Re: Popular Science

[paulb@arcadia.ece.nwu.edu (Paul Brazis)]

In article <1998081010335200.GAA01251@ladder03.news.aol.com>,
Supertimer <supertimer@aol.com> wrote:
> jim@calico.litterbox.com (Jim) wrote:
>
>>Couple things about cooling your chip.  First, doesn't the chip package have
>>to be designed for this so it doesn't shrink at the wrong speed and break the
>>leads to the die inside?  Second, if memory serves the 6502 family used to
>>be used in auto ignitions, where it was discovered that at subzero
>(farenheight) 
>>temperatures it would cease to work.  This may have been the condensate issue
>>someone brought up, however.
>
>I don't know about subzero farenheight, but a few degrees subzero
>celcius is doable.  I've tried putting a 65C816 into the freezer and
>trying to push it faster on a TWGS, but no go....I think that's because
>at the TWGS speeds, the 65C816 is not even breaking a sweat.
>CMD pushes these chips to a full 20Mhz and not one has failed.
>The reason TWGS tops out at 15Mhz or so is not because of the
>CPU, but because of the OTHER chips that make the accelerator
>work.
>
>I have not tried freezing the TWGS just yet because of the
>condensation issue....is there a way to put it in the freezer without
>it getting wet?  Just the CPU is ok because I can wipe it before
>plugging it back in, but a whole card...the moisture would ruin it!


The lab I work in often works with samples that are cooled all the way down
to liquid helium... to avoid condensation (which can be a problem), we place
the specimens under vacuum for a while, then fill the chamber with helium
gas. Nitrogen gas would work as well. I know that it really isn't too
convenient for what you're doing, but I'm not sure of a simpler way to keep
out condensation.