DIY LED wavelengths and total wattage

[MorganAtlanta]

Have people been successful with just 3W 660 nm reds, or is it worth adding some 630 nm reds and/or 455 nm Royal Blue?

Have people been using the same total number of watts of LED as the guidelines give for CFL, or can we back off the wattage a bit with LED since the light should be more efficient in terms of photosynthetically usable radiation?

[Ace25]

I am using a full screen and I am now only running 4 660nm reds per side, so with the driver, about 30w total to light both sides. Originally I was using 60w+ when I had 10 LEDs per side, whites though. I think 4 has the be the bare minimum though, but that was what I was trying to learn myself, what was the least amount I could use to achieve the same or better results than 26w CFLs. I think I am getting 2-3x better results with the LEDs than I was with the 26w 2700k CFLs even just using 4 LEDs per side (4 is needed to have good coverage on the screen), but I would bet if I added 2-4 more per side I would get even more growth. As it is now, I can only go 4 days tops between cleanings because the screen grows VERY fast and thick, even though I clean the screen very good each time, much cleaner than most people would recommend. On my other tank I also have an ATS and use 42w CFLs and I can easily go 2 weeks between cleanings if I wanted, it doesn't grow very fast at all.

Floyd is going to be able to tell us the answer on the other LEDs pretty soon. I really can't say if adding 630nm LEDs would really help that much, 640nm seems like it would help a lot more, but I have never seen a 640nm 3w LED. I don't think 455nm royal blues help at all, I think 420-425nm LEDs may help, but right now that is just a guess. Right now I think 660nm will provide a good 80% of the spectrum needed for algae, the other LEDs will add just a little.. "think" is the key word there, I do not know for certain at this time, but again, those will all be answered by Floyd pretty soon. I know I am anxiously waiting to see his results. I just know 660nm does indeed work very well by itself.

[tebo]

Ace25, a question:

You say that you are using 4 led per side, are those 3w leds?

What are the dimensions of your net?

Im asking because im about to buy the 3w Osram for my screen, my net’s dimensions are 46”x8”, i would like to know if 20 leds are enough putting 10 per side

thanks

[Ace25]

Screen is the standard 13.5"x10.5". Yes, 3w LEDs run at 750mA. I am thinking on a screen as big as yours, 12 per side is going to be needed.

[kerry]

Are you running a lens on these LEDs? If so what angle. I am planning a build to replace my CFL driven scrubber. Its 10G with a 2.75x6.75 horizontal sloped screen single sided. I was planning on two or three 660's with 80 degree lenses. I thought 3 might be overkill after reading you have 4 on you full size screen. I think I am going to be about 2 inches from my screen with what I have to work with.

[kerry]

After some more research I see the OSRM 660 does not have a lens offer at the rapidled site. ???

[MorganAtlanta]

Wow. Going by the standard guidelines, I'd expect a 13"x10" screen to use about 65W of CFL per side, but you are getting great growth with just one wavelength at less than half that wattage. Very cool. Now I've really got to go LED...

As for the other wavelengths, I was going by the chlorophyll numbers of peaks in A absorption at 453 and 660 and peaks in B absorption at 430 and 642. Based on this chart below, it looks like putting in at least some light in the 400-480 range would be helpful. I have some extra CREE 3W 455 nm LEDs laying around, so I'll probably throw them in.

[attachment=0:yeca7cib]absorption spectrum.jpg[/attachment:yeca7cib]

[Ace25]

Based on this chart below, it looks like putting in at least some light in the 400-480 range would be helpful. I have some extra CREE 3W 455 nm LEDs laying around, so I'll probably throw them in.

That was exactly my thinking as well... so I tried it, only to learn that Royal Blues not only did not grow algae, the algae seemed to die off with only royal blues. I was testing to see if Royal blues by themselves could grow algae, I tried mixing Red/Blue and noticed slightly less growth than all Red so I took it a step further and tried all blue. Algae turned to mush, didn't grow, and a gentle wipe of the finger would make it come right off. So after that experiement I came to the conclusion that algae does not have a preference for 455nm light, but I still don't have the answer for 430nm light.

As I learned there are a lot of other factors that go into photosynthesis besides just what the graph shows. Like in corals, corals have no way to regulate 660nm light, but it can regulate 455nm light, so while in theory 660nm light will grow corals, you have to be extremely specific on putting in the right amount of light without putting in too much or else the coral will die. There are 2 different areas for absorption, the lower and higher wavelengths, in corals if you give it too much 455nm it can take the extra light that goes beyond the max absorption and roll it over into a "storage" area in the 660nm area, kind of like an extra gas tank, but the flip side is corals do not have that same ability with 660nm, if you give it too much 660nm it won't put the extra into the 455nm area, it simply causes the zooxanthallea to leave the coral in turn bleaching it, and if nothing is done to correct it, the coral will die. The same thing can happen with too much 455nm as well in corals, but it takes a lot more since there is a mechanism in corals to put the extra light into storage so to speak. That doesn't mean anything with algae other than the fact there is obviously more to photosynthsis than just giving algae the light that the graph shows would be ideal.

[MorganAtlanta]

The 3W 430 nm LEDs are hard to come by, and expensive ($15/ea). If running all 660 nm works well, then I guess it doesn't make sense to bother with anything else.

[jnad]

Hello!

What about this panel (only 1 watt leds), will this panel work very close too the screen?
http://www.ledwholesalers.com/store/index.php?act=viewProd&productId=421

jnad

[MorganAtlanta]

Might work, but I can DIY exactly what I need for half that price. I'm going to convert an SM100 style scubber, so I need something more rectangular.

[jbmartelle]

I didn't see an answer to what lens might be utilized for the 3w leds? I'm assuming none since it's only 4-6" from the screen.
LOVE these results. I'm using six 23 watt 2700k to filter my 200 gallon according to feeding habits. But this LED recommendation would greatly reduce my watts and need to replace bulbs.

[Ace25]

http://www.rapidled.com/servlet/the-172 ... Red/Detail (http://www.rapidled.com/servlet/the-172/Osram-660nm-3W-Red/Detail)

No secondary lenses are available for this product (yet)

So, answer is, no optics since they don't make any so far that fit. ;)

[Floyd R Turbo]

I asked the horticulture guy about why there's no 640 LED and brought up the issue about the peaks and he pretty much scoffed at it, said there's a lot of misinformation out there. I don't know if they made the 630s and 660s because that's all they could 'hit' or because that's exactly what the horticulture industry demanded. All I know is that 630s grow plants a heck of a lot better than 660s so I don't think trying to hit 640 instead is going to make a hill of beans of difference, based on the conversations him and I have had.

[kerry]

I just bought some 660's from this ebay store and 120 degree lenses that I believe can be made to fit or they may not need any modification. Here is the link. http://www.ebay.com/sch/ledtoplights/m. ... ksid=p3984 (http://www.ebay.com/sch/ledtoplights/m.html?_nkw=&_armrs=1&_from=&_ipg=25&_trksid=p3984) anyway he has great prices, I got everything but the heat sink which I got at rapidled.

[Aeros]

I would love to see build pics and first few weeks results.

[kerry]

I will post some pics. I should have the parts middle of this coming week and hope to build it next weekend. Its a horizontal type modified from an Eclipse hood filter that I am going to swap first. Its now powered by a CFL 23W. Its a 10G Nano with a 2.75x6.75 screen. I am going to start here first as this one only grows brown algae. It still keeps my nitrate between just detectable to under 5PPM and the phosphates to zero-ish but, I think I can do better.

[tebo]

Screen is the standard 13.5"x10.5". Yes, 3w LEDs run at 750mA. I am thinking on a screen as big as yours, 12 per side is going to be needed.

how far are the net

thanks man, I appreciate your response :)

[Nazrac]

I have thought about building a turf scrubber for a while now but have been neglecting my husbandry for too long. But on topic I went to mouser and found a really interesting LED some might think is pretty cool. http://www.mouser.com/ProductDetail/LedEngin/LZ4-40R200/?qs=sGAEpiMZZMsgllGlynFdfkePTIzizdWGKTnrluW%252bEL Q%3d Now dont get me wrong its 16 bucks but its 4x 2.5 watt 660nm led dies in one package on a star. The other cool thing is that it's naturally (no optics) a 95º beam pattern. Using either a driver from rapid or steves you could do 2 of these on each side of a 13"x20" screen I would imagine. You can drive them up to 1200ma but at 1000ma it is 2.3W of actuall light output :o use the driver from rapid and you can go to 1050ma; one driver to rule them all :lol: Well that will be true if my math is right. 11.4Vf at 1000ma so if the driver can put out 48v at 1050ma should be fine to run 4 of them 11.4V*4=45.6? But a dual from steves would do the same but probably easier on it as it can do 2*25VDC on each channel. Here is the data sheet on the LED http://www.ledengin.com/files/products/LZ4/LZ4-00R200.pdf

[Ace25]

Since those are 10w LEDs you are correct, you can only run 4 of them per meanwell driver. Meanwell driver = 48w, 4x 10w is all you will get out of it, you could try and squeeze 5 out but you will reduce the overall output to each LED slightly. That would probably be a good thing for the LEDs, but hard on the driver and you would probably need cooling on the driver as well pushing it that hard. You would also need some serious cooling for the LEDs because the heat generated at the point of contact is going to be 4x as much heat to dissipate from a small area. I am sure they would work if designed correctly, but to me it seems like too much work trying to cool them if run at full power, or even 1/2 power. I have tried to "cheat" in the past, run LEDs in series in order to put more on 1 driver than is recommended, just ended up blowing up the driver in under 30 days doing that.

[MorganAtlanta]

I just pulled the trigger on 10 3W 660 nm and 10 3W 630nm. So we'll see how it goes. They won't be here for a month. I'll post some pics when I get the conversion from T5 done.