cause transmission zeros at finite frequencies if the couplings among the resonators are properly designed. From the result of the interstage-coupling effect, the designed BPF has two transmission zeros of finite frequency on the two sides of the passband in order to improve the frequency skirts of the passband edges.
The designed filter was fabricated on the FR4 substrate with filter dimensions 16.8 ϫ 11.6 mm and then measured using an Agilent 8753E Network Analyzer. Figure 5 also shows the measured results of the designed pseudo-interdigital BPF using the tapped I/O and SIR. The designed BPF at the center frequency f 0 ϭ 1.75 GHz has an insertion loss of Ϫ2.5 dB, a bandwidth of 0.32 GHz (with a 3-dB fractional bandwidth ratio of 18.5%), and return losses of Ϫ18.5 dB. The finite transmission zeros occur in the lower side of the passband edge at 1.5 GHz with Ϫ25-dB attenuation and in the higher side of passband edge at 2.25 GHz with Ϫ32-dB attenuation. In addition, the first and second spurious responses are shifted to higher than 3f 0 (designed center frequency) due to the SIR effect and effect of using the tapped I/O and fast passband-edge attenuation.
4. CONCLUSION
In this paper, a compact pseudo-interdigital BPF using tapped I/O and SIR in order to improve the skirt characteristics and stopband rejections has been reported. By placing transmission zeros near the passband edge, a designed BPF with higher selectivity and fewer resonators can be obtained. Using SIRs as the building blocks and the tapped-line I/O, the first and second spurious responses can be shifted to a higher frequency. This BPF was designed using EM simulation and then implemented in order to verify the prediction of our design.