Thanks BC for your reply.
Yes, the temp and fan speed readings are from the BIOS.
Thanks for the links.
Yes this problem started in the heat of the Summer but the fan didn't slow down when the temps went down at night.
Of course I have to make some assumptions about details that I don't know for sure what the facts are. I tried to find out what the expected behavior should be for RPM vs. temp. I was not able to find that out. From my experience with small speed controlled motors (radio controlled airplanes use CD ROM drive motors that have rewound coils to accept higher currents) I believe that 4300 rpm is FULL speed. If the controller in this computer is capable of incremental (not just two speeds - low and high) control, I should be hearing some modulation of the speed, not just a constant 4300 rpm.
Another assumption is that 27 C should not be hot enough to make the fan run at full speed. Does anyone know what the temp vs. speed specs should be?
Another assumption that I am laboring under is: the temp sensor for the CPU is a solid state junction that is part of the CPU wafer.
From those assumptions I would conclude that the 27 deg C that the BIOS sees, tells me that the CPU is not hot,and along with the ambient temp (about 20 deg C) and therefore the problem is probably with the controller circuit. This would rule out problems with the heat conductive paste between the CPU and the cooling fins.
Does anyone know where the controller circuit is? I would guess it is on the motherboard somewere.
This all leads me to think that the only fix is to replace the MB - eek.
Well I guess that noise canceling head phones would be another solution, eh? LOL
Okay, somebody please poke holes in my line of reasoning. Or better yet tell me that there is a way of fixing the controller - not likely, eh?