Re: RedHat 5.0 and Power Supply.

["Thomas E. Dodd /CSDC" <ted cypress com>]

Zoki wrote:
> 
> On Fri, 2 Oct 1998 hyprreal cdc net wrote:
> 
> ->At 01:08 AM 10/2/98 +0200, you wrote:
> ->>On Thu, 1 Oct 1998 hyprreal cdc net wrote:
> ->>*** Does the ventilo slow down when you start a program (when the CPU
> ->>works) and speed up when the computer (the CPU) is idle!??
> ->>
> ->>Zoki.
> ->>
> ->>Mailed with Linux and Pine.
> ->>
> ->>
> ->
> ->      Yes... Is there a way to correct this?
> 
> *** Daniel,
> 
> It seems that Linux is able to switch off the CPU when it's not used, thus
> freeing the power supply due to which the rest gets more power. I guess
> the HD's have a power regulator but the effect on the ventilo's is that
> they make substantial noise due to speeding up and slowing down. The
> problem doesn't apply to Win 95 because it doesn't control the CPU unless
> you put it on standby. The test we did was to use a little program written
> for the Cyrix CPU's that did the suspending thing on Win 95 and the result
> was that you could indeed hear the whining of the ventilo's going up and
> down. It has something to do with the wireing of the power supply. We ended

This is a problem in the powersupply, not the cpu/mother board.
Min has been doing it since I added the Cyrix patches to my
kernel. They are now in the RedHat kernels.
As above, Win95 and WinNT do this too with extra tools.
The patch/tools allow the CPU_HALT instruction to
cause a low power mode. The documentation with the tools
mention that the fan speed-up/slow-down is normal
and merley an annoyance.
If you can get a different power-supply/fan this can
be lessened, but it's not anything to worry about.

NOTE: This instruction keeps my 6x86L-PR150+ (120MHZ)
cool enough to run at 166MHz (PR233+ ?, the 200+ was
150MHz) using a standard heat sink and fan.

	-Thomas