Calibration is about being able to make meaningful, confident measurements
which are documented so they can be traced to national standards. Any manufacturer,
or in the case of the calibration laboratory, instrument repairer has to
be able to ensure that work is carried out to the required standards. Design
and development of any product depends on being able to prove /measure
performance characteristics. The reliability and stability of instrumentation
used for this purpose must be proved.
However the scene is relentlessly changing. New issues arise e.g. the
concern about electronic circuits interfering with one another, this is
known as electromagnetic compliance (EMC). More stable standards are continually
being developed by physicists at national laboratories. This reflects in
commercial instruments whose manufactures often use better specifications
as a selling point.
Occasionally the question gets asked why is all this accuracy necessary
? particularly when modern control systems are designed to accommodate
uncertainty in sensors by using artificial intelligence. Potentially thousands
can be saved by asking simple questions like " what do we need to measure
?" and "to what accuracy ?" Many users of measuring equipment prefer
to ignore this issue by insisting upon equipment being inside its commercial
specification when returned from a calibration facility.
This entails adjustment which may take some time to settle. User's
insisting upon adherence to the commercial specifications have to shoulder
the expense of sending the instrument to the manufacturer and then back
to the calibration laboratory for re-calibration. Here it is hoped that
there is compliance between the measurements made by the manufacturer and
the calibration laboratory. Effectively two calibrations and one adjustment
with a lot of transit costs. Asking for measured values is a more enlightened
approach provided the instrument is not too far out-of-tolerance.