air conditioner unit for a computer?

[greasepain37]

i started looking into cooling for computers. i was looking but i havent found any air conditioner units for inside a computer, if they even make those. would something like that be a good ideal. even if there not on the market yet i'm sure i could make one. has anyone seen anything like that and what temp. should the inside of my computer be?

 

[techfuzz]

The big problem with AC units is the amount of water vapor they generate. That obviously is gonna be bad for a computer parts.

techfuzz

 

[lastig21]

They do have refridgeration units just for the cpu. Vapochill and Promethia

 

[ChrisADuffy]

Why not just use a watercooling set-up?

 

[greasepain37]

becouse doing things different are better, and i'm about to trust water cooling setups just yet, lines might leak and so on..............and i want to know all the optins befor i decide

 

[beatle]

Different isn't always better. Refrigeration can get your cpu temps down further than watercooling, but I'd wager there is a bigger risk over watercooling, especially if you're not buying a ready made system.

 

[greasepain37]

i built my own system. and not to bash anyone here but i dont think any one that posted knows what there talking about. for starters ac units dont blow out moisture, and useing justa cpu cooler will work but what about the hard drives and cd drives that could be a few degs cooler. i could cool my whole unit in one shot or buy cooling units for every component and pay twice as much.

 

[SpeedTester]

Originally posted by: greasepain37
i built my own system. and not to bash anyone here but i dont think any one that posted knows what there talking about. for starters ac units dont blow out moisture, and useing justa cpu cooler will work but what about the hard drives and cd drives that could be a few degs cooler. i could cool my whole unit in one shot or buy cooling units for every component and pay twice as much.

Go ahead do it, watch how fast your PC either rusts away or shorts out from moisture.

 

[308nato]

Ever see that stuff that appears on a gallon of milk or your fridge shelves when you open the door ? Its MOISTURE.

What is that white stuff that rolls out of the inside of your freezer when it opens ? Smoke ? MOISTURE.

It might be possible if you could design a hermetically sealed case that held its temp +/- zero percent and you never opened it or shut it down.

 

[FenrisUlf]

Actually, this might not be too bad of an idea. It may not be too practical, but not impossible either. Water inside the computer will be less of a problem than most people think. The cold coil from the refrigeration unit will cause the water vapor in the air to condense to liquid. In a normal air conditioner unit, this water either drains outside the home, or is splashed onto the hot coil to aid in cooling. The air coming from the cold coil therefore contains LESS water vapor (humidity) than incoming air. Cold air can carry less water vapor than warm. This is also how dehumidifiers work. You would just have to account for water drainage from the cold coil. Also, you would have to make sure the cold coil doesn't get too cold - water could freeze and block airflow. A big problem would be finding a small enough air conditioner - most window units are designed for rooms with a bigger heat load than a computer. The physical size of the units might make the task daunting and the lack of a heat load might cause the coil to become too cold. If you could find a small enough air conditioner, you might be able to adapt it for use with a computer.

 

[techfuzz]

FenrisUlf, good ideas but one issue that still needs to be addressed. I'm not too familiar with the science behind it, but dew point is what it is called. When an object's temperature is cooled down to the dew point, condensation starts forming on the object from water vapor in the air. This could occur if you chilled objects inside the computer. Like I said I don't know the science behind the terminology but it is something that would need to be addressed and made sure that it didn't happen otherwise you'd have a very expensive, but quick fireworks show.

techfuzz

 

[steell]

Actually, the place where my brother works just disposed of several air conditioners that were used to cool rack mount computer systems. They mountedon the outside of the rack and blew cold air into it to cool the computers. If you are blowing cool air into a case, it is highly doubtful that you will have condensation problems inside the case. The outside of the case may be a different story tho

Steve

 

[greasepain37]

bla

 

[greasepain37]

i agree with keeping the case above dew point. i started think about electric coolers, igloo makes a thermal kool 35 which cost about $100, . which would be the only expence to make this compaired to liquid cooling my case. this unit i talk about is about 4in wide 8in long and maybe 6in thick. can easily be installed to the top portion of the case. i first thought about the air conditioner becouse it can cool everything inside your case unlike the liquid cooling that only works on the part it is hooked to. i will build it, most likly around april or may

 

[FenrisUlf]

Originally posted by: techfuzz
FenrisUlf, good ideas but one issue that still needs to be addressed. I'm not too familiar with the science behind it, but dew point is what it is called. When an object's temperature is cooled down to the dew point, condensation starts forming on the object from water vapor in the air. This could occur if you chilled objects inside the computer. Like I said I don't know the science behind the terminology but it is something that would need to be addressed and made sure that it didn't happen otherwise you'd have a very expensive, but quick fireworks show.

techfuzz

You would only have this problem if you cooled the components directly and exposed them to warm, humid air. Since you are cooling the air and then using that to cool the computer, it shouldn't be a problem. The coldest point in the system would be at the cooling coil. This is where the majority if not all of the condensation would occur. The coil temp should be below the dew point and cause condensation on the coil. The water then drips down the coil/radiator and out of the system. After that point, the cooled air has lost its water vapor, lowering its humidity and lowering its dew point temperature. The components of the computer all generate heat so they will always be warmer than the chilled air, and therefore couldn't reach the lowered dew point temperature.

If you were to suddenly open the case and expose cold components to warm, moist air, then you might have a problem. Of course, any case this well cooled should be insulated to prevent condensation on the outside. I would also recommend one of those devices that continue to run fans for a few minutes after the computer shuts down. You would want to slowly return the temp inside the case to that outside by continuing to run the air over the cold coil. The coil will slowly warm up (airflow will help speed this along) and the airflow will allow a controlled change in temperature rather than allowing warmer, moister air from outside to naturally find its way to still cold components.

Sufficient airflow should evaporate any water (especially since the chilled air is low in humidity already), and condensation is pure water, which isn't conductive (impurities on the components would get into the water and make it somewhat conductive, but evaporation should happen well before then).

At one job in the past, I worked in computer rooms where we had dozens of specialized racks of computers. We had a giant water cooled air conditioner for each room that would pump chilled air through the raised floor and into the bottom of each rack. We actually had the opposite problem - too low of humidity. In order to keep static down, the auto controls on the conditioner would add humidity to the air. Of course, this was in a fairly dry climate, so the humidity wasn't great to begin with.

 

[greasepain37]

this is for everyone that disagreed http://www.heatsink-guide.com/renegade.htm

 

[dzt]

you should check on prometeia's homepage. (try www.chip-con.com/tech-info.asp)
it seems cool and far enough from moist issue
temp reach down at 35 to 40 celcius

 

[FenrisUlf]

Of course the KryoTech and other units (Vapochill) have been on the market for years. lastig21 mentioned this in the third post. These units only cool the CPU, one of your previous posts mentioned that you wanted to cool all the components, not just the CPU. This would require some type of system that cooled incoming air, not the CPU. The KryoTech and vapochill type units use lots of insulation and sometimes heaters to make sure the outside of the CPU cooling device doesn't get below the dew point and cause condensation. These units are also really expensive. Unless the computer is in a hot environment or being radically overclocked, it should have no problem as long as there is a good CPU cooling soloution. You could build a cooling system that uses an air-conditioner like system to cool incoming air, but you would have to have a pretty good handle on HVAC design and construction. It would also be big and expensive.

The igloo cool isn't a good idea. Why? It uses a peltier thermoelectric element. (Also available here) Basically, when an electric current is applied, on side gets hot and the other cool. The items in an igloo cooler don't generate heat and the heat loss through the insulation of the cooler is fairly low, therefore the peltier unit doesn't have to do alot of work. This is basically a very low power peltier, not suitable for cooling the inside of a computer. To handle the heat load inside your computer, you would need at least half a dozen high powered peltiers. I won't go into the mathematics behind peltier performance - dig around on the Melcor site if you want to find out more. Peltiers also are VERY inefficient. For every watt of heat they transfer, they also produce at least one watt at the hot side. This means that your (for example) 200 watts of heat produced by the computer would now be at least 400 watts of heat dumped into the room. If your room temperature increases a great deal (let's say 20 degrees) due to this excess heat, then even if your case is 20 degrees below room temperature you would still have a net gain of 0. Plus you now need a 200 watt (minimum) power supply, and don't even think of using a computer power suppy, you'll need a high current linear type. Plus massive heat sinks and fans. Peltiers are best when used for targeted cooling - CPUs for example. Although this runs into the same issues with insulation and condensation as the Vapochill type units, as well as problems of its own.

I've heard of one guy who used a peltier to cool the water in his water-cooling rig to below room temperature, but it took alot of experimentation and tweaking to get any kind of gain (most times it actually cased the water temperature to slightly increase!). In case you're wondering, I've done the watercooling thing for years (I started about the same time as Jim at Benchtest, check out his site!) and was running a peltier back in 1997 on an AMD K6-200. I've got several pelts and the components for some power supplies sitting around at home - they're more trouble than they're worth, IMHO. That's why extreme cooling has gone to phase-change refrigeration and water. Water cooling can be done for (fairly) cheap and if you build it right, is leakproof.

 

[OulOat]

If you use cold air to cool your case, won't it eventually cause your room temp to reach the same temp?

 

[FenrisUlf]

Originally posted by: OulOat
If you use cold air to cool your case, won't it eventually cause your room temp to reach the same temp?

Depends. The heat from the air conditioner has to go somewhere. If it's dumped back into the room then the two negate. Actually you will end up with an increase in temperature due to the energy used by the air conditioner and computer. Plus you have people, monitors, and possibly other devices generating heat in the room. The air conditioner makes a temperature differential, but it doesn't "create" cold. Temperature is a form of energy, and therefore can neither be created or destroyed. You can re-organize it, which is what an air conditioner does, moving high energy states (heat) to one place and low energy states (cold) to another. The secret is to get that heat outside of the room where it can be transferred to the great outdoors.

The evaporator (cold coil) should (in any realistic scenario) be the coldest point in the room. After that heat is added to the air. Since heat is added, air exiting the case will always be warmer than inside the case (unless the computer is off and you accidentally leave the air cooler on). This air may still be cooler than air in the room, and may therefore provide cooling to the room, as long as the heat from the air conditioner is being dumped outside the room. Anywhere energy is being utilized, there will allways be heat created from that energy due to inefficiencies (energy changes forms from electricity to heat) and therefore there will be a natural energy (heat) flow from hottest to coldest. When heat into a system = heat out then you have a "steady state" system where the temps remain equal. I could get into a discussion on thermodynamics and heat transfer, but I think that is beyond the scope of this thread.

 

[Shack70]

They do make AC units that are small enough to use with a PC, but they are very noisey. I've seen them used in electrical panels on large industrial production machinery. They keep the large panels nice and cool and yes some of these panels hace PCs inside them. Like I said though they are very noisey.