Does the Pentium M Processor indirectly control fan speed ..

techman41973.RemoveThis@yahoo.com (Sanjay Punjab) wrote in
news:e8fde247.0401202240.3177967b@posting.google.com:

 > I own a fujitsu lifebook and the fan seems to have a life of its own.
 > Turning on and off randomly and racing periodically. It is hard to
 > explain, but the operating is erratic and my friend's identical laptop
 > does not suffer from this problem. The laptop has a Pentium M
 > processor. Fujitsu already replaced the motherboard (including CPU)
 > and the fan & heatsink, but still the problem exists. I am wondering
 > if the problem could be the hard drive, especially when the erratic
 > operation seems to coincide with hard drive operations.
 > The hard drive has that new S.M.A.R.T. interface that reports
 > temperature. I am wondering if the Pentium M processor is giving a
 > false signal to the fan controller chip based on faulty temperature
 > data from the hard drive via the S.M.A.R.T. interface.
 > I would appreciate any info on this. Thanks

On my Acer TravelMate 800 (Pentium-M) I noticed that fan activity depends
not on CPU temperature but on CPU activity/load. For example, using
scanner on USB raises CPU load during scan, and that kicks fan on
immediately. This is stupid design from manufacturer, but I learned to
live with it.

Alexei<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

"Alexei Boukirev" wrote in message
 > On my Acer TravelMate 800 (Pentium-M) I noticed
 > that fan activity depends not on CPU temperature but
 > on CPU activity/load.

Well, actually CPU temperature depends on CPU
activity/load. Especially in a portable environment
where pieces are slowed/shut down when not in use
to preserve battery life.

 > For example, using scanner on USB raises CPU load
 > during scan, and that kicks fan on immediately. This is
 > stupid design from manufacturer, but I learned to
 > live with it.

Not clear why this seems "stupid" to you? Would you
rather that it run at full speed/power all the time? Suck
the battery life for nothing? Run the fan all the time? Why?

I see the same effect in my big tower machine at home
that I use for editing video. As soon as I start rendering
DV, I can hear the CPU fan speed increase. When it
slows back down, I know it is done.<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

"Richard Crowley" <rcrowley7.RemoveThis@xprt.net> wrote in news:100t4hadcv0b061
@corp.supernews.com:

 > "Alexei Boukirev" wrote in message
  >> On my Acer TravelMate 800 (Pentium-M) I noticed
  >> that fan activity depends not on CPU temperature but
  >> on CPU activity/load.
 >
 > Well, actually CPU temperature depends on CPU
 > activity/load. Especially in a portable environment

It does, but not directly. Short rare spikes of high activity do not raise
temperature enough that turning fan on should be required. Yet, TM800
turns on fan each and every time.

 >
  >> For example, using scanner on USB raises CPU load
  >> during scan, and that kicks fan on immediately. This is
  >> stupid design from manufacturer, but I learned to
  >> live with it.
 >
 > Not clear why this seems "stupid" to you? Would you

See above.

Alexei<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

techman41973.TakeThisOut@yahoo.com (Sanjay Punjab) wrote in message news:<e8fde247.0401202240.3177967b.TakeThisOut@posting.google.com>...
 > I own a fujitsu lifebook and the fan seems to have a life of its own.
 > Turning on and off randomly and racing periodically. It is hard to
 > explain, but the operating is erratic and my friend's identical laptop
 > does not suffer from this problem. The laptop has a Pentium M
 > processor. Fujitsu already replaced the motherboard (including CPU)
 > and the fan & heatsink, but still the problem exists. I am wondering
 > if the problem could be the hard drive, especially when the erratic
 > operation seems to coincide with hard drive operations.
 > The hard drive has that new S.M.A.R.T. interface that reports
 > temperature. I am wondering if the Pentium M processor is giving a
 > false signal to the fan controller chip based on faulty temperature
 > data from the hard drive via the S.M.A.R.T. interface.
 > I would appreciate any info on this. Thanks

I owned a Toshiba laptop, and a BKPW Note Pro like this, its a
"feature" that keeps the processor at the same temperature while
maximising battery life (reducing the fan speed during low CPU usage,
and increasing it when the CPU is being thrashed, such as a lot of
3D).

You might try looking in the BIOS for anything like "Fan mode
settings" or similar, sometimes they give you the option of turning
off the demand based fan speed.

-A<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

On Wed, 21 Jan 2004 05:26:24 -0800, "Richard Crowley"
<rcrowley7.RemoveThis@xprt.net> wrote:

 >"Sanjay Punjab" wrote ...
  >> I own a fujitsu lifebook and the fan seems to have a life of its own.
  >> Turning on and off randomly and racing periodically. It is hard to
  >> explain, but the operating is erratic and my friend's identical laptop
  >> does not suffer from this problem. The laptop has a Pentium M
  >> processor. Fujitsu already replaced the motherboard (including CPU)
  >> and the fan & heatsink, but still the problem exists. I am wondering
  >> if the problem could be the hard drive, especially when the erratic
  >> operation seems to coincide with hard drive operations.
  >> The hard drive has that new S.M.A.R.T. interface that reports
  >> temperature. I am wondering if the Pentium M processor is giving a
  >> false signal to the fan controller chip based on faulty temperature
  >> data from the hard drive via the S.M.A.R.T. interface.
  >> I would appreciate any info on this. Thanks
 >
 >The processor (Pentium or whatever) only does what it is
 >told by the people who wrote the BIOS code, and the BIOS
 >does what it thinks is best based on the information it thinks
 >it is getting from the temperature sensors, etc.

Period. And it makes no sense to run the fan more than necessary. Do
you and your friend use the laptop identically? Do you run the same
programs, always having it resting on a hard surface so that vents are
not randomly blocked, say, by sitting in your lap? How attentive is
your friend to his computer? Does his "No, my computer never does
that" really mean it happens in situations where there are enough
distractions that he would never notice? Sit the two computers side
by side and perform a controlled experiment, if his computer really
*is* identical (same GHz, same memory, same disk)? No, you already
said the disk drives are different, and yes, disk operation can affect
the fan.

Nice of Fujitsu to replace all that stuff. No big surprise that it
didn't change anything. Find something else to worry about.

RM



Seems unlikely
 >that they would run what is essentially the CPU fan based on
 >heat information from the hard drive, but maybe it is designed
 >to provide cooling to both?
 >
 >Is this really a serious problem? What do Fujitsu say? If they
 >have already replaced the mobo/CPU I'd assume that you would
 >either have to live with it or sell/trade it if it really bugs you.
 ><!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

Howdy!

"Alexei Boukirev" <aboukirev.RemoveThis@blah.ameritech.net> wrote in message
news:Xns947770A35796Caboukirevblahamerite@216.196.105.138...
 > "Richard Crowley" <rcrowley7.RemoveThis@xprt.net> wrote in news:100t4hadcv0b061
 > @corp.supernews.com:
 >
  > > "Alexei Boukirev" wrote in message
   > >> On my Acer TravelMate 800 (Pentium-M) I noticed
   > >> that fan activity depends not on CPU temperature but
   > >> on CPU activity/load.
  > >
  > > Well, actually CPU temperature depends on CPU
  > > activity/load. Especially in a portable environment
 >
 > It does, but not directly. Short rare spikes of high activity do not
raise
 > temperature enough that turning fan on should be required. Yet, TM800
 > turns on fan each and every time.

So it should wait to turn it on? The fan control logic doesn't know
how long your task will run - and it could easily be a design criteria to
keep the temp down as low as possible concurrent with decent battery life.

Doesn't sound so bad to me,honestly.

RwP




----== Posted via Newsfeed.Com - Unlimited-Uncensored-Secure Usenet News==----
<a style='text-decoration: underline;' href="http://www.newsfeed.com" target="_blank">http://www.newsfeed.com</a> The #1 Newsgroup Service in the World! >100,000 Newsgroups
---= 19 East/West-Coast Specialized Servers - Total Privacy via Encryption =---<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed .. 

"Ralph Wade Phillips" <ralphp.DeleteThis@techie.com> wrote in
news:400f55a0_2@127.0.0.1:

 > Howdy!
 >
 > "Alexei Boukirev" <aboukirev.DeleteThis@blah.ameritech.net> wrote in message
 > news:Xns947770A35796Caboukirevblahamerite@216.196.105.138...
  >> "Richard Crowley" <rcrowley7.DeleteThis@xprt.net> wrote in news:100t4hadcv0b061
  >> @corp.supernews.com:
  >>
   >> > "Alexei Boukirev" wrote in message
   >> >> On my Acer TravelMate 800 (Pentium-M) I noticed
   >> >> that fan activity depends not on CPU temperature but
   >> >> on CPU activity/load.
   >> >
   >> > Well, actually CPU temperature depends on CPU
   >> > activity/load. Especially in a portable environment
  >>
  >> It does, but not directly. Short rare spikes of high activity do not
 > raise
  >> temperature enough that turning fan on should be required. Yet,
  >> TM800 turns on fan each and every time.
 >
 > So it should wait to turn it on? The fan control logic
 > doesn't know
 > how long your task will run - and it could easily be a design criteria
 > to keep the temp down as low as possible concurrent with decent
 > battery life.
 >
 > Doesn't sound so bad to me,honestly.
 >

It should turn on when temperature reaches certain level.
To prove my point let's see a different situation - medium CPU load for a
long time raises temperature significantly (much higher than after that
short spike of load), yet fan does not turn on. Is that a good design?
No way.

The idea is to keep temperature down, so that should be a controlled
parameter. You do not control engine temperature and cooling by
measuring how fast the car goes, you do it with a thermal sensor.

Alexei<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: Does the Pentium M Processor indirectly control fan speed ..