overheating Thinkpad X60s with 7.0-RC1

[Johannes Dieterich]

On 1/6/08, Jeremy Chadwick <koitsu at freebsd.org> wrote:
>
>
> All temperatures were compared when the systems were idling.
>
> The variance in temperatures was astounding.  In some cases, there was a
> almost a 20C difference between two machines, especially around the GPU
> area.  In other cases, batteries were reporting insane temperatures on
> one laptop (over 90C), while on another well within scope (~25-26C).

Sounds crazy...


My particular laptop is a problem child when it comes to noise -- that
> is to say, within literally seconds of the machine finishing POST, the
> fan kicks on, then 15-20 seconds later, increases speed.  My co-workers'
> laptops do not have this problem.  Cleaning out the heatsink area using
> a can of air made no difference.

It is rather clean, I'd say. Just recently I changed the keyboard from
German version to US and then tried to blow off all the dust.


I won't even bother mentioning what happens when I run something that's
> CPU or GPU intensive.  I haven't had any crashes, but in some cases,
> I've seen the GPU temperatures reach over 80C -- completely
> unacceptable, and bordering on insane.  It's gotten to the point where
> to use my T60p *quietly*, I'm forced to prop the rear corners up on
> little blocks or whatever, and then place a desk fan nearby, blowing
> cold air more or less underneathe the laptop.  This keeps the fan in low
> speed mode, which is semi-tolerable.

Unfortunately this is not an option for me since I am running it at home in
the docking station.


I *haven't* had any crashes or random system lockups, but many other
> co-workers of mine have, and it's safe to say heat is the cause.
>
> In my opinion, most of these laptops (the T60p series, and very likely
> related models!) are being assembled with improper amounts of thermal
> paste or TIM pads, without proper surface area contact.  Apple recently
> had a case of this happening as well with their Macbook Pros, where
> their assembly documentation stated they should use an *entire tube* of
> thermal paste between the CPU and the heatsink.

I don't know why it should then show right now. I mean: could it somehow?
Within those 1.5 years of having it, it ran Linux and FreeBSD. And
always without a problem. Until now. And I really haven't touched
anything in the ACPI configuration.



Lenovo should be ashamed at the lack of quality control used when these
> things are built.  Again, this is a laptop given to me by my workplace
> for work, so it's really not my choice (nor can I disassemble it to
> examine or fix what the problem might be) -- but if I ever am to buy a
> laptop for personal use, Lenovo would not be on my list of vendors.

I do agree with you that the quality has decreased dramatically since the
IBM/Lenovo deal, IMHO. Still they are the better ones on the market, IMHO.
Anyway...


I would urge those here to consider booting XP somehow (if possible) and
> running tp4xfancontrol to check actual temperatures, since FreeBSD's
> h/w monitoring capability is spotty at best (I think Linux wins out
> here, but at least there's room for growth...)

Sorry. XP is killed completely and I would need to fix myself a HDD from
somewhere and install it from scratch...

What remains for me, is that it has never been a problem. Until 7.0. And if
there is no hardware failure coincidence, it must be a mysterious (at least
for me... ;-)  ) software/configuration problem.

BTW: how certain can I be that the reported temperatures are REAL
temperatures?

Regards,

Johannes