Text

theses

The most important stage in a wireless transmitter is the power amplifier because this stage consumes a great deal of power in a wireless system and is a major factor in the battery life of portable equipment. Generally, switching RF power amplifiers have greater efficiency than their linear counterparts but they are also more difficult to analyze, have switching losses, and can introduce switching transients into the amplified signal. The research focus for this dissertation is on the Class E switching RF power amplifier. This circuit topology uses soft switching in order to minimize switching transients and achieve nearly 100% efficiency in the ideal case. In this dissertation, a short range wireless power transfer system, based on a Class E power converter, is investigated. It operates at 200 kHz and was constructed using commonly available components. A mathematical analysis of a Class E RF power amplifier, operating at 3.9 MHz, is then presented. The model is based on modern state-space techniques and is simulated using both MATLAB and Simulink. Class E RF power amplifiers are insensitive to signal amplitude variations so by itself, this system is only suitable for the amplification of constant envelope signals. This limitation can be overcome, however, with the addition of a supply modulator. One of the most efficient supply modulators is the Class G dual supply modulator. This device utilizes two supply voltages and automatically switches between them depending on signal amplitude. The process of switching from one supply voltage to the other produces transients or ”glitches” that introduce undesirable broadband noise into the signal output. The effects of these glitches, on the noise floor at the Class E power amplifier output, are investigated using a detailed simulation model run on MATLAB. A typical input data stream, glitch depth, and glitch duration are applied to this operating model. As a potential design aid, a purely hypothetical Class E amplifier state-space based model is analyzed based on same simplifying assumptions including a truncated Taylor series expansion of the matrix exponential. Two algebraic equations were derived to determine two Class E amplifier design components. Class E systems are often designed using Raab’s formulas which were derived based on a number of simplifying assumptions. Furthermore, electronic components have tolerances. As a consequence of this a Class E amplifier or power converter must be tuned for proper operation. In order to simplify this process a detailed sensitivity analysis of the Class E circuit topology is presented. This investigation is approached from multiple directions including a state-space based model simulated using MATLAB, a SPICE based simulation, and an experimental Class E prototype operating at 100 kHz.

ETD_8069

Ph.D.

Includes bibliographical references

by Sumati Sehajpal

doi:10.7282/T3125WKC

ETD doctoral

The author owns the copyright to this work.