![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
aquarium chillers
- Thread starter wtiger
- Start date
SledgeMakeGood
Weaksauce
- Joined
- Sep 18, 2004
- Messages
- 107
Search this area and you'll find at least a dozen threads on it. The short of it: temps go through the floor, electric bill goes through the roof, and you have to insulate the tubes and blocks to keep from getting condensation.
I tried it with a piece of lab equipment that does the same thing; it has a built in six liter reservoir, cooling down to -20C, and a 6lpm pump. On a side note, I'm suprised I never see anyone using these as they can often be had cheap used.
Anyway, with 0C water, I got a pretty marginal OC boost as compared to my Koolance Exos. Specifically, my 3.4 Prescott could get to 4.3 and change instead of 4.2. Not at all worth the hassle of having to insulate, and the noise and power consumption. I've never tried pushing it all the way down to -20C as I don't want to be bothered with insulating everything for a one-off experiment.
Anyway, with 0C water, I got a pretty marginal OC boost as compared to my Koolance Exos. Specifically, my 3.4 Prescott could get to 4.3 and change instead of 4.2. Not at all worth the hassle of having to insulate, and the noise and power consumption. I've never tried pushing it all the way down to -20C as I don't want to be bothered with insulating everything for a one-off experiment.
yeah I see your point. I've been tempted to get an 80 watt pelt with a waterblock on the cold side and a hsf on the other side inline before the cpu. Not as a cooling solution in of itself, but as a booster after the radiator before the cpu. Already have an extra waterblock, but no pelt or an extra cheap hsf. I was wondering what kind of cooling power an 80 watt pelt would have just hooked up the power supply? They're 14v correct, but can work on a regular 12v psu if it's big enough correct?
sounds nice, but the price and the fact that I already have a koolance exos al means I could do the peltier thing for like 50 bucks, it wouldn't weigh my case down much more and my case would still be portable for lan parties. That and I just feel like fiddeling. The aquarium cooler thing was just an idea until I saw the pricetag and the other drawbacks of it.
HeThatKnows
Gawd
- Joined
- Apr 19, 2004
- Messages
- 577
Your waterblock-pelt-heatsink will only cool your water down by a couple tenths of a degree. 
Maybe...maybe not
Depends on the implementation. It just so happens that I know someone who has been experimenting around with chilled H20. If done right it will cool a lot! For instance: Get one of those electrically operated (peltier) coolers at Wal-Mart that you can use to cool or keep hot beverages. Fill it with H20 and turn it on to cool. Next take a radiator and run a line from your reservoir and back to your CPU. Stick the whole unit in the water and fire up your computer. If you don't think it will work or don't want to try it you will probably be able to buy such a unit in the near future
Depends on the implementation. It just so happens that I know someone who has been experimenting around with chilled H20. If done right it will cool a lot! For instance: Get one of those electrically operated (peltier) coolers at Wal-Mart that you can use to cool or keep hot beverages. Fill it with H20 and turn it on to cool. Next take a radiator and run a line from your reservoir and back to your CPU. Stick the whole unit in the water and fire up your computer. If you don't think it will work or don't want to try it you will probably be able to buy such a unit in the near future
HeThatKnows
Gawd
- Joined
- Apr 19, 2004
- Messages
- 577
Ah, but TopNurse you're setting things up differently than wtiger.
He has the pelt inline with the rad and all. That means that as the water temperature drops, the radiator looses effectiveness. You end up with a balance, the rad dumps some of the loop's heat input, the pelt pulls out the remained. If the pelt was bigger (a lot bigger), then it would have the potential to move all the heat input and cool to water below ambient (at which point the radiator would have warm air and cool water -- the rad would be absorbing heat, not dumping it).
Your arrangement, with the rad in a chilled reservoir, also has weaknessess. First, if you're going to air cool pelts you need a lot of pelts and/or bigger heatsinks. The idea being that you don't want to waste your cooling capacity on have the hot side hot enough to dump it's heat. If a single pelt with a normal heatsink needs its hot-side 30° over ambient to dump it's heat load, then with a sink twice as big the hot-side (and thus the coldside) could be 15 degrees cooler. Add a second pelt, and each one only needs to pump half the system's heat, which mean the pelts can maintain a higher temperature difference between the hot and cold sides. The little Walmort cooler only have a single pelt with a pitiful heatsink.
Second, you're using water in a res as an intermediate heat-exchanger. Again, don't waste the limited temperature-difference capacity -- if the pelt(s) can cool the water they're in contact with down to X°, why have the coolant Y degrees hotter? (it has to be hotter to dump it's heat...)
A water-cooler pelt chiller is very possible, even an air cooled one it do-able. But you got to have the knowledge.
He has the pelt inline with the rad and all. That means that as the water temperature drops, the radiator looses effectiveness. You end up with a balance, the rad dumps some of the loop's heat input, the pelt pulls out the remained. If the pelt was bigger (a lot bigger), then it would have the potential to move all the heat input and cool to water below ambient (at which point the radiator would have warm air and cool water -- the rad would be absorbing heat, not dumping it).
Your arrangement, with the rad in a chilled reservoir, also has weaknessess. First, if you're going to air cool pelts you need a lot of pelts and/or bigger heatsinks. The idea being that you don't want to waste your cooling capacity on have the hot side hot enough to dump it's heat. If a single pelt with a normal heatsink needs its hot-side 30° over ambient to dump it's heat load, then with a sink twice as big the hot-side (and thus the coldside) could be 15 degrees cooler. Add a second pelt, and each one only needs to pump half the system's heat, which mean the pelts can maintain a higher temperature difference between the hot and cold sides. The little Walmort cooler only have a single pelt with a pitiful heatsink.
Second, you're using water in a res as an intermediate heat-exchanger. Again, don't waste the limited temperature-difference capacity -- if the pelt(s) can cool the water they're in contact with down to X°, why have the coolant Y degrees hotter? (it has to be hotter to dump it's heat...)
A water-cooler pelt chiller is very possible, even an air cooled one it do-able. But you got to have the knowledge.
I read your post with interest and if the truth be known I was thinking (just as I hit the enter key) that I was changing the conditions
But that Wal-Mart thing really does work as it was left on an entire weekend without incident. What I am curious to find out is how long it lasts when on for a few months. Though I think I would add a small submersible pump to just keep the H20 moving a bit.Now having to have the knowledge is sometimes yes and sometimes no. Many wondrous ideas have come about simply because someone wondered if something could be done and just tried it. As a child many moons ago I always wondered whether introducing a radioactive source to a laser would increase the power output due to an increase in the excitation of the atoms. Turns out that recently this has been shown to work very well with certain types of lasers. I guess I should have gone to school to be a physicist so I could have laid my hands on some radioactive material so I could experiment heh? Then I could be a famous laser physicist instead of a not-so-famous nurse
HeThatKnows
Gawd
- Joined
- Apr 19, 2004
- Messages
- 577
Works? Yes. Could work a helleva lot better? Oh yes!Top Nurse said:
My chiller uses six 226W/15V/24A pelts running at 4V
. It cools better than your Walmart thing while using less energy.It was just an idea, but you're probably right about it not cooling the water enough w/o having a much bigger pelt. Just something I was thinking about messing around with; because it wouldn't cost that much and it'd be entertaining to see exactly what kind of results I would get. I was more like 10° or so below ambient max so I didn't have to worry about condensation.. hell it probably wouldn't even help the oc all that much in reality. I just wanted to try it.
It's for fish nuts so they can keep their fish tanks at an ideal temp for their fish. It seems most common for saltwater tanks; since a 100 gallon saltwater tank seems to work best with 1000+ watts of lighting. This heats the water quite a bit and is bad for the fish and coral so you get a chiller to help keep your temps in check so your coral and such stays healthy.
They're also quite expensive as I found out when I started looking into using one after I posted.
http://www.aquadirect.com/store/customer/home.php?cat=5
They're also quite expensive as I found out when I started looking into using one after I posted.
http://www.aquadirect.com/store/customer/home.php?cat=5
HeThatKnows
Gawd
- Joined
- Apr 19, 2004
- Messages
- 577
Some day I'm gonna get a digital camera...
It's pretty simple -- a sandwich of hot-block/3 pelts/cold-block/3 pelts/hot-block.
For the cold block, I took two pieces of 3/8"x2.5"x8" copper and drilled lots of holes, with the holes overlapping a little, to make a pin-grid block. Then clamped them together and drilled an inlet and outlet in one end sized for 3/8 soft copper tubing. Finally, I soldered the peices together.
Yay for paint! Fear my massive waterblock!
The hot blocks used a pair of 3/8"x3"x8" aluminum bars and a pair of 1/8"x3"x8" caps. Again, lots of drilling to make pin-grids, inlets, outlets. The water channels in the hot blocks don't go all the way to the edge, so I could drill a series of hole along the edge and bolt the blocks to each other, clamping the pelts and coldblock in the middle. I don't have easy access to aluminum welding, so these blocks are sealed and held together with marine epoxy.
The pelts have quick disconnects on the wires. I can wire groups of three of them in series across the 12V rail, giving 4V per pelt (the usual); or in pairs across the 12V for 6V; or in pairs across the 5V for 2.5V; or all of them parallel across 5V or 3.3V; and so on and so forth.
A 4V, they keep the cold water nearly 20C under ambient while cooling a load of around 150W.
It's pretty simple -- a sandwich of hot-block/3 pelts/cold-block/3 pelts/hot-block.
For the cold block, I took two pieces of 3/8"x2.5"x8" copper and drilled lots of holes, with the holes overlapping a little, to make a pin-grid block. Then clamped them together and drilled an inlet and outlet in one end sized for 3/8 soft copper tubing. Finally, I soldered the peices together.
Yay for paint! Fear my massive waterblock!
The hot blocks used a pair of 3/8"x3"x8" aluminum bars and a pair of 1/8"x3"x8" caps. Again, lots of drilling to make pin-grids, inlets, outlets. The water channels in the hot blocks don't go all the way to the edge, so I could drill a series of hole along the edge and bolt the blocks to each other, clamping the pelts and coldblock in the middle. I don't have easy access to aluminum welding, so these blocks are sealed and held together with marine epoxy.
The pelts have quick disconnects on the wires. I can wire groups of three of them in series across the 12V rail, giving 4V per pelt (the usual); or in pairs across the 12V for 6V; or in pairs across the 5V for 2.5V; or all of them parallel across 5V or 3.3V; and so on and so forth.
A 4V, they keep the cold water nearly 20C under ambient while cooling a load of around 150W.
HeThatKnows
Gawd
- Joined
- Apr 19, 2004
- Messages
- 577
Not much of a change in power bill, since I already had four or five computers running 24/7. Plus I picked the number and size of the pelts and their operating voltage that I could meet my chilling goal with minimal power use. That said, a chiller made from a modified A/C or dehumidifier would be more energy efficient (but louder and bulkier).