tors. The quadrature ILFD uses the 3D inductors to save chip area and the ILFD has been successfully implemented in the TSMC 0.18-m CMOS process. At the supply voltage of 0.9 V, the core power consumption of the proposed circuit is about 9.36 mW. At the input power of Ϫ8 dBm, the total operation locking range is 2.45 GHz, from 3.6 to 6.05 GHz, in the Ϭ2 mode. The phase deviation between in-phase and quadrature-phase outputs is about 0.19°.
ACKNOWLEDGMENTThe authors thank the CIC of National Science Council, ROC, for the chip implementation. REFERENCES 1. H. Wu and A. Hajimiri, A 19 GHz 0.5 mW 0.35 m CMOS frequency divider with shunt-peaking locking-range enhancement, In: ISSCC Digest of Technical Papers, Feb. 2001, pp. 412-413. 2. M. Tiebout, A CMOS direct injection-locked oscillator topology as high-frequency low-power frequency divider, IEEE J Solid-State Circuits 39 (2004), 1170 -1174. 3. P. Andreani and X. Wang, On the phase-noise and phase-error performances of multiphase LC CMOS VCOs, IEEE J Solid-State Circuits 39 ABSTRACT: Parallel-coupled microstrip filters with MEMS switchable frequency responses are designed. The 15-GHz filter switches 5.3% with minimum insertion loss of 0.7 dB, whereas the 20-GHz filter has 12.9% switching range and minimum insertion loss of 0.6 dB. Both filters have the lowest insertion losses among MEMS tunable microwave bandpass filters to date. ABSTRACT: A fast integral equation method, termed IE-FFT, is applied with the combined field integral equation (CFIE) for solving electromagnetic scattering problems of three-dimensional (3D) electrically large targets. For closed conducting surfaces, the employment of CFIE eliminates the internal resonance problem suffered by both the electric field integral equation and the magnetic integral equation. Furthermore, it improves the efficiency of IE-FFT significantly by reducing the number of iterations for convergence. It is shown that the memory requirement and the computational complexity per iteration of the IE-FFT are O(N 1.5 ) and O(N 1.5 log N), respectively, for 3D perfectly electric conducting scattering problems. Some numerical examples are provided to demonstrate the accuracy and efficiency of the technique.