An Improved Methodology for Measuring the Conduction Characteristics of SiC MOSFETs

Developing accurate behavioral models for silicon carbide (SiC) mofsets requires measuring their conduction characteristics across their entire operating regime. However, commercial curve tracer instruments can only measure a limited range of conduction characteristics at low drain-source voltage biases. In addition, the long dynamic settling times intrinsic to commercial curve tracers introduce self-heating (SH) artifacts that compromise the accuracy of the resulting characterization data. These limitations prevent SiC mofsetmodels from accurately predicting dynamic switching events and short-circuit failures. Alternative characterization methods are difficult to implement, do not provide a holistic data set, and are susceptible to measurement and V<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TH</sub> hysteresis errors. This article proposes an advanced experimental conduction characterization approach that can accurately characterize SiC mofsets at high-voltage and high-current without SH or <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V</i><sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TH</sub> hysteresis errors. The method provides lower-error measurements than commercial curve tracers at the same operating conditions while also expanding the characterization range to high voltages. The measured results encompass the entire <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I–V</i> space and can be used to model SiC mofsets without combining data from multiple instruments. The proposed method is used to characterize and model a SiC half-bridge power module, which is shown to be more accurate than a model generated with commercial curve tracer data.