Dynamic drive for radio frequency power amplifiers with wideband active load modulation

• Applicant: Professor Dr.-Ing. Georg Fischer
• Subject Area: Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
• Term: from 2017 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 393269191

The objective of the project proposed here is to perform fundamental research on advanced techniques for the dynamic drive of radio frequency power amplifiers with wideband active load modulation. Unlike conventional, single-track power amplifiers, dual-input power amplifier architectures with active load modulation are not bound to a single, unique drive function. Their input vectors span a multidimensional load modulation space, which, in theory, contains an infinite number of drive functions that all lead to the same output signal, but at different instantaneous efficiencies and different bandwidth requirements, respectively. Thus, the goal of the proposed drive techniques is to maximize the average efficiency of the amplifier system while meeting the linearity requirements imposed by the communication standard and keeping the predistortion bandwidth within the limits imposed by the hardware. The idea of a dynamic drive function arises from the applicant`s previous work, which has shown that, under wideband operation, the maximum efficient drive function of a dual-input power amplifier with active load modulation exhibits multiple, characteristic discontinuities in its derivative. Since the available predistortion bandwidth is inevitably limited, the maximum efficient drive function can only be approximated poorly if the approximation is performed over the entire amplitude range and thus the amplifiers potential for efficiency cannot be fully exploited. On the other hand, it is observed that a) an approximation over the entire amplitude range is mostly unnecessary, due to signal statistics, and b) a transition between different drive functions does not increase predistortion bandwidth as long as it takes place within the low power region, before the onset of active load modulation. These findings give rise to the idea of dynamically adapting the drive function to the peak power within individual signal blocks, such that, on average, drive functions that produce significantly higher instantaneous efficiencies can be applied. In order to enable a practical implementation of the dynamic drive technique, novel predistortion models shall be developed within the proposed research project, addressing the current lack of predistortion models that genuinely support the single-input, dual-output architecture. In this context, it is important to point out the close similarity between the proposed dynamic drive techniques and the general problem of wideband/multiband predistortion. Due to this relationship, it can be assumed that a truly wideband predistortion model is able to incorporate the proposed drive function technique without significantly increasing its complexity. The potential of the dynamic drive techniques shall be verified and demonstrated within the proposed project based on hardware demonstrators of wideband dual-input power amplifiers which already exist at the applicant`s institute.

• DFG Programme: Research Grants