[Squiner]

Ok, here's a few pics of the tank. Should help visualize why I'm concerned with the PAR values from front to back.

 

[Mark Allred]

Like some other designs I've seen on here, I will combine all the tube ends with some sort of manifold and use a fan to draw air through the tubes for cooling. A though, has anyone ever tried to use stack effect to passively cool? In theory, the heat generated by the LEDs would induce a draft in a vertical pipe connected to the manifold. The greater the LED heat production the greater the draft. This would be much more reliable than a fan but may not be as effective. Any thoughts?
.

Chimney effect! Thinking out of the box! This might be worth doing some research on Making the stack "attractive" enough for the little dove, may be an issue...

 

[Squiner]

Chimney effect! Thinking out of the box! This might be worth doing some research on Making the stack "attractive" enough for the little dove, may be an issue...

Exactly Mark! That's my only concern with the chimney, that is it being ugly. Other than that, if I were to go tall enough, I should get plenty of draft I would imagine.

I'm now actually considering liquid cooling. I'd just tap into one of my return lines and connect each aluminum tube with some 1/2" vinyl tubing so the four LED sections would still be movable.

A concern with this is the water corroding the aluminum over time which may leak or kill the fish/ invertebrates. A workaround may be coating the inside of the tubes with either an epoxy or paint.

I really like the idea of supplementing my 800 watt heater with the LED waste heat and not having to use fans which will likely produce noise.

 

[Squiner]

I decided to go with water cooling my new lights. Today, I put a thin layer of fiberglass resin inside the aluminum tubes to prevent the tank water from touching the aluminum. Poured about 250 ml of resin in each tube, rotated it around for complete coverage and then let the extra resin run out. Heat formed some PVC elbows to the square stock. I'll epoxy them in place. I'll screw on some hose barbs to the PVC to connect each tube in series with vinyl tubing. They'll be connected to the sumps return line. Connecting them together with vinyl tubing so I can push the lights together to get them out of the way when needing to do tank maintenance.

 

[AaronT]

I've always wondered about water cooling LEDs. I'm interested to see this unfold.

 

[lamiskool]

I decided to go with water cooling my new lights. Today, I put a thin layer of fiberglass resin inside the aluminum tubes to prevent the tank water from touching the aluminum. Poured about 250 ml of resin in each tube, rotated it around for complete coverage and then let the extra resin run out. Heat formed some PVC elbows to the square stock. I'll epoxy them in place. I'll screw on some hose barbs to the PVC to connect each tube in series with vinyl tubing. They'll be connected to the sumps return line. Connecting them together with vinyl tubing so I can push the lights together to get them out of the way when needing to do tank maintenance.

man, thats intense!

 

[Squiner]

Well, finally an update!

I had to build a hood in order to get the LED light build going. Now that that's done, here are a few pics of where I stand. I built the lights on sliding rails to allow access to the tank when I open the hood door. The hood door is not seen on the pics.

Since I'm going with water cooling, you can see each aluminum tube is connected with vinyl tubing. Here's a pic of that.

Next update should be when it's installed on the tank. Thanks for looking!

Jim

 

[jeffkrol]

From an off the top of my head thought, I'd think about how much insulation the resin creates to heat transfer.........
Thermal expansion and contraction of the Al stock may also create a problem..

I have no "real" basis for these thoughts but it is only a possibility...

I've briefly contemplated water cooling and thought the "tube inside the tube" was more effective. Never built it but the main concept was al square tube thermal epoxied to the inside of the other sq tubing.. or flattened round w/ the possibility of using copper instead (flattened copper water line for example)., but too much work and the worry of galvanic corrosion stopped that one...

Oh and my next thought.. why not step the fixture to follow the shape of the slope in the tank.. Makes for a tall light but .......

Anyways JUST suggestions and food for thought... Results always trump theory..

 

[Squiner]

From an off the top of my head thought, I'd think about how much insulation the resin creates to heat transfer.........
Thermal expansion and contraction of the Al stock may also create a problem..

I have no "real" basis for these thoughts but it is only a possibility...

I've briefly contemplated water cooling and thought the "tube inside the tube" was more effective. Never built it but the main concept was al square tube thermal epoxied to the inside of the other sq tubing.. or flattened round w/ the possibility of using copper instead (flattened copper water line for example)., but too much work and the worry of galvanic corrosion stopped that one...

Oh and my next thought.. why not step the fixture to follow the shape of the slope in the tank.. Makes for a tall light but .......

Anyways JUST suggestions and food for thought... Results always trump theory..

Most people use air to remove the heat. I think I'll be good with the resin inside the tubing since it has an order of magnitude more thermal conductivity than air. It's an extremely thin layer also. As you say, we'll see what the IR gun says when it's up and running....

 

[Squiner]

Got the lights installed over the weekend. Everything is working wonderfully! The water cooling is keeping the LEDs nice and cool. I'd say about 80-90 degrees.

You can see the LED cooling water stream in back.

I've only got a few plants in there now but the plan is for the entire top half to be baby tears.

Let me know what you think. It was a fun project.

 

[discgo]

I wonder how much this will lower the operating cost of your heater? Each led probably only produces 20-30% of it consumed power as light and the other is wasted on heat. Great job trying to reclaim the heat!

 

[Squiner]

I wonder how much this will lower the operating cost of your heater? Each led probably only produces 20-30% of it consumed power as light and the other is wasted on heat. Great job trying to reclaim the heat!

That's a good question, I have a watt meter that I could measure energy consumption but think it would be hard to get good data over a short period of time. Plus I would need to run with the lights off for the same amount of time.

Maybe I could measure the delta between the input and output water temperature from the LED cooling system. Measure the water volume and from those variables calculate the calories of heat added by the LEDs. Should be able to convert calories into watts and from watts into $.

Sounds like fun evening, haha.

 

[genocdex]

im glad some one else is using the heat from the leds... what is the flow and wattage of the pump you are using for the liquid cooling?

 

[gus6464]

I assume you are using tank water to flow through the tubes? While this is good short term the use of non-distilled water without additives will eventually gunk up the tubes and cooling efficiency will plummet.

You might want to start thinking about running a closed system with a t-line or reservoir flowing through a small radiator. You can get a small 12V pump used for PC watercooling and contain the entire system in the canopy.

 

[Squiner]

im glad some one else is using the heat from the leds... what is the flow and wattage of the pump you are using for the liquid cooling?

I have a Fluval SP4 pump in my sump. It puts out 1822 gph. That runs the return water and the LED cooling. Is on a valve so I restrict the LED cooling to a slow trickle as you can see in one of the pics I attached before.

Are you running water cooling too?

 

[Squiner]

I assume you are using tank water to flow through the tubes? While this is good short term the use of non-distilled water without additives will eventually gunk up the tubes and cooling efficiency will plummet.

You might want to start thinking about running a closed system with a t-line or reservoir flowing through a small radiator. You can get a small 12V pump used for PC watercooling and contain the entire system in the canopy.

Yes, tank water. Is this based on evidence from your own water cooled system? I have had mine running for a month and a half and seen no degradation in cooling performance. I have algae growth in the vinyl lines connecting the aluminum tubes so I assume that there is growth in the aluminium tubes.

How would I use a closed loop system and reclaim the waste heat? The advantages of my design is three fold. One pump, reclaims waste heat and no need for a heat exchanger.

 

[gus6464]

Yes, tank water. Is this based on evidence from your own water cooled system? I have had mine running for a month and a half and seen no degradation in cooling performance. I have algae growth in the vinyl lines connecting the aluminum tubes so I assume that there is growth in the aluminium tubes.

How would I use a closed loop system and reclaim the waste heat? The advantages of my design is three fold. One pump, reclaims waste heat and no need for a heat exchanger.

It is based on experience from putting together PC watercooling systems. The gunk inside the aluminum tubes will get thicker and thicker which will eventually start to impede flow. On a PC based system it is recommended to flush and clean the system once a year and that is with distilled water with additives on a closed system. Most powerful system I've run is one with an Iwaki MD-20 and even after a while flow decreased quite a bit.

Since you are not running a radiator you want as much flow as possible going through the tubes. I know using a system like this is cool in theory because it cools your LEDs and heats up your water but it is not viable long term.

 

[Squiner]

It is based on experience from putting together PC watercooling systems. The gunk inside the aluminum tubes will get thicker and thicker which will eventually start to impede flow. On a PC based system it is recommended to flush and clean the system once a year and that is with distilled water with additives on a closed system. Most powerful system I've run is one with an Iwaki MD-20 and even after a while flow decreased quite a bit.

Since you are not running a radiator you want as much flow as possible going through the tubes. I know using a system like this is cool in theory because it cools your LEDs and heats up your water but it is not viable long term.

I see. I guess time will tell the story. I'll check the temps periodically and post again a couple months.

If need be ill pump a bleach solution through the tubing for a day to clean it out. It's easy to disconnect lines from the pump.

 

[mr one]

At our situation i would go with a separate water cooling loop and use in it a distiled water with some kind of PT nuke or silver kill coil. Been playing around a lot with PC water cooling So no algae on tubes or in other parts, also no gunk that can clog or lower heat transfer