The trend towards reconfigurable receivers requires on-chip flexible filters that can replace dedicated, bulky and non-tunable filters (e.g., SAW and BAW [1]). Although BAW filters are compatible with silicon processes, their center frequency is sensitive to thickness variation of the piezoelectric material and the achievable tuneability is limited [1]. Other techniques to make RF on-chip band-pass filters (BPFs) include Q-enhancement, gm-C and N-path. Q-enhancement approach has several disadvantages such as large area due to inductors which do not obey process scaling, limited tuneability and poor dynamic range [2]. Main drawbacks of gm-C filters are the tradeoff between power consumption, quality factor, center frequency and dynamic range and the need for tuning circuitry [3]. Recently there has been renewed interest in the translational impedance conversion of N-path filters [4-6]. Due to the “transparency‿ of the passive mixer, baseband impedance is translated to frequencies around the clock frequency flo [7]. The interesting features of these filters are their direct tuneability with flo, higher quality factor compared to on-chip CMOS LC filters [2], high linearity and graceful scaling with process.
展开▼
