Calibration Techniques in Phased-Locked Loops: Theoretical basis and practical applications

Digitally assisted analog circuits are becoming the mainstream solution in modern CMOS processes as they improve circuit performance and reduce implementation costs [1]. The most representative example of digital assistance is the compensation of the nonlinearity of an analog block, which is done by applying the digital inverse of the nonlinearity at the output. However, as the digital inverse is not known a priori, it has to be estimated, and the estimation needs to adapt in the background to the environmental conditions. This is typically accomplished by relying on adaptive filtering, i.e., by closing a system-identification engine in feedback that injects a modulation at the input of the nonlinear block. This article, after reviewing the basics of both PLLs and adaptive filters, will explore the most significant calibrations adopted in PLLs.