10,000 BTU A/C window unit to cool PC?

86 5.0L

Supreme [H]ardness
Joined
Nov 13, 2006
Messages
6,992
Summer is almost here, and my PC is heating the room well beyond comfortable levels, right now I have the PC under the window with a giant box fan in the window blowing out, helps somewhat especially on colder nights. but now its not enough, so heres my idea, install an a/c window unit, concoct some sort of ducting from the window unit to the PC intake fan(s) I believe this will be enough to keep the PC cool and my ambient temps in check.

thoughts, comments?
 

guitarslingerchris

Supreme [H]ardness
Joined
Oct 29, 2004
Messages
6,886
If the A/C is enough to drop the ambient that should be all you need. I wouldn't worry about ducting it into your case or anything.
 

heatlesssun

Extremely [H]
Joined
Nov 5, 2005
Messages
44,154
Yeah, I have a 180sq ft office with a 6000 BTU windows AC to agument and offset my central air and it works great. Looking for a better one, the one I have now is a little loud and its about 4 years old anyway, there should be better models out now, this was just a cheapo Walmart unit. You really can't find nice window ACs in stores these days since every new house had central AC now.

But yeah nothing special reall needed, it keeps the room as cold as I wan't unless is its REALLY hot, over 95F which isn't fairly rare in my part of North Carolina even in the summer, but even then if the central AC is on its still keeps everything pretty cold.
 

KublaKhan

Limp Gawd
Joined
Feb 27, 2004
Messages
222
I have a question regarding your question... :)

Wouldn't the cold air from the A/C cause condensation on the hardware??

I used to have my computer in a small bedroom cooled by a 8,000 BTU A/C, it kept the rig well cooled without any duct work.
 

86 5.0L

Supreme [H]ardness
Joined
Nov 13, 2006
Messages
6,992
If the A/C is enough to drop the ambient that should be all you need. I wouldn't worry about ducting it into your case or anything.

that was my first idea, but thought it would be alittle extra [H] to have a 10,000BTU CPU cooler :p

and +1 for the condensation, I guess my ducting idea is scraped :D

FYI, im sitting on a ~95*C CPU right now(F@H) and ambient is in the ~90*F range just in this room
 

guitarslingerchris

Supreme [H]ardness
Joined
Oct 29, 2004
Messages
6,886
I feel for you there, with just my server running in this room it stays at 83F pretty constantly, with the window open in the middle of the night I sometimes see it drop to 79F which is wonderful. Stupid Atomic Clock in here lets me know just exactly how miserable it is in here at any given time.
 

Zink

Gawd
Joined
Feb 9, 2010
Messages
693
I think that the air conditioner should take all of the moisture out of the air as it passes through, mine drips water in the summer. Anyone with a windows AC unit could turn it on and put a spoon or something in front and test, mines not installed now.

You could make a shelf to hold the case at window height with the side panel off and point all the case fans blowing out of the case. Air would flow in the side then out the back, (top) and front. You could make some ducting out of corrugated cardboard spray painted if you want it to look/be [H]arder but you wouldn't need any ducting.

The air conditioner already has a dust filter which is nice and it would keep your current cpu & cooler at 65C no matter the outside temperature. The air coming out should be about 10-15C which is about 25C cooler than you current temps and probably 30C cooler than your case ambient temp so that's 95C-30C = 65C.

Sounds like a good idea to me.
 

LOCO LAPTOP

[H]F Junkie
Joined
May 4, 2006
Messages
10,850
il have to find out with my old 12000btu unit that put away since I got a mini split so I don't worry about it anymore.
 

Rossi~

2[H]4U
Joined
Dec 8, 2008
Messages
3,479
So long as the hardware is warmer than the ambient still, condensation shouldn't be much of a problem. Get the A/C to cool the room and the hardware slightly, would be very nice for the summer :)
 

KublaKhan

Limp Gawd
Joined
Feb 27, 2004
Messages
222
So long as the hardware is warmer than the ambient still, condensation shouldn't be much of a problem. Get the A/C to cool the room and the hardware slightly, would be very nice for the summer :)

But he's talking about ducting the A/C into the case...that's going to lower the inside temps. to below what the air outside the case is, causing condensation, right?

Even if he ducted only one side of the A/C and left the other side to cool his room, it still wouldn't work, I don't think. He's using half of his available air to cool a (xx) by (xx) room and the other half to cool a (x) by (x) computer case. Of course, the case is going to be cooled quicker, which would again lower the temp. inside the case to below what is outside.

The way to do this, and be safe, is just have the A/C cool the entire room, get your 80 or 120's to pull the ambient air into your case and then exhaust it out the back/top, whatever...

Now, someone can come along and tell me why I'm an idiot.. :)
 

Rossi~

2[H]4U
Joined
Dec 8, 2008
Messages
3,479
Sorry, by "Get the A/C to cool the room and the hardware slightly, would be very nice for the summer" I meant cool the room as a priority, but try and use the A/C to cool the case as well. Just make sure that the ambient room temperature is cooler than that of the computer, which is easily done
 
Joined
May 11, 2010
Messages
14
Build a sealed duct system directly into the case and forcing air through the fans and the heatsink. make the exhaust system sealed to the outside of the case as well. Simply put, a closed loop system for cooling the heatsink. You can build a collection tray to sit below the duct and heatsink to port any condensation off the ducting out of the case.

I think that might be a pretty cool mod to try, just might get on that as well. I thought about setting up a water cooling system ported into the side of a small refrigerator as well.

I am always contemplating something crazy like that...
 

nissanztt90

2[H]4U
Joined
Apr 13, 2007
Messages
3,858
Refrigerators are not meant to handle heat loads...if you try and do something like that with a refrigerator you will overload it and burn it out. If you want to go crazy then build a water chiller out of an AC unit.

Also, as far as condensation with ducting the AC air to the PC...you guys are forgetting one thing...the "ambient" air is the air from the AC unit...the parts will never be colder than the air cooling them, which is how you get condensation. Condensation would not be a problem. An as someone else said...the AC removes water from the air anyway.
 

ghost6303

2[H]4U
Joined
Jul 24, 2004
Messages
2,291
Now, someone can come along and tell me why I'm an idiot.. :)

Il take a crack at that.... :p

condensation happens when a solid object is cold enough to permit the water vapor suspended in the surrounding air to condense back into a liquid, which forms on that object. the more humidity in the air, the faster this happens. it also has to do with the dew point which has to do with atmospheric pressure.
http://en.wikipedia.org/wiki/Phase_change

the already cooled air coming directly from the A/C is going to have less humidity in it then the air in the room. it will be blowing against a relatively warmer solid object (the computer), at the same time displacing the room temperature, higher humidity, air. If anything, this will prevent the water vapor from condensing on the computer.

the only way water vapor would condense is if the computer is of lower temperature then the surrounding air. if you surround the computer with low temperature air, from the A/C, you are making condensation harder to achieve.


getting back to the original post- i have run that sort of setup in the past with great results. my previous computer had a 120mm intake fan on the side of the case, so i just duct-ed up some flexible tube from the window a/c to that fan. in the winter i did the same thing except just stuck the tube out the window so the fan could draw in 0*F air.

overkill? sure. [H]? absolutely.
 

Hallis

Supreme [H]ardness
Joined
Oct 26, 2001
Messages
4,558
actually i just dug up this picture for another thread, but it illustrates the point nicely....

http://img.photobucket.com/albums/v239/ghost6303/DSCF0276.jpg

Am i missing something or is he blocking the whole AC output with a solid piece of metal?

To the OP:

With having several PC's in the bedroom that kick out some heat I also agree with the window unit to cool down the ambient in the room itself. Can be switched off and on as needed. I'll be looking into something similar as I can't leave my desktop rig on folding at night as it heats up the room so much i can't sleep.

If it's nice and comfortable for you in the room it will be for your computer as well :)
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
You could test your skills a bit...

Keep the control (thermostat, readout, buttons, etc), the compressor, and the outside radiator.

Bleed the system, let all of the phase change coolant out... or, somehow cut off flow on both sides of the cooling/inside radiator.

Machine or get your hands on a block to use just like a vapochill, and attach it to where the cooling radiator was.

Now you have a phase change unit... just need to get it refilled with coolant... maybe make some tweaks to the thermostat/temp probe.
 

vjcsmoke

Supreme [H]ardness
Joined
Dec 5, 2006
Messages
4,511
I wouldn't duct it directly to your PC. Because temps that are too low inside the case might cause condensation. If you have a water loop, you might be able to add the AC to your loop to keep the water chilled, but again watch your temps not getting too frosty. I think ambient temp is one of the worse problems during the summer and the AC in general should help with that.

Edit: Read a few of the above comments. Looks like the AC removes the moisture itself? I'll defer to people who have actually done this little experiment. :)
 

spine

2[H]4U
Joined
Feb 4, 2003
Messages
2,641
Just an idea...


Why not duct the output of the case air into the input off the AC unit and have that dump cool air into the room?

Since the only movement of air in the room would be going into the PC and out, it would naturally be the first thing to suck in the cooler air.

As long as the AC unit can keep the room as a whole cool (10,000BTU is enough surely?) then that'd be a simple hassle free solution.


Not as effective as the original idea, but simpler.
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
by that logic/method, simply using the AC to cool the same that the computer is in would result in the same thing.
 

compudocs

[H]ard|Gawd
Joined
Oct 5, 2009
Messages
1,213
If you duct the AC intake to the case it will pull any moisture from the case so you'd need to duct the cold air and the intake for the AC from the case. Kinda like a WC loop only with air o_O
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
If you duct the AC intake to the case it will pull any moisture from the case so you'd need to duct the cold air and the intake for the AC from the case. Kinda like a WC loop only with air o_O

An air loop with a computer case... hmm... let me think... I get this feeling that something like this already exists, but I just cant place it.

Im gonna have to go get a beer out of the fridge as I think about this one...
:D
 

ghost6303

2[H]4U
Joined
Jul 24, 2004
Messages
2,291
Why not duct the output of the case air into the input off the AC unit and have that dump cool air into the room?

everyone is just inventing a different way to slice the same apple. in water cooling- there is no difference in performance if you put your radiator between the pump and water block, or if you put it between the reservoir and pump.

the ambient room temperature would be the same if you piped hot air from the computer into the A/C or if you piped cold air into the computer. you are still removing the same amount of heat from the room. the difference lies with getting the coldest possible air over the heatsink, creating the largest possible temp delta. the most efficient way to cool your computer would be to pump the coldest possible air directly into the heatsink. any other method would be less efficient.

most people dont have a way to quantify "heat" in their heads, so if you use a different unit of measure (other then BTU or watts) for heat, its easier to picture thermal transfer.

I wouldn't duct it directly to your PC. Because temps that are too low inside the case might cause condensation.

as already stated, there is really no danger of condensation with any of these setups as the parts that you dont want condensation on are always hotter then the surrounding air, and the air coming out of the A/C is going to be dryer then any air that is not run through the A/C, so as long as you dont get condensation forming at normal room temperature, you would not get condensation with the A/C.
 

rive22

Supreme [H]ardness
Joined
Mar 10, 2004
Messages
4,640
I had a duct from my window to my pc with push pull fan config for the duration of the freezing winter and there was no condensation anywhere to speak of. Average ambient incoming temp was -10C - 1C
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
...in water cooling- there is no difference in performance if you put your radiator between the pump and water block, or if you put it between the reservoir and pump.

the ambient room temperature would be the same if you piped hot air from the computer into the A/C or if you piped cold air into the computer. you are still removing the same amount of heat from the room. the difference lies with getting the coldest possible air over the heatsink, creating the largest possible temp delta. the most efficient way to cool your computer would be to pump the coldest possible air directly into the heatsink. any other method would be less efficient.

most people dont have a way to quantify "heat" in their heads, so if you use a different unit of measure (other then BTU or watts) for heat, its easier to picture thermal transfer.

as already stated, there is really no danger of condensation with any of these setups as the parts that you dont want condensation on are always hotter then the surrounding air, and the air coming out of the A/C is going to be dryer then any air that is not run through the A/C, so as long as you dont get condensation forming at normal room temperature, you would not get condensation with the A/C.

NOT TRUE. I know what you are getting at, and in many small-loop water cooling systems, this is very true, but the circumstances do change for larger 'loops', as would be the case with a 'whole room' A/C or a water cooling setup with longer lengths and lower flow rates. Things are also very different because this is a compressor/phase-change system where the deltas between various points in the system are going to vary.

There is a very real possibility of condensation here... we aren't talking just heat transfer, but actual phase-change cooling being used... the A/C unit itself has condensation on its cooling coils, usually collected in a pan and ducted to the outside radiator and fan area for evaporation. If what you said was true, then this wouldn't be happening, now would it? Since the temperature of the air going through the cooling coils IS that much cooler than the surrounding air, there is a very real (and likely) danger of condensation. Sorry, but you have taken a 'rule of thumb' from another very specific application with a narrow set of qualifications in thermodynamics and totally misapplied it to this situation.

For example, if A/C was ducted into the case, the parts in the case would be very cold... much like in the fridge. When the A/C was shut off, the warm air on the outside of the case could rush in and presto... condensation on everything inside. This is much like when you open your fridge on a hot humid day... everything inside gets almost instant condensation on it. If you pull a bottle of liquid out, the botte/can it is in gets condensation on it almost instantly. In phase-change systems and larger closed loops, the temperature delta between points in the system is much larger... this is why systems like vapo-chill have to insulate the lines going to the blocks... or they would start dripping.
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
I had a duct from my window to my pc with push pull fan config for the duration of the freezing winter and there was no condensation anywhere to speak of. Average ambient incoming temp was -10C - 1C

I think you got lucky... that, and when outdoor temps are that low, the humidity is also very very low. If you did a similar setup by ducting cold air from a freezer locker into your PC in an otherwise humid environment, condensation would cause problems.

Think about how much water your window or central air system builds up on the cooling coils... or rather, a dehumidifier works by that very method. Unless you happen to live in the very dry and arid pacific southwest, or perhaps the rockies, condensation is a real concern.
 

Emission

Supreme [H]ardness
Joined
Dec 6, 2005
Messages
4,377
that was my first idea, but thought it would be alittle extra [H] to have a 10,000BTU CPU cooler :p

and +1 for the condensation, I guess my ducting idea is scraped :D

FYI, im sitting on a ~95*C CPU right now(F@H) and ambient is in the ~90*F range just in this room

95 *C CPU? That doesn't sound too healthy.
 

ghost6303

2[H]4U
Joined
Jul 24, 2004
Messages
2,291
There is a very real possibility of condensation here... {...} the A/C unit itself has condensation on its cooling coils, usually collected in a pan and ducted to the outside radiator and fan area for evaporation. If what you said was true, then this wouldn't be happening, now would it?

that is not a relevant example. if the CPU in the computer was absorbing heat instead of producing it, that example would be spot on. however the CPU/computer is less like the evaporating coil (cold side), and more like the condensing coil (hot side), which never gets condensation on it.

For example, if A/C was ducted into the case, the parts in the case would be very cold... much like in the fridge. When the A/C was shut off, the warm air on the outside of the case could rush in and presto... condensation on everything inside. [...]

in this situation that would equate to having your water pump die in a regular liquid cooled PC. sure if your computer was say 50*F sitting in a nice cold and relativelydry air conditioned room, a/c cranked up to max for hours prior, then you pick it up and bring it outside into 100*F 80% humidity, yes you will get condensation. that will happen on absolutely anything with mass, not just a computer.

however if you simply turn the a/c off in the already cooled room, the computer components that are no longer being fed cold air would heat up way faster then the surrounding air. preventing condensation. it would take much longer for the room to come back up to equilibrium, then for the computer. so the refrigerator reference is moot.

NOT TRUE [...] the circumstances do change for larger 'loops', as would be the case with a 'whole room' A/C or a water cooling setup with longer lengths and lower flow rates
i spent 4-5 years of my life running different water cooling setups (including chilled water) with capacities over 5 gallons, as a hobby. i would say my understanding of how large water cooling loops work is accurate. atleast this is what i have seen, with my own two eyes, to actually happen in real life.
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
^^^First Paragraph: Then why do computers that are mounted inside a fridge, for example, have problems with condensation? Assuming that the computer and A/C will turn on and off during their lifetime AND the A/C will not be running constantly (but on a thermostat), your assumption that the CPU (and surrounding components) will always be producing heat in error. Whenever the thermostat shuts down the compressor, the air flowing through the A/C, and therefore the case right after it will be warmer than the case's contents and also contain humidity. At the moment when the A/C and computer are shut off, the components inside the case will be cooler than the humid and warm air that backflows into the case. Beyond the CPU, there are several points inside the case that are not producing heat energy (or enough) to prevent condensation (its all relative, but I think we can assume that the heating power of many areas will not be able to keep up with the cooling power of say, a 10,000 BTU A/C unit). Every component inside the case that is biased towards heat transfer more than heat production will become a condensation point unless the A/C is made to run 24/7... which is impossible. Every capacitor (lossless component) and heatsink at least, if not the case itself would become a condensation collector. It is a very relevant example because it is exactly what would happen.

Your assumption that the room will be relatively cool in comparison to the computer is also in error. It isnt going to be the case all of the time. People open doors and let warm air in... the A/C might be turned on long enough for the case to cool, but may not be powerful enough to cool the room as well as the case contents. You forget that the wattage added by the computer will count against its ability to cool the room. By the A/C being ducted directly into the computer case, the case will therefore be the coolest point in the room, and the rest of the room... not so much.

Your assertion that the computer components that are no longer being fed cool air and would "heat up way faster then the surrounding air" is also completely false. The thermodynamic heat mass of air is many times less than the physical components inside the case. Otherwise, if you left the door open on your freezer, the solid contents and liquids inside the fridge would melt just as fast as the air warmed inside. Likewise, the fact that the contents of a freezer should need a 'defrost' cycle to remove the condensation inside on a regular basis, despite its rather infrequent opening and being sealed usually serves as good warning here. There is no way that you can claim that the contents of the case (solids) will warm up faster than the air... let alone that the air inside the case will remain still once the system is shut off. Many of the components in the case are insulators at that, so having them cool by convection to the surrounding air (I doubt much radiation or conduction is possible) means that the physical movement and heat capacity of air will with rare exception result in the air warming faster than the physical contents of the computer... something which aids us equal and opposite with air cooling! If components could somehow transfer heat (to cool or warm) as fast as air, then we would have no reason to duct air through our cases!!!

The only way I could see your statements to be true is if the turnover (fan size) on the A/C unit was very high relative to the size of the room... so that the equalization of temperatures remained very fast and maintained a narrow delta between temperatures. Such a high airflow alone would prevent most condensation in itself though... wicking away moisture just as fast as it settled.

As for my resonse to your last paragraph with the credentials: I will see your hobby experience of 4-5 years and raise you a degree in Electrical Engineering. I have designed and tested cooling systems at the professional level for a duration of more than double your hobby experience. I design phase change, heat-pipe, and liquid cooling with cold plates for everything including electric motors, LED lighting, electronics, lasers, VLSI, power systems, etc. I also built and maintain a home computer cooling system that runs two computers at once with a total of 4 gallons running through it and my own 8' tall floor standing passive radiator. Due to its external nature, and the lengths of hose between it and the computer cases, the usual rules of everything equalizing due to flow go out the window... Even with about 250 gph of flow through the loop, the temps AFTER the block are a good 3-6 degrees F above the water going into it (depends on room temps).

My experience tells me that your conclusions are from circumstance, and although possible, they are not likely to be repeated in many other circumstances. All it takes is ONE drop of water to kill a board if it falls into the right place... I've done that too, and so I know how easy it is get burned by science.

So sorry ghost, still not buying your armchair engineering.
 

ghost6303

2[H]4U
Joined
Jul 24, 2004
Messages
2,291
I also built and maintain a home computer cooling system that runs two computers at once with a total of 4 gallons running through it and my own 8' tall floor standing passive radiator. Due to its external nature, and the lengths of hose between it and the computer cases, the usual rules of everything equalizing due to flow go out the window...

all of what you said still has not been true, in my experience, as it is relevant to this discussion.
some proof of my degree in armchair engineering:
http://img.photobucket.com/albums/v239/ghost6303/my PC/P605s0029.jpg
http://img.photobucket.com/albums/v239/ghost6303/my PC/P6050041.jpg
http://img.photobucket.com/albums/v239/ghost6303/my PC/P6050030.jpg

you also must have missed the picture i posted earlier, of an air conditioner ducted directly into a water cooling radiator! no condensation after 8 months. thats 55*F, well below the due point for almost every single day in the summer where i live.

however, in the third link above, i used to fill that white bucket with ice until the water temp, measured from within the loop, was around the same temp as before. in this instance though, there was condensation forming on some of the metal bits of the loop. why you ask? because circulating a cold liquid thru a few warm components creates much better conditions for condensation. blowing cold dry air over all the components, does not.

http://img.photobucket.com/albums/v239/ghost6303/my PC/DSCF0286.jpg
here you can see an internal air conditioner i made out of a spare water block, heatsink, and 226w peltier. the case is sealed for the most part, everything inside that produces major heat had a water block on it, hooked to an external radiator. the hot side of the pelt was hooked up to the water block, and the cold side to the heatsink. the heatsink reached temps of 35-40*F at times, and internal case temps were in the low 50s at idle, with external ambient temps of 75-80F and fairly high relative humidity. this setup ran for around 18 months no problem. no condensation.


My experience tells me that your conclusions are from circumstance, and although possible, they are not likely to be repeated in many other circumstances. So sorry ghost, still not buying your armchair engineering.

my conclusions, regardless of their accuracy or relevance to other real world processes, are exactly what i have seen in real life. i work in a physics lab for a company that makes lots of household products, if you really want to compare cojones. products that i bet are even within 25 feet of where you sit right now. this doesnt have so much to do with electrical engineering, so your degree, however impressed I am by it, is moot. this is more of a thermodynamical argument, (which happens to be a branch of physics, btw ;)).

im not trying to be argumentative, but i couldnt care less about anyones educational qualifications on the internet. for all you know, i could have a masters in psyche, and this whole conversation was just to screw with you. or i could just be an 8 year old kid who knows how to use wikipedia.

the internet is serious business, after all.
 

undertheradar

[H]ard|Gawd
Joined
Oct 26, 2004
Messages
1,792
You are tossing around alot of examples here and there... a shotgun of experiences and different applications, but the application here only requires the application of one basic concept: "Water vapor will condense onto another surface when that surface is cooler than the dew point temperature, or when the water vapor equilibrium in air has been exceeded."

It is possible that your altitude could even be giving you varying results. Maybe as a consequence of your modified A/C system, the resulting temperature in the block was not as cool as some A/C's can get.

If the components inside a computer are cooled via A/C, they will be cooler than the dew point. Even if the room that the A/C is in is insulted and sealed, you can not claim that the temperatures in the room will be as low as in the case (simple logic, feel the cold air right at the outlet of any A/C, is the rest of the room that cool?). Depending on how cool and the gradient in temperatures (perhaps what you are also overlooking) from the cooler points in the case to the room air is, you have the potential to see condensation. You can not assume that the A/C will cool the room as well as the computer (in fact, it is unlikely since the cool air is being focused into the case, not spread into the room first). You should also not assume that the components in the computer will undergo a raise in temperature as fast as the air passing through them once the system is shut off.

IF there is a way to create a thermal buffer, or insulating layer between the cooled points and the warmer air, it is possible to avoid direct contact between the cold points in the case (maybe the case itself) and the outside room air. If greater thermal conductors are used (such as an aluminum case), there is the potential for greater condensation (or if the total cooling area is spread out over a larger area so there is less of a difference between the air temp and the cooled material).

There are several variables here, but some basic ideas (no degree required to understand) that cause concern when ducting an A/C's cool output directly into a PC case. FWIW, you might as well consider it to be similar to a phase-change unit. A phase change unit only creates a very focused temperature drop though, and beyond the CPU where it contacts, the exposed components (the pipe) are heavily insulated to prevent condensation. By ducting the cold air from an A/C into a PC case, you are cooling every component in the case, as well as the case itself. The cold air going into it will be warmed as it exits into the room by the PC, and the coolest air is from right where it exits the A/C, so the room will be warmer than the PC. Depending on how cold the A/C is, and how warm the room is, you can see condensation. This also depends on the heat properties of the materials used in the case... bare metals will transfer their heat to the cold air faster than the plastics, rubbers, etc... and so they become the primary condensation points. It is possible that with enough focus of cold enough air, the case itself would become the condensation surface.

I really cant say how you have lucked out and not had condensation problems... too many variables to consider. But I can say that condensation is still a very real concern, and proving it would be easily done. In winter, leave a computer outside to cool while off (you may argue that it needs to be on, fair enough, but there are several components that gain no heat from the computer being on). Then, go take a shower and warm up that bathroom to be something similar to summer... humid, warm, etc.

Then go outside and bring that case in quickly to the bathroom.

I dont think there is a soul here that would think the case wouldn't collect condensation both outside and inside as the warm air flows into the case (convection).

This simulates the condition where you turn the system off, then the A/C, in a humid/warm room (summer). Also, an A/C does not cool constantly... it has a thermostat. Even if the thermostat tells the unit that it still hasn't cooled enough, it will still shut down the compressor from time to time to preserve the motor. During these periods, warmer and more humid air will be pumped through the A/C than when the compressor is on.

You could also simulate this by pulling a cold can or bottle out of your fridge on a warm summer day. .. proving that no, the various things inside a computer case do not gain temps at the same rate (and in the case of the air, it doesn't stay still either).

Heck, I just took a shower and the humidity was enough to cause the mirror glass and metal fixtured to collect moisture, and its 75F here! Its all relative... humidity, exact temps, heat mass, gradients in heat, air speed, dew point, altitude, surface area... you might claim that you have never had humidity collect on your bathroom mirror, but that does not mean this is always the case. Something tells me that in your situation(s), your phase change system just never lowered the temps of any condensation surfaces of the computer to be lower than the dew point in the room(s) you happened to be working in, and the humidity levels you work in are low. I say that because your use of the A/C was to cool the water cooling radiator, not the components directly... this creates a very gradual gradient (an insulating buffer) that slows the transfer of heat from the CPU to the A/C through the water cooling system, not related to what we are talking about here. I too have done that... I spliced a 50' length of copper water line into my external cooling array and hung it out the window during cooler months. I had to be careful though, because if that copper shed heat (too cold outside) faster than my computer blocks and pump could supply it, the water inside the line would freeze... even with the warm water pumping through it. Now, since humidity is lower in cooler weather, I didn't experience much in the way of condensation inside, but I had to make sure that the water in the lines didn't drop below 32 degrees F. Still, one day, I took a hot shower (blowing humid air into the room where the computer radiator sits), and the floor standing radiator collected condensation even though the water in the loop was over 40 degrees F! Luckily, the water block did not get cool enough, but the tygon hose started to. The easiest solution at the time was to aim a desk fan at the hoses... increasing the air flow reversed the condensation... in this case. I wouldn't make any scientific observations based on those circumstances, but just an acceptance that there are more things going on than we might be controlling.

Any time you introduce a cooling component that can cool something in your computer to below the dew point temperature, you have cause for concern.

Thermodynamics is physics, sure... all branches of engineering are applications of physics. Thermodynamics is a huge field in itself that is touched on by some areas of physics, but it is a huge concern of electrical engineers, and although a specialty in such a field exists mostly in mechanical engineering, it is such a common concern of EE's that almost all electrical engineering education and practice includes thermo.

Otherwise, we could also argue that electrical energy flow is based on physics, so one more versed in physics would be better at electronics than an electrical engineer.

I can tell you have backstepped on several conclusions that you previously made and I pointed out were in error with the laws of thermodynamics, yet you still argue that somehow I am wrong based on your experiences that dont exactly apply.
 
Joined
Nov 17, 2010
Messages
5
Hey friends, how about strip the PC bared - meaning side panels out of the way - and enclose the PC inside an insulated enclosure - a bigger box that the PC case - and seal it tight, then pipe the cold air from the window A/C through it and back into the A/C input vent, in a closed loop. All cables can be routed through tight fitting holes, except a DVD drive, which can use an external USB drive. Would this solve the condensation problem, since the PC now would have no contact with outside warm air - room air, that is?

To overkill it, use water cooling rig... put a 3X360 radiator near the output of the A/C unit, so the CPU/GPU receive the chilled liquid through their water blocks.

To access the PC and its cooling parts, make one of the box's side panel hinged.
 

twelveparsex

[H]F Junkie
Joined
Nov 30, 2000
Messages
12,955
Refrigerators are not meant to handle heat loads...if you try and do something like that with a refrigerator you will overload it and burn it out. If you want to go crazy then build a water chiller out of an AC unit.

Also, as far as condensation with ducting the AC air to the PC...you guys are forgetting one thing...the "ambient" air is the air from the AC unit...the parts will never be colder than the air cooling them, which is how you get condensation. Condensation would not be a problem. An as someone else said...the AC removes water from the air anyway.

yeah but the ac cycles on and off...unless you crank the thermostat way down. when it cycles off is when you have to start worrying about condensation
 

EricFX1984

[H]ard|Gawd
Joined
May 21, 2010
Messages
1,122
To overkill it, use water cooling rig... put a 3X360 radiator near the output of the A/C unit, so the CPU/GPU receive the chilled liquid through their water blocks.

NO... all that does it make it catch hot air from the room

the best way to do this is to NOT run a rad at all

put the A/C units evaporator in a large res, insulated, fill it with winter washer fluid and automotive antifreeze (25%)

should give you temps of 0c or lower at full load

all a rad would do in a situation like this is pull heat out of the room and dump it back into the cooling system

also everything must be insulated

here is a link to some of my work

http://hardforum.com/showthread.php?t=1534223&highlight=madness
 

rick5127

Gawd
Joined
Feb 18, 2007
Messages
530
A few years ago I had the idea that I would buy a rack enclosure and install a window air conditioner in the back of it at a window height. Then back the whole thing in front of a window so the exhaust would be outside. I figured that I could easily keep the temps inside the enclosure below 70f or even lower. Now with all this talk of condensation I am not so sure.

Now it is winter here in Michigan and I thought about drilling a 3" hole through my floor to my crawlspace below the house and ducting that air to the intake fans on my PC. Figured the cold air from the crawlspace would really cool the machine down. Now I think maybe I would get condensation and maybe I would be better off just leaving it alone.

All this reading is making my head hurt...
 

KatalDT

2[H]4U
Joined
Jul 28, 2010
Messages
2,567
I tried this once on my old E8400 4850x2 system. Took a drier tube and ran it to the intake fan on the side of an Antec 900 case. Running at 4.2GHz with load temps lower than base temps without the AC on was cool.

I took it out due to condensation concerns, also entirely unnecessary and inefficient.
 
Top