The Planted Tank Forum banner

DIY LED array for my 40gal tall

3640 14
I've been really enjoying everyone's LED project and the discussions very much. I've been giving some thought to building one of my own. It can't be cheap quality, but it really needs to give me a very good bang for the buck.

I'm going with CREE LEDs as they are the best bang for the buck IMHO.
The Mean Well LED driver also looks pretty good to me.

Font Rectangle Parallel Circle Screenshot


Font Rectangle Screenshot Parallel Auto part


I just don't have the budget to get too fancy, but I don't want to paint myself in a design corner either.

That Mean Well will drive a single string of 12 CREE XM-L @1300ma or two parallel 12 XM-G strings @650ma ea. I can always add another MW later as an upgrade for more light if I go high-tech.

I'd like to alternate WW & CW XM-G LEDs or one string of the NW XM-L.

I'd love to go with 2x12 of the XM-Ls, but @$180 that breaks the piggy bank.

I'm planning on 3" o/c spacing & 4"-6" between rows.
I don't plan on any optics as I want to have them almost touching the top glass cover, 1" above the waterline and 22" above the substrate.

I'll use one heat-sink channel per string. Inverted C with the LEDs mounted inside.

Tank is 36"x12"x24"

What kind of PAR levels should I expect with either option?

I'm going medium tech with modest ferts, but no CO2 (yet).

Thanx for looking
1 - 15 of 15 Posts

·
Premium Member
Joined
·
1,904 Posts
Looks like a good part list!
So are you planning on 12 or 24 LEDs? I think if you do 12 over a 40 gallon, you'll have a lot of spot lighting issues, but this will mean much much more shimmer.

If you do 24, sure it won't be as bright but you'll get a better spread.

With 24 LEDs at 650ma at 22 inches, you'll probably be into the med/med-high range. (i'm basing this off my old set up, 24 epistars [which are less efficient than the crees, but ran at 700ma] and at about 24 inches off the substrate, I was hitting about 40umols of par at the substrate.)

if you're not going to do co2, i would suggest lifting the light up higher.
 

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #3 ·
I have a pretty good ripple going on the surface from my Marineland 350 discharge that will give me some good shimmer too. I have a lot of fish to help generate some good ferts and CO2 LOL. If the light is still too high, the MW is dimmable.

Pity that I've just upgraded to T5HOs this week.
 

·
Registered
Joined
·
667 Posts
I can't comment upon the PAR, but after experimenting with a mixture of cool and warm white LED's, I'd definitely opt for the neutral white.

I don't understand though why lifting LED's a little higher would decrease their intensity. They are very directional of lights, so are far from being a point source to which an inverse square falloff would be applicable. Besides, with that driver you could just dim them slightly to get them to last longer and run more efficiently instead.
 

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #5 ·
I can't comment upon the PAR, but after experimenting with a mixture of cool and warm white LED's, I'd definitely opt for the neutral white.
What was it about the WW/CW combo that made you prefer the NW lights?
 

·
Registered
Joined
·
667 Posts
What was it about the WW/CW combo that made you prefer the NW lights?
Neutral white would have more accurate and natural colour rendition more similar to daylight, while cool white and warm white can each be a bit too blue and yellow respectively even in combination, in my own opinion. The cool white in particular really seem to wash the colours of fish out, although that is what many people nevertheless seem to use, which seems odd to me unless you're running a reef tank.

If you look at the spectral outputs of the those three LED colours, they all seem to lack much light in approximately the turquoise part of the spectrum. So when I do another tank I may even look at supplementing those wavelengths with a few low power coloured LEDs. I'm by no means an expert on LEDs however.
 

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #7 ·
Here's the XM-G spectral power distribution curves as per the data sheet.

Product Azure Rectangle Slope Font


Pity that blue-green gap isn't covered with colored LEDs either.

 

·
Registered
Joined
·
667 Posts
CREE don't seem to have them, but you can resort to other suppliers and can of course cheaply procure some coloured LED strips on Ebay. It's not as if you require a great deal of intensity to fill in just that one part of the spectrum. Green LEDs are around 525 nm, and Aqua LEDs are about 505 nm, so with a combination of those two you could potentially fill out the lighting to better mimic the spectrum of sunlight. Your plants don't really require that, but I would personally expect it to overall make your tank and your fish look better.
 

·
Registered
Joined
·
1,121 Posts
Sounds like you have a good plan.

FWIW you can actually run 14 XPG's (in series) or 28 XPG (in parallel).

You can run 16 XML's in series or 28 XML in parallel.

This is on the Meanwell ELN60-48D. The XML's have a lower forward voltage meaning you can run more of them up to the 48v max.



And here is a quick glimpse of the spectral data we collected over the weekend of a bunch of LEDs. This is the cool white XML at 700, 1500, and 3000mA.
 

Attachments

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #11 ·
Hmmm... 16 XM-L will give me 2.25" o/c spacing.

Looks like there is more red in the spectral curves than what the data sheet indicates.

I wonder what a single row of the NWs would look like?
 

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #12 ·
What is the PAR of the NW vs the CW?

Seems the blue spike makes the difference as the WW have a lower blue spike and are lower PAR, but the NW & CW have about the same sized higher blue spike.

Since I've recently invested $95 in two individual T5HO fixtures with quality reflectors, I don't have any spare bux this month to pursue the LED's. However I know where to get 'em when the time is right. Meanwhile I'll head over to the scrap/surplus metal dealer and procure some decent extruded Al stock suitable for the heatsinks.

I'm planning on a single row of 15 CREE XM-L 2.25" on center. No optics so I hope there is good overlap. I'm not sure how well they will cover the 12" wide tank floor. Hopefully the glass refraction will reflect most of the dim wide angle beam back in and offset any dimming of the beam at the front and back tank substrate. I'll just add a second string later if needed.
 

·
Registered
Joined
·
667 Posts
Sounds like you have a good plan.

FWIW you can actually run 14 XPG's (in series) or 28 XPG (in parallel).

You can run 16 XML's in series or 28 XML in parallel.

This is on the Meanwell ELN60-48D. The XML's have a lower forward voltage meaning you can run more of them up to the 48v max.



And here is a quick glimpse of the spectral data we collected over the weekend of a bunch of LEDs. This is the cool white XML at 700, 1500, and 3000mA.
Sounds like a pretty wild weekend you had! ; ) What tool did you happen to use to collect this? And does it give any readings such as the actual total PAR and/or lumen at each current? Helpful though, thanks.
 

·
Registered
Joined
·
667 Posts
What is the PAR of the NW vs the CW?

Seems the blue spike makes the difference as the WW have a lower blue spike and are lower PAR, but the NW & CW have about the same sized higher blue spike.

Since I've recently invested $95 in two individual T5HO fixtures with quality reflectors, I don't have any spare bux this month to pursue the LED's. However I know where to get 'em when the time is right. Meanwhile I'll head over to the scrap/surplus metal dealer and procure some decent extruded Al stock suitable for the heatsinks.

I'm planning on a single row of 15 CREE XM-L 2.25" on center. No optics so I hope there is good overlap. I'm not sure how well they will cover the 12" wide tank floor. Hopefully the glass refraction will reflect most of the dim wide angle beam back in and offset any dimming of the beam at the front and back tank substrate. I'll just add a second string later if needed.
Judging from the graphs, and as PAR emphasizes wavelengths both in the red and blue areas, the WW may even have the highest PAR being that the blue spikes of each are relatively narrow and considering that the WW have a great deal of area under the line at the red end of the graph. Similarly, the NW should have higher PAR than the CW as their blue spikes are equal but considering that they have red as well. However, I'd personally be more inclined to pick light on the basis of what looks good rather than simply PAR.

In my opinion the light bouncing off the front pane should help illuminate your fish. I personally prefer to see the colors and iridescence of my fish, rather than seeing a backlit X-ray look to them. So I wonder if you'll eventually end up adding another string so that you can have one of them closer to the front for primarily that reason even.
 

·
Registered
Joined
·
2,220 Posts
Discussion Starter · #15 ·
Got my two T5HOs w/reflectors set up and they are pretty bright.

Tank is 36"x12"x24" and lights are 22" from substrate going through glass cover & 21" of water column.

How will the PAR of my proposed 15 CREE XM-L 6500 K string @1300ma w/o optics compare to one of my T5HO 6500 K lamps?

I've shortened the photo period to 11 hours with a 3 hour mid day overlap burst period.
 
1 - 15 of 15 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top