TABLE OF CONTENTS Before we can do any high-frequency designs, we need simplified high-frequency models for active devices; BJT (bipolar junction transistors) and FET (field effect transistors) that are (1) generic, i.e., applicable to just about any active device, (2) simple, in order to allow algebraic manipulation that is easy enough to perform, and (3) a model that provides easy-to-understand results. Generalized, hybrid- ? models are too complicated to meet this need, although that is where we will start. They contain information needed for analysis at low frequencies, which is not needed for high-frequency analysis. We want to create a model that contains only the information needed for high-frequency analysis and nothing more. The model we will create is based on frequency-dependent sources with DC gains set to infinity. It will work so long as the active device can be modeled with one pole in its frequency response, and it will work for frequencies above the point where the gain begins to roll off with frequency (sometimes called the f ? frequency).
Once created, the model will first be applied to follower circuits. Follower circuits provide unity voltage gain from input to output while keeping the input impedance at ? and the output impedance at zero. At first glance, these circuits seem simple. Unfortunately, these circuits tend to oscillate freely at high frequencies when loaded into certain kinds of loads (principally capacitive). In some cases, the only sign that the circuit is oscillating is a distorted waveform that changes...
