Microwave FET Modeling and Applications
Abstract: This thesis deals with three distinct topics within the areas of modeling, analysis and circuit design with microwave field effect transistors (FETs).First, the extraction of FET small-signal model parameters is addressed. A method is presented where the model parameter uncertainties are derived from S-parameter measurement uncertainties and uncertainties in the parasitic elements. This allows a method to be presented where each model parameter is determined with minimum uncertainty, thus being optimal in a statistical sense. Accurate extractions can thereby be performed independent of the specific device characteristics or bias point. Thereafter, analysis of intermodulation distortion (IMD) in power amplifiers (PAs) is treated. A new analysis method is described where the large-signal IMD behavior of PAs can be analytically predicted. The method allows the IMD generating mechanisms to be identified, thus providing a tool for tailoring device characteristics for IMD reduction. The method has been used to predict the large-signal IMD behavior of LDMOS and CMOS PA circuits. The IMD prediction capabilities of common large-signal transistor models have been evaluated, which led to modifications to an industry-standard LDMOS model being proposed. Finally, FMCW radar transceivers are described. A FET transceiver suitable for FMCW radars is presented. The transceiver utilizes the FET simultaneously to amplify the transmitted signals and as a resistive mixer to down-convert the received signal. Therefore, unlike most existing FMCW radar transceivers, the transmitted and received signals do not need to be separated. Since AM noise rejection is important in FMCW radars, a similar balanced FET transceiver is also presented. Compared to the unbalanced transceiver, the AM noise performance is substantially improved. The transceivers? simple topology and the elimination of the need for separation of transmitted and received signals make them suitable for integration in MMIC technology.
This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.