I am trying to find what reasonable values would be for the actual flow rate through a water block and what a typical temperature rise (Tin - Tout) would be. Has anyone measured these values in their system?
I water cooled a couple of years ago and was very pleased with the ease at which the water cooling system kept the CPU temps reasonable. Although the water cooling worked great, it was too much machinery to stuff in an already crowded case so I reverted to air cooling. I was also displeased with the amount of noise the supposedly quiet 120mm fan made.
I now want to go back to water cooling because I am having much difficulty providing adequate cooling for my new 2.4c P4 while keeping the noise level to some acceptable level. The integrated factory solutions such as the Koolance Exos or their water cooled case seem like an excellent way to go. I would expect the Koolance system to do a more than adequate job since my experience with my custom designed system two years ago indicated that it does not take much water cooling to be very effective.
Before I blow a lot of money on this, I want to prove to myself that the Koolance system will indeed limit the CPU to reasonable temperatures when it is under 100% load as it is when I am rendeing video DVDs. Koolance told me that I can expect a flow rate of approximately 1000 cc/min (.26 GPM) and a temperature rise across the CPU water block of about .5 to 1.0 C (~1.0 to 1.5 F) degrees. If one cubic foot of water is 7.48 gallons, then .26 GPM represents two cubic feet per hour. At 80 degrees one Lbm water is .0161 cubic feet. This means the flow rate is about 130 Lbm per hour. If the specific heat of water is 0ne btu/Lbm-degree, and there is a one degree rise across the water block, then the Koolance system will only remove 130 BTUs of heat per hour. If one BTU of energy is equivalent to .293 watts for one hour, then the Koolance system is only removing energy at a rate of 38 watts. This seems to be very inadequate for a 2.4c P4 at 3.25 GHz. Therefore, I have to assume the flow rate and/or temperature rise I used based on the information from Koolance is incorrect. I suspect the actual flow rate is much greater but, when I water cooled, there was a barely perceptable temp rise across the water block so one degree F might indeed be reasonable.
BTW, my goal is not to achieve extremely low CPU temps. I only want to keep the CPU sufficiently cool to prevent problems or abnormal operation.
I water cooled a couple of years ago and was very pleased with the ease at which the water cooling system kept the CPU temps reasonable. Although the water cooling worked great, it was too much machinery to stuff in an already crowded case so I reverted to air cooling. I was also displeased with the amount of noise the supposedly quiet 120mm fan made.
I now want to go back to water cooling because I am having much difficulty providing adequate cooling for my new 2.4c P4 while keeping the noise level to some acceptable level. The integrated factory solutions such as the Koolance Exos or their water cooled case seem like an excellent way to go. I would expect the Koolance system to do a more than adequate job since my experience with my custom designed system two years ago indicated that it does not take much water cooling to be very effective.
Before I blow a lot of money on this, I want to prove to myself that the Koolance system will indeed limit the CPU to reasonable temperatures when it is under 100% load as it is when I am rendeing video DVDs. Koolance told me that I can expect a flow rate of approximately 1000 cc/min (.26 GPM) and a temperature rise across the CPU water block of about .5 to 1.0 C (~1.0 to 1.5 F) degrees. If one cubic foot of water is 7.48 gallons, then .26 GPM represents two cubic feet per hour. At 80 degrees one Lbm water is .0161 cubic feet. This means the flow rate is about 130 Lbm per hour. If the specific heat of water is 0ne btu/Lbm-degree, and there is a one degree rise across the water block, then the Koolance system will only remove 130 BTUs of heat per hour. If one BTU of energy is equivalent to .293 watts for one hour, then the Koolance system is only removing energy at a rate of 38 watts. This seems to be very inadequate for a 2.4c P4 at 3.25 GHz. Therefore, I have to assume the flow rate and/or temperature rise I used based on the information from Koolance is incorrect. I suspect the actual flow rate is much greater but, when I water cooled, there was a barely perceptable temp rise across the water block so one degree F might indeed be reasonable.
BTW, my goal is not to achieve extremely low CPU temps. I only want to keep the CPU sufficiently cool to prevent problems or abnormal operation.
