Fecha de publicación: 14/01/1997 Fuente: Wipo "digitalization" A analog calibration signal including at least one frequency component is generated by a very pure signal source, such as a digital oscillator and a digital-to-analog converter (DAC) that has been calibrated to be ultralinear. The analog calibration signal is converted by an analog-to-digital converter (ADC) to a digital signal. The digital signal is digitally compensated in accordance with compensation coefficients to produce a compensated digital signal. The compensated digital signal is digitally processed to isolate and measure distortion components, and the compensation coefficients are adjusted in response to the distortion components in order to reduce the distortion components. Feedback causes the distortion components to be minimized so that the compensation coefficients correct the nonlinearity in the analog-to-digital converter. Preferably the analog calibration signal is a two-tone signal, the digital signal is compensated by computing a polynomial, the compensation coefficients are coefficients of the polynomial, and the distortion components are intermodulation components. After calibration, the calibration coefficients can be adjusted to provide a specified non-linear transfer function. The method can also be used to compensate for nonlinearity in analog signal conditioning circuits that condition the analog signal prior to conversion. Moreover, the calibrated ADC could be used to re-calibrate the DAC, which could be used to re-calibrate the ADC, for enhanced linearity.