Tom McKay

Linear cellular antenna switch for highly-integrated SOI front-end

A CMOS-compatible SOI SP6T cellular antenna switch achieves linearity heretofore requiring more costly sapphire or GaAs substrate materials. The prototype TX path P-o.ids is 41 dBm and harmonics are 79 dBc at 34 dBm output power. A low insertion loss of 0.8 dB and isolation of 40 dB is obtained at 900 MHz. A CMOS-compatible SOI SP6T cellular antenna switch achieves linearity heretofore requiring more costly sapphire or GaAs substrate materials. The prototype TX path P-o.ids is 41 dBm and harmonics are 79 dBc at 34 dBm output power. A low insertion loss of 0.8 dB and isolation of 40 dB is obtained at 900 MHz.

Advances in Silicon-On-Insulator Cellular Antenna Switch Technology

Silicon-on-insulator technology utilizing very high resistivity handle wafers demonstrates sufficient performance for many cellular handset requirements. Technologies with either thin or thick device layers show promise. Thick-SOI prototype single pole six throw switch (SP6T) P-0.1dB of about 40 dBm and -75 dBc harmonic at 35 dBm output have been demonstrated at 900 MHz. Thin-SOI Ron-Coff product similar to that of pHEMT has been demonstrated experimentally. Obstacles such as distortion caused by Si-SiO2 interfaces remain a fascinating engineering challenge to fully exploit the opportunity.

Reverse noise measurement and use in device characterization

We review the concept of reverse noise measurements in the context of on-wafer transistor noise characterization. Several different applications of reverse noise measurements are suggested and demonstrated. Reverse measurements can be used to check measurement results, to significantly reduce the uncertainty in |Gammaopt |, to reduce the occurrence of nonphysical results, and possibly to directly measure or constrain parameters in models of transistors

Inter-laboratory comparison of CMOS distortion measurements

We describe a measurement comparison of distortion in a complementary metal-oxide semiconductor low-noise device operating under weakly nonlinear conditions. Issues that commonly arise in performing and interpreting nonlinear measurements are discussed, such as power and wave-based representations and the effects of terminating impedance on intermodulation distortion. We demonstrate that the increased confidence provided by a measurement comparison can help to diagnose issues with a device model that was initially derived from DC I/V curves and their derivatives and then compared to RF measurement.