A novel large input range source-follower based filter architecture

Abstract

Abstract:A novel large input range source-follower based filter architecture is proposed offering a increased 1-dB compression point, without increased power consumption. An alternative feedback mechanism enables single-ended use and simplifies the bias scheme. Simulation has confirmed a 250 MHz second order filter stage consuming 40 uA @ 1.2 V supply in a 0.13 um CMOS technology with 1-dB compression point at a differential peak to peak input amplitude of 1.4 V, more than doubling the input range of previous implementations with similar power dissipation.

A large input range source-follower based bi-quad filter cell

“…The filter offers 2 complex poles and 2 complex zeros in a single biquad stage, without increasing the power consumption with respect to previous work [1,2,3]. A 4th order elliptical filter was designed based on this new topology and simulations were performed to confirm correct…”

“…In [2], the authors modified the topology presented in [1] to increase linearity for large input signals, increasing the dynamic range while maintaining the same lower power consumption (see Fig. 1.c).…”

“…1.c). The topologies of both [1] and [2], however, clearly lack the possibility to offer zeros in their transfer functions, hindering implementation of e.g. Chebyshev II and elliptic filter stages.…”

“…Furthermore, tuning the transconductance of pMOS and nMOS transistors separately is hard to realize in the improved topology of [2]. This letter proposes a new source-follower based filter architecture with large input range, that offers complex zeros, without relying on transconductance inbalance.…”