---

There aren't too many choices out there for solid copper heatsinks. In fact, I only know of one manufacturer in Asia that makes them, so what to do? Well, this is FrostyTech, so we made one :-)  Enter the Frosty CopperSinkr. With the finished version shown, let's go through all the steps we used to construct the Frosty CopperSinkr. The construction process took us few hours one weekend, and required only a few specialized tools (see last page).

Why Copper? The question many might wonder, is "why copper"? Why not simply go out and buy a cheap aluminum heatsink, or a more expensive Alpha then go to all the trouble of making ones' own heatsink. The answer to this is simple; copper has a thermal conductivity of 394 W/m*C, where as 6063 aluminum has a thermal conductivity of only 193 W/m*C. A difference of 201 W/m*C means the trouble to make a solid copper heatsink is well worth it. It also means a much smaller copper heatsink can do the job of a larger aluminum one, but we ignored that. ;-)

What kind of heatsink?

Cast: The very first idea we had was to cast a heatsink from pure copper. But casting an amazing heatsink - possibly based on the current high performers out there - was out of the question. That's not to say we didn't try. Before settling on the plate and fin design we tried our hands at casting a very simple copper shape. Making the mould for the heatsink design was easy, melting the copper wasn't. 

At our disposal was an oxy-propane torch. While oxy-acetylene may have made this possible, we had no access to such a rig. Anyhow, we found it almost impossible to melt a small chunk of copper and keep it in its liquid state. From what I understand, this difficulty is partially due to the rate at which copper sheds its heat to the surroundings. An induction furnace is probably needed to melt the volume of copper needed for a small heatsink.....

Extruded: Extruding the heatsink would have been nice, but that wasn't ever an option. :-(

Assembled: Eventually, a trip down to the Metal Supermarkets was made and we returned with a 6" x 12" x 1/4" plate and 36" x 12" x 1/32" sheet of copper. Both of the pieces together weighed in at around 15 pounds and cost us nearly $70 CDN. While that may seem like a lot of money, we got much more copper then was necessary for this heatsink. There is enough material here for many more heatsinks. I'm sure it will all get used as we have many more heatsink variations on the drawing board already!

Finally, we settled on a design determined by what we could make. Backtracking from the tools at hand, to the materials available, we were able to come up with a suitable fin-based design. The goal was to make a heatsink capable of mounting atop a CuMine 500E. With that in mind, a heatsink for the socket 370 form factor began to take shape. As a bonus, the heatsink would also be backwards compatible with the Celerons - which is good cause we have a 366A that does 616mhz ;-) The only constraint we had, involved the clip to attach the heatsink to the socket. We salvaged the clip from an old Coller Master heatsink, and were sure to leave enough room between the center fins for it - more on that later though.