Guys,
As an enigneer, the use of proper terminology is paramount, which defines how text and specifications are written in technical manuals, and keeps discusions from being misunderstood. Terminology is simply a fundamental part of language which transcends all topics. Reading through this thread, the nature of your questions and comments are obvious, but some of the terminology is incorrect and misleading.
The title of this thread "How do you measure your CPUs temp?" is inappropriate, and confusing. CPU temperature (Tcase) can be measured by SpeedFan and Hardware Monitor, which are the only correct answers in the poll, since Core Temp and Real Temp only measure Core temperatures (Tjunction).
If you read the Real Temp documentation -
http://www.techpowerup.com/realtemp/docs.php - then you know that Real Temp offers 2 calibration points; Tjunction Max and Tjunction Idle. The author of Real Temp,
unclewebb, has performed extensive research, testing and analysis on key Core 2 processor variants in order to create accurate Tjunction Max tables, so even though a Tjunction Max calibration feature is provided, the default value shown is already correct. The authors of Core Temp and Everest haven't performed any testing. They simply adopt Intel's documented Tjunction Max tables for mobile variants, which differ in thermal behavior from desktops, and in some instances, aren't even close.
Real Temp is the only utility that provides a separate Idle calibration to compensate for known problems with sensor linearity, since Intel's Digital Thermal Sensors were never intended for temperature monitoring from Idle to Load. The DTS sensors are only designed to provide accurate thermal protection for Throttle (overtemp) and Shutdown. This shortcoming is more obvious on 45 nanometer processors, which are often reported with DTS sensors which "stick", and won't report actual Idle temperatures below a certain value. Only the Tcase sensor (CPU temperature) is linear from Idle to Load, and can be accurately calibrated, as well as Tjunction (Core Temperatures), using my Core 2 Quad and Duo Temperature Guide. A detailed explanation of the differences between these sensors is in the Guide.
If you use Real Temp's default Tjunction Max value and perform the Idle calibrations, then Real Temp will be "dead on". To further validate the accuracy of Real Temp, consider the following; since the techniques I developed in my Temp Guide for calibrating SpeedFan don't use Tjunction Max values to achieve accuracy, then why, after SpeedFan calibrations have been completed, does the Core temperatures agree with Real Temp within a degree or two? Check it out for yourself. Try my SpeedFan calibrations then compare your Core temperature results with Real Temp.
And yes, the B3 is a pair of B2 Cores, so Tjunction Max is correct at 85c, which is Real Temp's default Tjunction Max value for the Q6600 B3.
I hope this explanation helps to clarify and answer your questions.
Comp