On the Calibration of DA Converters Based on $R/\beta R$ Ladder Networks

In this paper, we discuss the possibility of realizing high-accuracy digital-analog(DA) converters by exploiting the properties of R/betaR (beta > 2) ladder networks. Extensive simulations demonstrate that, if a proper value of beta is chosen, high accuracy can be obtained by resorting to a very simple self-calibration algorithm. This is true, even in the case of large tolerance values for the resistors that make up the ladder network. Such a result may allow the design of a high-accuracy very low cost converter. The effect of the offset of the comparator needed for the self-calibration algorithm is also discussed.