how impractical is water cooled LEDs

[7808]

spent some valuable time today thinking this through lol

so if I have a 40b fwlr and I want to make a "energy efficient" tank, switch from t5 to LEDs, and actually pump the tank water through a cooling block the LEDs are mounted too in turn using the say 40-80watts of LED power to heat the water... with no control over it.. of course I still use a normal heater..

anyone tried it? did it transfer too much heat to your water? how much watts of LEDs did you have? did 78 degree water keep LEDs cool enough even? did you filter the water going into the cooling blocks to keep them clean inside for maximum thermal transfer? these are all questions im asking because I asked them to myself

http://www.ebay.com/itm/2-PCS-CPU-W...d-Cooler-Aluminium-122-41mm-W14-/281125241077

http://ecx.images-amazon.com/images/I/41vS2jwFpqL._SL500_AA300_.jpg plastic y strainer, think 80 mesh screen

I was thinking something like that for two 10 or 20w multichips each

idealy the tank would maintain temp during the day with the heater only running to keep the temp stable, but the LEDs would stay cool and the tank would have supplemental heat from the LEDS

 

[Boston21]

Doubt anyone's done it, but it would be cool to see. My 120WLED does a decent job of keeping my heater off sometimes, its not hot, but it isnt cold. lol

EDIT: Dang, those water blocks are sick. I wonder if those were meant for some sort of blade server.

 

[mcgyvr]

Impractical is not the word... The word is "overkill"
There is really no need and its just adding complexity to the system.

 

[SGT_York]

The cost of the additional pumping would be more than the air fan. let alone the corrosion and salt creep issues it would cause.

 

[hllywd]

spent some valuable time today thinking this through lol

so if I have a 40b fwlr and I want to make a "energy efficient" tank, switch from t5 to LEDs, and actually pump the tank water through a cooling block the LEDs are mounted too in turn using the say 40-80watts of LED power to heat the water... with no control over it.. of course I still use a normal heater..

anyone tried it? did it transfer too much heat to your water? how much watts of LEDs did you have? did 78 degree water keep LEDs cool enough even? did you filter the water going into the cooling blocks to keep them clean inside for maximum thermal transfer? these are all questions im asking because I asked them to myself

http://www.ebay.com/itm/2-PCS-CPU-W...d-Cooler-Aluminium-122-41mm-W14-/281125241077

http://ecx.images-amazon.com/images/I/41vS2jwFpqL._SL500_AA300_.jpg plastic y strainer, think 80 mesh screen

I was thinking something like that for two 10 or 20w multichips each

idealy the tank would maintain temp during the day with the heater only running to keep the temp stable, but the LEDs would stay cool and the tank would have supplemental heat from the LEDS

Probably pretty impractical as the other respondents have indicated. In warm weather you probably don't want much, if any of the heat back in the tank. A big part of my switch to LEDs was to get rid of the lighting induced heat. Further, you wouldn't want an aluminum block in contact with your salt water for at least a couple different reasons, it would need to be titanium.

IMO water cooling is almost always overkill, and mostly used for the "look how complex, and expensive my system is" cool factor. I'm not opposed to that thinking, however here you spring a leak and you likely nuke your tank, and kill everything in it. Maybe burn the house down...

Like an old Type 1 VW, air is the best was to get the heat away, figure on heating the tank some other way!

 

[richiero]

i second the salt creep issue and corrosion issue

 

[7808]

i think if air was the best way to get heat off an engine all modern engines would not have radiators lol, they do it because its cheaper and simple on mopeds etc..

i dont really see salt creep being an issue on a sealed system, and i didnt have any wow or cool factor in mind, actually thought it would be simpler to just use an old powerhead and a couple tubes instead of multiple cpu fans wiring and another power brick hanging off my power strip

just a thought, i may or may not attempt it.

 

[hllywd]

i think if air was the best way to get heat off an engine all modern engines would not have radiators lol, they do it because its cheaper and simple on mopeds etc..

i dont really see salt creep being an issue on a sealed system, and i didnt have any wow or cool factor in mind, actually thought it would be simpler to just use an old powerhead and a couple tubes instead of multiple cpu fans wiring and another power brick hanging off my power strip

just a thought, i may or may not attempt it.

Most cars aren't anymore, todays vehicles use more power, and convert more of it to heat, therefore the need to move it away from the vehicle. A lot of motorcycles are air cooled to this day. In lawn mowers, chainsaws, other small engine equipment, air conditioners, refrigerators, computers, stereos... air cooling is very effective, it works pretty good. In an aquarium situation, more often than not you need to get equipment heat out of the equation in large part because it's uncontrolled. Simply dumping heat from your lights into your 40G breeder will make for very unstable daily temperature control, or a fairly elaborate control system to dump heat elsewhere would likely be required. Then you haven't addressed aluminum's reactivity to a salt water environment.

An old power head, tubes, and a cooling block that reacts with salt water actually seems much less elegant than a couple fans, and a power supply to me.

 

[7808]

ya, i know, ive owned air cooled and water cooled bikes. i was replying to the 'look at old VW, air is the best way to get heat away" comment, i was saying they dont air cool engines because its better. lawnmowers are air cooled. high tech race bikes are water cooled...

i'd say the powerhead and tubes vs fans and wiring are kind of a stupid argument. at the size of my tank and lights, its pretty much not going to matter.. powerhead power consumption vs fans and running heater more often.. = so small of difference who cares. theres already going to be wiring going right there to the lights anyways.

i was looking more or less to see if anyone has done it and how it went for them. and how much the heat from the LED chips really effected water temp gallon per watts etc... just curious

 

[FullBoreReefer]

My hydras put off some heat. But they also have their own fans in them. Unless you built your own led fixture i would say 99% already have fans. I like DIY but to me there are more Cons than Pros.

We already put tons of electricity by saltwater, but putting saltwater on "top" of or mounted to something is alittle more risky IMO.

 

[Boston21]

If the block happens to crack it would be a day from hell.

 

[hllywd]

ya, i know, ive owned air cooled and water cooled bikes. i was replying to the 'look at old VW, air is the best way to get heat away" comment, i was saying they dont air cool engines because its better. lawnmowers are air cooled. high tech race bikes are water cooled...

Air cooling has it's place. They didn't stop producing air cooled VWs because the cooling systems didn't work. There are a lot of low and high tech air cooled race bikes too... certainly a LED system for most of our tanks doesn't rise to the level of a Ducati... Air cooling works in the LED lighting application. I've been running 96 Cree XR-Es on 4 aluminum heatsinks, with nothing but a fan in the end of a 6' x 2' canopy to remove the trapped heat for over 3 years now. I've seen 100% reliability from the LEDs to date.

Do some searches, I know the water cooling question has come up here before. I couldn't tell you if anyone has ever tried it, but the cons of trying to water cool an LED system seemed to far outweigh the pros in those discussions. The big obstacle, besides the heat issues, to building something like you want to is finding a non-reactive heat exchanger, not to mention how much that may cost. I did a quick search and aluminum was the only thing I found, you can't use aluminum if you intend to run tank water though it.

 

[Tornader]

Bad idea. Aluminum and salt water - just bad idea. That's why you don't want to buy an outboard motor that has been in the ocean without proper flushing after each use.

 

[mussel and hate]

I have an experimental watercooled LED luminaire however I do not employ tank water for coolant and I'd advise against doing so. If you feel the need to attempt such a system I recommend heating through the tank glass with a secondary block on a dedicated coolant loop. The risk of copper contamination isn't worth the potential energy savings, indeed in my case the additional heat would be a problem.

 

[7808]

Air cooling has it's place. They didn't stop producing air cooled VWs because the cooling systems didn't work. There are a lot of low and high tech air cooled race bikes too... certainly a LED system for most of our tanks doesn't rise to the level of a Ducati... Air cooling works in the LED lighting application. I've been running 96 Cree XR-Es on 4 aluminum heatsinks, with nothing but a fan in the end of a 6' x 2' canopy to remove the trapped heat for over 3 years now. I've seen 100% reliability from the LEDs to date.

Do some searches, I know the water cooling question has come up here before. I couldn't tell you if anyone has ever tried it, but the cons of trying to water cool an LED system seemed to far outweigh the pros in those discussions. The big obstacle, besides the heat issues, to building something like you want to is finding a non-reactive heat exchanger, not to mention how much that may cost. I did a quick search and aluminum was the only thing I found, you can't use aluminum if you intend to run tank water though it.

ya it has its place in lawnmowers, old VW, mopeds, basic dirtbikes and scooters lol ya im sure people race some bikes that have air cooled engines.. maybe even cars on the drag strip.. never said it didn't work. dumb argument.

 

[Duvallj]

It could be done. Aluminum/copper are good at heat transfer, but due to saltwater it wouldn't be advisable to use these materials. A high grade of saltwater resistant stainless steel work work and still transfer heat well enough for the led's. So if it would work? Sure just pump water through something like 1" squarestock with 1/8" walls, prefurably custom extruded to include fins on the inside of the pipe.

Now the real question is would it be worth it? Nope. Even though the large led chips are able to generate "high" temperatures there isn't really a whole lot of energy as heat. This is why air cooling led's is plenty. A nicely designed heatsink with proper temperature heat tubes would be far enough overkill but still more practical than watercooling.


Now back to what I think you really want to do: capture heat produced from the leds to heat your tank and maximize efficiency.
Well this one is simple math if you use a simple model. Lets say best case senario is that you can capture all of the heat produced and dump it into you tank. Also that the efficency of the led is such that any inefficency is actually heat (probably pretty close to it too).
So if you used 85% eff leds with a total of 80watts, you only harness only 15%, or 12watts of heat to your tank. In real life you'll more than likely produce even less than this.

Also to answer your question, will 78*F be able to cool the leds? Yes, very much yes. The real cooling power comes from higher specific heat capactiy.

You can look up watercooling guides for cpu's, the math is very handy and would be easy enough to translate to led cooling.

So you could do it, but I don't think it would be useful enough/a good idea.

 

[jerpa]

Now the real question is would it be worth it? Nope. Even though the large led chips are able to generate "high"temperatures there isn't really a whole lot of energy as heat. This is why air cooling led's is plenty. A nicely designed heatsink with proper temperature heat tubes would be far enough overkill but still more practical than watercooling.

Now back to what I think you really want to do: capture heat produced from the leds to heat your tank and maximize efficiency. Well this one is simple math if you use a simple model. Lets say best case senario is that you can capture all of the heat produced and dump it into you tank. Also that the efficency of the led is such that any inefficency is actually heat (probably pretty close to it too). So if you used 85% eff leds with a total of 80watts, you only harness only 15%, or 12watts of heat to your tank. In real life you'll more than likely produce even less than this.

I'm not arguing whether or not a system like this is worth the trouble but LEDs give off closer to 70-75% of the energy consumed as heat straight into the heatsink. There is no UV or IR so effectively the rest of the energy produces visible light. This does leave a large percentage of heat to be scavenged.

 

[Sander_Bos]

would it be an option to 'lead' a hose through the blades of the heatsinks?

I recently bought some old heatsinks from my University, and the sellers asked me why I needed those because I could cool the leds with the water....

There must be some way to cool leds with water which is cheaper than running a fan....

 

[7808]

good info. i guess someone would just have to try it. i did find some crazy high priced water cooled LED lights designed for aquarium use, but nothing that used tank water.

like this one for $1200:uhoh3: http://www.bluemoonaquatics.com/BMA_P100.php pretty cool though

 

[ReeferTJ]

It works great! I built a 2 100w LEDs on a aluminum waterblock with a small laser radiator and small pump, total cost was around 100$ and much lighter than my old 250mh and Ballast. EBay is your friend! Mines a closed loop, no metal touches SW