Technidyne Header Image

Monday, September 15, 2014

Why have multiple ranges of calibration standards?

Whether it is a balance, caliper, brightness, tensile, moisture, consistency meter, many people calibrate equipment to only one range of measurement.  This is simple, however, it does not verify that their calibration is correct over a range.  It may be as simple as a zero-range in addition to another range, but the more ranges that we can verify the better.

As TAPPI Test Method T 452 (Brightness of pulp, paper and paperboard) states:
  • "Obtain reflectance standards at monthly intervals and calibrate the test instrument with them as soon as they are received."
  • "Check the test instrument readings at least monthly against the assigned values of all the standards."
Graph 1 - Good Calibration
Notice that it says to check instrument readings against the assigned values of ALL the standards. This is to ensure that their is proper linearity in the calibration.

In Graph 1, the calibration was performed with the 70-range standard. Then it was compared against the other standards at the different points e.g. 50, 60, 80 and 90. It is right on all of these points. This verifies calibration across a large range.

Graph 2 - Bad Calibration Using 70-range
In Graph 2, the calibration was again performed with the 70-range standard. Then it was compared against the other standards at the different points e.g. 50, 60, 80 and 90. It only matches at the 70-range point. The measured values show a linear response, but it is off at every range other than the 70.

Graph 3 - Bad Calibration Using 80-range
Many customers will choose to do the primary calibration in the range in which they normally measure.  In Graph 3, the calibration was performed with the 80-range standard. Then it was compared against the other standards at the different points e.g. 50, 60, 70 and 90. It only matches at the 80-range point. The measured values show a linear response, however, it does not match at any range other than 80.

If single-point calibration is used and there is no means to verify the linearity of the measurement, we may run into a situation like Graph 2 or 3. In these cases the instrument responds in a linear manner, but the calibration is not correct.  This could be due to a number of other issues such as spectral response, failing electronics or failing optics, to name a few.

Graph 4 - Bad Calibration with Non-linear Response
If the instrument is responding poorly enough, the values could be questionable even over a very tight range around the calibration range. If we calibrated to the 80-range, but were measuring in the high 80's, the values may not be reliable. 

Also, as seen in Graph 4, the calibration was performed with the 80-range standard. Then it was compared against the other ranges.  This example shows a non-linear response. This would not be seen if only the 80- and 50-range standards were measured.

ALWAYS use more than one range to verify your measuring devices. If you see responses like those in Graphs 2, 3 and 4:
   1. Recalibrate you device
   2. Get fresh calibration standards (or get them recertified)
   3. Contact a factory-certified service technician for support