In an effort to resume previous work [A, B, C, D] on developing a crystal oven for radio frequency transmitter / receiver stabilization purposes, the first step for me was to create a device to accurately measure and log temperature.
Check out the video after the break.
Are we confusing uncertainty and resolution?
Being able to measure the temperature down to 0.01F isn’t much use if the calibrated temperature error is around 0.3C according to the datasheet.
If you are stabilizing a crystal, you probably care much more about temperature drift than the absolute temperature accuracy. In that case, the more resolution the better, even if the absolute calibration is not great.
You might enjoy looking at Jim Williams’ Application Note 12 October 1985
Circuit Techniques for Clock Sources
http://cds.linear.com/docs/en/application-note/an12fa.pdf
Some help on this subject and additional system considerations.
http://www.edn.com/design/analog/4368344/High-accuracy-temperature-measurements-call-for-PRTDs-and-precision-delta-sigma-ADCs