This project is aimed at the design and test of an
innovative RF Self-Oscillating Power Amplifier (SOPA) in which the feedback loop is not made at the RF carrier, as in the usual approach, but now between the output signal envelope and the input baseband signal. This topology is a novel, and very promising, approach to the linearity/efficiency compromise in RF transmitters as it is capable of greatly reducing switching losses but also compensate for the nonlinearities of the AM and PM modulator, of the
RF PA and of the output band-pass reconstruction filter.
Furthermore, the complex stability and ultimate injection locked behavior analysis of these SOPA structures (which is a determinant problem for any highly nonlinear dynamic feedback loop) will be thoroughly performed through bifurcation analysis across input envelope amplitude.
In addition, this new SOPA based Tx architecture, but also many other coming digital Tx’s, suffer from the inexistence of proper simulation tools. As they involve baseband AM and PM signals, a digital self-oscillating clock and a RF carrier - of three disparate time-scales - exciting strongly nonlinear low-frequency broadband but also RF narrow-band circuits, no circuit simulation tool is able to handle it. So state-of-the-art simulation techniques for systems with autonomous components (as is the case of SOPAs) will be studied.
In summary, the expected results from this project are: (i) system-level analysis of SOPA behavior, its stability and injection locking conditions; (ii) the design, implementation, test and validation of a novel, highly efficient and linear, SOPA based Tx
topology; and (iii) a new simulation method dedicated to device-level simulation of SOPA circuits (or circuits that suffer from similar multirate problems).