Abstract—This paper presents a capacitor mismatch calibration technique in multibit discrete-time sigma-delta modulators
based on a capacitor error model, including nonideal integrator gain errors. This model enables the compensation of mismatch-induced nonlinear memory errors in conversion using a simple finiteimpulse-response structure. Single-bit pseudorandom noise (PN) is utilized to identify the error coefficients, and an analog-domain PN removal technique is devised to minimize the input signal dynamic-range loss due to the PN circulation in the loop. The behavioral simulation demonstrates that the proposed scheme effectively compensates for the multibit capacitor mismatch errors in the first- and second-order modulators