Forum Professor
Joined: 13 Apr 2007
Posts: 1247
Location: Pennsylvania
|
As I understand the two terms, the tolerance is a parameter used by the manufacturer, and the uncertainty is the accuracy of the thermometer in its end use, considering all sources of error.
As an example, a manufacturer of resistance temperature detectors might have a criteria that the RTD has a resistance of 100 ohms plus or minus some value, at a given temperature, and the linearity must be within certain limits. This is what is meant by tolerance.
When the end user places that RTD in service, he will have to consider a lot of other factors. How accurate is the electronic measuring circuit in which the RTD has been connected? What is the insulation resistance of the cables? What is the error due to stratification of the process fluid he is measuring? Is there any instrument drift? Maybe he can have the RTD calibrated over its range of measurement, and he can compensate for the nonlinearity. Then he might even end up with less uncertainty than the manufacturer's tolerance. |
|
