Another LED DIY project... questions to ask, details to show

dacianb

Active Member
Sorry to everybody for my long absence, but I am back finally and ready to show up my lights :).

Screen Shot 02-20-15 at 09.55 PM.JPG Screen Shot 02-20-15 at 09.56 PM.JPG

The boards as revision 1 had some aspects which I did't like (too much power in a single spot) and hard to assembly connectors.
So I redesigned the boards - using less LEDs, increasing the thermal transfer area and also regular 2.54mm pitch connectors.

There are 4 channels with footprints like:

Channel 1 - 2 x Luxeon Z
Channel 2 - 12 x Osram
Channel 3 - 12 x Osram
Channel 4 - 6 x Osram
Size of it 90x90 mm with 32 LEDs on 4 channels (approx 100W):

LED array.jpg
 

StirCrayzy

Well-Known Member
My thoughts...
Three 100w units on a 42-48" tank means you will have customers concerned about center brace shadowing.
I should hope a prototype raised a bit higher could prove that wrong, but it's what jumps out at me when I look at your model.
Have you tested/measured how the light spreads at different heights .?
 

dacianb

Active Member
My thoughts...
Three 100w units on a 42-48" tank means you will have customers concerned about center brace shadowing.
I should hope a prototype raised a bit higher could prove that wrong, but it's what jumps out at me when I look at your model.
Have you tested/measured how the light spreads at different heights .?

The model I show there is a RS Reefer - so no central brace. Maybe even 2 modules will be enough to cover the tank. Or 4? Hahaha... will see once are on the tank. I have ordered enough parts to be able to build as many lights I need. Light coming out of leds is very wide distributed (50% at 120°).
The idea is to have a rimless tank with almost invisible lights above... This is why I want them at low level.
 

StirCrayzy

Well-Known Member
I was looking at it from a mass appeal standpoint. Rimless is awesome, but reality is that there are a higher qty of braced standard tanks out there on the market, thus from a business perspective it's worth consideration.
If you aren't running optics, there probably won't be a need for three on 48" tank, making my previous statements all but moot.
I think you will lose efficiency though, and run higher power to get good PAR #'s at depth towards the tank edges. Keep it up, this is an interesting project .
 

dacianb

Active Member
StirCrayzy, I dont see myself with a factory to produce such lights, but would be nice.
Here are 2 concepts I designed - first picture is 200 mm long with one led board into (same as I put on previous image above tank)

Second image is longer (400mm) with 2 LED boards into (so 200W each). Main difference is that for short version fan is ON the light, for longer version fan is IN the lights, so not visible anymore.

With 2 long versions I can cover a long tank with central brace. Happy??

ScreenShot3.jpg


hdfhxfgh.jpg
 

dacianb

Active Member
And some details about the LEDs on the board:

- channel 1 - 2xLuxeon UV at 420-425nm - max current 1A

- channel 2 - 12 x Oslon Square Deep Blue - powered at 1A, but those leds can handle 1.8A - so high efficiency and increased lifetime

- channel 3 - 12 x Oslon Square White - 4500K with high CRI =85 - powered at 1A, but those leds can handle 1.8A - so high efficiency and increased lifetime. Using high CRI leds makes other colors (red, green, etc) useless

- channel 4 - 4xOslon Signal Blue + 2xOslon Signal Verde (505 nm) - powered at 1A too
 

dacianb

Active Member
I was looking at it from a mass appeal standpoint. Rimless is awesome, but reality is that there are a higher qty of braced standard tanks out there on the market, thus from a business perspective it's worth consideration.
If you aren't running optics, there probably won't be a need for three on 48" tank, making my previous statements all but moot.
I think you will lose efficiency though, and run higher power to get good PAR #'s at depth towards the tank edges. Keep it up, this is an interesting project .

No, I will not using any optics on LEDs and there are few reasons for that.
- I need the lights placed at "water level" and have a good coverage
- I don't want to have hotspots into tank and then just to add T5s to make things a bit more uniform. Not using collimators, light distribution is close to a T5.
- actually, due to laws of physics any collimator will lower the total efficiency of a led thanks to Fresnel losses and internal absorption. Such optics take light distributed on a wide area and focus them on a smaller area, but with some losses. If you measure on central points, indeed collimated light is higher than uncollimated, but if will measure average on large surfaces it is more efficient without any optics.

I see really stupid to have colimated lights and placed them on the ceiling to have a decent coverage. Is a way for manufacturers to say how high PAR their lamps have in a point, but everything else in the tank to be dark. Honestly I am more interested by total energy on the bottom of the tank and the ratio between lower and higher energy point.

First tests I run on the board shows how great coverage I can have, no hotspots at all and also no "disco effect" even at few cm away from light.

Penetration into water - small cluster of LED with high power - will do it easily, don't worry.
 

dacianb

Active Member
Something about mounting bridge...

It is made using industrial AL profiles black anodized and with dedicated connection systems, making a very rigid and durable system.

ScreenShot.jpg
On the inside there are channels to place wires inside and then to cover them with plastic profiles - so no wires visible.
ScreenShot1.jpg

For the moment I installed it on my existing tank (Juwel Rio 180). What is not visible is that on back side (where profiles touches the glass edge) there are silicone profiles.
The 2 clamps block the bridge to fall accidentally into tank, but also if I turned them upside down the holes are perfect for hanging system.
Making them L shape allow to mount the bridge on top cover.
Anytime I can add some feets from same profiles to rise the lights higher if necessary.
 

dacianb

Active Member
Some pictures of the test setup...

20150320_095248.jpg
LED board is mounted in the case and all connections made. Picture is taken with LEDs at 1%, but still camera is saturated - colors are not actual. One DeepBlue LED is off, probably due to a short under it (too much flow paste)

20150320_095424.jpg
Rising the current to 10% it is already impossible to take a picture. I think that the lines on image are the PWM frequency.
 

dacianb

Active Member
Short version of light without any covers yet.

20150130_065754.jpg

And here the final version of the power board. 4x Meanwell drivers + a DC/DC converter for the fan (48 to 12V) - adapted to fit on Coralux board

20150327_104128.jpg
 

dacianb

Active Member
Side cover and splash screen mounted, here is how things looks like:
- on first image it is visible the thermostat for fan and the fan connector
- second image - black cable is the 48V power in and the 2 jacks are for the 4 channels coming from Storm controller (or any other PWM controller in the future)
- third image is the bottom part with splash screen mounted, but not quite visible. I dont like the screws (ugly) and also the wires have to be thinner. For tests I will keep them as they are.

20150417_133041.jpg 20150417_133050.jpg 20150417_135740.jpg
 

dacianb

Active Member
And light is mounted on the tank and set to 50% on all 4 channels. Here is how things looks like.
I have a very ugly tank for the moment (please dont lough about it) - hope soon to change things.
Only one light mounted and still cover more than 50% of the tank from almost water level high (tank is Juwel 180 - 1 meter long)
01.jpg 02.jpg 03.jpg 04.jpg 05.jpg 06.jpg
07.jpg
 

sirrealism

Well-Known Member
I like this a lot. I have always wanted to DIY LEDs but all of the ones I have seen others do look like crap. I know you do this for a living so I am sure that makes a big difference. But I really like that I can chose what colors I want in a DIY set up.
 
Top