Johannes Krauser II
n00b
- Joined
- May 12, 2020
- Messages
- 18
Hello,
--Before starting, I won't answer to any details right now, in order to avoid off-topics, but I will post everything once my project is complete.--
The for-now details, my questions and my preliminary testing:
I want to liquid-cool my system WITHOUT water. I am using oil, and I need some expert advice on selecting mainly the pump, and a heat exchanger (if possible). The oil is @ 20 cst and I need a pump to push that oil through a cpu and gpu waterblock, be small (max size about as the pump below in my test) and silent. Second, I need a compact all copper heat exchanger, brazed, 10 plates max, 2kw max or the nearest) More on this later.
So far, I got an old P4 @ 2.6 to be the testbench, a simple single radiator (with 140mm fan), a basic generic waterblock (alluminium, 40x40), the head/waterblock from a faulty antec kuhler h2o 650 (yeah, the frankenstein AIO, just took the part that goes to the cpu) and a generic TMC-06301 noisy AF pump.
The P4 @ 2.66ghz has 66.1w tdp. Overclocking it @ 3ghz (that's the best the generic mobo could do), I linear project about 75w tdp, but I know that there is NO linerar corellation, so I hope that the tdp is higher than this.
At about 32c ambient temp (outdoors, shadow), using first the generic block I got 51c under load (using furmark's cpu burner, I know its "light" vs prime95, but whatever). Swapping the basic block with the antec one, I got 38c-41c !!! (core temp didnt show values between 38 and 41). Without the fan on the radiator I got max 48c. All readings for a period of half an hour.
I believe that the testing was success! Keep in mind that the oil in the test flows through the radiator. In my system I will have it pass through a heat exchanger and cool it via water. Also, the antec waterblock has huge restriction: 90deg for inlet, another 90 to turn right and pass through the microfins that are parallel to the inlet/outlet, another 90 to turn again right, and another 90 to go to the oulet. The flow of the antec block was very low compared to the basic block, but the results are beyond my expectations.
About the pump I used in the test, it has 378LPH flow rate. I have my thoughts for a D5 pump, that has 1350LPH flowrate, but here I need your thoughts.
--Before starting, I won't answer to any details right now, in order to avoid off-topics, but I will post everything once my project is complete.--
The for-now details, my questions and my preliminary testing:
I want to liquid-cool my system WITHOUT water. I am using oil, and I need some expert advice on selecting mainly the pump, and a heat exchanger (if possible). The oil is @ 20 cst and I need a pump to push that oil through a cpu and gpu waterblock, be small (max size about as the pump below in my test) and silent. Second, I need a compact all copper heat exchanger, brazed, 10 plates max, 2kw max or the nearest) More on this later.
So far, I got an old P4 @ 2.6 to be the testbench, a simple single radiator (with 140mm fan), a basic generic waterblock (alluminium, 40x40), the head/waterblock from a faulty antec kuhler h2o 650 (yeah, the frankenstein AIO, just took the part that goes to the cpu) and a generic TMC-06301 noisy AF pump.
The P4 @ 2.66ghz has 66.1w tdp. Overclocking it @ 3ghz (that's the best the generic mobo could do), I linear project about 75w tdp, but I know that there is NO linerar corellation, so I hope that the tdp is higher than this.
At about 32c ambient temp (outdoors, shadow), using first the generic block I got 51c under load (using furmark's cpu burner, I know its "light" vs prime95, but whatever). Swapping the basic block with the antec one, I got 38c-41c !!! (core temp didnt show values between 38 and 41). Without the fan on the radiator I got max 48c. All readings for a period of half an hour.
I believe that the testing was success! Keep in mind that the oil in the test flows through the radiator. In my system I will have it pass through a heat exchanger and cool it via water. Also, the antec waterblock has huge restriction: 90deg for inlet, another 90 to turn right and pass through the microfins that are parallel to the inlet/outlet, another 90 to turn again right, and another 90 to go to the oulet. The flow of the antec block was very low compared to the basic block, but the results are beyond my expectations.
About the pump I used in the test, it has 378LPH flow rate. I have my thoughts for a D5 pump, that has 1350LPH flowrate, but here I need your thoughts.
Last edited: