I've begun Stage II of my project to supercool my AthlonXP-M. My project goal is to achieve 3.2GHz or better, but I don't want to kill any components. That's why I'm practicing on lesser equipment first.
Today's project goal: AMD K6-II 333MHz at 550. I've got my heatsink made up already. It is a two by two inch block of aluminum that is an inch and a quarter tall. On top of this is a sheet aluminum tray (custom built) screwed into the aluminum block with two self-tapping screws. The tray is designed to hold dry ice, while the aluminum provides an easy way for this temperature difference to reach the CPU core. There will be Arctic Silver 5 between the aluminum block and the CPU.
The heatsink itself is surrounded with 5 layers of double-sticky sided foam tape. This stuff is only half as thick as what I used in the peltier experiment, so that's why I doubled it up. Hopefully this will prove thick enough to prevent condensation. Since dry ice is only -78.5C, I didn't take extreme measures to ensure that no condensation occurs. In my future liquid nitrogen experiments, there will be massive insulation since it is at an impressive -195.5C. I've recently found where to obtain LN2 here =)
Once I conduct the dry ice experiment, I will move to LN2 and then ultimately to LN2 on an XP-M/nForce2 solution. The only things left between me and that goal are:
1) Money for a new motherboard for my main machine
2) Perfection of the technique on a lesser rig
The reason I am using old hardware to test my cooling methods is quite simple. Not only is it cheap, but more importantly, it puts off less heat. If it doesn't condensate there, with less load, it won't condensate when more heat is being pumped into it.
Results and pictures will be posted today or tomorrow, hopefully.