3D Printed DIY LED Lighting with Dual Ramp Controller

[ReeferDan246]

Hello all,

 

First time poster

 

Anyway, thought I would share my work on my DIY LED lighting system for my 5 gallon fluval chi (tank only). I'm a mechanical engineer with access to 3D printers, so I designed my own housing for the lights and then printed it out.

 

I designed this as a 36W system that can independlty control blue and white LEDs. I made the wattage ridiculously high so I could run the LEDs dimmer, and therefore cooler for a nice long life. I also have a PC cooling fan.

 

Stuff needed:

 

(6) 3W 10K LED diodes $7

(6) 3W 455 nm LED diodes $7

26 gage wire $4

laptop cooling fan $5

3D Printer with PLA (Mine were printed on a Makerbot, Taz, and Printerbot)

Generic 12V Power Supply (Mine was for a laptop, Make sure it is rated for enough watts) $8

(8) 1.5 ohm resistors (high wattage) $5

Solder and soldering iron

Dual Ramp LED Controller $35

(2) female 2.1mm x 5mm DC plugs (To plug LEDs in to Dual Ramp) $5

Shrink sleeving or electrical tape ($5)

Wire strippers

 

Depending on what you already have available, you could probably do this for about $70, which isn't half bad for a fully controlled system.

 

First I designed the LED circuits. Because I used a constant voltage power supply, I needed a current limiting resistor. Without these resistors, the LEDs would experience a massive current and quickly burn out. To do this, google "LED array resistor calculator" and then you don't have to do any math. My resistor was 3 ohms, but I had to put (2) 1.5 ohm resistors together because I couldn't find any 3 ohm resistors rated above 5W.

 

Next I designed the housings and printed them out. I glued the LEDs onto the housing and wired them up by soldering.

 

Then I wired the cooling fan parallel with the blue LED circuits. I chose this circuit over the white because I like bluer tanks, and will be running the blue channel higher. This means the fan will receive more power than if it were wired with the white circuit.

 

I bought a 4-wire cable so only one cable had to come out of the back. This looks a lot cleaner than having 4 cables come out the back. I covered all solder joints between wires with either shrink sleeving or electrical tape.

 

Then I wired the led terminals to the DC plugs, and was good to go. The system looks terrific and all indications are that the coral will be happy. I put a GSP in as a test and he seems happy.

 

If you want my CAD files to print out your own similar system, let me know. Hope you enjoyed!

 

Dan

 

***Pursue your similar venture at your own risk, and with sufficient knowledge of electrical safety. I'm not responsible for any injury or damage caused by you attempting this project.***

 

[Horerczy]

Is the red thing a heatsink?

[Sunstar]

I might be interested in your cad file.

 

I would have to look into making/using a heat sink, but I am resourceful in this manner.

[ReeferDan246]

The red part is just a platform that I put the lights on before I screwed them into the black piece. This helped me tuck all of the wires back behind the lights, and let's me replace them with another rack later.

 

The red piece would sink some heat, but besides the star base I'm not actually using a metal heat sink. They run at only 50% (white) and 60% (blue) with a fan directly overhead, which has kept the diodes nice and cool to the touch.

[braaap]

The red part is just a platform that I put the lights on before I screwed them into the black piece. This helped me tuck all of the wires back behind the lights, and let's me replace them with another rack later.

 

The red piece would sink some heat, but besides the star base I'm not actually using a metal heat sink. They run at only 50% (white) and 60% (blue) with a fan directly overhead, which has kept the diodes nice and cool to the touch.

 

That is cool. I'm surprised though that you can make it function without a heatsink. Could you run them higher then the current percentages and not melt the plastic? What is the melting point of the plastic?

[jedimasterben]

Without a heatsink, despite active cooling, the LEDs won't last long. That particular LED package is already poor at thermal transfer away from the diode, and the LEDs being cheap themselves is compounded by that.

 

That is cool. I'm surprised though that you can make it function without a heatsink. Could you run them higher then the current percentages and not melt the plastic? What is the melting point of the plastic?

PLA melts at around 200C. The LEDs will fail when they get in the 70-100C range.

[braaap]

Without a heatsink, despite active cooling, the LEDs won't last long. That particular LED package is already poor at thermal transfer away from the diode, and the LEDs being cheap themselves is compounded by that.

PLA melts at around 200C. The LEDs will fail when they get in the 70-100C range.

 

I'm convinced your brain is going to explode sooner or later.

[ReeferDan246]

Haha. I'll give you a thermocouple reading tonight. I haven't measured it because it's cool to the touch. I doubt it's above 40C.

[ReeferDan246]

Without a heatsink, despite active cooling, the LEDs won't last long. That particular LED package is already poor at thermal transfer away from the diode, and the LEDs being cheap themselves is compounded by that.

 

PLA melts at around 200C. The LEDs will fail when they get in the 70-100C range.

 

 

 

Thanks for the reply jedimasterben. It's good to have discussions about these things.

 

I personally chose not to have a passive heat sink for my application for the reasons below.

 

1) PC fans are typically cheaper. I didn't want to spend $10-$15 for a second cooling method for LEDs that cost $7 for pack of 10. Especially since I only use 6 out of those 10.

 

2) It was really easy in the design and build to just glue the LEDs directly to the overhang. With a heat sink this would have been a little more complicated and would have taken up a lot of room. Most heat sinks I saw were .5" to 1.25" high. My housing's interior is 1" high. It would have had to be probably twice as tall to accommodate a passive sink.

 

3) In a longer package (such as overhanging a 36" long tank) passive sinks would be the best option. Your wattage is pretty spread out, so the housing can dump the heat from your heat sink into your room better because it has more surface area. Fans would also be poor at getting flow across such a huge surface.

In my case, I have 36 watts in a 5x5x1 inch space. Heat would build up rapidly in the housing, which would negatively affect the performance of the sink (the temperature difference between your LEDs and housing air temperature is smaller, and heat flows from high to low temps).

This meant that I needed a fan, regardless of whether or not I had a heat sink.

 

4) If you take a look online you'll see an overwhelming number of articles that believe active is more efficient than passive from a heat transfer perspective. The biggest disadvantages are the moving parts that could fail or start to get noisy, and the dust accumulation.

 

 

Hope this addresses some of your concern. Let me know if you have any other questions. I'll get you that temp reading tonight.

 

Take care,

Dan

[ReeferDan246]

The diode temps ranged from 108-113 F. Room temp was 75F. Should have a nice long life.

[jedimasterben]

How did you get the temperature of the diode?

[ReeferDan246]

How did you get the temperature of the diode?

Probed around with my thermocouple for highest temperature on entire LED. The lens in front of the light emitting side yielded the hottest measurement. The back of the LEDs was 85-87 depending on the LED. The housing ambient air was 80F.

 

I'm sure the actual diode enclosed in the lens is probably 5-15F hotter than what I could measure. But still definitely at a safe temperature.