Reflection type Q-factor measurement using standard least squares methods

Abstract

Methods to determine resonant cavity Q-factor from vector network analyser reflection data are often quite complex. This study presents a simple method which makes use of two least squares algorithms applied sequentially. Firstly, a circle is fitted to the data, which is then projected onto a linear frequency scale using the Ginzton technique. A straight line is fitted to the projected data, and the loaded Q-factor and the resonant frequency are obtained from the line coefficients. The method has been tested on both synthetic and experimental data, and compared with the Kajfez QZERO software, and the results are shown to be accurate to a fraction of 1%.