Channel temperature estimation in GaAs FET devices

Abstract

Operating channel temperature has an important influence on the electrical performance and reliability of GaAs FET amplifiers but this parameter is difficult to determine experimentally or analytically. Simplified closed-form and numerical models are commonly used although both are subject to errors in the case of sub-micron heat sources surrounded by epitaxial layers. Some assumptions about the heat source size and location are explored using finite-difference simulation. A simple measurement technique making use of the temperature coefficient of gate metal resistance is described, and an improved procedure for spatial averaging correction of infrared measurements of sub-micron heat sources is proposed.

Indirect techniques for channel temperature estimation of HEMT microwave transistors: Comparison and limits

“…Channel temperature estimated with IR thermocamera, as already shown by other authors [3,5], showed a lower R TH value.…”

“…Over the last years, several techniques to evaluate junction temperature have been proposed in the literature [1][2][3][4][5][6][7][8]. Channel temperature and thermal resistance (R TH ) can be evaluated by means of electrical measurements, through the analysis of the variation of a temperature sensitive parameter (TSP) as a function of temperature and of the operating conditions.…”

“…of regions where temperature reaches a local maximum. All these techniques have limited lateral resolution and require special set-up; the same limitations affect infrared thermography [3][4][5], which is, however, easier to apply. Optical techniques usually represent the only viable solution for experimental thermal analysis of multistage MMICs, where substantial thermal interaction among stages can take place; however, they are limited by the presence of metallization on the chip surface.…”

Characterization of Dynamic Self-Heating in GaN HEMTs Using Gate Resistance Measurement

“…Literature reports on an electrical method (G-RTD) using gate resistance temperature dependence to determine device temperature in a straightforward manner [1][2][3]. A comparison of this method with micro-Raman spectroscopy has proved this method to be reliable with a net uncertainty of ±5% [4].…”