Test Platform
The test platform's work was done by the same AMD set which was used in the previous cooler roundup. The hardware was placed on a table through the whole testing as the temperature was ~22C. For comparison I chose the winner of our couple of week old cooler roundup, Thermalright's very efficient SLK-947U. There was a Papst 8412NGML cooling the cooler all along. Thermalrights results are from the previous roundup but are still fit for comparing with Asetek because the test platform remained the same and I took care that the temperature remained the same also. The platform looked like this:
- AMD Duron 1400MHz 1,5V @ AMD XP 2000MHz (10x200) 1,7V
- Epox 8RDA3+
- 2 x M-Tec PC2700 256Mt Dual Channel 5-3-2-2
- Maxtor D740X-6L 40Gt
- LG 52x
- HEC 300W-LR-PT
- Asetek WaterChill Complete Set; Thermalright SLK-947U + Papst 8412NGML
- Inno3D Tornado GeForce 4 Ti4200-8x 64Mt
The NB- and GPU-blocks were compared to the hardware’s original coolers. Epox's northbridge had a quite effective and big aluminium cooler with a 40mm fan. There is a rather big aluminium cooler with a small fan also on the graphics card.
Testing Methods
While testing the CPU- & NB-blocks the platform was teased with a CPUburnK7 program (High priority) until the temperature didn't rise any more. After the teasing I let the computer idle since the temperature didn't come down anymore. This was when the idle temperature was marked. The graphics card's exhausting temperatures were marked after running 3D Mark 2001 SE's Draghotic test many times in a row with High settings.
The processor temperatures were measured in two ways, when the radiator's fan blew to the radiator, and when the fan sucked out air. While testing the NB- & GPU-blocks the fan sucked air out as Asetek recommends. The graphics cards & northbridges temperatures were measured with an external meter. The sensor was taped on the backside right in the middle of the core. So this makes the results are comparable and trend-setting, but not the real temperatures of the objects. The processor's temperature was measured with MBM5 from the socket's sensor. Epox's sensor is placed right on the processor's bottom surface.
Results: CPU-block
As the fan sucked air out of the radiator the WaterChill beat SLK-947U in both 7V and 12V. In exhaustion and 7V the WaterChill wins Thermalright with 3 degrees and while idling with 3 degrees. With 12V the same differences are 7 and 6 degrees. The results were as expected, nothing special.
The radiator's fan loses capacity clearly when blowing to the radiator. The airflow doesn't seem to matter since the results with 7V and 12V are the same in practise. After all, the temperatures raised several degrees after switching the blowing direction. The SLK-947U isn't able to beat the WaterChill anyway, even though the results are pretty even.
Results: NB-block
The northbridge's temperature results were as expected. The WaterChill beats the original cooler with several degrees in both idle and exhaust. The differences may be even bigger because of the measuring technique.
Results: GPU-block
The GPU-block had a harder time with Inno3D's original cooler. The difference was only 1,5 to 2,5 degrees for Asetek. The tiny difference made me surprise so that I even ran the tests several times, checked the sensor and the blocks installation. No measuring mistake did occur any way. Let's still remember that the results may not be 100% reliable!
Conclusion
The WaterChill is a rather worthy kit, but on the other hand it contains anything you can imagine. Considering that quality always costs, the WaterChill's quality costs an acceptable price. You should also notice that WaterChill's products are available as DYO-versions (Do-Your-Own) which are even 50% cheaper than normal products. In practise this means that DYO-products are sold as parts of WaterChill items which are supposed to be assembled by the consumer. More information about these here.
The blocks represent first-rate products and they are very classy. The bottoms of the blocks are generally very smooth and good work, even though I think the last finishing hasn't been done. Every block's attachment was working, stable and easy. The quick connectors, Perspex covers and the additional cover for incompatible main boards left a great impact.
The Control Unit, the heart of the kit, is a genius invention; very working and nice package. The power cables route to the case has clearly been planned. The cable going through the back panel is a small but important detail. The Black Ice radiator used in the kit seems first-rate and compact. The weak paint surface gave me a negative impact anyhow. But as I wrote before, Asetek has given feedback for the manufacturer and this should be improved in the future. A nice feature is the possibility to attach the radiator to your case or use two 120mm fans. The fans attaching hasn't been designed very well because of the horrible screws. Asetek promises that any of these kinds of screws shouldn't be sent away from the factory anymore.
The Hydor Seltz which was to work as the pump handled it's duty well even though the small grumbling might annoy the worst silence freaks;) This is probably because the pump is used as a pipe pump, not a sunken one which are always more silent. The possibility to place the pump in any position is also a positive thing. The pumps and tanks attachment can be considered as a plus as well. The massive Sunon fan used with the radiator does its duty well, even though the noise with 12V is annoying in long-term use. In 7V the noise is acceptable.
The biggest problems occurred with the stiff pneumatic pipes but this is understandable on the other hand. Asetek has chosen lasting and high-quality pipes to the WaterChill which have their negative side, the stiffness. The biggest singular problem appeared with the graphics card where the card bent because of the stiff pipes. But this wasn't only the pipes fault. In the end the pneumatic pipes with its credits and minors are a rather good choice for the job. After cutting the pipes to their right size the problems decrease.
Overall the WaterChill is a kit of quality designed and manufactured water chilling, which end up as a rather working system with few flaws. So if you’re in a need of a first-rate water cooling kit I recommend Asetek's WaterChill Complete Kit. So off you go to the store :)
Good
- Quality
- Quick connectors
- Extra cover for the CPU-block
- Blocks' attachments
- Classy looking blocks
- Perfect Kit
- Users Book ;)
- Radiator's attaching holes
- Tank
- Pump's installing possibilities
- WaterChill Control Unit
Bad
- Stiff pipes
- Attaching the fan to the radiator (has been improved now)
- Noisy with 12V
Translated by barneyz
Asetek WaterChill Complete Kit
Post a comment