1-10v, PWM, or internal POT).

This is the cheapest place I have found Meanwell drivers in the US (located in the SF Bay area in California).
http://store.bravoelectro.com/other-c-1 ... 14&sort=2a (http://store.bravoelectro.com/other-c-1.html?page=14&sort=2a)

The ELN60-48D/P is $22.60 vs $36 at rapidled.
The HLG-185 is $85 vs $115 at rapidled

[SantaMonica]

Another update, after removing screen. Had to turn off LED for pic...


http://www.radio-media.com/fish/CleanedGrowthDark.jpg

http://www.radio-media.com/fish/CleanedGrowthLight.jpg