Automated Characterization Platform for Comprehensive Dynamic R_dson Assessment of GaN HEMTs From 50 K to 400 K

Gallium Nitride (GaN) HEMTs demonstrate superior performance at cryogenic temperatures, making them promising candidates for lunar missions where temperatures vary from 50 to 400 K. However, the current collapse effect increases dynamic on-state resistance (R<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dson</inf>) during switching, leading to additional energy loss, especially in high-frequency applications. To conduct a comprehensive study with high efficiency, we developed an automated characterization platform for dynamic R<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dson</inf> extraction under multi-stressor conditions. The design considerations include 1) programmable stressors such as temperatures, voltages, currents, etc., 2) electrothermal-mechanical compatibility with the cryocooler across a wide temperature range, and 3) software framework for full automation. Results for a commercial GaN HEMT show that dynamic R<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dson</inf> varies non-monotonically with temperature increment, as it increases at 50 - 180 K and then decreases a 180 - 400 K. Dynamic R<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dson</inf> follows this trend at both hard- and soft-switching, but is more sensitive to stressors at hard switching, while remaining nearly unchanged at soft switching, indicating that, under this testing plan that applies low stressing voltages, hot electron effect makes the major contribution to dynamic R<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dson</inf>.