1. Purpose and scope
Mechanical structures vary more from part to part than most electronic circuits. A spring a little wider, a gap a little smaller or a little more package stress changes a sensor's offset and sensitivity. Rather than reject every part that differs, sensors are calibrated one by one: each unit is stimulated with known inputs, its response measured, and coefficients computed and stored on the chip so the output is corrected. Calibration therefore becomes part of the product, and its method, equipment and coefficients are controlled like the design.
This stage defines that method. It decides which parameters are calibrated, at how many temperatures and stimulus points, at wafer level or after packaging, with which model, and how the coefficients are quantised into the trim registers. It specifies the stimulus equipment, with its accuracy traceable to references, and verifies that the compensation in the ASIC or firmware reproduces the fitted result. Calibrated parts must meet the accuracy targets across temperature, and the residuals after calibration must be understood.
In scope: the calibration strategy decision, the calibration model, the calibration procedure, the calibration stimulus equipment specification, compensation algorithms, and the characterisation data the model is fitted on. Out of scope: the production calibration program and its release, which are completed in Test & Characterisation.
12 more sections for registered members.