question about T43 fan noise

[danda821]

I just talked with a IBM technical support on the phone. I told him I can clearly hear the fan noise in normal sitting position, the other machine (T42) I have does not have such kind noise. He said it should not be like that, the fan should not borther you for normal use. He can send me a new fan or I can return and get an exchange of the machine. Does anyone have similar experience? Will a new fan or a new machine help? My machine is 2668-AJU , which is a CTO model. Thank you.

[GomJabbar]

Look right above at the Sticky: ***FAN CONTROL UTILITY***. Why do you think it was made a Sticky?

EDIT: I just checked, and in less than a month since the original thread was started by Shimodax, it has been viewed over 20,000 times Must be a new record.

[danda821]

Thannk you. I have tried Shimodax's tool. It works. But still T43 is not as quiet as T42.

[christopher_wolf]

On my T43 using the Fan Control Utility, I have gotten my T43 a tad quieter than my friend's T42.

[pipspeak]

The Fan Control utility is indeed useful but does not solve the problem, not least because no-one seems to have discovered exactly how fan speed is set by the machine.

For example, my T43 fan usually runs at level 2 all day on bios control, which is quiet enough and keeps CPU and GPU temps under about 47 degrees. Occasionally, and for no discernable reason, the fan speed will kick up to level 4 despite the fact that no sensor temperatures (incl. GPU and CPU) are over 43 degrees. It will then stay at level 4 the rest of the day even if sensor temps drop under 40 unless I intevene with the fan control utility.

It's very odd behaviour, seemingly random, really annoying, and I wish Lenovo would tell someone how the heck fan speed is set!

[christopher_wolf]

We can only try and RE how that is done; I actually think the fan uses a PID controller and some rather aggressive gains to achieve a quicker *rate* of cooling rather than steady state as it does seem to be approach some sort of a limit; at least, that is what I think the "fan pulsing" could be an indicator of. It is almost as if it switches into PD control, all of a sudden, then overshoots the steady-state, throttles down on the next query cycle, then over-compensates, etc. I have seen that kind of behavior in cooling system before, although they weren't for computers or Thinkpads(lab cold rooms, had a problem with one last month, same pulsing behavior). Without a good layout of the logic setup that is used, we can't really be sure. The Fan Control Utility, does however, seem to mange it quite well.