Informative calibration of the instrumentation

The calibration cost of an instrument is so expensive that is necessary, to reduce the number of calibrated points, in the scale. There is the necessity to exploit as well as possible the metrological valence of few calibrated points to extend all over the instrumental range. The choice of the number and the position of such few points is the problem faced in the work. Utilizing the Shannon sampling procedure in the transposed domain of the input entities of the instrument, with respect to the time, the truncation error of the cardinal series has faced utilizing the Helms and Thomas upper bound evaluation. To enhance the efficiency of the procedure, has been utilized the minimax procedure searching the optimal indication on better determination. In every case the efficiency of the procedure is lower on the edge of the instrumental scale. To upgrade the edge calibration, making as even as possible the calibration, upon the scale, one has been utilized the generalized sampling theorem by Papoulis that allows to chose the calibration points more frequent at the edge, with respect of the central part of the scale. An algorithm to guide the calibration procedure is tempted.