Factorizational synthesis of SAW bandpass filters

The design of surface acoustic wave (SAW) filters with linear phase (LP), prescribed nonlinear phase (NLP), minimum phase (MP), and optimum minimum phase (OMP) characteristics by factorizing (splitting) the filter frequency response is considered. A universal factorization algorithm based on Z-transform root searching, lexicographical sorting, and sharing between the transducers is proposed. A feature of the algorithm is that, irrespective of the overall phase specifications, both weighted IDTs tend to have MP characteristics that are less sensitive to the tap-weighting inaccuracies. Design examples of LP, NLP, MP, and OMP SAW filters are presented which confirm the flexibility and efficiency of the design procedure. Given the design specifications, the factorization design leads uniquely to minimum-length SAW filters that are usually 20-30% shorter compared to the conventional design where two identical weighted IDTs are used