Frequency conversion analysis of noise in heterojunction bipolar transistor oscillators including periodically modulated noise sources

Abstract

<p>A procedure which may be used to analyze the noise characteristics of HBT oscillators is presented. A large signal model of a HBT is developed based largely on the Gummel-Poon transistor model. A new base-emitter diffusion capacitance model is also generated which shows improved accuracy over conventional bipolar transistor models in characterizing HBTs. The large signal characteristics of the oscillator are then established using a time domain simulator. Conversion matrix theory is applied to the large signal model to generate an equivalent linear oscillator model which accounts for all of the frequency conversion effects which occur due to circuit nonlinearities. A bias dependent noise model of the HBT is also developed. It is found that the dominant sources of noise within the device are due to thermal, shot, and flicker noise. The modulation of the shot and flicker noise sources are also analyzed. The amount of frequency conversion which occurs as a result from the modulation of these noise sources is established. The modulated noise sources within the HBT are then inserted into the linear oscillator model to establish the overall noise characteristics of the HBT oscillator. Experimental results are also presented which support the validity of the analysis procedure;It Is found that the effects which result from the modulation of shot and flicker noise sources within oscillators are very significant. The upconversion of noise due to the modulation of flicker noise sources is found to be as significant as the upconversion which results from circuit nonlinearities. Thus, the upconversion of flicker noise cannot be associated with any one nonlinear mechanism. Rather it is a result of a complex interaction between the circuit nonlinearities and noise modulation mechanisms;The noise within many other periodically driven nonlinear networks may also be characterized using the analysis procedure presented in this paper. Examples of such networks include mixers, limiters, and frequency multipliers.</p>