A top-down optimization methodology for SC filter circuit design

Email Print Request Permissions The manual design of Switched-Capacitor (SC) filters can be a laborious process. When these filters use low gain amplifiers or voltage followers instead of high gain opamps, this task becomes even more complex due to the loss of the virtual ground node, requiring the compensation of the parasitic capacitances during the design phase. This paper proposes an automatic procedure for the design of SC filters using low gain amplifiers, including the automatic compensation of the parasitic capacitances. This design methodology is based on a Genetic Algorithm (GA) using hybrid cost functions. The cost function first uses equations to estimate the filter transfer function, the gain and settling-time of the amplifier and the RC time constants of the switches. This reduces the computation time, thus allowing to use large populations to cover the entire design space. Once all specifications are met, the GA uses transient electrical simulations of the circuit in the cost functions, thus resulting in a high accuracy in the determination of the filter's transfer function and allowing to accurately compensate the parasitic capacitances and obtain the final design solution within a reasonable computation time.