Re: [iopener] Major heat reduction on IO processor

New Message Reply Date view Thread view Subject view Author view

From: tomatom@juno.com
Date: 04/15/00-12:36:25 PM Z


      on the altcpu-list@topica.com , 2 yrs ago they discovered

 that using "silver-based conductive/thermal paste " can lower

 the cpu temp as much as 13 degrees C. !!! but i suspect

 that in the high (13) case, the previous heat sink paste was

 improperly applied ( too much/or too little )...expect 8-10...

 you can also polish the cpu top and heat sink bottom with

 600 w/d emory paper---when 2 metal surfaces come into contact,

 there are pits and valleys that cause the TOTAL surface contact

 to result in only 1% to 10% !!! that's why you NEED somekind

 of paste to fill the air pocket/insulating voids !!! i've tried the

 al foil and copper sheet fixes, but didn't have the thermosisters

 for serious tests---when heat is a problem, like over-clocking ;) ,

 go for the silver-paste ( $6-$11 for a tiny tube, but use a single-edge

 razorblade to apply a super thin coat, so many applications from

 a 1/8 oz. tube )

        speaking of o'clocking, is there a 75 mhz jumper on the

 iOpener ???

On Sat, 15 Apr 2000 11:03:45 -0400 Guy Gordon <gordon@white-crane.com>
writes:
> The thermal engineering of the iopener seems way out of line with the
> electrical and mechanical engineering -- both of which are well done.
>
> The circuit board is a thing of beauty and economy. The physical
> design makes you want to own one at first sight. But why do the
> heatsink grooves go horizontal instead of vertical? Why does the
> heatsink stand off so far from the cpu that it needs a spacer. Why
> did they use a plastic pad for a spacer instead of metal?
>

_______________________________________________
iopener mailing list iopener@scsiboy.com
To change your mailing options, please go to http://snoopy.net/mailman/listinfo/iopener
Read the FAQ, please! http://fastolfe.net/misc/i-opener-faq.html


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29.