I was originally introduced to Cool-Cases CF1 microstructure water block, whilst scanning reviews at
8Dimensional.com. I came across what I thought to be an esoteric cooler review at
Kuehlertest.de. The CF1 performed extremely well in several reviews, out performing most other European Wasserküehlers. Due to the language barrier, I'd originally thought the block was manufactured by a company known as
Watercooling.de. This company is located in Wienheim Germany, and is known to the Euro market via they're Online store
Watercooling.de. They offer an extensive line of water cooling products, and for very reasonable prices.
I contacted Florian Jakob proprietor of
Watercooling.de, who was kind enough to send me this block. Whether you’re in Germany, or anywhere in Europe, I highly recommend visiting this store which offers a full line of Wasserkuehling, VapoChill and such interesting Enthusiast products as the water-cooled
Pro-Silence 450W PSU, and the Speed-Strip
SSA-1. The SSA-1 unlocks multiplier's 5x - 12.5x on specific AMD Barton, and most Thoroughbred CPU's, via the simple application of a strip to the back of the processor.
The
CF1-Microstructure CPU-Kuehler at its heart, features 324 glass-ground copper pins as the primary impingement surface. Its design is well executed, epitomizing CNC precision engineering. The mounting hardware supplied for AMD and Intel is straight forward, solidly constructed, and makes for a secure mount.
The inlet barb screws into what seems to be a high-pressure extrusion moulded ceramic, in which a compression channel increases water velocity through a copper accelerator plate. The plate features a 9-stream inlet, drilled to produce swirling, and forcing multiple impingements upon the 324 glass-ground copper pins. This creates a very high turbulence effect, ensuring thorough heat absorption occurs. Overall thickness on the CF1's highly polished copper base plate is approximate 6.5mm. Where the milling creates channels between the pins, the thickness drops from to 6mm to 4mm. This is an ideal amount of material between heat source and water impingement, and should make for an excellent performer (photo's below were borrowed from
CaseUmbau.de);
The benefit of a multi-pin design is heat travels up through the pins, while 9-streams forced through the copper acceleration plate (seen inserted in the top portion of the CF1 above) flows down, around, and between the pins. As described above this method creates a high amount of turbulence, and the resulting stagnation from accelerated impingement allows the water to remain just long enough such that all the energy (heat) is absorbed. One will often hear of a "stagnation" effect when the Cascade design is mentioned. In the Cascade, "cups" are milled into the primary impingement area, and "tubes" partially submerged into these cups, forces the turbulence resulting from "jet impingement" to stagnate so absorption can occur without interference. Basically Cascade’s turbulence is analogous to the base of a waterfall where water turns "inside out," so all the molecules are exposed to the heat, even heat travelling up the cup sides.
I do, however; believe the pin design may have an advantage over Cascade or the cup design, because the copper surface areas between cups results in untreated "hot zones." In the pin design such "zones" are non-existent. The photo below may better emphasize the pin-impingement area which, in the CF1 is centrally located over the CPU core (the photo below was also borrowed from Caseumbau's CF1 review);
While it may seem like an optical illusion, there is a slight concave in the center of the pins effecting slight centrifugal impingement in which the CPU core location beneath becomes the axis. The finish on the CF1's base plate was excellent, and there was no need for lapping. This is another indicator of a company's commitment to quality. I believe it's imperative to check the surface of your water block (or heatsink) by placing either a razor-blade edge, or ruler to determine its flatness. Placing the base plate on a mirror or glass plate will not suffice as any concave imperfections in the base plate’s surface will not be revealed using this method. You should also check your processor with a flat edge.
It's important to understand the surfaces you're mating. If for example, both are concave, the resulting gap will be detrimental to heat transfer. I've read in numerous forums where people purchase new heatsinks, and wonder why they're temps haven’t dropped, and in some cases temps increased? This is due to the fact some Intel CPU Integrated Heat Spreaders, as well as custom heatsinks may have a slight concave or convex anomaly in their surface. In such circumstances it may have been the case your processor had a slight concave surface, while your stock heatsink had a slight convex protrusion. Ergo they mated well, albeit a relationship based on imperfection, it was none-the-less, functional. In purchasing your expensive perfectly flat, or slightly concave custom heatsink, that thermal transfer is now rendered moot. I tested the CF1's base plate surface, and was impressed to find near perfection. My Intel 2.4C minus its IHS has a flat surface as well;