RE: [cobalt-users] Raq3 running temps

[Gerald Waugh <gwaugh@xxxxxxxxxxxxxxxxxxxxxxx>]

On Wed, 14 Apr 2004, Phil Beynon wrote:

> description above that's where it should be anyways! <grin>
> In theory the processor and associated circuitry is doing more work as it
> has the refresh cycles to run in more memory, plus address line loading etc.

No two CPUs will have identical temperatures, although it could be loading
and refresh as you suggest. It would be ingteresting for the original
poster to swap the CPUs between the units and see what the results are.

> I'm not sure on the raq where it gets the cpu temp from anyway, if the K6
> has an internal sensor or if it is just using a thermistor in the immediate
> vicinity of the cpu to collect the data, in which case it could just be the
> amount or lack of thermal paste on one sensor that accounts for the
> difference.

The RaQ gets it's temperature from an IC under the CPU. there is a small 8
pin IC (National Semiconductor LM77) under the CPU. So the thermal
coupling is very poor, and the temperature read is a guesstimate.

Gerald
--
http://frontstreetnetworks.com | http://raqware.com
  Front Street Networks LLC, 229 Front Street, Ste.#C
  New Haven, CT 06513-3203 | phone: +1-203-785-0699