The Radiators
This content was originally featured on Amdmb.com and has been converted to PC Perspective’s website. Some color changes and flaws may appear.
The old:
The original Aquacoil radiator I used in my last review I bought over a year ago. It has 2 x 90 degree bends and 19 x180 degree bends with 3/8” tubing. The cube takes up approximately 7”x5”x5” of space and has roughly 40 aluminium fins that cool the water traveling though the radiator .This design is very restrictive to flow rates as every 90 degree bend cuts your flow rates down, resulting in the water sitting in your block longer then it should .This can cause the CPU temperatures to rise. As well, because the radiator is so deep, it requires a fair amount of airflow [cfm] across the fins to cool the incoming water. Set back from the radiator’s fins about an inch is a nice mounting bracket built on to the one side of the cube to attach a 120mm fan. This does the same job as a shroud.
In order to show how drastically flow rates can affect temperatures, I have substituted the cube radiator I used in the last review with the Heater Core radiator. While there are many other radiator‘s I could have used had I had access to them, this was the only radiator I had in my possession besides the cube at this time. This radiator, unlike the cube radiator I was using, does not have a lot of bends and utilizes a much less restrictive flow design. This results in better flow rates throughout the water cooled setup which in turns helps to lower the CPU temps.
The new:
The heater core came equipped with a shroud and an Enermax adjustable rpm fan [not shown] and was supplied by CPU-FX. It has ½” fittings to allow greater flow rates through the radiator instead of the 3/8” fittings that the cube and waterblock use. The 3/8” tubing with some hot water and a little bit of elbow grease, can be fitted on the ½” fittings quite nicely. The core is approximately 6.5”x6”x2” and the equipped shroud which came with it, allows for the 120mm Enermax adjustable rpm fan to draw air across the entire radiator‘s surface instead of just the 120 mm area where the fan would sit. The adjustable rpm control on the Enermax fan allows you to turn the fan speed down to a suitable noise level [less airflow] or up for higher airflow [louder]. Since this radiator is going to be installed in another test system for a CPU-FX review later I did not install the radiator inside this case and will be doing this review with an open case. [Note: when I closed the case on the cube radiator my temps went up by about 1/2 c] Other then an increased vcore from 1.775v to 1.85v, a half inch return line and a different radiator the rest of the test rig remains identical to the last review.
Testing
The test bed:
The Black Edge water block and the Lexan clamp from BeCooling.
Via Aqua 1300 pump [370 gph]
256 Megs of Samsung pc2700
Msi kt3 ultra
Xp1700@1667mhz 1.850 Vcore and 1.775 Vcore
Heater Core
The testing procedures I used are the same as the last review and were done days after the initial installation of the water-cooling equipment. This allows for the system to purge any air that may be trapped in the lines, radiator or block, it also allows for the Arctic Silver II to cure. All CPU temps are taken at the die of the CPU by a DigitalDoc5, which was graciously supplied by Firestorm in Calgary. To ensure that the CPU was under a full load I used SETI which can warm up a CPU quite well, and I listened to mp3s. I took the temperatures every ½ hr for 4 hrs [8 readings in all]on the water, the cpu and the ambient temperatures, and at the end I added each of categories temperatures up and divided them by 8 to get the average temperature overall.
Note: all temps are in Celsius
| Radiator | VCore | Ambientº | Waterº | CPU Loadº | Deltaº=Loadº-Waterº |
| Heater Core | 1.775 | 22.3 | 24.1 | 27.7 | 3.6 |
| Heater Core | 1.775 | 23.1 | 24.9 | 28.5 | 3.6 |
| Aquacoil | 1.775 | 23.1 | 28.1 | 31.9 | 3.8 |
| Heater Core | 1.85 | 23.1 | 25.3 | 29.3 | 4.0 |
| Heater Core | 1.85 | 22.1 | 24.3 | 28.3 | 4.0 |
Now we can clearly see the difference between the two test systems with the different radiators. Because the heater core radiator is removing the heat better due to faster flow rates we can see that the Black Edge is also responding to the higher flow rates with a .3ºC drop in temperatures between the water and CPU. [3.6º for the heater core and 3.8º for the Aqua cube when using the same voltages] We can also see that the water temps between the two radiators has a difference of 3.2ºC in the heater cores favour. In fact, it even does a better job then the Aquacoil when we increase the Vcore from 1.775v to a 1.850v.
Although the heater core has performed better in this case, the Aqua cube still performs admirably and won’t be removed from this test setup quite yet .It would in fact probably perform better then this test shows with a 500 gph or higher pump, since it requires a much higher flow rate then the heater core. The .3ºC drop in the Black Edge‘s performance was a bonus in my opinion as I didn’t expect a drop in temperature on the water block, although it makes perfect sense as the higher flow rates can affect the entire system not just one part of it.
