Characterization and modeling of memory effects in switched GaN HEMT devices
Final Report Abstract
GaN-based transistors are increasingly used in power and microwave electronics, since this material system can operate at higher voltages and currents and achieve shorter switching times than commercial transistors in other material systems. What is new is that GaN HEMTs for these very different tasks have a conceptually identical structure, so that they should be described in the circuit design also by a uniform model. At the same time, the operating modes converge when the switching operations in switching converters have to be calculated with the same high accuracy as in analog circuits, and when analog microwave power amplifiers increasingly use the transistor as a switch. This development calls for a new, uniform approach to transistor modeling. The memory effects typical for GaN HEMTs pose a particular difficulty, as they significantly worsen the efficiency of switching converters and amplifiers. As there is no standardized method for characterizing trapping effects in the context of power electronics, a robust and accurate measurement methodology was developed in the project that is capable of single-factor analysis of trapping and detrapping in modern high-voltage power transistors. The measurement circuit developed in the project thus enables a much more flexible and accurate characterization than is possible with conventional measurement methods. This means that the measurement methodology obtained can also be used for further investigations and follow-up projects. The modeling work is limited to the accurate description of the trapping and detrapping dynamics of low-voltage components. By superimposing Gaussian weighted trapping and detrapping time constants, the model obtained enables a more precise description of the observed behavior without requiring an excessively large parameter set. However, the model can be extended in further work for hard-switching applications and validated with high-voltage components. With regard to the modeling of high-frequency components, physical simulation (TCAD) was combined with the derivation of analytical trap descriptions for the compact ASM-HEMT model and pulsed measurement technology. In this way, the influence of deep traps in the buffer and on the surface could be identified and modeled. One result of the project is that the physical simulation could be extended to the highly non-linear radio-frequency application. TCAD is used to extract the model parameters of an ASM-HEMT model that can represent the essential physics. The numerically complex simulations can then be carried out using a circuit simulator. This approach was used to clearly assign the degradation observed in the RF power range due to trapping to surface states for the present case. A subsequent optimization of the passivation confirms this prediction. Last but not least, a trap description based on the Shockley-Read-Hall theory was derived, which also allows the gate-side slow trapping to be simulated as a function of temperature, which is important for the simulation of switching processes.
Publications
-
Insight into Buffer Trap-Induced Current Saturation and Current Collapse in GaN RF Heterojunction Field-Effect Transistors. IEEE Transactions on Electron Devices, 67(12), 5460-5465.
Tripathi, Durgesh C.; Uren, Michael J. & Ritter, Dan
-
A Novel Approach for the Modeling of the Dynamic ON-State Resistance of GaN-HEMTs. IEEE Transactions on Electron Devices, 68(9), 4302-4309.
Weiser, Mathias C. J.; Hückelheim, Jan & Kallfass, Ingmar
-
Extension of ASM HEMT Model with Trapping Effects in GaN Power Devices. 2021 23rd European Conference on Power Electronics and Applications (EPE'21 ECCE Europe), P.1-P.9. IEEE.
Weiser, Mathias & Kallfass, Ingmar
-
Design Considerations for Fast On-State Voltage Measurement Circuits,” 2022 24th European Conference on Power Electronics and Applications (EPE’22 ECCE Europe), Hanover, Germany, 2022, pp. 1-9
M. C. J. Weiser, M. Rueß & I. Kallfass
-
A Fast ON-State Drain-to-Source Voltage Amplifier for the Dynamic Characterization of GaN Power Transistors. 2023 IEEE Applied Power Electronics Conference and Exposition (APEC), 637-644. IEEE.
Weiser, Mathias C. J.; Barón, Kevin Muñoz; Fink, Tobias & Kallfass, Ingmar
-
Dynamic RDModeling by Exploiting Gate Current Dependency of Virtual Gate Effect. 2023 18th European Microwave Integrated Circuits Conference (EuMIC), 1-4. IEEE.
Beleniotis, Petros; Zervos, Christos; Schnieder, Frank & Rudolph, Matthias
-
The role of gate leakage on surface-related current collapse in AlGaN/GaN HEMTs. 2023 18th European Microwave Integrated Circuits Conference (EuMIC), 297-300. IEEE.
Zervos, Christos; Beleniotis, Petros; Krause, Sascha; Ritter, Dan & Rudolph, Matthias
-
A Flexible Setup for Dynamic On-State Resistance Measurements of GaN HEMTs With One-Factor-at-a-Time Capability. IEEE Transactions on Power Electronics, 39(6), 7086-7095.
Weiser, Mathias C. J.; Köhnlein, Viktor & Kallfass, Ingmar
-
A Physics-Based Model for Slow Gate-Induced Electron Trapping in GaN HEMTs. IEEE Transactions on Electron Devices, 71(7), 4058-4065.
Beleniotis, Petros; Krause, Sascha; Zervos, Christos & Rudolph, Matthias
-
Investigation of Traps Impact on PAE and Linearity of AlGaN/GaN HEMTs Relying on a Combined TCAD–Compact Model Approach. IEEE Transactions on Electron Devices, 71(6), 3582-3589.
Beleniotis, Petros; Zervos, Christos; Krause, Sascha; Chevtchenko, Serguei; Ritter, Dan & Rudolph, Matthias
