Here’s the page we think you wanted. See search results instead:


Discutez avec un expert

DynaFET: Advanced model for GaN/GaAs HEMTs from NVNA measurements and ANNs Webcast

1   Hour | Webcast - enregistré | Where & When


DynaFET is more than just a novel nonlinear simulation model for advanced compound semiconductor transistors (e.g. GaN and GaAs FETs); it is a complete end-to-end characterization, modeling, and validation flow. The DynaFET solution consists of three major system components designed to fit together for overall success.

• The DynaFET model, implemented natively in Keysight EEsof EDA Advanced Design System (ADS), accounts in detail for dynamic self-heating, drain-lag (“knee walk-out”), and gate-lag phenomena that manifest themselves as critical long-term memory effects in III-V FETs.

• The DUT characterization approach features an active source injection data acquisition system built around a Keysight nonlinear vector network analyzer (NVNA). The measured, fully calibrated large-signal waveform data are more indicative of the conditions in which the device will operate in actual applications, and embody DUT behavior sampled at faster time-scales than are typical of more conventional pulsed I-V measurements.

• Advanced model generation (extraction) methods, implemented in Keysight EDA's IC-CAP software, are used to identify the key dynamical variables (e.g., junction temperature and multiple trap states) and train artificial neural networks to model their detailed effects on the DUT characteristics.

The built-in DynaFET component in ADS reads the generated neural network representation of the device constitutive relations, simulates efficiently (quickly), accurately, and robustly (excellent convergence properties) in all simulation modes (transient, harmonic balance, circuit envelope, etc.). From the designer’s perspective it has the same simple use model as any modern compact transistor model. Unlike many other GaN/GaAs models, however, the scalable DynaFET model produces a general, global model with no need for application-specific parameter tuning. The model can therefore be used with confidence to design power amplifiers and circuits with GaN and GaAs transistors for all applications where the device is working in the active mode.


The webcast is targeted to GaN/GaAs HEMT technologists, device modeling specialists, R&D transistor characterization experts, RF and microwave circuit designers in III-V FET technologies, device specialists, RF PA designers, and CAD managers.

The webcast introduces each of the three components of the DynaFET solution and their arrangement in the overall flow. It takes the participants through the basic model theory and the major steps in the characterization, model generation (extraction), and validation of the DynaFET model. Detailed comparisons with independently measured data on a wide range of actual GaN HFET and GaAs HEMT devices will be presented.


Dr. David E. Root, Keysight Technologies Research Fellow
David E. Root received B.S. degrees in physics and mathematics, and, in 1986, the Ph.D. degree in physics, all from MIT. He joined Keysight Technologies in 1985, where he is now Keysight Technologies Research Fellow. His current responsibilities include nonlinear behavioral and device modeling, large-signal simulation, and nonlinear measurements for new technical capabilities and business opportunities for Keysight. He is a pioneer of measurement-based active device modeling, for which he was elected an IEEE Fellow in 2002.



Dr. Jianjun Xu, Senior Device Modeling R&D Engineer
Jianjun Xu received the B.Eng. Degree from Tianjin University, Tianjin, P.R.China in 1998, and the Ph.D. Degree in Electrical Engineering from Carleton University, Ottawa, Canada, in 2004. He joined Keysight Technologies in 2005, where he is now Senior Device Modeling R&D Engineer. He was the key developer of the Keysight NeuroFET model and is the technical lead on the entire DynaFET modeling flow. He is currently working on advanced measurement based nonlinear device modeling for GaAs and GaN.

Where & When

Price Location Pour plus d'information
Free At Your PC Enroll to view the September 4, 2013 recorded broadcast 

Prix susceptibles de modification sans préavis.