Cool-Cases CF1 microstructure water block review

Cooling/Water Cooling by KeithSuppe @ 2004-03-21

I thoroughly tested this block several times, and this review ended up taking three days. I positioned, and re-positioned the TTF-103 thermistor at least 15x, to ensure optimal contact was achieved. I\'m very impressed with Cooled-Cases.org, I can honestly say they may be the most innovative H20 PC-cooling company out there.

  • prev
  • next

Test Setup

Our test system will feature the Abit AI7 (BIOS 16) utilizing Abit's new µGuru overclocking and system monitoring microchip, with corresponding Abit EQ utility. For our test catalyst I used the P4 2.4C L310A735-0101 with IHS removed. Thermal Interface Material used throughout, shall be Arctic Silver AS5, certainly the most effective thermal paste I've used to date. The Abit AI7's µGuru utility is more then a simple software/BIOS utility.

Via the Guru microchip, FSB, Vcore, VDIMM, and VDDQ can be adjusted in Windows. Currently only Windows is supported, Linux lovers will have to await and see what Abit comes up with. Below is a photo of the µGuru microchip (note its location in proximity to the ICS clock generator), the second photo exemplifies its real-time monitoring through Windows;

Madshrimps (c)


Madshrimps (c)



The temps above were taken running the 2.4C overclocked; 250FSB on 1:1 aspect ratio, for 3000MHz. It's apparent from jumpstart; the CF1 is an adroit performer. In fact its performance exceeded that of Danger Den's venerable RBX on many levels. In all fairness, however; the RBX was tested with accelerator nozzle-5. And the combination of the Hydor L-30, with accelerator nozzle-5 had a deleterious effect on RBX performance. Nozzle-5 is ideally suited for a high velocity pump. In fact I look forward to mating the RBX with the Hydor L-40 and L-45 shortly, as I'm awaiting shipment from Italy.

For the monitoring of temperatures I'll utilize a combination of several utilities/devices. In addition to the Abit AI7 µGuru temp/voltage monitoring facility, I'll be using the Coolermaster Musketeer. It features an external thermistor, the VOLTTS TTF-103. I tested the thermistor's accuracy to within +/-2C. The 2nd photo down taken of the Musketeer's TEMP meter corresponds to the software reading above;

Madshrimps (c)


Madshrimps (c)



The next photo reveals thermistor placement with the CF1 installed on the Abit AI7. I want to share briefly what I gleaned from a discussion with Nevin of Arctic Silver, regarding thermistor's. First of all, what a thermistor essentially measures is itself. If we use the analogy of a tennis ball, thermistor being the ball, only 5% of its total surface would make contact with the item it's pushed up against, in this case the exposed P4 core. Let's say 10% makes contact with the water-block above, and another 10% contacting the socket beneath.

This would leave 75% of the tennis-ball (thermistor's) surface area exposed to ambient air temp, albeit the air between socket and water-block (or heatsink). One can now begin to see the conundrum. In this case even the Delta-T rule may not suffice in determining an accurate temp. In addition many thermistor makers state their products are affected by EMI, and temp variations along the wiring. With the IHS removed, and my ensuring the lead made no contact with others devices, our temperature readings should be more accurate;

Madshrimps (c)



Per my usual set-up the case is located in the corner of my room on a stone floor, perpendicular from two windows which are partially open. This allows cool air from outdoors to waft in, down, and into the open case (I live approximately 200-meters from the Atlantic Ocean, Newport RI USA). As cold air sinks to the stone floor where the case sits, a combination of Fourriers Law, and the Thermal Equilibrium rule have devices in the tower maintaining some pretty low temps.

The case used was TTGI-USA's TT-201T3 brushed-aluminium, 5x80mm/fan. The case features three windows, removable motherboard-tray, and top-mounted 80mm LED-fan. Without a doubt the coolest (figuratively and literally) case I've owned. I have the side panel removed to exacerbate cooling, and for ease of access. This also eliminates an additional variable in the formula, namely the temperature differential between case/ambient and room/ambient;

Madshrimps (c)



I'll be running the CF1 with the P4 2.4C with its IHS removed. As I stated earlier Arctic-Silver AS5 will be used throughout.

Test system:
  • CPU - P4 2.4C SL6WF L310A735 (IHS removed)
  • Motherboard - Abit AI7 (BIOS 16)
  • RAM - Corsair Twin-X4000PRO (2x512MB DIMM's 0,2)
  • VGA - Asus V9950 (Det. 53.03)
  • HDD - Maxtor DiamondMax Plus 9 (SATA150, 120GB)
  • Optical - TDK VeloCD/RW
  • PSU - TTGI USA TT-550TS (Quad fan aluminum 550W)
  • OS – Windows XP

    Onto the results ->
    • prev
    • next