TABLE OF CONTENTS In the general analysis, no assumptions are made about the relative magnitudes of the LO or input signal(s). To analyse this general case, a full non-linear analysis must be performed where both signals are present. A theoretical implementation is difficult, and such analyses are computationally intensive compared with the restricted case. It requires a robust convergence algorithm using time-domain or harmonic-balance methods, which makes use of a special type of Fourier transform for multiple non-harmonically related signals. Such a general analysis can be simulated with a range of different CAD software packages described in Chapter 4. Mixer products are produced with the form in eqn. (7.7).
In this analysis, it is assumed that the LO is much larger than the RF signal and that it is solely responsible for the time-varying circuit elements. The analysis can then be made in two parts: first, a large-signal analysis is performed using only the LO, and second, a linear analysis is performed using the time-varying circuit elements derived from the large-signal analysis. The former is performed using time-domain or harmonic-balance techniques. The latter is usually performed in the frequency domain using the Fourier components of the time-varying circuit elements to set up a conversion matrix. The mixing products generated have the form in eqn. (7.6), and there is a linear relationship between the RF and IF signal amplitudes. This is commonly referred to as linear mixing.
