Has anyone tried making DIY high power LED fixtures using the nanometer speific LEDs from Cree or Chanzan, or any other manufacturer?

[BeanAnimal]

I built my water cooled Cree LED fixture.



It had water running through it and a radiator.





Even the algae scrubber light was water cooled with Cree LEDs.

My goodness! I just don't know what to say..

Other than it just makes me want a Paul B Fish even more.

 

[BeanAnimal]

Well the half empty approach..
Yea you probably won't be able to build as cheap as say the Nicrew or Noopsyche but certainly can build for cheaper than the $9/watt Radions, Mitras, Geiseman.

Def won't look as pretty.

100W of leds is still about $100 even for quality.
Coralux, Steves controllers are still out there.
Ldd's are still cheap as are power supplies.

The MAIN advantage would be to tailor the output at 100% to the spectrum you like. Throwing out 25% of a radions potential is annoying.
And being self constructed, if you screw up the spectrum it is cheaper to modify it than buy new lights .

So a bit of glass 1/2 full..

No probably at best break even unless one goes really dirt cheap like fleabay $$2/10 "3w" leds.

Throwing the light in a hood and cosmetics are inconsequential.

I bet you never thought you would hear me say this...

WAY TOO MUCH tinkering for me. I have a small wooden crate if high output LED parts, lenses, heatsinks, drivers, etc. Another small wooden crate full of microcontroller parts, half finished boards, etc. and bins full of components. It is not that I don't enjoy electronics anymore. I just don't have the time, esp for something that I can buy. I think much of the allure (then) was building what we couldn't buy

When I get the urge to dig in, i have two monoblock tube amps that I need to finish wiring. I had the chasis black chrome plated and the transformer shells copper plated, etc. I built the base enclosure out of chechen with hand cut dovetials. I will finish the project one day.


I think the Noo Psyche K7 fits the bill for the price/trouble. But that is me. I know other folks still enjoy DIY.

 

[T Diddy]

If you have an Apex or other controller, you can control many drivers with 0-10v. That's what I do for my diy led strips.

View attachment 32393504

At the time, rapidled offered a pot with 3 different daylight settings ranging from 8-12 hours sunrise/sunset. It’s currently controlling six different circuits/drivers. They don’t have it on their website anymore, but it is perfect for my needs.

 

[Timfish]

I built my water cooled Cree LED fixture. . .

Cool microscope! College days?

Awesome fixture!

Did you ever look the temp of the water or air coming off the radiator? Have you seen any corrosion issues? I'm looking at converting a lot of my LEDs in my fish room to water cooled. Mostly to dump heat outside in the summer to lower my AC load and partly to redistribute heat in the winter.

This first prototype using a CPU cooling block is to see what's needed to keep a DIY LED disk cool. Plan is to build light bars using rectangular aluminum tubing and insulated vinyl tubing running to a radiator outside.

 

[kharmaguru]

At the time, rapidled offered a pot with 3 different daylight settings ranging from 8-12 hours sunrise/sunset. It’s currently controlling six different circuits/drivers. They don’t have it on their website anymore, but it is perfect for my needs.

I used this Coralux controller from them for years. I'll probably work it back into something eventually.

 

[Paul B]

Cool microscope! College days?

You are assuming I went to college. I didn't, there was that war thing and I got drafted right out of high school when I was 18.

I went to apply to a college so I could get wisdom but they told me I was enough of a wize guy already.

Besides, If I went to college I would have been an electrical engineer working for half my salary for twice the hours and would not have retired 20 years ago while enjoying a nice pension and free medical.

Did you ever look the temp of the water or air coming off the radiator? Have you seen any corrosion issues?

The thing stayed very cool as long as the water was flowing with that tiny pump. In the first picture of it I posted you can see a clear tube sticking up on the left side. I built a flow switch in there with a floating magnet. If the pump stopped, the water would go down in that tube shutting off the power to the lights. It also had a temp sensor that would shut it off if it got hot. It worked perfectly but when I moved here the room the tank was in had a new concrete floor and it was so damp in there the exposed wiring and LEDs kept shorting out.

The entire thing was sprayed with a clear plastic coating so no corrosion.

 

[T Diddy]

Ever since this thread popped up, my gears have been turning. The light I built 12 years ago is still quite functional, but LED technology has come a long, long way. Limitations included a lack of wavelength specific diodes, varying maximum current capabilities between different colors of the same manufacturer/models, finding efficient drivers that could run a high number of LEDs, etc…just to name a few. As kharmaguru pointed out, part of the allure associated with DIY was making a better fixture than was commercially available.
After doing some research on the various “reef LED lighting” systems that are currently available, I still think I can build a better, more powerful fixture. Looking at the prices of the off-the-shelf fixtures, I realized that another DIY project was on the horizon. I don’t care about connectivity, and I don’t use any reef management software or apps. All I need is a high powered fixture with programmable or preset sunrise/sunset, and possibly a dedicated moonlight circuit.
The original fixture was built at a time when a mix of blues and whites was as good as it was going to get. Today, I’ll be able to incorporate virtually any color combinations, and all at virtually identical voltages and currents…something that was impossible in years past. I’m not sure how long it will take me to pull the trigger on this project, but I definitely intend to make it happen. Thanks for the inspiration, OP
Wish me luck !

 

[griss]

Ever since this thread popped up, my gears have been turning. The light I built 12 years ago is still quite functional, but LED technology has come a long, long way. Limitations included a lack of wavelength specific diodes, varying maximum current capabilities between different colors of the same manufacturer/models, finding efficient drivers that could run a high number of LEDs, etc…just to name a few. As kharmaguru pointed out, part of the allure associated with DIY was making a better fixture than was commercially available.
After doing some research on the various “reef LED lighting” systems that are currently available, I still think I can build a better, more powerful fixture. Looking at the prices of the off-the-shelf fixtures, I realized that another DIY project was on the horizon. I don’t care about connectivity, and I don’t use any reef management software or apps. All I need is a high powered fixture with programmable or preset sunrise/sunset, and possibly a dedicated moonlight circuit.
The original fixture was built at a time when a mix of blues and whites was as good as it was going to get. Today, I’ll be able to incorporate virtually any color combinations, and all at virtually identical voltages and currents…something that was impossible in years past. I’m not sure how long it will take me to pull the trigger on this project, but I definitely intend to make it happen. Thanks for the inspiration, OP
Wish me luck !

Tagging in here for when you get around to starting this project.

 

[Misled]

Tagging in here for when you get around to starting this project.

yup

 

[T Diddy]

Tagging in here for when you get around to starting this project.

I’ve been mulling it over ever since I restarted my aquarium, but now I’ve got new motivation. Yesterday, my wife said that the colors of everything looked washed out and dull…and she’s right. I had previously supplemented the LEDs with T5. Nowadays, my LFS doesn’t even sell T5 bulbs anymore. They don’t even have Mg test kits, which raises other concerns, but the fact remains that I need a better light fixture, and refuse to pay the asking price for the available models.
It’s going to be expensive, but nowhere near as much as buying two radions. Currently, I have two arrays of 36 LEDs each. I’ll replace them with two arrays of 42 LEDs, and the intensity will be inherently greater based upon the difference between the technology and maximum current capacity of yesteryear and the modern LED options. Now what I need is to borrow/rent a PAR meter

 

[oreo57]

I’ve been mulling it over ever since I restarted my aquarium, but now I’ve got new motivation. Yesterday, my wife said that the colors of everything looked washed out and dull…and she’s right. I had previously supplemented the LEDs with T5. Nowadays, my LFS doesn’t even sell T5 bulbs anymore. They don’t even have Mg test kits, which raises other concerns, but the fact remains that I need a better light fixture, and refuse to pay the asking price for the available models.
It’s going to be expensive, but nowhere near as much as buying two radions. Currently, I have two arrays of 36 LEDs each. I’ll replace them with two arrays of 42 LEDs, and the intensity will be inherently greater based upon the difference between the technology and maximum current capacity of yesteryear and the modern LED options. Now what I need is to borrow/rent a PAR meter

Whats another $200..

Amazon.com

What is your color plan?

 

[T Diddy]

Whats another $200..

Amazon.com

What is your color plan?

That’s interesting. There weren’t a ton of reviews, but it’s worth looking into.

As for the new lights, I’m having trouble deciding on the exact configuration, but after playing with the radion at the lfs, I’ve pretty much narrowed it down to this…

6 clusters per array
Each cluster consists of:
(1) Violet 400-420 nm
(2) Royal Blue 450-465 nm
(1) Blue 465-485 nm
(1) Cyan 490-510 nm
(1) Cool White ~5000 kelvin
(1) Warm White ~2700 kelvin

This combination will greatly broaden the spectrum of usable light, as well as give me more of the visible reds and purples of the fish and corals. My original light from 12 years ago only had ( essentially) royal blues and cool whites, making for a very monochromatic view, as well as limited useful photosynthetic spectrum. The result is a bunch of washed out corals, and dull coloration on the fishz.

 

[kharmaguru]

That’s interesting. There weren’t a ton of reviews, but it’s worth looking into.

telegraham reviewed it as part of a comparison piece:

LED - PARwise vs Hopocolor vs Apogee

 

[T Diddy]

telegraham reviewed it as part of a comparison piece:

LED - PARwise vs Hopocolor vs Apogee

That’s a great thread! Thanks for sharing…I’m going to look a little deeper.

 

[T Diddy]

I’ve already convinced myself…there’s so much overlap between the cool white and royal blue, I’m going to go with this:

Each cluster consists of:
(2) Violet 400-420 nm
(2) Royal Blue 450-465 nm
(1) Blue 465-485 nm
(1) Cyan 490-510 nm
(1) Warm White ~2700 kelvin

 

[BeanAnimal]

They don’t even have Mg test kits,

Randy Holmes Farley made a pretty sound argument that you should never have to test Mg and given that most Mg tests are way off... it is pointless anyway because if you dose it at the appropriate ratio to Cal/Alk (which can be measured) it is almost impossible to get it out of balance.

 

[T Diddy]

I always refer back to an article by Randy

here . I also don’t add anything to the aquarium without testing it. Salt mixes vary, as does my consistency with kalk

 

[T Diddy]

My plans were shattered when I dig a little deeper. I wanted to use the Cree XEG series, as they are all very similar in forward voltage, and capable of a 3 amp current. I was going to use a meanwell 2100 mA driver giving a big time output in terms of light intensity, but still well below the maximum range of the LEDs.
Unfortunately, no one currently has all of the available colors on individual starboards. Nonetheless, I’m tired of my twelve year old blue/white rig that is actually over clocking the LEDs. So, I ordered combination of more modern LEDs and a more suitable driver. As a result, the specific wavelengths are a little different than what I originally imagined. The good news is that the voltages are all very similar, and nothing will be pushed beyond their maximum current. Anyway, it’s going to look like this…each cluster will consist of:

1400 mA meanwell drivers
(2) luxeon C violets 420-430 nm
(2) Cree xpg3 royal blue 440-455 nm
(1) Cree xeg blue 465-485 nm
(1) Cree xeg cyan 490-510 nm
(1) Cree xpg3 warm white 4000k

again, the spectrum is a tad different than what I envisioned, but I can still pull it off without stretching any components beyond intended limits.

 

[oreo57]

My plans were shattered when I dig a little deeper. I wanted to use the Cree XEG series, as they are all very similar in forward voltage, and capable of a 3 amp current. I was going to use a meanwell 2100 mA driver giving a big time output in terms of light intensity, but still well below the maximum range of the LEDs.
Unfortunately, no one currently has all of the available colors on individual starboards. Nonetheless, I’m tired of my twelve year old blue/white rig that is actually over clocking the LEDs. So, I ordered combination of more modern LEDs and a more suitable driver. As a result, the specific wavelengths are a little different than what I originally imagined. The good news is that the voltages are all very similar, and nothing will be pushed beyond their maximum current. Anyway, it’s going to look like this…each cluster will consist of:

1400 mA meanwell drivers
(2) luxeon C violets 420-430 nm
(2) Cree xpg3 royal blue 440-455 nm
(1) Cree xeg blue 465-485 nm
(1) Cree xeg cyan 490-510 nm
(1) Cree xpg3 warm white 4000k

again, the spectrum is a tad different than what I envisioned, but I can still pull it off without stretching any components beyond intended limits.

Where do you get CREE cyans from? In years of puttering around in DIY I've never seen them.

 

[kharmaguru]

Where do you get CREE cyans from? In years of puttering around in DIY I've never seen them.

Cyan is a new product from Cree so who's got them on starboards? Probably nobody. Mouser has 5 cree cyan products in a few formats but they're all surface mount from them.