The Art of H20 Cooling



Watercooling has certainly evolved from a purely DIY overclocking hobby into a multi-million dollar industry. With high quality pre-packaged watercooling systems those considering H20-cooling may choose from a plethora of options at various price points. It's been several years since that infamous DIY threads about the over zealous PC-Enthusiast who went about watercooling his PC by sealing all the holes and filling the case with water. Tales such as this have now become part of watercooling folklore. As far fetched as that thread may have seemed submergence cooling is a tractable concept using state of the art solutions such as 3M's Flourinert FC-77. Watercooling has afforded true artisans an opportunity to express their craft with the marriage of thermodynamics and CADCAM metallurgy. Perhaps the pinnacle of art and CNC waterblock construction can be found at the Spanish Enthusiast site DevilMaster. The example below of a graphic card cooler which uses direct impingement to the VPU and GDDR is one of many articles which explain the entire process from planning to performance.





Of course it's not economically viable to produce blocks of this magnitude for the masses, although I wouldn't hesitate to pay for such craftsmanship if I could. The majority of us demand two characteristics from our watercooling hardware, performance and reliability, anything else is an ergonomic icing on the watercooling cake. In the case of Viscool and in particular the V2 we're testing today we get much of the former attributes and some of the latter.

Viscool Eastern Euro CNC



Although Viscool might be considered a small manufacturer located in the Slovak Republic, what they lack in numbers they compensate for in innovation and precision machining. Back in 2005 Viscool released their Extherm CPU cooler (seen above) which as far as I can tell, had only seen one independent test bench via Watercoolplanet whom seem to find and test most every waterblock ever built. eXtherm was a decent performer and long before I had a chance to contact Viscool they had moved on to improve their design.



The block is well constructed and it's obvious the quality control and precision machining are there as well as decent ergonomics. The block is constructed using a copper baseplate, steel inlet/outlets which are encased in derlin. The mounting brackets are most definitely universal so long as the included through-board hardware is used. This is invaluable given the number of supported Socket's on the market at this time. I tested the Viscool on the relatively new AM2 system which did concern me until I realized Viscool has implemented what has to be the most versatile mounting hardware on the market. We'll discuss this in detail later on. Suffice it to say I was honestly impressed with the craftsmanship and exacting standards evident in the V2.



From the photo above its evident Viscool has done something different in the "impingement" area and as we'll see this design is an amalgamation of several concepts....

Baseplate
Whether your waterblock baseplate is constructed of silver, copper, aluminum or steel (although I've yet to see the latter) if the base surface is flawed contact between the waterblock and processor will be jeopardized. In this respect the formula is simple, the percentage of heat conducted is in direct proportion to the amount of surface area contact.



Viscool may not have the smoothest finish; however, it's definitely above average and very flat. Laying a aluminum ruler across the baseplate is crude but an acceptable method. Machining pattern marks can be seen with the light angle below as the reflection in the first photo only shows how "polished" a surface is not how flat it is.



Viscool has kept the price of the V2 low by their effective use and distribution of materials. The base-plate is made of copper where that material is most beneficial for heat conduction with the top portion of the waterblock made of aluminum. A separate inlet/outlet section seems to be pressure inserted into the black portion where the barbs thread. Due to the difference in metals an additive preventing dissimilar metal corrosion should be used. In fact this is always a good idea since 95% of all H20 systems employ different metals at some point. To view an Anodic Chart for different metals click on the following Link.



Since the inlet is offset from center an arrow has been placed on top of the waterblock to indicate the inlet barb. Where the inlet is offset for mounting reasons a cross over channel is machined into the upper portion of the waterblock to ensure the incoming water-stream flows directly onto/into the center of the impingement zone. Even with the advent of Dual Core processors, ensuring the impingement zone is centered above the CPU is integral to the design. In theory this is where the majority of the processor's heat will be conducted through the copper where the fluid carries that heat away.



Due to the offset and the degree of the angles the water stream encounters as it's re-routed down to the impingement zone, the V2 will most likely benefit from a high flow rate pump. In his review of the V2 Joe at Overclockers.com recommends a minimum 300GPH pump. Unfortunately in Viscool's desire to appease all experience levels, the center of the top portion of the V2 must be free for Ziff tab type mounting hardware. This prevents what is referred to as direct impingement explained above. At first I was confused since I'd thought the universal brackets were the only mounting method, until I looked further into Viscool's site finding evidence of a Ziff tab mounting hardware.



Viscool's V2 now earns the title as the most universal mounting system. the previous title owner had been the X20 CPU-waterblock from XSPC which we reviewed here. Although the brackets supplied do take some getting used to and are a little difficult to align, once you get used to these they are a blessing. the photo below was taken from Viscool's site.



Onto the heart of the V2 and any waterblock. it's impingement zone....

Impingement Zone
This is where the V2 differs from any other waterblock I've seen to date. As seen in the photo below Viscool began with a 5mm thick copper plate machining that plate from 5mm at the edges to 1mm thickness at its center. By cutting channels at right angles the remaining material leaves "pins" approx 4mm in height. This is an extruded pin design and just like air-coolers the goal is surface area. What Viscool has done was to bevel the base leaving an "impingement zone" at the center where the extruded pins stand. A truly unique feature of the V2 is the ring at the center circumventing and enclosing the upper portion of the extruded pins forcing the incoming water-stream into the center of the base. Although the inlet fixture or barb is offset a channel forces the water down through the ring over the pins and to the base. From below the ring channels can be seen cut into the beveled base where the water exits to the outlet.



In the V2 design the ring surrounding the pins at the base center come into contact with a slightly protruding center section from the toplate. This rough seal contains the water-stream forcing down through the "ring" into the deepest section of the base. The thumbnail below the photo shows a close-up of the toplate.





Using a small penlight the channels extending out below the ring can be seen more clearly. This is the first time I've tested any waterblock with this type of impingement zone. To be honest I was a bit perplexed as to how Viscool engineers integrated the "ring" onto/into the base. Thanks to MaxxRacer (Brian) of XtremeSys Inc who infomed me the ring is "pressed" on.





Undoubtedly V2 is a unique block, and while my description of its inner working's may seem verbose I only hope to communicate the thought put into this design. Many waterblock's come and go and the extruded pin design takes many forums most of which begin to resemble one another save their overall attributes. In this respect the V2 won't make centerfold of the year in Waterblock Voyeur magazine, what is does do; however, is perform as we'll find in the next section.

Testing - Corsair Nautilus 500



In July I tested the Corsair Nautilus 500 external cooling system. In that review I had considered the possibility of mating other waterblocks with the external cooling unit. The Nautilus 500 is supplied with a waterblock made especially for Corsair based on the Swiftech MCW6000. The external unit is based on a single aluminum radiator mated to a Vette 120MM fan rated at 74.4CFM at 1800RPM when the dual speed switch is set on "high." All tests were run at 1800RPM. The pump used in the Nautilus 500 is the Laing DDC-1 running 420l/h, 3.7m H at 12V (approx). Comparative testing began with AMD stock HSF and the Nautilus 500's included CPU block, then switching to the XSPC X20 an above average performer we tested in 02/06. Finally the Viscool V2 was installed as seen in the photos above and below.



The V2's mounting hardware is so simplistic as to be ingenious. While it does take a few moments to align and re-tighten all the screws and hand fastener?s, the ability to mount the V2 on just about every motherboard in production with holes for through-mounting is well worth it. It took me approximately 5-minutes to install the waterblock. As with all my H20 reviews I re-mount the waterblock several times and compare results determining the best mount by the lowest temps (all things (ambient room temp) being equal). As I began recording temps I realized how far behind I am in the Dual Core department. It wasn't until I found an extraordinarily large discrepancy in my water-temp measurements and on-die results I realized how important it is to utilize a monitoring program capable if reading both on-die thermal diode readings in a Dual Core environment. Asus AI-Booster and PC-Probe fail to differentiate since the software doesn't distinguish in DC. Therefore I chose the Beta program Core Temp, thanks to the knowledgeable Enthusiasts at thecoolest.zerobrains.com a very interesting forum website. The current version may be downloaded from the following; Core Temp v0.9.0.91 (zip). This program accurately measures both Conroe and K8 DC processors simultaneously which is explained here.

Test System

AMD Test System
CPU	Althlon64 3800X2 AM2 Retail
Mainboard	Asus M2N-SLI Deluxe
Memory	Crucial PC2-8000 2x1024MB DC CL5-4-4-12
Graphics	Leadtek 7950GX2
Power Supply	PCPower&Cooling Turbo Cool 1KW
Cooling	Corsair Nautilus 500 XSPC X20 Viscool
Operating System	Windows XP

To summarize testing consisted of running our AM2 3800X2 at default speed and voltage 2.0GHz = 10x200FSB / 1.35Vcore, and overclocking the CPU to 2.85GHz = 10x285FSB / 1.40Vcore. CPU Temperatures were recorded with AMD's stock HSF, Corsair's Nautilus 500 waterblock, XSPC's X20 waterblock and Viscool's V2 waterblock all waterblock's shared the Nautilus 500's external cooling unit. To simulate processor LOAD the stress test utility S&M was used. The original website for this program can be found at testmen.nm.ru and the current version may be downloaded directly at S&M 1.8.1 (160). Maximum temperatures were recorded running S&M at 100% and as stated above the program Core Temp allowed me to record both temps in the results chart. Thermal paste used throughout was Arctic Silver Luminere a special electrically non-conductive thermal paste with a rapid cure rate made specifically for product testing. In the chart below the results for air-cooling at 2850MHz were left blank due to the system crashing at 100% LOAD under stock cooling. Ambeint temps were maintained between 18C ~ 20C throughout testing.



Summary
I was able to find just three tests on the Viscool V2; Watercoolplanet and as I stated earlier Joe at Overclocker's who recommended a pump with a minimum flow rate of 300gph or 1135l/h. Admit tingly the pump I chose was under powered in the flow-rate department at 110gph. While I most definitely concur with Joe's assessment, based on the block's design I chose to use the LAING DDC-1 for two reasons:
1st.) flow-rate does not account for pressure as in DOOMPC's waterblock roundup an XSPC DC1000 pump was used which is most likely a Hydor L25. While this pump's flow-rate is around 1000l/h it's Height (pressure) is very low at 1.35 m/Height. The DDC-1 specifies a 3.7m/Height.
2nd.) based on the waterblock's low price it's most likely the case this block will end up in a system using a "budget" type pump such as the Laing DDC-1 which is relatively inexpensive. The popularity of the DDC-1 based on its size alone has made the pump an attractive alternative to OEM manufacturers, not to mention off-the-shelf sales.
In the very near future I'll be retesting the V2 in my next waterblock comparison against XSPC's new direct impingement version of the X20. In that test the pump used will be the LAING D5 Strong.

PRO
Price!
Innovative design.
Perhaps most universal mounting system available.
All copper baseplate high quality machining/finish.
Solid performance.

CON
Channel from barb restricts flow (direct impingement is best).
Limited availability.

The V2 can be purchased through Viscool for just $24 (USD) Viscool V2.

Conclusion
Viscool has given us a truly exceptional waterblock in the V2 for so many reasons. I decided to mate the V2 with the Laing DDC-1 simply because it's so widely used and given Viscool's limited "watercooling kit" option I would expect those looking at the V2 would be switching out a waterblock they're not satisfied with for the V2. Even with a flow-rate handicap the V2's performance speaks volumes. At $24 Viscool has produced what is most likely the best "budget" waterblock available, only further testing will tell. I would like to thank Viscool for submitting their product, stay tuned for the next test.

Questions/Comments: forum thread