To achieve a high IIP2 level on a mixer, static calibration techniques have been developed. Most of them are based on an intentional introduction of a calibrated mismatch in the structure of the mixer. They are performed at production stage. It is also possible to automate them but their activation is strongly limited in portable devices because of system restrictions. Furthermore, IIP2 is sensitive to system variations, thereby degrading the calibration operation. The challenge is so to make the calibration system dynamic, i.e. performing an online calibration. This paper presents a perturbance-based algorithm as part of an automatic calibration system to track the optimum IIP2 level. Measurements validate the algorithm behavior and indicate the feasibility of using it in a complete calibration system for a future on-chip implementation.
