IIR filter scaling for real-time signal processing

Students generally have difficulty implementing infinite impulse response filters (IIR) using fixed-point arithmetic. Most of the trouble is with amplitude scaling the filter and representing the coefficients in the fixed-point format. The authors present a systematic procedure for scaling a second-order IIR filter which is used as the basic building block of higher-order filters. A simple simulation program is used to estimate the size of the signals at the summing nodes of the filter section. Once the estimates are known, the filter section can he reliably scaled. The procedure is clarified by implementing a second-order resonant filter on a digital signal processor: the TMS320C25.<
>