Thermal memory effects modeling and compensation in RF power amplifiers and predistortion linearizers

Memory effects, which influence the performance of RF power amplifiers (PAs) and predistortion-based linearizers, become more significant and critical in designing these circuits as the modulation signal bandwidth and operation power increase. This paper reports on an attempt to investigate, model, and quantify the contributions of the electrical nonlinearity effects and the thermal memory effects to a PA's distortion generation, as well as how to compensate for these effects in designing baseband predistortion schemes. The first part of this paper reports on the development of an accurate dynamic expression of the instantaneous junction temperature as a function of the instantaneous dissipated power. This expression has been used in the construction of an electrothermal model for the PA. Parameters for the new proposed behavior model were determined from the PA measurements obtained under different excitation conditions (e.g., small-signal and pulsed RF tests). This study led us to conclude that the effects of the transistor self-heating phenomenon are more important under narrow-band signal (e.g., enhanced data for global evolution of global system for mobile communications) than for signals with wide modulation bandwidth (CDMA2000, Universal Mobile Telecommunications System). In the second part of this paper, the newly developed model has also been used to design a temperature-compensated predistortion function to compensate for these effects. The linearized PA output spectrum and error vector magnitude show a significant performance improvement in the temperature-compensated predistortion function over a memoryless predistortion. The results of these measurements that have been conducted on a 90-W peak lateral double-diffused metal-oxide-semiconductor PA are in agreement with those obtained from simulations using the developed PA and the predistorter models implemented in an ADS environment.