People still diy leds?

[peterj66]

I built mine two years ago and would not hesitate to do it again for another tank. Total cost for a 200 gallon tank was 400$. LEDs are all Cree and Philips Lumileds. Arduino for light schedule. To buy something similar would cost me many times more to get the same quality. What you have to decide is if you want to put in the time to build it yourself. Lenses and spread are not an issue when you control how many LEDs to use and where to place them. I have 12 groups of LEDs which are spread out evenly over the tank, no lenses. If something breaks its easily fixed.

 

[Nano sapiens]

I built this passively cooled wide dispersion LED array for a small nano over 5 years ago and it's still going strong:

https://www.nano-reef.com/forums/topic/370285-nano-sapiens-12g-diy-pwm-led-build/#entry5256957

The only repair I had to make was to replace the violet/hyperviolet emitters since I over drove the channel at one point and wore them out prematurely (the ease of changing/replacing the solderless emitters is a big plus point).

High end commercial arrays now have many of the features that people were looking for many years ago, so it's no wonder that there are a lot less DIY builds now.

 

[Stigigemla]

I noted the coolers in the thread.
It has been a great development in led efficiency.
In 2008 the blue leds were about 40% and white about 34%.
That means for a radiated light power of 100w You needed 250w input for blue and 203 w for white.
The Heat power was 150w for blue and 203w for white.

In 2021 the efficiency is 65% for blue and about 53% for white.
That means for a radiated light power of 100w You will need 154w for blue and 189w for white.
The heat power will be 54w for blue and 89w for white.

If we mix 50/50 the heat power has gone from 176.5 to 72.5w.
We need much less cooling today and passive cooling on a plate formed as a T5 light will dissipate enough for a tank of
40 gallon = 18 x 18 x 36 inches. A plate 5 inch wide and 36 inch length will transfer enough heat for 70w input and give about 100 par at the bottom. Just 1/16" aluminum.

 

[flozfzx1]

Diy used to be all the thing. Rapid led was my main site...people dont do this anymore?
I would think a diy led unit with the new chip style leds. The big 50w ones would be great.
Lol.
Idk im going t5. I just ask cuz im gonna diy build moonlights i think.
D

i recently built 60 led fixture for my 30 gal, (i know it’s a bit overkill but looking to upgrading my tank to a 60gal soon controlling it with a TC421 wifi timer running them at 60% just blues, whites @ 10%. par i am getting at top 268 @ sandbed 98.

Was fun building them but cost wise almost the same as pre made. i think most people are just buying premade it instead of DIY.

 

[flozfzx1]

.

 

[Rjramos]

My first experience with DIY LEDs, came with retrofitting a biocube 29 in 2014. This light is still going and not 1 of the Cree LED’s has failed, if 1 ever does, I can change it out. RapidLED had a retrofit kit with instructions precisely to do this. They no longer sell this retrofit kit. It came with heatsink, 24 Cree LEDs of your choice, power pack, drivers, screws and instructions.
Years later, while looking for a good quality refugium light, I put 2 of these lights shown above together. 15 LED’s (45 watts) . These were by far not a kit. 2 heatsinks from heatsink USA, which I cut a circle in the back cooling fins to accommodate a fan. The LED’s, I selected were 8 reds, 4 blues and 3 white. I added 60 or 80 degree optic lens to these. Dimmable driver with dimmer. The cases were made with clear and black plexiglas by me.
The cost for all components was considerable and my skills and labor came to play.Overall, I am really happy with these DIY lights. Although not UL inspected, I know, they are put together well and they work great! However, I will DIY for nobody.

 

[coralbeauties]

Thought I would add my set up to the thread. I built this years ago and have done several updates along the way. I first started out when leds were first getting started when it was just white or blue. Along the way I updated to ldd drivers that are controller by an ardurino with a jarduino program. I dont think I would do this again now that black box leds are do cheap. I have ocean revived over my frag tank that do a wonderful job.

 

[reddogf5]

My.moonlights for a 48" tank in the checkout right now is $67. Dimmable and will probably provide the shimmer im looking for during daylight at full power.
If i like this 1...i will make another with whites to see what i like better.
Ive got two old diy fixtures that need a run through so these moonlights will be a good step stone. Lol.
D

Not sure what you mean by “moonlight”, but an actual moonlight in a tank would be one low output cool white LED - moonlight is about 1/400 of sunlight. If you are just talking about an all blue light to see fluorescence, keep in mind that is nearly as bright to a coral as a regular light is to you, and they might not like it on all night. Just FYI

 

[monkeyCmonkeyDo]

K reddog votes white! Lol
D

 

[monkeyCmonkeyDo]

They offer me a $28 heatsink now. Lol man theyre good...
D

 

[MaddyP]

I'm in the process of building a DIY unit now, utilizing MakersLED and Nanobox Reef components. Not necessarily cheaper, but will allow control from my DIY controller via serial communication.

Still awaiting the second driver. Should have it operational in the next few weeks.

 

[elysics]

I do wonder, if it doesn't save any money when you go for the best components, are the best components really necessary or what you would get with commercial options? Bulk discounts aren't that big, and the bulk in the reefing industry isn't that big, not to mention software development and assembly costs and profit of course.

 

[Reefs and Geeks]

I do wonder, if it doesn't save any money when you go for the best components, are the best components really necessary or what you would get with commercial options? Bulk discounts aren't that big, and the bulk in the reefing industry isn't that big, not to mention software development and assembly costs and profit of course.

Good point, high end lights may not be using all of the best components available to make their lights. Could just be they have the best marketing for their lights and the herd follows along to buy the latest generation every 5 or 6 years. Could a better than the best commercially available light be built is sparing no expense? perhaps. I'm personally no good at making DIY electronics look good, and if spending top dollar I'd want the lights to look as nice as they work. Others are more skilled than me though and probably could make that happen.

 

[monkeyCmonkeyDo]

The housing and presentation of all the wires is something yes...
D

 

[dedragon]

they work fine and can get overall color right the problem i ran into was disco ball effect on mine were terrible, maybe with a light diffuser in place of the acrylic would work but would prob drop the par a bunch. i like maddyp approach of using the nanobox led groupings that could be switch out down the line if/when things change. How are the thermals on that light though? the heat sink you have looks like it was made for a different type of led so just curious

 

[Andrewcor]

Just ordered leds today for a diy supplemental light bars. Custom blend of violet, royal blue, blue and turquoise. Half the price when comparing to name brand led bars of equivalent power.

Where did you purchase the leds from? I've been looking to do the same thing.

 

[MaddyP]

they work fine and can get overall color right the problem i ran into was disco ball effect on mine were terrible, maybe with a light diffuser in place of the acrylic would work but would prob drop the par a bunch. i like maddyp approach of using the nanobox led groupings that could be switch out down the line if/when things change. How are the thermals on that light though? the heat sink you have looks like it was made for a different type of led so just curious

Not sure yet about the thermals, the heatsink was designed for individual leds certainly. Once the second driver comes in I will run a few tests to ensure these don't overheat.

 

[monkeyCmonkeyDo]

I just ordered.
Honestly i had a great debate on the color of my moonlighting.
Lol
D

 

[dedragon]

steves leds have this fixture as well, anyone ever try to incorporate this into their lighting ever? at 50 watts it might make a nice shimmer with some t5s idk

 

[Biokabe]

I do wonder, if it doesn't save any money when you go for the best components, are the best components really necessary or what you would get with commercial options? Bulk discounts aren't that big, and the bulk in the reefing industry isn't that big, not to mention software development and assembly costs.

It's the same stuff, the difference is that the commercial fixtures are paying wholesale prices for their components, you're paying retail prices when building a DIY fixture. Depending on the markup for LEDs, that can be a significant difference. Just FYI, there's a HUGE difference between what something costs to make at the factory, vs. what the retail customer pays for it. Obviously each industry is different, but in my industry, the retail consumer is usually paying 10x or more above the factory price. The farther back in the supply chain you are, the lower that cost is.

So a big manufacturer like Phillips, for example, is paying rock-bottom prices for their core components, so they can pack their light with high-end emitters, package it in an attractive heat sink, slap a fancy lensing system on it, and still come in competitive with an equivalent DIY fixture. The same will be true for all of the manufacturers, though Phillips likely gets the biggest benefit from it. The others are likely buying their LEDs from the factory, but Phillips actually owns the factory that produces their LEDs.

And that's why I stopped building DIY LED fixtures. To match the components of the manufactured units, I wasn't saving much for the time I invested in researching, designing, and building my fixture, along with putting together the control options and coming up with a mounting solution. There are doubtless still situations where a DIY solution would come out better - that fixture for a 1000g system, for example - but these days I'm more willing to just buy a manufactured unit and save my time for other things.