Hybrid method for estimating peak channel temperature in AlGaN/GaN HEMTs using resistance thermal detector thermometry and a simplified analytical model

Abstract

In this work, we introduce an analytical approach for accurately determining the peak channel temperature (Tpeak,ana) of AlGaN/GaN high electron mobility transistors from experimental measurements. A thermal model based on the device structure estimates the temperature profile by applying the carrier transport in conjunction with the heat conduction in the transistor channel region. This profile provides the temperature difference (δT) between the experimentally measured average temperature (Tavg,meas) and peak junction temperature (Tpeak,ana′). This difference is used to determine the Tpeak,ana value in the saturation region. Finally, Kirchhoff's transformation addresses the nonlinearity in thermal conduction, resulting in higher accuracy. The validity of this model is verified through gate resistance temperature detector thermometry and technology computer-aided design simulations with various device geometric parameters. Overall, this extraction methodology is a powerful tool for rapidly and precisely determining the peak channel temperature from the measured data, making it suitable for device modeling, reliability analysis, and thermal management.