DIY LED thoughts for Biocube 29

[saf1]

Both reto kits from steves and rapid will keep any corals. I have owned both. Steve's had more royal blues than rapid which was a 50/50 split.

Both run around 250....

 

[aznflyfisherman]

IMO i think if your going to spend 200 plus, i would just buy the AI Prime n install it in the hood. Then your done n you can keep most corals in there.

 

[bblumberg]

If you did not buy the lights already, you can't go wrong with the systems sold by RapidLED. No doubt Steve's is also good. I have a RapidLED upgrade (12 RB, 12 CW) in my Nanocube 24 and it works great for LPS and a few SPS.

I built a new light based on their 24" enclosure and individual LEDS for my 75g corner tank (30" deep). I used a mixture of (4) 400-410, (4) 410-420, (4) 430, (26) RB, (9) cool whites, 2 lime green and 2 photo red - a total of 51 LEDS on 4 channels with 80 degree optics. The light looks and works great and I get ~200 PAR with only the violet and RB lights on at the sand bed at 50% power. I can bump that to more than 600 with all lights (including the white/green/red) at 100% on the sand bed. More than 1200 12" below the light.

If I had it to do over again, I'd go with the MakersLED gear that you can get from the other vendor you mentioned because their bits seem much better integrated (more space in the enclosure, drivers mount directly on top, controller connects by simple cable). I wasted a lot of time searching for enclosures for the drivers, wiring the whole mess up, etc. The violet/uv leds are not very bright either. Next time, I'd just sub 430 nm for them.

Good luck.

 

[window]

I am still leaning away from any of the retrofit kits for a couple reasons. It seems like I have more options and better LEDs if I don't go with those. At this point, already having a heatsink and drivers, just buying LEDs is going to be cheaper.

Similarly, I don't much want to DIY retrofit a prebuilt LED system because it will end up costing more than just buying the LEDs it has even if I were to buy used and I think it would likely not work as well for me being much closer to the water than it was designed for and maybe not fitting in such a tight place.

If I had it to do over again, I'd go with the MakersLED gear that you can get from the other vendor you mentioned because their bits seem much better integrated (more space in the enclosure, drivers mount directly on top, controller connects by simple cable). I wasted a lot of time searching for enclosures for the drivers, wiring the whole mess up, etc. The violet/uv leds are not very bright either. Next time, I'd just sub 430 nm for them.

I saw the MakersLED and was pretty tempted by how clean it is, but I feel like spreading out the light over the large heatsink I have would give better light coverage and thermal properties. If I were not making a system for the biocube, I would probably go with that.

I appreciate all the advice.

 

[zachts]

I see that has very high CRI, but in what other ways is it better than using multiple single LEDs? It seems like the light distribution of multiple smaller points of light would be better considering how close the lights are to the top of the water in the Biocube. The Vero 10 has a 120 degree lens, which would certainly cover the bottom of the tank, but it seems like there would be much more shadows. I will look into that more.

To me I think one would plentyand give some nice shimmer. They have a much "softer look" and a more even spread with a lot less shadowing even from a single vero chip as they have a very large (comparatively speaking) phosphor area with many small emitters below it creating less of a distinct point source light than single die leds. The color distribution is also very even given it is has a flat phosphor layer unlike single die emmiters which tend to have color bands because of the domed lens and single blue light source, being more blue under the LED and more yellow the further to the side you get.

If you wan't to go for a more flat fluorescent look no reason you can't use more, no more than 4 of them I would think(all I have for whites over my 55gal is 4 veros), wired all in parallel to a single driver, make sure it's dimmable!

The Nichia is comparable from a CRI stand point just more costly as you need many more to get the same amount of light and coverage.

you can't go wrong with the vero though, whether you decide to use one or many, pick up several and try one, if you don't like it add more. Cost wise I think you'll still come out ahead compared to single 3w type leds.

 

[window]

you can't go wrong with the vero though, whether you decide to use one or many, pick up several and try one, if you don't like it add more. Cost wise I think you'll still come out ahead compared to single 3w type leds.

I think you are right about them being cheaper. I will probably go with that or the vero 13 to be on the safe side since my drivers are right at the max current for the vero 10. They are PWM drivers, so dimming isn't a problem. Thanks for the advice.

 

[window]

Can I get suggestions on lenses? Right now I think I don't need them due to how close the LEDs will be to the top of the water. Is that logic sound?

 

[zachts]

Can I get suggestions on lenses? Right now I think I don't need them due to how close the LEDs will be to the top of the water. Is that logic sound?

Correct, you should not need lenses. I'd still stick with the vero 10, better value and same footprint as a normal LED star. The 13 is much larger in diameter but not really all that much more powerful about an extra 1/3 the power at twice the price. Just run more than one vero 10 in parallel and don't worry, you'll never have them running anywhere near full power anyhow. The four chips I've got over my 55gal are all in parallel and only sharing around 350mA and it's plenty bright.

Driving them at full power of the driver (1.4A on an old meanwell ELN) gives each chip around 350mA and drowns out the 42 other blue and violet LEDs making the tank look like a 6500k planted aquarium! and that's only half power of what they can handle. These little beauties are crazy powerful. (technically 4 vero 10s are capable of providing more light than 36 regular Cree or luxeon whites when compared at the same drive current.)

On a side note: I REALLY, REALLY, REALLY wish bridgelux would release a vero with no phosphor that would just be royal blue! Someday, I hope.

 

[window]

Driving them at full power of the driver (1.4A on an old meanwell ELN) gives each chip around 350mA and drowns out the 42 other blue and violet LEDs making the tank look like a 6500k planted aquarium! and that's only half power of what they can handle. These little beauties are crazy powerful. (technically 4 vero 10s are capable of providing more light than 36 regular Cree or luxeon whites when compared at the same drive current.)

Do you see any issues with thermal runaway running them in parallel?

 

[zachts]

Do you see any issues with thermal runaway running them in parallel?

No, since they are running at such low current, Total current to the four is only half what a single vero could handle. If you were to run them at higher current in parallel you'd want to employ fuses or a current mirror, etc as thermal run away would be more of a concern at that point though just use a separate driver on each.

Running them in parallel is just practical since they need such low current in my setup and comparably over a biocube.

 

[window]

Great, thanks for all the information.