15 Apr 2003, 00:05, JIM HOLSONBACK (1:123/140), wrote to MARK LEWIS:

Hi JIM.

 JH> But about that shorting out - - I see where a number of vendors 
 JH> sell a "copper shim" for use between the heatsink and
chip.  I wonder 
 JH> how those work without shorting things out?

Intersting!  The concept itself seems flaky, as that would introduce _two_
additional surfaces with 'air-gap' potential.  If it is actually made of
copper (which is both conductive and relatively hard), that would give it
three strikes in my NSHO.

 JH> But Pentium II's, in the original SEC packaging - - some of those 
 JH> came with passive heatsink apparently permanently riveted on, but for 
 JH> those which had a removable heatsink or heatsink/fan, I can't recall 
 JH> ever seeing evidence of any thermal compound being used between the 
 JH> heat plate of the SEC cartridge and the clip-on heatsink or 
 JH> heatsink/fan.

I don't recall any thermal compounds on those either, but those were less
than fully integrated and had several discrete chips spread over a rather
large area.

The packaged Intel P-4s come with a "shim" also, which appears to
be about a .003" thick lead strip covered with a thin brown 'powdery'
coating glued to the heatsink by it's edges.  Those heatsinks are NOT for
thin MBs though, as they are cam-latched to the MB, and deflect the MB
about an eighth inch when they are locked down.  Intel warns that cheap MBs
could become sufficiently distorted to contact the metal case beneath the
MB. 


     Good luck...  M.

--- Msged/386 TE 06 (pre)