Guidelines for reduced-order thermal modeling of multifinger GaN HEMTs

Robert Pearson, Bikramjit Chatterjee, Samuel Kim, Samuel Graham, Alexander Rattner, Sukwon Choi

Research output: Contribution to journalArticlepeer-review

15 Scopus citations

Abstract

The increasing demand for tightly integrated gallium nitride high electron mobility transistors (HEMT) into electronics systems requires accurate thermal evaluation. While these devices exhibit favorable electrical characteristics, the performance and reliability suffer from elevated operating temperatures. Localized device self-heating, with peak channel and die level heat fluxes of the order of 1MW cm~ and 1 kW cm~, respectively, presents a need for thermal management that is reliant on accurate channel temperature predictions. In this publication, a high-fidelity multiphysics modeling approach employing one-way electrothermal coupling is validated against experimental results from Raman thermometry of a 60-finger gallium nitride (GaN) HEMT power amplifier under a set of direct current (DC)-bias conditions. A survey of commonly assumed reduced-order approximations, in the form of numerical and analytical models, are systematically evaluated with comparisons to the peak channel temperature rise of the coupled multiphysics model. Recommendations of modeling assumptions are made relating to heat generation, material properties, and composite layer discretization for numerical and analytical models. The importance of electrothermal coupling is emphasized given the structural and bias condition effect on the heat generation profile. Discretization of the composite layers, with temperature-dependent thermal properties that are physically representative, are also recommended.

Original languageEnglish (US)
Article number021012-1
JournalJournal of Electronic Packaging, Transactions of the ASME
Volume142
Issue number2
DOIs
StatePublished - Jun 2020

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Mechanics of Materials
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Guidelines for reduced-order thermal modeling of multifinger GaN HEMTs'. Together they form a unique fingerprint.

Cite this