Solid-state microwave amplification into millimetric frequencies
The work presented in this thesis is published research work undertaken by the author over a long period of time 1975 to 2004, and to assist the reader it is split into three periods. The first period covers 1975 to 1979 and concentrates on the research and development of the silicon TRAPATI diode for use in an X-band power amplifier suitable for airborne application. The work includes the development and measurement of the thermal operation of the diode, the microwave circuit design required for the class - C reflection amplifier operation and an in depth analysis of the technical problems associated with the design of high-frequency TRAP A IT circuits, which final led to the decline of the technology at frequencies beyond S-band. The second period covers 1980 to 1987 and describes in detail the authors' contribution in the development of a milli-metric GaAs MESFET giving state of the art minimum noise performance in the mid 1980's. The work covers research into the design aspects of the transistor, and in particular the RF characterization and the development of an equivalent circuit model. The minimum noise figure measurements were compared with Fukui model and deviation at the high frequency was identified and was attributed to distributed effects along the electrode metallization patterns. The work also led to the invention of a new travelling wave structure, the LGT (linear gate transistor) which was fabricated and fully RF characterized. The work describes the LGT structure in detail which was ahead of its time with respect to the available technology for fabrication. The third period covers 2000 to 2004 where the author extended his research work into microwave transistors fabricated on wide band-gap materials, for example Gallium Nitride (GaN). The work includes noise analysis of the GaN high electron mobility transistor (HEMT) using the Fukui analysis. The prime part of the work has been in 'extracting the intrinsic device parameters over bias conditions, which has led to novel method of extracting the parasitic source resistance (R.) and the intrinsic saturation velocity (V.i). The extracted intrinsic parameters are used for both large signal and minimum noise analysis.