A novel method for stochastic nonlinearity analysis of a CMOS pipeline ADC

An analytic approach is presented for estimating the nonlinearity of an analog to digital converter (ADC) as a function of the variations in the circuit devices. The approach is demonstrated for the case of a pipeline ADC with digital error correction. Under some mild assumptions on the expected variations, the error probability is expressed as a simple explicit function of the standard deviations in the components' parameters: gain errors, comparator offset errors and resistor errors. The analytical expression is verified for Integral Non Linearity (INL), and its limits are studied using Monte-Carlo simulations of a 10 bit pipeline ADC structure.