A linearisation and compensation method for integrated sensors

Abstract A new linearisation and compensation method for integrated sensors is presented. It is based on a reduction of a variable number of input pulses from an integrating analog/digital converter to a fixed number of output pulses. This method can be interpreted as a linearisation of the derivative of the sensor transfer function and results in a simple implementation and in extremely modest memory requirements. It is quasi self calibrating and therefore very well suited for batch calibration in mass production. The algorithm has been implemented in a field-programmable gate array and has been applied successfully to different sensor transfer functions.