air conditioning for pc?

Q

quakefiend420

Limp Gawd
Joined
Sep 22, 2004
Messages
224
was thinking about moving a windows unit ac into another room to cool my comp...here's the idea...ac sits in window, 12"X3" duct goes from vents on front of ac unit into some kind of case about 4 times the volume of my comp case...ac set on low w/ thermostat control to keep temp inside outer case about 50F...is this a good idea? or will i have condensation problems? or some other problems that i haven't thought about, perhaps? would appreciate your input!

edit: o yeah pc sits inside ducted case...i'm sure you figured that out though... :D
 
icehokplyr said:
I would think condensation would give you the biggest problem.
Click to expand...

Condensation won't be a problem. If any condensation would happen it would be on the outside of the "outer" case.

You will need a return air duct to complete the loop. If you don't have a return airduct, put a vent hole in the "outer" case so that air can exit it.
 
that desperate huh? water cooling maybe? or u could bo with the A.C's specifically designed for computers look under Vapor Cooling
 
The window mount AC's can be bought for as low as only $100 for a cheap one and some good ones for $200 - $300.

Watercooling and Vapochill systems cost much much more than that.
 
my cooling system: a 9" windmere clip-on fan. does wonders for my toshiba satellite 1415-s173... which is 2 years old and you could bake a chicken on it under load.
 
condensation would be a problem and also a big electricity bill...
 
darktiger said:
condensation would be a problem and also a big electricity bill...
Click to expand...
No it will not be a problem.

He is going to make a cabinet that will have in interior cooled by the air from the AC. The PC will sit inside that cabinet.

Air from the AC is LOW humidity along with being cool. The only condensation that could possibly happen will happen on the OUTSIDE of the cabinet because the exterior surface of the cabinet may be cooler than the dew point of the air from the room it's in. I highly doubt it will ever happen.

So you tell me how you think that condensation will be a problem? Also, the AC units while cooling entire rooms usually only run $20-$30 a month on electricity, this should be consideralbly less...should be.
 
a hearty good luck then. it will be interesting to see how this goes...please get back then with an update. I too think condensation will be the enemy there but I am unsure of that proposed design but long hours of operation will determine...
 
Met-AL said:
Condensation won't be a problem. If any condensation would happen it would be on the outside of the "outer" case.

You will need a return air duct to complete the loop. If you don't have a return airduct, put a vent hole in the "outer" case so that air can exit it.
Click to expand...

that was why i was thinking of using a second case that the pc actually sits inside of, eliminate, or at least greatly reduce the risk of condensation giveing me issues...the reason i am thinking about this is my gf is very cold natured...ie wants the heater on a lot now that its winter...means my temps go through the roof...so i was trying to come up with a cheap solution to keeping my rig running cool...ive gone from 38-39 idle on my cpu to 45...akk...abd before you say it, i know my temps are still within acceptable limits, i just like it running as cool as possible...only problem with a exit hole is it will make the room where my pc sits a cold sob after a while...maybe i can duct it to the outside back through the window the ac sits in...i'll come back with an update and maybe some pics of the construction...once i do get this setup complete, im going to run an old box in it for a few days and see how it fares, so far as condensation...even though i know i shouldnt have a problem with it due to the outer case...it still makes me nervous :D

i'll update when i have more info...in the meantime, suggestions or input is more than welcome!

EDIT:same shitty typing
 
Scheizekopf said:
The window mount AC's can be bought for as low as only $100 for a cheap one and some good ones for $200 - $300.

Watercooling and Vapochill systems cost much much more than that.
Click to expand...

yeah since ive already got the ac unit, really all this is going to cost me is materials for the outer case and ducts...bet i can get out for under 50 bucks!

EDIT: my shitty typing
 
was just thinking would i need an exhaust fan at the beginning of my exit duct you think? or would pressure in the outer case be sufficent?

i am also thinking a temp probe inside the outer cabinet would be a good idea...what do you tthink would be a good temp to shoot for? go as low as possible? or just try for about 50F or so?

also any suggestions for materials for the outer case, ducts?
 
quakefiend420 said:
was just thinking would i need an exhaust fan at the beginning of my exit duct you think? or would pressure in the outer case be sufficent?

i am also thinking a temp probe inside the outer cabinet would be a good idea...what do you tthink would be a good temp to shoot for? go as low as possible? or just try for about 50F or so?

also any suggestions for materials for the outer case, ducts?
Click to expand...

You should have a return air duct that goes back to the AC unit so it doesn't run nonstop. The AC units thermostat's probe is behind the grill, so if your return air from the cabinet went over that, the AC unit could cycle on and off.
 
I use a window AC to keep my computer room cool. I also use it as a water chiller. I simply mounted my radiator inside the AC right after the evaporator. Water temp stays 17*C while the AC is running and due to the dehumidifying properties of air conditioning, I haven't had the slightest problem with condensation, even though my waterblocks are cold to the touch. As a matter of fact, I have a fresh cold soda sitting on my desk right now and the can is dry as a bone.

BTW, I live in FL where the humidity is always fairly high.
 
Met-AL said:
You should have a return air duct that goes back to the AC unit so it doesn't run nonstop. The AC units thermostat's probe is behind the grill, so if your return air from the cabinet went over that, the AC unit could cycle on and off.
Click to expand...

if the ac unit cycles on and off, when it cycles off wont its dehumidifying properties be lost for the off cycle? it still blows on thermostat control, just not nearly as cold...also anyone have ideas what i should try to go for as far as temps in the outer case? cold as possible or should i try to keep it at a certian temp, for condensation purposes?
 
50!? that's a bad ass room a/c i'd be afraid of blowing out the fuses on the wall sockets.
 
how did my thread get moved to the same place it was before...???

and now it wont go to the top...what gives?

lol i'm retarded...just realized thread was moved fom cooling to xtreme cooling...w00t i'm extreme :D

had a new idea though...i picked up a digital thermometer with a external probe...now i want something to check moisture levels inside the inner case...i got a moisture level meter (for plants, hey dont laugh i thought for 5 bucks it was worth a shot!)from wally world thinking it might do the trick but it seems to not be nearly sensitive enough for my needs...anyone got any ideas?

my buddy and i are starting on this project tomorrow...if he can get his dads digicam i'll have some pics and stuff for ya of what we've got so far...i dont think im going to get too cold though...stuck the temp probe in the ac vent and left it there for a couple hours to make sure temp was gonna stay stable...looks to be about 10C in the vent...im guessing i can keep about 13-15C in the outer case...so i might not get any really insane overclocks from this, but i bet i can get higher at least a bit...as long as my CPU stays at the right temp i might give it another tenth or so of a volt...might get me up a little higher, should at least let me get back to where i was before the gf and the heater..... ;)

the basic idea now is to take a table and line the underside of the top and make walls out of that styrofoam insulation board stuff(dunno what its called, exactly, hitting the hardware store tomorrow) to make the outer case...someone suggested getting a mini fridge, taking the door off and just setting it inside the outer case instead of using the ac...anyone got any input on that? also any kind of special prep i should perform on my comp to get it ready for colder temps?

almost got the test box ready to go...300w cheapo psu, p2 400, 128mb pc100, no hdd yet...gonna hit up the local pc shops tomorrow and try to find a 2 or 4 gig for cheap, no optical drives either...would kinda defeat the purpose of this little endeavor if i had to open the case all the time to insert and remove discs...gonna stick my dvd burner in my htpc(gotta love network drives :D ) gonna run this box in here for about a week under prime torture test and see if it survives...if its still ticking im going to put the good rig in there...wish me luck...

another question i had was should i try to run this on thermostat control? the reason i ask is when the compressor kicks off it sitll blows, just not nearly as cold...im wondering if i should do this or have it running constantly due to warmer air entering the outer case might cause condensation to form...if i did this i would have to run my exit duct back to the intake on the ac unit...someone had suggested i do that anyway for thermostat control but i'm kind of leary......................on another topic, though somewhat related, what do you guys think about taking silica packs (you know, that stuff they put in shoes and beef jerky) all over the place and maybe line the vent with it...just to further reduce the risk of moisture...time to go rob my local shoestore ;) also how long do those things last? well i guess it would depend on the amount of moisture in the enviroment to begin with, huh? scratch that last ?

another thing i was thinking was the dust filter in the ac's intake might cut down on the amount of dust entering my case....kind of a nifty bonus! ok enough with the long winded post...i apoligize for making you read all that, lol

thanks for all you guys' thoughts and opinions on this...makes me a little more secure doing this with my pc, should have an update for ya tomorrow!
 
This is very true the only condensation that are you going to have is outside to avoid that problem I would recommend to keep the temp. in the room low within 10 to 15 degrees of the inside of the cabinet or insulate the case and trust you don't want to do number 2.
Met-AL said:
No it will not be a problem.

He is going to make a cabinet that will have in interior cooled by the air from the AC. The PC will sit inside that cabinet.

Air from the AC is LOW humidity along with being cool. The only condensation that could possibly happen will happen on the OUTSIDE of the cabinet because the exterior surface of the cabinet may be cooler than the dew point of the air from the room it's in. I highly doubt it will ever happen.

So you tell me how you think that condensation will be a problem? Also, the AC units while cooling entire rooms usually only run $20-$30 a month on electricity, this should be consideralbly less...should be.
Click to expand...
 
with the moisture sensor thing:

the way that those plant sensors work is similar to a classic burglar alarm. electricity tends to choose the path of least resistance, in this case thru a small ballast resistor (say, 10kohm) rather than thru the led and its 470kohm resistor(in the case of a 12v circuit). because no moisture is directly shorting out the contacts of your plant sensor, it's not working... and it would be extremely difficult to make a working system. you would basically need a sheet of something that is hydroconductive (conducts electricity only when water is on it) and you'd have to adapt that to the plant sensor.

i know the stuff exists because it's used on the winshield of some cars to trigger "automatic" windshield wipers, but i can't tell you where to find it, tho american science and surplus (www.sciplus.com) usually stocks stuff that's wierd, unusual, and helpful.
 
starhawk said:
with the moisture sensor thing:

the way that those plant sensors work is similar to a classic burglar alarm. electricity tends to choose the path of least resistance, in this case thru a small ballast resistor (say, 10kohm) rather than thru the led and its 470kohm resistor(in the case of a 12v circuit). because no moisture is directly shorting out the contacts of your plant sensor, it's not working... and it would be extremely difficult to make a working system. you would basically need a sheet of something that is hydroconductive (conducts electricity only when water is on it) and you'd have to adapt that to the plant sensor.

i know the stuff exists because it's used on the winshield of some cars to trigger "automatic" windshield wipers, but i can't tell you where to find it, tho american science and surplus (www.sciplus.com) usually stocks stuff that's wierd, unusual, and helpful.
Click to expand...

i think i am going to buy a cheap humidor and rip the humidity sensor out of that...see how that works out!
 
ok we have it built...changed the design a bit from the original...there is a celeron 633 running in the case right now...got it under prime load...temps at idle were 17.7, we are at 18.9 under load right now...not getting the temps we expected...we think the problem is airflow...right now we are just using case pressure to drive the air out of the same window the ac sits in...the exhaust vent is about 2" by 12"...we have been taking video and a few stills throughout the construction this evening...we are gonna make use of my isp webspace and put up a site...we are amateur web designers so not a whole lotta critique on that, please :D if you want the whole story though, watch the video...it has a lot more info about our construction process, since we kinda neglected the still cam in favor of the camcorder......so i am open to suggestions again...do you guys think that the lack of forced exhaust is what is hurting me? another thing we were contemplating was the thermostat sensor...but the way the heat is rising...i don't think that is the problem at the moment...btw the ac is on medium right now...going to bump it up to high tomorrow probably...gonna let this run all night and see if its still alive tomorrow...also anyone who understands how a thermostat sensor in an ac works please pm me, i have some questions about maybe putting a variable resistor inline with it to fool the ac into thinking its warmer than it is...just in case we get to the point where it is limiting what i can do i can have a variable temp at a lower level...any feedback is much appreciated!

UPDATE: temp inside the case is now up to 20.6! yikes...definitley something wrong here...what do you think about ducting the pc's exhaust out...just so it doesn't create something else in the case for the ac to cool...
 
exhaust is my problem...just opened the top of my door for a little while and temps dropped from 20.8 to 18.8 in about 10 mins...now its time for forced exhaust...think i should strive for negative pressure in there?
 
ok my problem isn't exhaust its the thermostat...wrapped it in a heating pad and got the box down to 4C...can't keep it there though...gets back up to about 18 before the thermostat kicks the compressor back on...got some temp screens though!

idle_full_cool_13-29-5C.bmp


if i could keep my temp in the outer case steady i know i could get better than this...

load_full_cool_19-44-6C.bmp


need thermostat ideas!!!!
 
temps in the outer case are at -6.6C case temp shows to be -1C cpu is at 13C...kickin much ass!

e.bmp


thanks to drewmeister at extremeoverclocking for helping me in the final stages of this project!
 
Tats really neat man. Gratz.

Now for the optical. What U need to do is to mount yout optical thru the outer case.
 
thanks! yeah thats how i have my hdd running right now...im just gonna pop my dvd burner in a different comp on my network
 
starhawk said:
with the moisture sensor thing:

the way that those plant sensors work is similar to a classic burglar alarm. electricity tends to choose the path of least resistance, in this case thru a small ballast resistor (say, 10kohm) rather than thru the led and its 470kohm resistor(in the case of a 12v circuit). because no moisture is directly shorting out the contacts of your plant sensor, it's not working... and it would be extremely difficult to make a working system. you would basically need a sheet of something that is hydroconductive (conducts electricity only when water is on it) and you'd have to adapt that to the plant sensor.

i know the stuff exists because it's used on the winshield of some cars to trigger "automatic" windshield wipers, but i can't tell you where to find it, tho american science and surplus (www.sciplus.com) usually stocks stuff that's wierd, unusual, and helpful.
Click to expand...

If we're thinking of the same thing (soil moisture meters that have a metal prong that you stick into the dirt), then you're a little off on how they work. They have no electronic circuitry, just a pair of 2 different metals that create an electric potential when water contacts them. More water in the soil = better electrical conduction and more voltage is produced. It works the same way as galvanic corrosion occurring in a watercooling system - if you leave the probe in the soil long enough, you'll see that one of the metals corrodes and the other gets gunked up with crusty stuff.
 
:D whoops. i never did use one of those d*** things. i figured on a logical electric circuit with what i know of electronics, which is plenty. but not much on metals and chemicals...

/off topic
tho i did build a magnesium-copper battery once. :p boy would that have kicked @$$ if i had bothered to do a professional job... heh, nostalgia. sorry, i digress way too much.
 
got it down to -10.9C in the outer case, mobo temp was -8C CPU 8C!

having problems with the evap icing over...need to find my leakage...
 
quakefiend420 said:
another thing we were contemplating was the thermostat sensor...but the way the heat is rising...i don't think that is the problem at the moment...btw the ac is on medium right now...going to bump it up to high tomorrow probably...gonna let this run all night and see if its still alive tomorrow...also anyone who understands how a thermostat sensor in an ac works please pm me, i have some questions about maybe putting a variable resistor inline with it to fool the ac into thinking its warmer than it is...just in case we get to the point where it is limiting what i can do i can have a variable temp at a lower level...any feedback is much appreciated!


...
Click to expand...
Is this a window unit? A thermostat is just a simple swicth on/off there is nothing special about it, what you need to do is take out the t-stat sensor that is in front of the unit at the evaporator coil without braking the line and dunk it in hot water and try that to see how it works, because the other problem that you might encounter is that by fooling the t-stat the compressor is not going to stop running and you might well actually you will freeze the unit and make things worst. I haven't work in a/c units in a couple of years so I'm little rusted but I can still remmember my a/c days. That is my best recommendation for now, don't make anything permanent changes yet, and then if it works then do something more permanent. I hope this can help with something.
 
quakefiend420 said:
got it down to -10.9C in the outer case, mobo temp was -8C CPU 8C!

having problems with the evap icing over...need to find my leakage...
Click to expand...
Like I said in my previous post, there is a chance in freezing the coil. Anytime you try to bypass or fool the tstat you are going to have that problem, you are going to find another way to lower your temp. without fooling the t-stat to avoid this kind of problem. The only way I can help is looking at your setup, how can I see that video that you were talking about.
 
no leak problem. if it was leaking freon or whatever coolant they use these days (it ain't water, as sure as i'm a geek!) it'd just get a crappier and crappier effect until one day it didn't work at all. this does actually happen due to minor imperfections or such (hence the occasional need for a "recharge") but not at a rate where you actually notice it.

what is actually happening is similar to an iced drink by a pool on a scorching day. water vapor has plenty of energy, the coils do not (refer to the kinetic theory of matter) and thus the vapor is attracted somehow. when the vapor contacts the coils, it (vapor) dumps a veritable ton of its(vapor's) energy into the coils. this causes the vapor to condense. energy is still flowing, so eventually it freezes. right on the coils. this happens to insulate the coils somewhat and thus make them somewhere between mostly ineffective and utterly worthless.

my father, who is selling his mobile home (having actually gotten a house) used to go after the thing with an old, dull cold chisel. wouldn't ding up the a-frame piping, but it sure got the ice out of there! that's the best suggestion i can give.

alternatively, if you were able to work out a way to do this, you might think about running some of that piping right through the computer.
 
can you post some pictures of what this thing looks like?? it sounds like a really interesting project and i'd like to see how it looks/ how much room it takes up...

pretty awesome cooling at a discount price :D
 
the leakage i was referring to was air leakage...moisturized air getting in and making the evap ice...i know im not losing much freon...i already have the thermostat pulled off...compressor runs constantly...what i really need is a decent electronic thermostat that i can program to keep it at about -8 to -5C in there...i'll try to get the video up tonite along with some stills if i can...
 
quakefiend420-loaded_10_below0C.jpg


quakefiend420-h.jpg


quakefiend420-h_edit.jpg


getting it cold is no problem...keeping it there is...need suggestions on how to keep my evap from freezing...any ideas on how to remove humidity from the air in there? right now its really humid outside and its frezzing even quicker...argh
 
Aside from my PC pasion, i also have a koi pond. Really high-end koi ponds use ozone (o3) systems to super-oxygenate the water. At anyrate, ozone generators' efficancies are hughley lowered by humid air. they make products that remove the humidity in the air (or greatly reduce it) before it enters the ozone. im not sure how you could implement one of these devices in your system, maybe between the AC and case? anyway, hope that helps. Try searching up pond ozone systems on google, a little poking around should get you what you need.
 
You must log in or register to reply here.
Back
Top