2.6.26: thinkpad x60 is cooled passively

2.6.26: thinkpad x60 is cooled passively

[Pavel Machek]

Hi!

While verifying that -rc2 regression is gone, I noticed alarming
stuff:

1) thinkpad x60 relies on passive cooling. You have nice 1.8GHz cpu,
which goes away as soon as you load it.

2) after some time it decides that 100% cpu in 1.0GHz is fine for you,
and it takes quite long to recover.
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: 2.6.26: thinkpad x60 is cooled passively

[Theodore Tso]

On Wed, Aug 13, 2008 at 10:06:18AM +0200, Pavel Machek wrote:
> While verifying that -rc2 regression is gone, I noticed alarming
> stuff:
> 
> 1) thinkpad x60 relies on passive cooling. You have nice 1.8GHz cpu,
> which goes away as soon as you load it.
> 
> 2) after some time it decides that 100% cpu in 1.0GHz is fine for you,
> and it takes quite long to recover.

There is no fan in the X60?  I didn't think that was the case.  I do
know the fans have a tendency to go bad within 12-18 months, and so
the CPU heat sink and fan unit needs to be replaced perdiocially.
What does /proc/acpi/ibm/fan report? 

Or are you saying that the fan isn't keeping keeping up with the heat
output of the CPU?  Are you sure this isn't due to a hardware problem?
As mentioned previously on this list, Lenovo has had a relatively bad
track record with fan failures and with incompetent installation of
the heat sink with bad applications of the thermal compound; it's the
one failings of an otherwise very nice laptop design.

    	     	   	     	       	      - Ted

Re: 2.6.26: thinkpad x60 is cooled passively

[Henrique de Moraes Holschuh]

On Wed, 13 Aug 2008, Theodore Tso wrote:
> On Wed, Aug 13, 2008 at 10:06:18AM +0200, Pavel Machek wrote:
> > While verifying that -rc2 regression is gone, I noticed alarming
> > stuff:
> > 
> > 1) thinkpad x60 relies on passive cooling. You have nice 1.8GHz cpu,
> > which goes away as soon as you load it.
> > 
> > 2) after some time it decides that 100% cpu in 1.0GHz is fine for you,
> > and it takes quite long to recover.
> 
> There is no fan in the X60?  I didn't think that was the case.  I do

It has a fan, yes.

> know the fans have a tendency to go bad within 12-18 months, and so
> the CPU heat sink and fan unit needs to be replaced perdiocially.
> What does /proc/acpi/ibm/fan report? 

ThinkPad fans have a overdesign range of about 2x, they almost never stop
because they're too old.  Clogged in dirt, on the other hand... a clogged
fan often spins FASTER and cools less.

Want to see how much your fan has as far as a safety margin goes? use
thinkpad-acpi to measure RPMs in level 7 and also when you switch PWM off
(i.e. set it to 100% duty cycle).  Mine gets faster by 1500RPM or so that
reflects the extra torque the EC can apply to get the fan to run at the
speed it wants it to.

But the thermal sink interface to the CPU and GPU are sometimes crap [when
compared with the quality of the rest of the machine].  Open it, remove the
crap, and replace it with a very high quality compound.  Dependin on
ThinkPad model, you also need to repair thermal sink design defects to make
it fit properly (read: press really tightly) on all chips.

> As mentioned previously on this list, Lenovo has had a relatively bad
> track record with fan failures and with incompetent installation of
> the heat sink with bad applications of the thermal compound; it's the
> one failings of an otherwise very nice laptop design.

Indeed.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: 2.6.26: thinkpad x60 is cooled passively

[Theodore Tso]

On Wed, Aug 13, 2008 at 05:31:37PM -0300, Henrique de Moraes Holschuh wrote:
> > know the fans have a tendency to go bad within 12-18 months, and so
> > the CPU heat sink and fan unit needs to be replaced perdiocially.
> > What does /proc/acpi/ibm/fan report? 
> 
> ThinkPad fans have a overdesign range of about 2x, they almost never stop
> because they're too old.  Clogged in dirt, on the other hand... a clogged
> fan often spins FASTER and cools less.
> 
> Want to see how much your fan has as far as a safety margin goes? use
> thinkpad-acpi to measure RPMs in level 7 and also when you switch PWM off
> (i.e. set it to 100% duty cycle).  Mine gets faster by 1500RPM or so that
> reflects the extra torque the EC can apply to get the fan to run at the
> speed it wants it to.

Hmm, maybe I've just had bad luck, or I live in a very dusty
environment (four cats, including two longhairs in perma-shed mode,
will do that to any house), but I've had at least 3 fan failures in
the past 5-6 years where the fan stopped entirely, and had to be
replaced.  When I took it out, it was quite dirty, but the way it spun
why I tried moving the fan blades with my finger, it felt just like
the failure modes I've had with standard power supply fans that have
gone bad due to old age.

       	    	      	       	      - Ted

Re: 2.6.26: thinkpad x60 is cooled passively

[Pavel Machek]

On Wed 2008-08-13 16:43:38, Theodore Tso wrote:
> On Wed, Aug 13, 2008 at 05:31:37PM -0300, Henrique de Moraes Holschuh wrote:
> > > know the fans have a tendency to go bad within 12-18 months, and so
> > > the CPU heat sink and fan unit needs to be replaced perdiocially.
> > > What does /proc/acpi/ibm/fan report? 
> > 
> > ThinkPad fans have a overdesign range of about 2x, they almost never stop
> > because they're too old.  Clogged in dirt, on the other hand... a clogged
> > fan often spins FASTER and cools less.
> > 
> > Want to see how much your fan has as far as a safety margin goes? use
> > thinkpad-acpi to measure RPMs in level 7 and also when you switch PWM off
> > (i.e. set it to 100% duty cycle).  Mine gets faster by 1500RPM or so that
> > reflects the extra torque the EC can apply to get the fan to run at the
> > speed it wants it to.
> 
> Hmm, maybe I've just had bad luck, or I live in a very dusty
> environment (four cats, including two longhairs in perma-shed mode,
> will do that to any house), but I've had at least 3 fan failures in
> the past 5-6 years where the fan stopped entirely, and had to be
> replaced.  When I took it out, it was quite dirty, but the way it spun
> why I tried moving the fan blades with my finger, it felt just like
> the failure modes I've had with standard power supply fans that have
> gone bad due to old age.

Okay, seems like vojtech fixed my notebook with a bit of compressed
air... a lot of dust came out, and machine works okay now.
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: 2.6.26: thinkpad x60 is cooled passively

[Lennart Sorensen]

On Mon, Sep 08, 2008 at 03:29:12PM +0200, Pavel Machek wrote:
> Okay, seems like vojtech fixed my notebook with a bit of compressed
> air... a lot of dust came out, and machine works okay now.

Amazing what that does.

My wife's Asus R1F was giving occational "throtled CPU due to
overtemperature" and I took the cpu fan out and removed the dust from
the heatsink using a q-tip, and the cpu temperature dropped 16C and the
fan is much quieter now.  I will have to remember to do that more often.

-- 
Len Sorensen

Re: 2.6.26: thinkpad x60 is cooled passively

[Jeremy Fitzhardinge]

Henrique de Moraes Holschuh wrote:
> ThinkPad fans have a overdesign range of about 2x, they almost never stop
> because they're too old.  Clogged in dirt, on the other hand... a clogged
> fan often spins FASTER and cools less.
>
> Want to see how much your fan has as far as a safety margin goes? use
> thinkpad-acpi to measure RPMs in level 7 and also when you switch PWM off
> (i.e. set it to 100% duty cycle).  Mine gets faster by 1500RPM or so that
> reflects the extra torque the EC can apply to get the fan to run at the
> speed it wants it to.
>
> But the thermal sink interface to the CPU and GPU are sometimes crap [when
> compared with the quality of the rest of the machine].  Open it, remove the
> crap, and replace it with a very high quality compound.  Dependin on
> ThinkPad model, you also need to repair thermal sink design defects to make
> it fit properly (read: press really tightly) on all chips.

Yes, this seems like a widespread and chronic problem with these little 
Thinkpads.  I've had both an X31 and X60 go bad this way, and a friend 
has had his X60 fail twice.  A pity given how nice the rest of the 
machine is.

    J

Re: 2.6.26: thinkpad x60 is cooled passively

[Pavel Machek]

On Wed 2008-08-13 11:24:31, Theodore Tso wrote:
> On Wed, Aug 13, 2008 at 10:06:18AM +0200, Pavel Machek wrote:
> > While verifying that -rc2 regression is gone, I noticed alarming
> > stuff:
> > 
> > 1) thinkpad x60 relies on passive cooling. You have nice 1.8GHz cpu,
> > which goes away as soon as you load it.
> > 
> > 2) after some time it decides that 100% cpu in 1.0GHz is fine for you,
> > and it takes quite long to recover.
> 
> There is no fan in the X60?  I didn't think that was the case.  I do
> know the fans have a tendency to go bad within 12-18 months, and so
> the CPU heat sink and fan unit needs to be replaced perdiocially.
> What does /proc/acpi/ibm/fan report? 

3700 on level 7, more (like 4700) on disengaged, but still not enough
to keep it from going to 1GHz.

> Or are you saying that the fan isn't keeping keeping up with the heat
> output of the CPU?  Are you sure this isn't due to a hardware
> problem?

Yes, I'm trying to say that the fan isn't keeping keeping up with the
heat output of the CPU.

> As mentioned previously on this list, Lenovo has had a relatively bad
> track record with fan failures and with incompetent installation of
> the heat sink with bad applications of the thermal compound; it's the
> one failings of an otherwise very nice laptop design.

Hmm, ok, maybe. It is true that nearby X60p behaved slightly
differently.
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html