Quote:
Originally Posted by harlequin2
That should be 5/9 not 5/8. Typo? So that's 1.25% conversion error. Now, if you had used proper units in the first place, the conversion error would be from Celsius to Fahrenheit (:-)
|
Yes, was a typo. Happens when I try to squeeze these posts in during work
Won't even get into the C vs F debate. Suffice to say, 49 years raised on the Fahrenheit system is a hard habit to break. Everything from home thermostats to the weather reports are in F here in the states. And this did start as a personal project after all
If you want to eliminate the 1.25% conversion error you could always redo the lookup tables for C. Don't know if it's really worthwhile though in light of the other tolerance errors.
A very popular temperature IC that I and other engineers have used over the years is the national semi LM50. That was labeled as "precision" and had an error of +/-2degC at room temp and +/-4degC over full scale.