Towards a Dispersion Free GaN HEMT (TOPGAN)

Start date 01/01/2015
End date The project is closed: 31/12/2016
Motivation
The GaN HEMT is today a standard product offered by several commercial suppliers. However, long term memory effects are preventing the use of GaN HEMTs in e.g. mobile communication systems. Although long term memory effects in GaN HEMTs has been a major research area for many years, a clear coupling between e.g. EPI specification and system performance has not been established. This project aims at connecting EPI growth and device processing to amplifier and oscillator performance.

Furthermore, the growing interest for mm-wave bands in future mobile communication systems set new demands on the amplifiers, and in the end the transistor. The power level will be significantly lower, especially in MIMO systems, and the bandwidth will be significantly wider (~1 GHz). These requirements preclude the use of digital pre-distortion, and hence the transistor needs to be inherently linear, and highly efficient.
To achieve this, new characterization methods together with new and improved models need to be developed.

Objectives
  • Correlate system performance to material properties and device fabrication
  • Optimize GaN HEMTs for linear transmitters for mm-wave applications
  • Optimize GaN HEMTs for low phase noise oscillators
  • Develop system like characterization methods at device level
  • Develop improved models enabling transient simulations in TCAD and circuit simulators
Project outcome
  • 5 articles in high impact journals
  • 3 co-authored (industry/academia) scientific articles
  • 1 test-bed for system like characterization
  • Updated roadmaps for future GaN technology
  • 1 growing start up company
  • 1 success story related to improved EPI
Research Partners
Project Leaders
Mattias Thorsell, Chalmers, Deputy Project Leader: Hans Hjelmgren, Chalmers
This page is only available in english

Published: Mon 18 Apr 2016.