DIY Full Spectrum LED build..

[jedimasterben]

i am now left with drivers,

 

what would you suggest, ELN 60-48P or D, i would be controlling with a typhon controller, but read that 'P' drivers wont even turn on without PWM.

 

 

regards

Zoaguy

Correct, with PWM drivers, you need a PWM controller for them to even turn on, but that's where the Typhon comes in.

 

what length heat sink should i consider for a tank of length 30'', and what width ? tank width is 21''

 

i will be placing LED's 6'' above water line,

 

 

Regards

Zoaguy

I would do two heatsinks, maybe 12"x4".

 

guys,

 

at what current are you running rebel ES ? i think max is 1A, am i right ??

 

 

Regards

Zoaguy:)

NW, RB, CB, Cy all can run at 1A, DR is 700ma.

[uglybuckling]

i am now left with drivers,

 

what would you suggest, ELN 60-48P or D, i would be controlling with a typhon controller, but read that 'P' drivers wont even turn on without PWM.

 

 

regards

Zoaguy

 

As far as I know, the Typhon controller puts out PWM and therefore you need the "P" drivers. The "D" drivers use 1-10v dimming, which the Typhon does not support. Unless somebody knows something about the Typhon that I don't, I think you'd better go with the "P" drivers.

 

 

what length heat sink should i consider for a tank of length 30'', and what width ? tank width is 21''

 

i will be placing LED's 6'' above water line,

 

 

Regards

Zoaguy

 

A 30" tank requires an 18-24" heatsink depending on how much loss of light you're willing to tolerate at the edges. A six inch wide one should be fine. Makers (available from Milad's company, LEDgroupbuy), galvanized aluminum, or whatever else you want to use will work; a fan may become important particularly if you're going to drive your Luxeons at around 1000mA, and so a heatsink with the ability to easily mount a fan on it is really nice (though I'm sure you could MacGyver something, it would probably not look as pro as one that is machined for the purpose...).

 

 

guys,

 

at what current are you running rebel ES ? i think max is 1A, am i right ??

 

 

Regards

Zoaguy:)

 

Jedi runs his at 998 mA. I will resist the urge to give him a hard time about it on your thread.

[jedimasterben]

Jedi runs his at 998 mA. I will resist the urge to give him a hard time about it on your thread.

As long as they're kept cool, you can run them even higher. 1200ma isn't out of the question, but it takes some good heatsinking (and gonads ).

[zoaguy]

As long as they're kept cool, you can run them even higher. 1200ma isn't out of the question, but it takes some good heatsinking (and gonads ).

 

solved most of my confusions,

drivers ELN-60-48P

 

controller based on Arduino.

[jedimasterben]

solved most of my confusions,

drivers ELN-60-48P

 

controller based on Arduino.

Good deal. While you can use the 60-48D, it's not as smooth with dimming (since it is a different type of signal), there is no long-term data on whether or not it's actually good for the driver to do it that way.

 

 

And I would still run them at 1000ma or less, don't push your luck.

[zoaguy]

Good deal. While you can use the 60-48D, it's not as smooth with dimming (since it is a different type of signal), there is no long-term data on whether or not it's actually good for the driver to do it that way.

 

 

And I would still run them at 1000ma or less, don't push your luck.

 

 

i would start @ 400mA and ramp it upto 900mA - MAX over a period of a few weeks, so that everything gets acclimated to change in light

 

edit : i would be getting a typhon to do the dimming.

 

Regards

Zoaguy

[loyalhero90]

Hello,

I almost had my spectrum build perfect. The "daylight" lights which are 6xNW and 6xRB work perfectly with the 48P Meanwell which works well with the PWN controller. The lights actually came on and I did my happy dance but then I tried with the colors. When I plugged in the 48P there was a loud pop and the lights in my room went off. I checked the LEDs and they work fine. Assuming I tested the driver with the MM correctly it still works. Could it be a plug malfunction; as in the plug connected to the 48P malfunctioned or the wiring is off? The plug was actually kind of melted and I'm pretty sure the socket is pushing up daisies (good thing it was on a surge protector); the other sockets on the protector work though...

So far only half work which is really cool but still sucks that the whole fixture can't work. The seagrasses and macro look great and are growing well though. It could still just be wiring since the LEDs are really close together to minimize the disco effect. I don't have disco on the ones that work but the wiring was a PITA. Everything else works fine even the color LEDs but the driver/plug combo just isn't doing well.

Any thoughts and suggestions?

[zoaguy]

Correct, with PWM drivers, you need a PWM controller for them to even turn on, but that's where the Typhon comes in.

 

 

I would do two heatsinks, maybe 12"x4".

 

 

NW, RB, CB, Cy all can run at 1A, DR is 700ma.

 

 

jedi,

 

what do you think would be a good option,

 

2 heatsinks - 12''X4''

or

2 heatsinks - 12''X7''

 

LED's arranges in 1:2 ratio of NW:RB - like AI fixtures

BTW using 80 or 90 degree optics,

 

Regards

Zoaguy:)

[jedimasterben]

Hello,

I almost had my spectrum build perfect. The "daylight" lights which are 6xNW and 6xRB work perfectly with the 48P Meanwell which works well with the PWN controller. The lights actually came on and I did my happy dance but then I tried with the colors. When I plugged in the 48P there was a loud pop and the lights in my room went off. I checked the LEDs and they work fine. Assuming I tested the driver with the MM correctly it still works. Could it be a plug malfunction; as in the plug connected to the 48P malfunctioned or the wiring is off? The plug was actually kind of melted and I'm pretty sure the socket is pushing up daisies (good thing it was on a surge protector); the other sockets on the protector work though...

So far only half work which is really cool but still sucks that the whole fixture can't work. The seagrasses and macro look great and are growing well though. It could still just be wiring since the LEDs are really close together to minimize the disco effect. I don't have disco on the ones that work but the wiring was a PITA. Everything else works fine even the color LEDs but the driver/plug combo just isn't doing well.

Any thoughts and suggestions?

Check all wiring and test each LED with a multimeter for continuity.

 

jedi,

 

what do you think would be a good option,

 

2 heatsinks - 12''X4''

or

2 heatsinks - 12''X7''

 

LED's arranges in 1:2 ratio of NW:RB - like AI fixtures

BTW using 80 or 90 degree optics,

 

Regards

Zoaguy:)

Either heatsink will work fine, with smaller being cheaper. What do you mean like AI fixtures? They do not use neutral whtie.

[zoaguy]

Check all wiring and test each LED with a multimeter for continuity.

 

 

Either heatsink will work fine, with smaller being cheaper. What do you mean like AI fixtures? They do not use neutral whtie.

 

 

jedi,

 

i mean AI in the sense of LED arrangement

 

--------------------------------------------

RB NW RB

NW RB RB RB NW RB

 

 

TV/OCW TV/OCW TV/OCW

 

 

RB NW RB

NW RB RB RB NW RB

--------------------------------------------

 

what heatsink to go with this kind of arrangement or do you have some better ideas on arranging led's ?

 

12'' X 4''

or

12'' X 7''

 

 

 

thanks

Zoaguy:)

[jedimasterben]

jedi,

 

i mean AI in the sense of LED arrangement

 

--------------------------------------------

RB NW RB

NW RB RB RB NW RB

 

 

TV/OCW TV/OCW TV/OCW

 

 

RB NW RB

NW RB RB RB NW RB

--------------------------------------------

 

what heatsink to go with this kind of arrangement or do you have some better ideas on arranging led's ?

 

12'' X 4''

or

12'' X 7''

 

 

 

thanks

Zoaguy:)

Either one is fine. You do not really need to space them out like that. You can make two clusters at the end of each heatsink and it would spread just as well.

[loyalhero90]

Sorry if this is a repeat but which LED color addition is better for seagrasses and corals: Violet UV or Red? And could you say why?

 

P.S Thanks for your previous answer!

[uglybuckling]

Assuming you have NW already (and therefore some red, just by virtue of the spectrum of NWs), I think that supplementation with true violet (around 410-420nm wavelength) will add the most PAR.

 

Deep red would be a close second, and might oustrip true violet when it comes to macroalgae. The photosynthetic spectra of Zooxanthellae and (some) macros are kicking around this site somewhere. Photosynthetically-useful spectra change somewhat depending on what color things are (since different pigments are present in different amounts), so it's tough to draw out a universal photosynthetic spectrum, but there are some pretty useful "general" ones around.

 

Actual ultraviolet (below 400nm wavelength) is essentially harmful to everything. There's some debate about whether it makes corals color up (like a sun tan), but generally it is not photosynthetically useful.

 

 

<<<<warning--extreme physics / biology geekery starts here>>>>

 

 

Since you asked for a "why" here's some data stolen from Bitts' posts on his macro thread heeyah: http://www.nano-reef.com/forums/index.php?showtopic=234487

and from Milad's LED thread ova heeyah:

http://www.nano-reef.com/forums/index.php?showtopic=294733

 

 

As you can see from the black line in the below diagram, chlorophyll B is pretty much taken care of by our royal blue and neutral white LEDs. The idea behind supplementation of these is to hit some of the peaks associated with chlorophyll A. As you can see from the white line, true violet hits the shorter-wavelength peak for chlorophyll A, while deep red hits the longer-wavelength peak for cholorophyll A:

 

 

Another way of putting it...chlorophyll A has peaks at 417 and 655 nanometers:

 

Hitting the other pigments like phycocyanin and whatnot can, of course, help, but is largely unnecessary, since they are far less efficient for photosynthesis than the chlorophylls, and (in higher organisms like macroalgae) are never *the only* pigment present (i.e., chlorophylls A and B are always present in some amount).

 

As such, the PAR spectrum (again this is general and can vary depending on amounts of each pigment present!) looks something like an overlay of chlorophylls A and B:

[jedimasterben]

+1 to all of that. And a word of caution, you can definitely overdo it with either LED, it's just more prominent to our eyes when using red colors. We aren't particularly good at seeing below 430nm or so, the LEDs just look dim, but are still putting out a ton of light.

[loyalhero90]

+1 to all of that. And a word of caution, you can definitely overdo it with either LED, it's just more prominent to our eyes when using red colors. We aren't particularly good at seeing below 430nm or so, the LEDs just look dim, but are still putting out a ton of light.

 

I'm using CW. Would that make a difference?

I was thinking of trying 6RB:4CW: 2DR or 2UV

Sorry that probably would have helped earlier.

[jedimasterben]

I'm using CW. Would that make a difference?

I was thinking of trying 6RB:4CW: 2DR or 2UV

Sorry that probably would have helped earlier.

Never use cool white LEDs.

[Reef Breeders]

Never use cool white LEDs.

I disagree, and find nuetral whites can give corals a pastel color if over used. Cool whites are MUCH more beneficial, as in they hit the nail on the head with the 450nm grow spectrum. Adding reds, violets, and greens/cyans will supplement the colors just fine. For the OP's tank, I would use 12 7500k CW XT-E (or XP-G), 18 455nm royal blue XT-E, 3-6 420nm violet, 3 660nm reds, and 3 500nm cyans. That should be plenty for sps in that tank, and should be dialed down as mentioned before. Good luck with your build!

[jedimasterben]

Let me break all that down.

 

I disagree, and find nuetral whites can give corals a pastel color if over used.

Cool white LEDs should never be used in an situation where accurate color rendering is needed, such as a reef tank, the technology just isn't there yet. Corals present a very wide variety of colors, and cool white misses so much of the spectrum (even when adding in cyan and deep red to help fill in the gaps) that it will reduce the amount of colors you actually see, because the light isn't coming in for the corals to reflect. Corals aren't more pastel under neutrals - what you are seeing is getting closer to the true colors. I'd much rather have more colorful coral than be limited to blue and green corals, with reds, oranges, yellows being dulled down.

 

Check out this thread on Candlepowerforum.com for more information on why cool white LEDs are bad at color rendition: http://www.candlepowerforums.com/vb/showth...Club-Here/page8

 

See the replies on that page from user GaAslamp, they will clear things up far better than I can.

 

If the phosphors were better and more spectrum could be covered with cool white while still looking cool, then they would be better to use. Citizen has made a white LED (CITIZEN CL-L233-HC13N1-C) with a 5000k color temperature that covers more spectrum than I've ever seen from a neutral white chip, but so far, that's the closest we've come at this point in time with LEDs.

 

 

However, it's still not 'cool' enough to be a cool white chip, and it is very expensive (about $40 each for a 13w, 890lm).

 

Cool whites are MUCH more beneficial, as in they hit the nail on the head with the 450nm grow spectrum.

Then why not light a tank with only cool white, or only royal blue at 450nm? There is much more to the photosynthetic action spectrum than that.

 

See here:

 

And here:

 

And here:

 

 

And compare those to the spectral output of the cool white XT-E:

 

Adding more royal blue around 450nm really isn't going to help much, only taking that initial spike even higher. It takes more spectrum to truly thrive in our systems.

 

Adding reds, violets, and greens/cyans will supplement the colors just fine. For the OP's tank, I would use 9 7500k CW XT-E (or XP-G), 18 455nm royal blue XT-E, 3-6 420nm violet, 3 660nm reds, and 3 500nm cyans. That should be plenty for sps in that tank, and should be dialed down as mentioned before.

A 1:2 ratio of cool white to royal blue? That's pretty blue, and that's not even adding cool blue (which, again, is a part of the action spectrum that cool white skips out on, hence my original recommendation).

 

With XT-Es, multiples of 9 are best, as the most common drivers, meanwell ELN 60-48 Ds, are best suited for 9 XT-Es, which would be 27v at 700ma, and 45v at 1300ma, well within their range.

At 700ma, an XT-E needs ~3.1v, and at 1300ma it needs ~3.3v. An ELN-60-48 driver can support up to 14 XT-E chips, not 9. For 9x LEDs to use up 45v at 1300ma, they would be pulling 6 watts apiece, which they don't.

[uglybuckling]

I've never used cool whites in an LED build, so my opinion means...not much. That being said, from a sheer theoretical perspective, I think the cool white DOES hit the 450nm a lot harder than the neutral white, but royal blue LEDs, which we use in very high numbers in our systems anyway, hit 450nm pretty effectively too.

 

Making a tall peak taller (at 450nm) is valuable--chlorophyll B is very important and pushing it harder is always a good thing--but so is hitting a (useful) area of the spectrum where you lack coverage, such as 620-ish nm, in chlorophyll A territory.

 

I think there's more than one way to get good growth--if neutral whites are adequately supplemented by royal blues, or if cool whites are adequately supplemented at the high end of the spectrum (do you recommend or use orange LEDs?), I think you will get similar spectral curves and similar growth under either.

 

That being said, I believe that given current technology, from a sheer theoretical perspective, neutral whites should have higher color-rendering index. What this means to your individual eye may vary.

 

I will say that I am very happy with my full-spectrum build, which uses neutral whites. I don't say this to brag, but rather to give you some idea of how much of a difference full-spectrum can make when done in one of the probably-many-correct-ways that it can be done: my light pretty much wins when compared with any lights I've ever seen at a LFS (including ecotech radions, which it cost less than half the price of, though mine is nowhere near as adjustable as they are), and any lights I've seen over the tanks of other local reef enthusiasts.

 

Jedi is correct about the drivers.

 

Then why not light a tank with only cool white, or only royal blue at 450nm?

 

I have a friend who does exactly this. Straight LED-lit royal blue ONLY 450nm tank. It looks like Joseph's Technicolor Dreamcoat and has for two years. In fact, it hasn't changed much in two years. The mushrooms and palythoas and zoanthids stay alive, but do not grow.

[jedimasterben]

I have a friend who does exactly this. Straight LED-lit royal blue ONLY 450nm tank. It looks like Joseph's Technicolor Dreamcoat and has for two years. In fact, it hasn't changed much in two years. The mushrooms and palythoas and zoanthids stay alive, but do not grow.

Precisely my point. While having a nice spike at 450nm is pleasing to our eyes, it isn't the only important wavelength to our corals.

 

My motto (and recommendations) have always been to provide as full of a spectrum as humanly possible, not only to bring out colors, but to increase growth and therefore health. You can only give a coral so much blue light before it stops absorbing it and has to reflect the rest, but if you can also give it violet and red, you can get it to absorb significantly more light and get better growth because of it.

[loyalhero90]

Thanks for the replies. I'm not really building I was just looking at some Par38 that used the CW, RB, DR ratio. I was thinking of buying a lamp while I fix my own full spectrum fixture... I just can't get the wiring right for the color parts. Are there any Full spectrum lamps by Par38 with NW?

I like the fixture I'm building on though which does include turquoise, DR, CB, TV and NW/RB though. I just need to finish it.

[jedimasterben]

There aren't any PAR38 like that that I know of, not at this point. LEDtric still won't use NW chips, no matter how much I ask.

[loyalhero90]

So no on CREE par38 lamp with CW/RB/DR?

[jedimasterben]

So you're just needing a temporary light while you put your main build together? Just use a clip-on shop light with a 23w CFL (or more) from Home Depot. Will provide plenty of light (they give pretty awesome PAR) and while not looking the best, it is only temporary.

 

 

What's holding up your build?

[topjimmy]

So you're just needing a temporary light while you put your main build together? Just use a clip-on shop light with a 23w CFL (or more) from Home Depot. Will provide plenty of light (they give pretty awesome PAR) and while not looking the best, it is only temporary.

 

 

What's holding up your build?

That is exactly what I am using now. Hope to have my build done this week too so that I can join the club. I am having a problem with condensation building up in the hood. The back is open, but the lid gets really wet. I don't know if the fans for the cooling of the lights will help this or if I should put a PC fan in the top, or side to move air, maybe even some vents just cut in the top would help.