17.1 Half-Octave Bandpass Filter Design Program

The program HALFOCT.EXE converts the five-element, all-pole, equally loaded (1-ohm) lowpass prototype filter shown in Figure 17.1A, with a bandwidth of 1.0 rad/sec, into the classical bandpass filter shown in Figure 17.1B, and then into the filter of Figure 17.1C. The program uses the Norton transformation in order to control the values of the inductors. This overcomes a common problem, that these values often vary over a wide range and can be unreasonably large or small for the particular frequency range of the filter. Large values have too much self-capacitance, and small values require very large values of tuning capacitance and tend to have low Q at low frequencies.

Figure 17.1: The half-octave filter: (A) lowpass prototype; (B) prototype bandpass filter; (C) filter after the Norton transformations.

For correct operation, the inductor values for L2 and L4 should be the "effective" values, measured at the geometric mean of the passband. These values should not change much due to self-capacitance effects across the passband, therefore coils with large distributed C should be avoided. The self-capacitances of L1, L3, and L5 are part of C1, C6, and C11, respectively. C1, C6, and C11 should therefore be...