An A/C system for my computer case... can it be done ?

**abcdefg1675** · 22-10-2002, 01:04 AM

Originally posted by Prof Sporlan

Your mixing energy units (Btu's) and power units (Btu/min). A Btu will cool one pound of water one degree Fahrenheit. A Btu/min will cool one pound of water one degree Fahrenheit in one minute...

Mmmm... an overachieving a/c unit 102 gal * 8.33 lb/gal * 10°C * 1.8°F/°C = 15,294 Btu. If cooled in one minute: 15,294 Btu/min = 76.5 tons...

yes. when i was doing the calculations, it seemed that something was hokey.
on my airconditioner, the BTU's are in BTU's/hour. so it can cool 8500lbs of water by 1 degree in an hour. that seems right, or is it? thats only about 1020 gallons of water.

water problably wouldnt be a good thing to measure your a/c's btus. the only way you could get it to cool 1020 gallons of water is to lift the unit up on a crane and have the evaporator sitting on the surface of a pool. i suppose maybe a space heater would work, if you have it blow the hot air on the evaporator, and have the cool air ducted to the back of the heater, and take a temp reading of the air in the back of the heater.

how many BTU's are equal to a Watt of heat?

**aenigma** · 22-10-2002, 08:32 AM

Uhmmmm what are you smoking?

P3's can make well over 40 watts.Actually I think somewhere around 80 watts overclocked.
Some P4's make over 150 watts overclocked.

Oh by the way an 8500btu a/c unit should move about 2500 watts of heat.

**Hubris** · 22-10-2002, 01:53 PM

That would be quite the overclock indeed. AMD chips make more heat than Intel....and when the chips shrunk to .13u die, the heat production dipped slightly.

150W output from the cpu alone....I'd think you'd have to be 50% overvolting the chip, with a serious overclock as well.

Regardless....your challenge isn't simply trying to remove the heat - a system is plenty capable of doing that....the challenge is to remove the heat, keeping the CPU cold enough for your overclock. You need something specific (like a TEC) to pull the heat directly away from the cpu.

**aenigma** · 22-10-2002, 09:19 PM

Uhm or refrigeration.You know something effecient that actually removes heat.

Yeah a P4 will put out 150watts.AMD's will do it pretty easily also, I havent checked on them lately.
I just remember talking to some guy that calculated his cpu's wattage, and it was 150 watts.

Alright according to a cpu wattage calculator, a p4 2.0 @ 3.0 stock voltage is 150 watts.This is for a .18 micron cpu, not .13
The .13 do run cooler that is for sure, they only have a voltage of 1.525.

My cpu, when I get around to fixing my evapblock, will be putting out 150 watts of heat if I hit 2ghz@2.3v.

**Hubris** · 22-10-2002, 09:28 PM

You can download a program called Radiate that will calculate the heat output, based on the chip, the speed, and the voltage. I was first put onto it by Tom Leufken...one of the pioneers of using TEC's on computers.

I'm pretty sure 150W is really high. The Palomino core Athlon 1400 is about the highest there is.... (the 2200 and later is lower, due to the die change) and at 1.75V at stock speeds it makes 85W.

Regardless...the challenge of the original poster (wanting to refrigerate a case and aircool the cpu within that case) is not to keep the case cool....but to prevent condensation. With aircooling, your CPU is always going to be X hotter than the case temps. The cooler your case, the cooler your cpu.

**aenigma** · 22-10-2002, 09:31 PM

I edited my post, I already used a cpu wattage calculator that actually works.By the way, tom leufkin is a retard hahaha

Of course 150 watts is high, it is an overclocked cpu.I am not talking stock speeds here, who would refrigerate their case to have stock speeds?

**abcdefg1675** · 23-10-2002, 01:06 AM

i bought a really tall networking tower 386 computer at goodwill. if i took out the rack thing that mounts 6 hard drives (i can hold 1 hard drive without it) i can cram in the evaporator, condensor and compressor of a refridgerator that i recently gutted.

is electro still an active member? its been a VERY long time since he/she has used this forum.

i did ruin the 386 computer with a 50W peltier. the whole cpu was a block of ice. its going to have a tiny atx slot 1 athlon mobo put in it.

the evaporator of the refridgerator is an "ice chest". its a tin box that holds meat and ice. the motherboard can fit in it along with the video and networking card. i also have plenty of insulation salvaged from the refridgerator. unfortunatly, its about 40 years old and the compressors very weak. the evaporator doesnt get very cool when its just sitting outside plugged in.

**Hubris** · 23-10-2002, 01:55 AM

You are going to have to completely seal the case against the outside air....at least any portion that is getting cold. If any moist air from your room comes into contact with the cold components, moisture, water, frost, and ice will become the end of your components.

**abcdefg1675** · 25-10-2002, 12:22 AM

i remeber looking at a link posted at www.hardforum.com about the heat output from processors released from their manufacturers. its been awhile, and i dont know what i was smoking.

theoretically, if the ice chest was fully insulated and sealed, the slow transfer of heat would probably dehumidify and create frost on the inside of the box, before the water in the air would have a chance to condense on the cpu and motherboard.

**abcdefg1675** · 25-10-2002, 12:27 AM

im not going to put the refridgerator inside my pc. it wont cut the mustard with an overclocked amd K7-500. it can go up to ~800 and stay stable. im thinking of using water pumps to circulate cooled water from 2-3 55W peltiers, and have the peltiers at the top of the case where they can be cooled with a regular heatsink.

**Hubris** · 25-10-2002, 05:26 AM

A chiller system....what you've described here....is very capable of yielding low cpu temps - however it typically isn't used with a single water system. The chillers cool water in one circuit, which keep a waterbock cold - and this waterblock cools the water for an additional separate circuit which actually cools the cpu.

Very complicated....takes up a lot of space, requires a lot of fabricating (or expensive purchases).

My suggestion if you intend to watercool.....is to build a chiller that acts like an office watercooler - chill a resevoir of water outside the pc, and circulate the water to cool the cpu.

Using peltiers as individual chillers makes things overly complicated, in my opinion.

I've seriously considered literally buying an office watercooler. About the only change you might need to make, is a fan blowing through the condensor...since they by default are statically cooled in the air.

**aenigma** · 25-10-2002, 06:10 AM

Ok, pelt chillers do not work for chilling water.I remember someone used about 4-6 226 watt pelts and he still didnt get very good temps.Pelts are very uneffecient.

You can make a waterchiller very easy just by gutting old freezers/refrigerators, a/c units, or office water chillers like hubris said.If your refrigerator you gutted is in good working order, it will handle the heat of your cpu piece of cake.

But it doesnt sound like it is working if its only getting a little cold.Need to go scrounging for some stuff like a/c's and refrigerators

Another thing you can use is a dehumidifer.

Check out http://www.phase-change.com

**abcdefg1675** · 25-10-2002, 11:58 PM

i noticed when i was trying to freeze a tablespoon of water with my peltier, a few degrees of temp difference on the hot side really changed the temp of the cold side.

i would use one of my 3 8500btu airconditoners, but they use 1300W. that doesnt look good on my parents electric bill.

i was thinking of builing a steel frame for one airconditioner, and making it cube shaped like a dehumidifyer, and have 360* rotating wheels on the bottom. the hot air would be blown out the back, and the cold air could either be sucked in the front and blown over the condensor like in a dehunidifyer, and i could have a lever on the top to open/close a door that lets the air either go to the condensor, or through a 10' long 5" diameter ducting.

im going to wait untill spring to do this. i dont need an airconditioner when its 20 degrees in the morning and 59 in the afternoon.

through, when its snowing outside, i could just use the -10 degree air from outside and have it go to the pc, and have the psu blow dry, fridged warmer air back ouside.

my house had the kitchen extended, and it has a daylight basement. my room is directly below the kitchen, and when the kitchen was extended, it left this 4.5' tall 8' wide 9' deep "cave" between my room and the rest of the house. our lawnmower, snowblower are stored there. so i could move my computer desk over toward my window and just set my computer out in the "cave, which the cement floor of it is level with my desk table. but during the summer i will try tweaking on my refridgerator guts with a kinked return line and oriface. (i had to bend the 4' long line a few times to get it out of the refridgerator, and it bent a few more times on its way out to my grandpas house)

**aenigma** · 26-10-2002, 03:29 AM

My direct die system (when I get it running finally) uses an 8000btu compressor and a 13,500btu compressor

Air conditioners work great, might as well use one and just see what your electrical bill goes to.1300 watts seems like alot for that system...Also if you gut it and recharge it to handle less heat, power consumption should drop.I dont know if it would be alot, or barely noticable.

Where in washington do you live anyways?

**abcdefg1675** · 27-10-2002, 03:43 AM

Im hooked up to avista utilites in stevens county.

the airconditioners compressor is 3/4 horsepower (i think). i have found 2 a/c compressors that problably weigh 90+lbs each. one is almost 3ft tall, and 8 inches in diameter, and one is about 14" in diameter and 20" tall. they have their pipes cut, and exposed to air, so i am not sure if their siezed or not. the tall skinny one would be great to charge with R22 for 7$/lb and make a 20,000+btu window unit.

**aenigma** · 27-10-2002, 07:38 AM

muahha yeah I have a big one like that and access to several others, but 220v is the problem.I dont have a cord and I am lazy haha
Ok that skinny one I havent een one like that, you sure it isnt a scroll compressor or something?If it is grab it!
I have got a fre compressor that had their pipes cut and they work, but that big one I have has crud in the pipes, wonder what would work to flush it all out....

I want to make a cascade system using methane with that compressor and a 1.5hp copelmatic 2 cylinder compressor.I think they are used in autocascade ultra low temp freezers....

**Gary** · 27-10-2002, 06:29 PM

Keep in mind that BTU's drop as evaporating temperature drops. An 8000btuh A/C system run at low temp isn't going to put out anywhere near 8000btuh. Systems are rated under their design conditions.

**aenigma** · 27-10-2002, 10:32 PM

Yeah I just give btu's as an example of what power rating the compressor is.In my case I only need to it move about 500BTU at a very low evap temp.Now if I had a thermometer, I could find out what evap temp would be, but for the life of me I cant find something that goes below -100f.Unless I want to pay out the nose.

**Gary** · 28-10-2002, 04:24 AM

I think you're dreaming. Worry about that if/when you achieve -100F.

**aenigma** · 28-10-2002, 04:34 AM

I think my "dream" is a reality.
My thermometer only goes to -58, so I dont know my evaporator temp.I never said anything about hitting -100f, I just want a thermometer that goes there so I have a nice range.

I know for a fact it will hit -80f(unloaded) because someone I know gets -80f with a 1/4hp compressor using R22, also no load on the evaporator.

By the way, do you have a p/t chart that gives the boiling point of R22 in a vacuum?

<edit>
How would I know if I got to -100f if I dont have a temp probe that goes that low?

**Gary** · 28-10-2002, 04:38 AM

i noticed when i was trying to freeze a tablespoon of water with my peltier, a few degrees of temp difference on the hot side really changed the temp of the cold side.

If I were building an experimental system, I think it would be interesting to refrigerate the hot side of a pelt. Sort of a hybrid cascade. The downside would be the extra heat generated by the pelt, necessitating a larger refrigeration system.

**aenigma** · 28-10-2002, 04:43 AM

I have heard that peltiers lose effeciency the lower they go.Now I dont know if this is true or not.I have had my P3 700 at -49c using an 80w pelt and 2c liquid cooling it.But the mounting was horrible, so it should have been lower than that.

**Gary** · 28-10-2002, 04:47 AM

Why does it matter how low it gets with no load? That's like saying the speedometer on your car will go 200mph if the axle is up on blocks. Load is where the rubber meets the road.

A vacuum P/T chart for R22? Yes, I have one right here in my hot little hands.

**Gary** · 28-10-2002, 04:53 AM

All types of refrigeration systems lose efficiency the lower they go.

**aenigma** · 28-10-2002, 04:54 AM

Originally posted by Gary
Why does it matter how low it gets with no load?

A vacuum P/T chart for R22? Yes, I have one right here in my hot little hands.

GIMME GIMME

Why does it matter what it gets with no load?I am assuming you mean the 2 stage system.
It doesnt matter, other than braggin rights.It is driving me crazy, I really want to know what it gets to.

I was just pointing out that his system running in a vacuum got -80f.Now my system will also run in a vacuum, but it will handle a heatload also.

I want a thermometer so I can tell how far below -58f it actually is, and I am changing capillary to get much lower.I have low resistance (short and big) capillary right now.
I am curious what does R22 boil at in 20 inches?With the low resistance capillary I was getting 18 inches, and it was overfeeding.

**aenigma** · 28-10-2002, 04:56 AM

Originally posted by Gary
All types of refrigeration systems lose efficiency the lower they go.

Well by losing efficiency, I meant the temps actually start to climb.(even with no load).Now I dont know if this is true or not, I dont do much with peltiers.

**Gary** · 28-10-2002, 05:00 AM

I can't help you there. I've never seen a pelt.

**Gary** · 28-10-2002, 05:08 AM

R22 at 18in is about -74F. That's evaporating temperature (aka saturated suction temperature or SST), which is not the same as evaporator temperature. It doesn't mean much by itself.

**aenigma** · 28-10-2002, 05:13 AM

Yeah it is just a rough example.Gives me an estimate.I am sure the evaporator was somewhere around that temp, considering it was definately saturated.
Hey do you use anything like msn or icq/aim?

**Gary** · 28-10-2002, 05:22 AM

I am fairly certain that it was not somewhere around that temp.

I have icq and aim, but unfortunately my scanner isn't working, so I can't scan the P/T chart for you. Got a fax machine?

P.S. To avoid embarassment, restrict bragging about no-load temps to the uninitiated.

**aenigma** · 28-10-2002, 05:27 AM

Originally posted by Gary
I am fairly certain that it was not somewhere around that temp.

I have icq and aim, but unfortunately my scanner isn't working, so I can't scan the P/T chart for you. Got a fax machine?

P.S. To avoid embarassment, restrict bragging about no-load temps to the uninitiated.

Oh the chart doesnt matter, I just wanted to know what it boils at around 20 inches...

BTW what do you mean about restrict bragging to the uninitiated?

**Gary** · 28-10-2002, 05:39 AM

R22 at 18in is about -80F SST.

People in the refrigeration trade will tend to chuckle if you mention temperatures achieved with no load. It doesn't mean a thing. How high can you jump if there is no gravity? The load is where the rubber meets the road.

**aenigma** · 28-10-2002, 06:42 AM

Ok I see what you mean.I wasnt talking aobut bragging to people like that

**abcdefg1675** · 05-12-2002, 12:53 AM

this seems to be a "dead thread". electro hasnt even replied to this thread. electro hasnt even loggen on in months.

i remeber seeing a website, which was about "direct die" evaporative cooling. their was a link their for a case made my a company with a refridgerator in the bottom part of it. their were insulated lines going from the bottom part to where you would mount an evaporator block onto the cpu. i think the company that made the case was promentaria, or somethig that sounds similar.

**Hubris** · 05-12-2002, 02:53 AM

Yes...Prometria make an addon-cooler that you put under your case, and extend the evaporator up into where your cpu sits. This is very much the same design as I've been taking in trying to make my cooler...unfortunately I've had a bit of a problem finding someone to finish everything (solder the last tubing, charging with refrigerant) and more recently it's been called into question whether the setup I'm building is powerful enough to handle a cpu.

www.vapochill.com is one manufacturer of complete cases, and
www.kryotech.com was the original commercial computer cooler.

**aenigma** · 06-12-2002, 06:55 AM

Yeah, prometeia, its a real disgrace.Overpriced and underperforming.But people still pay for it.

**abcdefg1675** · 10-12-2002, 02:24 AM

as long as their isnt much competition, people will pay money and get ripped off for a crappy evaporative cooler.

i still think it would be a better more engergy efficent idea if someone were to buy two 270W peltiers, and have the hot side watercooled. you could easily have the power variable to the both of them.
im waiting for some peltiers someone is giving me, and i will just "gang" about 3 of them, and have everything controlled with the parrellel port and some homemade software.

**herefishy** · 10-12-2002, 03:08 AM

........................sheesh...............!!!!!

**aenigma** · 11-12-2002, 08:14 AM

Originally posted by abcdefg1675
as long as their isnt much competition, people will pay money and get ripped off for a crappy evaporative cooler.

i still think it would be a better more engergy efficent idea if someone were to buy two 270W peltiers, and have the hot side watercooled. you could easily have the power variable to the both of them.
im waiting for some peltiers someone is giving me, and i will just "gang" about 3 of them, and have everything controlled with the parrellel port and some homemade software.

Peltiers are extremely unneffecient.People usually run 1 226w pelt, and always get lousy temps.Back in the day with low wattage p3's peltiers worked ok though.

The current pelts (226 watt) need high amperage PSU's to run them.So say your running 2 then your drawing over 500 watts, assuming they are effecient.About 20 amps at 12 volts times 2 pelts.Alot more than a little 1/4hp compressor.

A refrigeration system is much much better, just don't buy a premade system.

Although someone is coming out with these things called "cool-chips" and they say it is more effecient than refrigeration...We will see....

**abcdefg1675** · 15-12-2002, 07:03 PM

Peltiers are very small compared to an Ac system, and they dont make noise.

But with compressors, they take up nearly half of your case. with peltiers, you can vary the power to them with a circuit similar to a dc motor speed control circuit.

**aenigma** · 16-12-2002, 12:00 AM

Yes but you see, compressors are almost silent.The fans are pretty quiet also.With a pelt, they are dumping a very large ammount of heat into the water, and you use loud fans to get rid of that heat.
Direct die refrigeration unit = quiet + good temps.
Pelt cooling = noisy + lousy temps.
I think we see the winner here...

Ah by the way, I was talking about average 1/4hp compressors from refrigerators/freezers, not A/C compressors.They are kind of noisy, but not bad at all.

**abcdefg1675** · 17-12-2002, 12:11 AM

i gutted an old frigidair refridgerator, with the icebox thing that was the actual evaporator. it has a 1/8hp compressor and the only noticible noise is when it first powers up, and the hissing of the *****. I kinked up the line pretty bad (the return line from the evaporator) it looks like someone took pliers and squashed the line. I cant afford ***** at 60$/lb, so i doubt i will use this system in my computer anytime soon. R22 is 7$/lb, and i have to buy it in 30lb containers (the servicing place doesnt do business with domestic home refridgerators, only the huge walk in ones that need 30lbs or more of refridgerant, so if they open the can they charge me for the whole thing).

If i did plan to convert the system to R22, will i need to change only the restriction capillary tubing (the really tiny piping)?

ive seen a chart for different refrigerants, and ive seen 2 different refridgerants of R11, one of them boils at -100*C. i would like to have a tiny crappy fridgerator that would run on that stuff. (it would be my luck thats its 500$/lb)

**aenigma** · 17-12-2002, 05:31 AM

Originally posted by abcdefg1675
i gutted an old frigidair refridgerator, with the icebox thing that was the actual evaporator. it has a 1/8hp compressor and the only noticible noise is when it first powers up, and the hissing of the *****. I kinked up the line pretty bad (the return line from the evaporator) it looks like someone took pliers and squashed the line. I cant afford ***** at 60$/lb, so i doubt i will use this system in my computer anytime soon. R22 is 7$/lb, and i have to buy it in 30lb containers (the servicing place doesnt do business with domestic home refridgerators, only the huge walk in ones that need 30lbs or more of refridgerant, so if they open the can they charge me for the whole thing).

If i did plan to convert the system to R22, will i need to change only the restriction capillary tubing (the really tiny piping)?

ive seen a chart for different refrigerants, and ive seen 2 different refridgerants of R11, one of them boils at -100*C. i would like to have a tiny crappy fridgerator that would run on that stuff. (it would be my luck thats its 500$/lb)

If your refrigerator can keep low side at around 0psi or so, and a highside pressure of probably 1000psi then you could use it, sure

That is a cascade refrigerant.And who said you need R22?Use R290(propane), I use it in all of my system, don't be scared away by it being flammable as a sealed system won't just magically set itself on fire.Alot of people think propane is just death in a can.

I would go with at least 6 feet of .026" capillary, maybe more....

**abcdefg1675** · 17-01-2003, 01:09 AM

I would prefer a refridgerant that is most efficent, which would be R12, but that isnt an option. If propane is better than R22 at conducting heat and boiling easier, then i will use it.

**aenigma** · 17-01-2003, 02:17 AM

Propane is much better than R12, lower boiling point, better capacity etc.
But it is not better than R22.

The main advantage of propane is the cost and availibility.

**AndrOvr** · 18-01-2003, 12:58 AM

well.. I finished my phase change cooler

using r409a and 1/4hp.

I tested just for some mim, in bios.. I got readings of -18c w stock (1.8a)
Its not all that great.. I know ... not like prometeia... but it costed 1/4 of it.. and Im a beginner in this :P

**MRW** · 18-02-2003, 10:04 PM

there is an tower base unit that would fit under a desk,and
is not much bigger than a normal base unit.
It has a cold plate and runs on r134a.
The operating temp is -17c,or if the gas charge is altered,it can go to - 25,but you run the risk of blowing up the compressor.
I think the name of the kit is asiacs,or something like that.
I will try and locate the correct spec of the kit,the guy that had this piece of kit was running an 2.6 chip,heavily clocked.

**subzero*psia** · 19-02-2003, 03:32 PM

I'm not really much of a computer guru so bear with me... why is there such an interest in increasing the speed? If your Internet Service Provider only has the capability to deliver at a set rate of speed... you still have to wait for graphics to load if the internet is bogged down with users. If a person is running graphic intense program off their own PC or performing mathamatical computations for NASA... I guess then I could see why. Is this for using on the internet or home based programs?

I could be wrong... wouldn't be the first time....

**Robearbam** · 19-02-2003, 03:53 PM

Actually, I believe that the cooling concerns are more prominent today because of the almost microwave freq's of the processors. Surely games are the most proc demanding therefor heat effects are a problem. Just by keeping an electric motor or transformer cool, you can up it's capacity rating. I believe everything evolves around temperature, even the globe.

Shane Rupe · 19-02-2003, 06:29 PM

aenigma
Did you see the news story a few years back about the guy in Spokane that blew his car up useing propane in his ac?

Thread: An A/C system for my computer case... can it be done ?

Thread Tools

Display

a/c for computers

Similar Threads

Daikin fault codes

System Burn Out Troubleshooting

Burn out System, Clean up procedure. & Troubleshooting.

What is System balance?

Refrigeration system proposal; comments please.

Posting Permissions