Hello Subscriber
Last weekend I've bought a watermelon. The fact itself is not very spectacular, but it's a good example for different aspects of metering.
|
I did a calibration by measuring the weight. The weighting scale has a resolution of 1 g.
|
So we have a measurement uncertainty of +/- 0.444 g. (We skip the remaining decimals because a 1 ppm test result would be OK for us.)
|
Measuring the weight twenty times hasn't shown any difference, so we have zero standard deviation. So, for a measurement uncertainty budget calculation, I would continue with a coverage factor of 2, which means 95% probability that all measurement results are within the same range.
|
Since the supermarket scale was showing 5 grams more and with a valid calibration mark, we have to deal with additional uncertainties.
|
a) my scale is not calibrated
|
b) we have no reference temperature for comparison
|
So I can't use the melon as a transfer standard to calibrate my weight scale.
|
Also, the long time stability of a melon is just so so.
|
Conclusion: Keep your measurement equipment calibrated and traceable to international standards. Only this makes sure, that your measurement results are accurate and repeatable.
|
|
|
|