Abstract

We report for the first time the simulation of the large-signal dynamic load-line of high-Q matched mm-wave power amplifiers obtained through a Monte Carlo particle-based device simulator. Due to the long transient time of large reactive circuit elements, the time-domain solution of power amplifier high-Q matching networks requires prohibitive simulation time for the already time-consuming Monte Carlo technique. However, by emulating the high-Q matching network and the load impedance through an active load-line, we show that, in combination with our fast Cellular Monte Carlo algorithm, particle-based accurate device simulations of the large signal operations of AlGaN/GaN HEMTS are possible in a time-effective manner. Reliability issues and parasitic elements (such as dislocations and contact resistance) are also taken into account by, respectively, exploiting the accurate carrier dynamics description of the Monte Carlo technique and self-consistently coupling a Finite Difference Time Domain network solver with our device simulator code.

Original languageEnglish (US)
Title of host publication2011 International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2011
Pages87-90
Number of pages4
DOIs
StatePublished - Nov 1 2011
Event2011 International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2011 - Osaka, Japan
Duration: Sep 8 2011Sep 10 2011

Publication series

NameInternational Conference on Simulation of Semiconductor Processes and Devices, SISPAD

Other

Other2011 International Conference on Simulation of Semiconductor Processes and Devices, SISPAD 2011
Country/TerritoryJapan
CityOsaka
Period9/8/119/10/11

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Modeling and Simulation

Fingerprint

Dive into the research topics of 'Large-signal full-band Monte Carlo device simulation of millimeter-wave power GaN HEMTs with the inclusion of parasitic and reliability issues'. Together they form a unique fingerprint.

Cite this