Thinking about my first Watercooling project

[govtcheez75]

can someone lead me into a good direction on where to get the stuff, and what might be a easy/good "starter" type of kit or parts? I would like a full kit that cools my 939 processor, X800XT PE and maybe the Northbridge too.

 

[NotquiteanooB]

Corsair has an easy to use kit; good instructions (on a CD unfortunately) CWC100-1000 for AMD 64 and early P4/P3 models and CWC100-1001 for AMD XP and P4 775 sockets.
Kits are about $230 Cdn. Knock off 20% for US $; that should put the kit under $200 US$. Google it and read some of the reviews. I've just finished installing 2 of the kits in my P4 PC's; and have ordered 2 more kits for the AMD PC's. They also have optional GPU and Northbridge cooling blocks available (CWC100-1003 & 1004) Temps on my OC'd P4's have dropped from 50*C to 39*C; mobo temp 33*C and room temp is 24*C. After the other PC's are done, logic tells me that the room temp should drop again and in turn the case temps and maybe the CPU temps.
There other kits available .. Swiftech and others; but I'm impressed with the Corsair units.
Corsair has a forum where you can get the good and bad news!

http://www.houseofhelp.com/forums/forumdisplay.php?f=155

 

[racinjimy]

check out the "guide to watercooling" sticky at XS

XS liquid cooling

all information is up to date, it is how I got my start in WC

 

[Turtle 1]

http://www.frozencpu.com/

 

[Greenman]

Originally posted by: NotquiteanooB
Corsair has an easy to use kit; good instructions (on a CD unfortunately) CWC100-1000 for AMD 64 and early P4/P3 models and CWC100-1001 for AMD XP and P4 775 sockets.
Kits are about $230 Cdn. Knock off 20% for US $; that should put the kit under $200 US$. Google it and read some of the reviews. I've just finished installing 2 of the kits in my P4 PC's; and have ordered 2 more kits for the AMD PC's. They also have optional GPU and Northbridge cooling blocks available (CWC100-1003 & 1004) Temps on my OC'd P4's have dropped from 50*C to 39*C; mobo temp 33*C and room temp is 24*C. After the other PC's are done, logic tells me that the room temp should drop again and in turn the case temps and maybe the CPU temps.
There other kits available .. Swiftech and others; but I'm impressed with the Corsair units.
Corsair has a forum where you can get the good and bad news!

http://www.houseofhelp.com/forums/forumdisplay.php?f=155

How could room temp drop? Unless you put your rad outside or in another room, you're putting more heat into the room. (Everything you had plus a pump).

 

[NotquiteanooB]

Before 2 of my 4 PC's were watercooled, the ambient room temp was 27-28*C. After 2 were watercooled the ambient room dropped to 24*C. If one air cooled PC produces 2*C of additional heat in the room; by watercooling all 4 should drop the ambient temp to 20*C (I hope) For some reason the air coming thru the rad is far cooler than the air coming from the exhaust fans on the rear of the air cooled cases. Obviously mobo or CPU temps of watercooled units will never get cooler than the ambient of the room, but in theory; the lower the room temp. the cooler the air will be that is blowing thru the rad. The pump, tubing etc. are the same temp as the fluid running thru them. 39*C, same temp that the CPU is reporting. The pump produces no heat. Maybe I should say; little or no heat.

4 PC's in a room that is about 8'x9'; produce a lot of heat and noise. Not only has the room temp dropped... so has the noise level.

Edit: I think the reason the air is cooler coming thru the rad than from the air cooled CPU cases is: The air cooled CPU's are showing 50*C therefore the air coming out of the case will be 50*C or slightly less than 50*C; whereas the watercooled CPU units are showing 39*C, therefore the air coming thru the rads will be 39*C or slightly less. With 11*C less hot air coming into the room; the ambient temp has to drop.

 

[NotquiteanooB]

Bump Edit:

 

[Jimmah]

Originally posted by: NotquiteanooB
Before 2 of my 4 PC's were watercooled, the ambient room temp was 27-28*C. After 2 were watercooled the ambient room dropped to 24*C. If one air cooled PC produces 2*C of additional heat in the room; by watercooling all 4 should drop the ambient temp to 20*C (I hope) For some reason the air coming thru the rad is far cooler than the air coming from the exhaust fans on the rear of the air cooled cases. Obviously mobo or CPU temps of watercooled units will never get cooler than the ambient of the room, but in theory; the lower the room temp. the cooler the air will be that is blowing thru the rad. The pump, tubing etc. are the same temp as the fluid running thru them. 39*C, same temp that the CPU is reporting. The pump produces no heat. Maybe I should say; little or no heat.

4 PC's in a room that is about 8'x9'; produce a lot of heat and noise. Not only has the room temp dropped... so has the noise level.

Edit: I think the reason the air is cooler coming thru the rad than from the air cooled CPU cases is: The air cooled CPU's are showing 50*C therefore the air coming out of the case will be 50*C or slightly less than 50*C; whereas the watercooled CPU units are showing 39*C, therefore the air coming thru the rads will be 39*C or slightly less. With 11*C less hot air coming into the room; the ambient temp has to drop.

This doesn't make any sense. In order to have the P4's cooler it would mean the water needs to take away MORE heat, even if the water is a cool 39*, its still dumping the same if not more heat into the room. The heat just can't 'go away', it needs to go into the surrounding atmosphere, aka your room. Air and water cooling do the exact same thing, just water has an extra step before being air cooled.

And now to get back to the OP's question: Build your own system if you can, its extremely easy if you go the general route everyone else does (ie: buy the same parts as the most popular budget setups). As for a kit, I'm a fan of the reserator, its quiet and works good enough for my tastes.

Still, IMO, water cooling isn't all that worth it compared to air, I've used both and the cost savings alone have made me stay aircooloed, I just can't justify 200$ per machine when 30$ in fans and hsf's do the same job only 10c higher.

 

[NotquiteanooB]

The fluid is a better cooling medium than a finned heatsink. The fluid never lets the CPU GET to 50*C. If the CPU gets to 50*C again, the fluid will be at 50*C also. At this time it's the fluid preventing the CPU from hitting 50*C. The proof is in the thermometer. It is cooler in the room; the Cpu's are running cooler. The final proof will be after I add 2 more watercooling units to the remaining PC's and see if the room temp drops again.
Another analogy is the use of anti-freeze in a car rad. A 50/50 mixture of water/anti-freeze is good to -44*F; but in addition it raises the boiling point of the water in the rad from 212*F to about 245*F. Thereby actually running to a hotter point before boiling and allowing the rad to disipate more heat. It's not just to stop freezing; it's to adjust the boiling point higher!
I don't know what NASCAR Nextel cars are using, but they boil at about 260*F.

IMO watercooling is worth every penny. Less heat when OC'g = less heat damage to components; CPU and others; AND how many less db's from hi-spd Heatsink fans !! Maybe as much as 40db.

 

[Greenman]

Originally posted by: NotquiteanooB
The fluid is a better cooling medium than a finned heatsink. The fluid never lets the CPU GET to 50*C. If the CPU gets to 50*C again, the fluid will be at 50*C also. At this time it's the fluid preventing the CPU from hitting 50*C. The proof is in the thermometer. It is cooler in the room; the Cpu's are running cooler. The final proof will be after I add 2 more watercooling units to the remaining PC's and see if the room temp drops again.
Another analogy is the use of anti-freeze in a car rad. A 50/50 mixture of water/anti-freeze is good to -44*F; but in addition it raises the boiling point of the water in the rad from 212*F to about 245*F. Thereby actually running to a hotter point before boiling and allowing the rad to disipate more heat. It's not just to stop freezing; it's to adjust the boiling point higher!
I don't know what NASCAR Nextel cars are using, but they boil at about 260*F.

IMO watercooling is worth every penny. Less heat when OC'g = less heat damage to components; CPU and others; AND how many less db's from hi-spd Heatsink fans !! Maybe as much as 40db.

That is incorrect. The amount of energy entering the room increases when you water cool, because of the added pump. The other components are putting out the EXACT same amount of heat, weather it's air cooled or water cooled does not matter.

 

[NotquiteanooB]

Originally posted by: Greenman

Originally posted by: NotquiteanooB
The fluid is a better cooling medium than a finned heatsink. The fluid never lets the CPU GET to 50*C. If the CPU gets to 50*C again, the fluid will be at 50*C also. At this time it's the fluid preventing the CPU from hitting 50*C. The proof is in the thermometer. It is cooler in the room; the Cpu's are running cooler. The final proof will be after I add 2 more watercooling units to the remaining PC's and see if the room temp drops again.
Another analogy is the use of anti-freeze in a car rad. A 50/50 mixture of water/anti-freeze is good to -44*F; but in addition it raises the boiling point of the water in the rad from 212*F to about 245*F. Thereby actually running to a hotter point before boiling and allowing the rad to disipate more heat. It's not just to stop freezing; it's to adjust the boiling point higher!
I don't know what NASCAR Nextel cars are using, but they boil at about 260*F.

IMO watercooling is worth every penny. Less heat when OC'g = less heat damage to components; CPU and others; AND how many less db's from hi-spd Heatsink fans !! Maybe as much as 40db.

That is incorrect. The amount of energy entering the room increases when you water cool, because of the added pump. The other components are putting out the EXACT same amount of heat, weather it's air cooled or water cooled does not matter.

Energy ... maybe ...But; btu's are less and db's are less. If I remove a HSF of 2.5 watts and replace it with a pump of 2.5 watts ... how can that be more energy.

I will hang a thermometer in front of the rad to check the air temp coming thru; then in front of the exhaust of an air cooled case. I know what I'll see. Temps in the watercooled case exhaust will be lower than the air cooled case exhaust.

 

[Jimmah]

Originally posted by: NotquiteanooB
Energy ... maybe ...But; btu's are less and db's are less. If I remove a HSF of 2.5 watts and replace it with a pump of 2.5 watts ... how can that be more energy.

A pump is NEVER 2.5w and efficient enough to cool a computer. 10w maybe, 15w more likely.

First thing, heat is energy. Second, cpu's produce a ton of concentrated heat. Third, both cooling systems are there to remove this unwanted cpu heat, both having different efficiencies of doing such. A water cooled system will do a better job of removing this heat, thus making temps lower on the cpu die, air cooling does a decent job but isn't as efficient as water. Also, water will ALWAYS add more heat compared to a HSF assembly (eg: a 20w pump will produce approx. 10w of heat + whatever frictional losses). The rad must then remove this unwanted heat from the water loop, and if the cooling loop is cooling the cpu to equal or less than air, it therefore must remove more heat than said HSF assembly.

The quantity of joules simply cannot disappear, especially when more are being removed faster, so your idea of less heat on die heat = less overall temps isn't possible, basic physics determines this.


Also, about the water vs. air thing. It has been beated to death, over and over again.

Cost savings - air
Lower temps (even if tiny) and sometimes less noise - water

To the OP: Dude, wander over to www.frozencpu.com and take a gander, do some google and see whats the most common setup, and do some research on what you want out of it all, and if you honestly can afford it. Water is a PITA at times, but if done right can be well worth it.

My .02$

 

[NotquiteanooB]

Thermometer is a very accurate probe type for mesasuring film and negative developing fluids. 15 minutes in front of rad exhaust = 38 *C CPU by ASUS probe =39*C; mobo @ 33*C; room at 24*C

15 minutes in front of rear fan exhaust of an air cooled 2800 Barton. = 46*C; CPU by ASUS probe 51*C; mobo 34*C and room 24*C.

Only 1*C difference on the air coming from the rad as opposed to 5*C on the air cooled Barton. IE: the Hotter CPU is putting hotter air into the room. Reduce the heat on the Barton and the room should get cooler.

 

[Greenman]

Originally posted by: NotquiteanooB

Originally posted by: Greenman

Originally posted by: NotquiteanooB
The fluid is a better cooling medium than a finned heatsink. The fluid never lets the CPU GET to 50*C. If the CPU gets to 50*C again, the fluid will be at 50*C also. At this time it's the fluid preventing the CPU from hitting 50*C. The proof is in the thermometer. It is cooler in the room; the Cpu's are running cooler. The final proof will be after I add 2 more watercooling units to the remaining PC's and see if the room temp drops again.
Another analogy is the use of anti-freeze in a car rad. A 50/50 mixture of water/anti-freeze is good to -44*F; but in addition it raises the boiling point of the water in the rad from 212*F to about 245*F. Thereby actually running to a hotter point before boiling and allowing the rad to disipate more heat. It's not just to stop freezing; it's to adjust the boiling point higher!
I don't know what NASCAR Nextel cars are using, but they boil at about 260*F.

IMO watercooling is worth every penny. Less heat when OC'g = less heat damage to components; CPU and others; AND how many less db's from hi-spd Heatsink fans !! Maybe as much as 40db.

That is incorrect. The amount of energy entering the room increases when you water cool, because of the added pump. The other components are putting out the EXACT same amount of heat, weather it's air cooled or water cooled does not matter.

Energy ... maybe ...But; btu's are less and db's are less. If I remove a HSF of 2.5 watts and replace it with a pump of 2.5 watts ... how can that be more energy.

I will hang a thermometer in front of the rad to check the air temp coming thru; then in front of the exhaust of an air cooled case. I know what I'll see. Temps in the watercooled case exhaust will be lower than the air cooled case exhaust.

Just tell me where the heat is going please.

 

[Jimmah]

Originally posted by: NotquiteanooB
Thermometer is a very accurate probe type for mesasuring film and negative developing fluids. 15 minutes in front of rad exhaust = 38 *C CPU by ASUS probe =39*C; mobo @ 33*C; room at 24*C

15 minutes in front of rear fan exhaust of an air cooled 2800 Barton. = 46*C; CPU by ASUS probe 51*C; mobo 34*C and room 24*C.

Only 1*C difference on the air coming from the rad as opposed to 5*C on the air cooled Barton. IE: the Hotter CPU is putting hotter air into the room. Reduce the heat on the Barton and the room should get cooler.

You just don't get it do you. I will re-try to explain.

You cannot simply 'reduce' the heat on the barton, the heat is a constant in this equation, its the heat exchanging that is the variable. Your hsf cannot remove the heat fast enough, so it stays at 50c, but the water cooling does it faster keeping the heat generated at 39c (this is efficiency at work). Now take the total area of your rad vs. the total area of your heatsinks fins, probaby a 5/1 ratio in rad/fins, overall creating less of an heat change due to the larger surface area, which is also increase heat dissipation thus causing lower net cpu temps.

Now where does all this heat go? Greater cooling capacity requires greater dissipation, physics determins such an event.

Here try this sometime: Fill 2 glasses with 25c water. Now put your left hand around one glass without anythign in between, and on your right one put a cloth in between your hand and the glass. Leave it there for 10 mins. You will notice your left hand feels cooler and your right one feels warmer (your hand is the cpu, the glass of water the ambient room temp, and the cloth is there to creat an innefficiency). Now measure the water temp. You will notice the water from the glass from your left hand is warmer, possibly even significantly, than the one from your right. This is the exact same scenario as what happens with water vs. air cooling, increase the efficiency of the cooling system and you decrease the original outputs temperature, but increase the heat exchangers temperatures (ie: surrounding air).

There, thats as plain as I can make it.