PCBUILD Archives

Personal Computer Hardware discussion List

PCBUILD@LISTSERV.ICORS.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Reply To:
PCBUILD - Personal Computer Hardware discussion List <[log in to unmask]>
Date:
Sun, 4 Oct 1998 19:36:44 EDT
Content-Type:
text/plain
Parts/Attachments:
text/plain (87 lines)
In a message dated 10/4/98 2:40:50 AM Eastern Daylight Time,
[log in to unmask] writes:

<< > Date:    Tue, 29 Sep 1998 13:41:37 -0400
 > From:    "Max A. Lupton" <[log in to unmask]>
 > Subject: Re: cooling amd
 >
 > I was just wondering what the issues with "too much cooling" might be.
 > Max
 >

<snip>

 > >On 23 Sep 98 at 17:08, Roberto Safora wrote:

<snip>

 >  > The older K6s may be okay with the slightly smaller fans.  [You
 > >*can* put too much cooling on a CPU, but only with a Peltier
 > >device....]
 > >David G

> Liquid nitrogen, etc. will also stop electronics cold - it has been
> used by bomb squads (it also kills the battery).
>
>Referring to SAMS "Reference Data for Engineers" - 1993, p.18-10,
>  "Equation 21 is called the diode equation, which describes the total
>   current through the diode for either forward or reverse bias."
>
>
>                                              qV/kT
>      I = q A [ (D  /L )p   +  (D  /L )n  ] (e       - 1)
>                  p   p  n       n   n  p
>
> Assuming V>>kT/q, qV/kT>>1.0 and the equation becomes essentially
> exponential.
> As the Absolute temperature T increases from T1 to T2, qV/kT decreases
> and I decreases by a factor of e raised to the power of T1/T2
>
>                 Range          T1/T2      factor
>  commercial    0 to 70 C       1.256       1.292
>  military   -55 to 125 C       1.825       2.281
>
> If you push the temperature range too far either way, the "resistance"
> of the components change too much, and the device goes "out of spec."
> and may even stop working.  You can think of a typical bipolar
> transistor as two "back-to-back" diodes, although this is an over-
> simplification.
>
> Do you have a rough idea how many million transistors there are in
> your computer?  If something gets too hot, your first sign of
> trouble may be "garbage" on the screen.  If you are lucky, you may
> be able to switch off in time to save the system, and replace the
> burned-out fan or take the lint out of the heatsink.  Other times,
> the first symptom is "sudden death" of some chip.
>
> Boyd Ramsay
>
> [log in to unmask] >>


Hi,
  While what you say may be applicable to bipolar transistos, modern
microprocessors are constructed with CMOS transistors (a type of FET), and
these actually work better (faster) at lower temperatures. There is at least
one company that takes advantage of this fact, Kryotech. They sell systems
complete with a refrigerated system to cool the CPU down to -40C. At this
temperature, the CPU will run approximately 30-50% faster than at room
temperature. They have in the works a system that will cool the CPU down to
liquid nitrogen temperature, it will run 100% faster at these temps (i.e. a
450Mhz PII could run at 900Mhz).
  You can find more information about this at
www.tomshardware.com/kryotech.html
or at their web page at www.kryotech.com
  The second post describes "too much cooling" with a peltier device. The
problem is not from too much cooling, but rather from the condensation that
will form on the cold parts, unless special measures are taken to prevent
this.

Hope this helps,
Peter Hogan
[log in to unmask]

                                  -----
                PCBUILD mailing list -  http://nospin.com
         Bob Wright:[log in to unmask] - Drew Dunn:[log in to unmask]

ATOM RSS1 RSS2