A charge function identification procedure for GaN-HEMTs is proposed. This is based on a frequency-domain integration of displacement current waveforms obtained from an auxiliary model extracted from multi-bias S-parameters. The method is compared with a similar technique recently proposed, which is instead based on direct acquisitions of large-signal waveforms at the transistor ports by means of a nonlinear vector network analyzer (NVNA). Comparisons between the two approaches are provided by using a 1-mm GaN-on-SiC HEMT, leading to conclude that thermal and trap-induced dispersion on charges have an impact quantified in ∼ 4% - 18% normalized mean square error on the displacement current prediction, depending on the waveforms considered.
Gian Piero, G., Santarelli, A., Filicori, F. (2018). A Procedure for GaN HEMT Charge Functions Extraction from Multi-Bias S-Parameters. Institute of Electrical and Electronics Engineers Inc. [10.23919/EuMIC.2018.8539947].
A Procedure for GaN HEMT Charge Functions Extraction from Multi-Bias S-Parameters
Gian Piero, Gibiino;Santarelli, Alberto;Filicori, Fabio
2018
Abstract
A charge function identification procedure for GaN-HEMTs is proposed. This is based on a frequency-domain integration of displacement current waveforms obtained from an auxiliary model extracted from multi-bias S-parameters. The method is compared with a similar technique recently proposed, which is instead based on direct acquisitions of large-signal waveforms at the transistor ports by means of a nonlinear vector network analyzer (NVNA). Comparisons between the two approaches are provided by using a 1-mm GaN-on-SiC HEMT, leading to conclude that thermal and trap-induced dispersion on charges have an impact quantified in ∼ 4% - 18% normalized mean square error on the displacement current prediction, depending on the waveforms considered.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
