CPU heating!

[Walt Pawley]

At 7:53 PM -0300 2/17/11, Mario Lobo wrote:
>I replaced the thermal grease (as advised here) with a new one but that didn't
>change those figures.

I've not attempted to keep up with changes in thermal
conductivity of heat sink compounds for something like 40
years. About that time, National Semiconductor published a
paper about the effectiveness of various means of mounting
power semiconductors to heat sinks. As I recall, the only
method of obtaining truly efficient heat transfer was soldering
the case to the heat sink - clearly not all that practical for
many purposes.

All of the thermal compounds tested did comparatively poorly,
especially when applied too thickly. They worked best when
applied just heavily enough to fill voids in the two surfaces
being mated.

Perhaps things are different today but I have my doubts. If
they were, we'd make heat sinks out of thin walled finned
copper cans containing the thermal compound.