A power amplifier (HPA) is most often composed of several MMIC circuits, and each of these monolithic circuits combines, in its output stage, the power of multiple transistors.

At the MMIC HPA level, the matching network of each individual transistor, as well as the power combining of the different paths, are implemented on the semiconductor material itself. This technology penalizes the losses of this passive circuitry, and therefore degrades the overall DC to RF efficiency.

The output powers from the different MMIC HPAs are then combined into a single output waveguide. Interfaces between the MMIC HPAs and the output waveguide therefore integrate transitions between microstrip lines and the waveguide. To achieve this second level of power combining, the MMIC circuits are mounted on a PCB, which thus provides the interface between the monolithic circuit and the waveguide, without integrating additional RF functions, even though this substrate exhibits lower losses than the semiconductor material.

As part of a collaboration between Thales Alenia Space and XLIM (PhD thesis of Ghofrane Mastouri – October 2022 to March 2026), we have been working on an innovative architecture:

- Each drain access of the MMIC circuit is connected to the PCB, and the output matching and power combining networks of each MMIC HPA are implemented on the PCB, thus reducing losses.

- The matching and combining functions of the different MMIC HPA paths are integrated without requiring a 50-ohm reference, which provides greater design flexibility.

- A common antenna for the HPA output paths is implemented on the PCB and transfers the signal to a rectangular waveguide combiner.

However, this approach is currently limited to two HPA MMIC output paths in order to maintain proper loading symmetry between transistors, given that only one antenna is used for the microstrip-to-waveguide transition. This study is also restricted to the Ka band. Thus, while the concept is validated, it remains far from the actual power requirements, given the limited output power level and the number of transistors in the MMIC HPA output stage (target up to 16 GaN transistors per MMIC).

Spatial power combining, through the use of multiple antennas, would make it possible to increase the number of MMIC/antenna interface accesses while maintaining load symmetry between transistors. This would enable an increase in HPA output power without the need for an on-chip output combiner.

This is the innovation we propose to study in this PhD work. Our main objective remains to reduce the matching and combining losses of MMIC HPAs, thereby lowering their power consumption and preparing for the development of HPA MMICs at higher frequency bands, such as Q and even E bands.

We intend to explore two implementation approaches:

- A PCB-based antenna implementation.

This is likely to be the preferred solution for the Ka band, considering the relative dimensions of the MMIC and the output waveguide. In this case, the PCB will handle the output matching of the different paths and their combination—partly on the PCB, and partly directly within the output waveguide via multiple antennas.

- An on-chip antenna implementation.

The more ambitious approach involves integrating the antennas directly on the MMIC and embedding the MMIC within the waveguide. This solution becomes feasible in the Q and E bands. As in the Ka band, the goal will be to develop optimized architectures, likely involving partial combining on the MMIC and spatial combining within the waveguide.

These approaches are highly innovative and have the potential to significantly enhance HPA power efficiency. However, they also raise critical questions beyond matching and combining, particularly concerning thermal management of the MMIC and electrical stability. These aspects will also be analyzed in the course of this PhD.

The following steps are planned:

• Literature review

• Evaluation of the amplifier/transition/combiner co-design concept developed in Ka band in Ghofrane Mastouri’s PhD, extended to up to 16 GaN transistors per MMIC

• Assessment of spatial combining potential across different frequency bands (K, Q, E)

• Design and realization of a demonstrator in the frequency band identified as the most relevant in terms of requirements and potential

This work will be carried out within the framework of the AXIS joint laboratory, linking Thales Alenia Space and XLIM. This research topic has been selected by the laboratory’s steering committee, among a set of other proposals submitted by both partners.

================

For more Information about the topics and the co-financial partner (found by the lab!); contact Directeur de thèse - serge.verdeyme@xlim.fr

Then, prepare a resume, a recent transcript and a reference letter from your M2 supervisor/ engineering school director and you will be ready to apply online before March 13th, 2026 Midnight Paris time!