Evil Cluster - No Holds Barred

wmilas

Keep magnets 9" apart!
Hello,

Some of you may remember me. I' built my current system about 8.. maybe its been 10? years ago. I've cycled through all the lighting options (including LED from the major manufacturers) and have always come back to 400W MH (Radiums) for both look and ease of maintenance on my main display (120x36x30). I use 4 with high quality eballasts and reflectors. I supplement with VHO.

I've been far from thrilled with the commercial led fixtures out there. Most provide way too much low CRI white and not nearly enough blue. They miss important fluorescent spikes and the shadowing can be terrible.

I've been playing around with Led in my home shop. I'm at the point where I want to try and build either an all LED or a LED W/ T5. Here are the specs:

1) Each fixture needs to replace 1 of my 400W MHs. I want 4 fixtures to cover 120x36x30. I don't run sand, and I do keep sps at the bottom so it really needs to be 30". The tank is actually more like 34" so 30" is when the main circ is off. Wave makers move the average depth between 28" and 32" all day long. So each "unit" needs to throw 30x36x30. I'm not sure Led's can penetrate this far realistically and NOT discoball.

2) Mount the fixture 12" above water.

3) Minimum amount of shadows.

4) No disco.

5) As many fluorescent spikes covered as possible.

6) Thinking one remote power supply for the LED's not on fixture, the rest on fixture.

7) Full reflective color support. This means lime and reds.

8) Multiple channel control but don't want to go nuts. As many channels as needed for tweaking and no more.

9) Less wattage than the MH setup. This means I'd like to stay under 400watts when tuned (not max).

10) Heat sink(s) will more than likely be actively cooled. That kind of wattage on a passive sink would make the sink too big to be practical.

11) Ease of maintenance. That means I'd favor tight clusters with smaller splash shields over massive slabs of light.

So any ideas on what the led selection would look like? I plan to get feedback here then start building and tweaking. I'm not looking for someone to completely lay this fixture out, I'm willing to do that. I just need a solid place to start cluster wise with the highest end emitters. Hopefully be someone that's done a bit of this before. I don't want to re-invent the wheel if there is knowledge already out there.

Thanks!
 
4) No disco.

I'm with you. I abhor the "disco effect."

I don't think it can be avoided unless a multi-chip is used. Even Radions with their tight LED spacing are subject to it.

I'd love to see someone (wink wink nudge nudge) try a few of these out: Borealis Luxeon Z Array

I bet three of these grouped in a triangle would match a 400w halide and give good punch and spread without harsh shadowing. The trick is cooling them.
 
I don't really see the point of using those Rapid pucks. You'd giving up almost all the benefits of the DIY route (vendor independence, upgradability, ability to choose LEDs, etc), and not much gain since you can pack as many LEDs into the same area using multi-chips like the ES 3ups or Luxeon K8/16 RBs, etc. The whites they use also seem inferior to the Veros.

I'm about to start my own build with similar requirements to yours. My clusters will consist of 7 stars, and will include the Luxeon K16, Vero10, and Lime/Cyan/Violet 3ups. In terms of reflective color support, I think you don't need red since the limes will cover the spectrum well and so will a high CRI white.

I think the biggest issue with LEDs would be finding a good balance between coverage, intensity, and lack of disco.
 
I would wait till reefll.com starts selling the projectors based on their assembly. You could buy one and see how it compares to your MH then order more if you like.
 
I'm with you. I abhor the "disco effect."

I don't think it can be avoided unless a multi-chip is used. Even Radions with their tight LED spacing are subject to it.

I'd love to see someone (wink wink nudge nudge) try a few of these out: Borealis Luxeon Z Array

I bet three of these grouped in a triangle would match a 400w halide and give good punch and spread without harsh shadowing. The trick is cooling them.

The issue I have with these is the lack of a lens. I'm thinking pretty much everything I'm going to use will need a 90 degree lens, at least for the blues. even a whole chip 120 would be good.

The second issue I have is that the uv and the red are not on separate channels. I don't understand the reasoning behind this.

I have to admit though that the puck pretty much does have everything you would need. I don't think heat dissipation if actively cooled is that big of an issue with the right sink.

The price is very attractive.
 
I don't really see the point of using those Rapid pucks. You'd giving up almost all the benefits of the DIY route (vendor independence, upgradability, ability to choose LEDs, etc), and not much gain since you can pack as many LEDs into the same area using multi-chips like the ES 3ups or Luxeon K8/16 RBs, etc. The whites they use also seem inferior to the Veros.

I'm about to start my own build with similar requirements to yours. My clusters will consist of 7 stars, and will include the Luxeon K16, Vero10, and Lime/Cyan/Violet 3ups. In terms of reflective color support, I think you don't need red since the limes will cover the spectrum well and so will a high CRI white.

I think the biggest issue with LEDs would be finding a good balance between coverage, intensity, and lack of disco.

The attraction is ease of assembly and maintenance. If you have 4 fixtures, 3 picks per fixture, 12 pucks total. If one craps out, replace it. Its only $120.

You are right about being able to do "better" be increasing the granularity. I believe for prototyping I'll be forced to go stars to get the right mix.
 
I would wait till reefll.com starts selling the projectors based on their assembly. You could buy one and see how it compares to your MH then order more if you like.

Do you have a link to these?

The issue I have with Reefll is their cost. For instance the 12xV3.0 is $79. This seems steep. The bins are excellent though, I will give them that. The lack of Lime is an issue. It looks like it was designed before the limes came out.

The interesting thing is the balancing of blue to violet. I really like the ratio. This setup looks to maximize fluorescence. I'm wondering about growth though.
 
The other main "ideal" I'm struggling with is to explore the lime /pc amber route with no NW at all, or to blend lime with NW. I'm not convinced that the NW adds anything to the coral growth at all. If that is indeed true, why include it? Dunno.

I see most builds are going with a Vero 10 for the NW. As jedi posted on another thread I think he prefers 50% NW and 50% lime.

I should probably just buy a crap load of all the Rebel colors and play with them :)
 
I'm not convinced that the NW adds anything to the coral growth at all.
Sure, you may use only PC Amber and Royal Blue LED to mimic NW LED.
By the way - corals may use the most of visible spectrum. Depend of exact coral species, exact absorption spectrum will vary, but in any case big part of spectrum of NW LED will be used for coral growth.
 
The lack of Lime is an issue. It looks like it was designed before the limes came out.
Yes, you right, this assemblies was designed before Limes is come out. But Lime is near to useless for coral fluorescence and not good for the corals growth. But Lime is extremely bright. Because of this, we will use Lime in our projectors. But I not sure in necessity of using it in 12up assemblies.

Maybe you'll be curious to read our study: http://www.advancedaquarist.com/2014/4/lighting
 
ReefLedLab: I've read that study. Great work btw. It is what has started me thinking that LED, if designed correctly, could rival my beloved 400w MH Radiums.
 
So despite the fact that the peak wavelength for ES Limes may not be in the most useful range, they cover a wider spectrum than any of the other color Rebels. So saying they are not good for coral growth may be a bit misleading.
 
So saying they are not good for coral growth may be a bit misleading.
I say "not good", but I had not say "completely useless" :)
Definitely, corals can be using for their growth the most part of visible spectrum. But don't forget - Lime has contain, in simple words, two part of spectrum.
First part 510-580nm is the most poorly used for the most corals part of spectrum.
Second part 580-750nm is suitable for shallow water species, but may be useless or even harmful for some deepwater species. By the way, all radiation above 700nm is completely useless.
 
Back
Top