F. Udrea's research while affiliated with Cambridge Consultants Ltd and other places

What is this page?


This page lists the scientific contributions of an author, who either does not have a ResearchGate profile, or has not yet added these contributions to their profile.

It was automatically created by ResearchGate to create a record of this author's body of work. We create such pages to advance our goal of creating and maintaining the most comprehensive scientific repository possible. In doing so, we process publicly available (personal) data relating to the author as a member of the scientific community.

If you're a ResearchGate member, you can follow this page to keep up with this author's work.

If you are this author, and you don't want us to display this page anymore, please let us know.

Publications (573)


Gate Robustness and Reliability of P-Gate GaN HEMT Evaluated by a Circuit Method
  • Article

May 2024

·

59 Reads

·

1 Citation

IEEE Transactions on Power Electronics

·

·

·

[...]

·

Yuhao Zhang

The small gate overvoltage margin is a key reliability concern of the GaN Schottky-type p-gate high electron mobility transistor (GaN SP-HEMT). Current evaluation of gate reliability in GaN SP-HEMTs relies on either the dc bias stress or pulse I–V method, neither of which resembles the gate voltage ( V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GS</sub> ) overshoot waveform in practical converters. This article develops a new circuit method to characterize the gate robustness and reliability in GaN SP-HEMTs, which features a resonance-like V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GS</sub> ringing with pulse width down to 20 ns and an inductive switching concurrently in the drain-source loop. Using this method, the gate's single-pulse failure boundary, i.e., dynamic gate breakdown voltage (BV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DYN</sub> ), is first obtained under the hard switching (HSW) and drain-source grounded (DSG) conditions. The gate's switching lifetime is then tested under the repetitive V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GS</sub> ringing, and the number of switching cycles to failure (SCTF # ) is fitted by Weibull or Lognormal distributions. The SCTF # shows a power law relation with the V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GS</sub> peak value and little dependence on the switching frequency. More interestingly, the gate's BV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DYN</sub> and lifetime are both higher in HSW than those in DSG, as well as at higher temperatures. Such findings, as well as the gate degradation behaviors in a prolonged overvoltage stress test, can be explained by the time-dependent Schottky breakdown mechanism. The gate leakage current is found to be the major precursor of gate degradation. At 125 °C and 100 kHz, the V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GS</sub> limits for a 10-year lifetime are projected to be ∼6 V and ∼10 V under the DSG and HSW conditions, respectively. These results provide a new qualification method and reveal new physical insights for gate reliability and robustness in p-gate GaN HEMTs.

Share

Fig. 1. Schematic of the five-channel AlGaN/GaN heterojunction PSJ device with long p-type GaN cap.
Fig. 2. (a) Schematic of the CV test structure. (b) Experimental and simulated C-V curves for a five-channel structure with pGaN cap. (c) Schematics of the multichannel AlGaN/GaN SBD. (d) Experimental and simulated forward I-V characteristics of SBD. (e) Two-dimensional cross section of the AlGaN/GaN MC 2 -HEMT device. (f) Experimental, simulated, and model OFF-state I-V curves of E-mode MC 2 -HEMTs with different L MC GD . The 103-µm device did not breakdown at experimental test limits.
Fig. 8. (a)-(d) X -direction electric field contour for different V D (with AlGaN doping). (e) X -direction electric field in profile at the bottom AlGaN/GaN interface of the device for different V D (with AlGaN doping).
Numerical Simulation and Analytical Modeling of Multichannel AlGaN/GaN Devices
  • Article
  • Full-text available

March 2024

·

70 Reads

IEEE Transactions on Electron Devices

In this article, we delve into the intricacies of carrier population and depletion in AlGaN/GaN multichannel structures through a combination of TCAD numerical simulations and theoretical modeling. Here, we take into account the unintentional or intentional n-type doping commonly seen in experimental multichannel devices. The impact of doping and layer thickness on the carrier population in the multichannel, as well as the carrier depletion and C – V characteristics, is investigated. The simulation results demonstrate that the doping in the AlGaN layer disrupts the balance between the 2-D electron gas (2DEG) and 2-D hole gas (2DHG), but the charge balance can be re-established with an additional pGaN layer on top of the multichannel structure. This pGaN layer can deplete the excessive electrons at high bias, resulting in a nearly flat electric field distribution. Moreover, an analytical model of the blocking electric field and conduction resistance in multichannel devices is built and validated by simulation. With this model, the impact of channel quantities on the device figure-of-merit (FOM) is studied and the optimization guideline of structure parameters is provided. This study provides key insights into the physics, performance space, and prospects of the multichannel gallium nitride (GaN) power devices.

Download

Fig. 1 | System description. a, Cutaway CAD image showing the key components of the 24-channel odor delivery device, including the air filter, diaphragm pump, solenoid valves and odorant cartridge. b, Pneumatic diagram illustrating the airflow through the components of the odor delivery device. Air from the surrounding environment is filtered, pumped and channelled by a bank of 24 solenoid valves to an odor reservoir where it is odorized before reaching an outlet. c, A system diagram of the electronic control circuitry. A Bluetooth / microprocessor module (Raytac, MDBT50Q-1MV2) is used for system control and communication. Digital lines are interfaced to higher voltage (12 V) drivers for activation of the solenoid valves and diaphragm pump. A pressure sensor is used for diagnostic tests during system operation. d, An image of the assembled device and its interaction with the user and mobile application. The device is housed in a 3D printed plastic case and contains an odorant cartridge which slides onto fixing posts and is screwed into place within the assembly. Individual outlet pipes are held using an adapter. A mobile app is used for system control and for recording perceptual data from users (stored on a cloud database). e, A screenshot showing the mobile app used for system control and recording perceptual data from users of the smell delivery device. After each odor exposure, users are asked a series of questions including inviting them to rate the smell intensity, character and familiarity.
Fig. 2 | System performance. a, Transient response of the photo ionization detector (PID) to an odor exposure generated by the odor delivery device with a 3 s activation time (activation times indicated by the dotted lines). b, Temporal stability of the odor intensity generated by the odor delivery device over a 1 hour time window with 3 s activation times. c, Thermal response, showing the variation in measured odor intensity from the odor delivery device with ambient temperature. d, Spatial distribution of odor intensity, measured over a distance of 40 mm from the outlet of the odor delivery device, in the direction of odor flow. e, Spatial distribution of odor intensity across the path of odour flow, measured at a distance of 100 mm from the outlet of the odor delivery device. f, Measured odor intensity for the SMELL-S odour stimuli. g, An example of a subject's performance during the SMELL-S test, for a patient with smell loss treated at the Geneva University Hospital (red line) and a subject with an intact sense of smell (green line). We used an adaptive staircase threshold paradigm to measure olfactory threshold. The task is becoming more and more difficult as the level is approaching 10. A reversal occurs when the direction in which the concentration is changed reverses (blue dots). The score is defined as the average of the last four reversals (grey dotted lines). h, Violin plots showing the duration's for the Sniffin' Sticks and SMELL-S threshold tests.
Fig. A4 | Cost breakdown. A pie chart showing the relative cost of the different system components making up the odor delivery device. The cost of the custom machined parts dominates.
Odor delivery system approaches.
Odor delivery device specifications.
Multi-channel portable odor delivery device for self-administered and rapid smell testing

February 2024

·

65 Reads

To improve our understanding of the perception of odors, researchers are often required to undertake experimental procedures with users exposed to multiple odors in a variety of settings, including to diagnose smell loss in clinics and care homes. Existing smell tests are typically administered using multiple sniffing pens manually presented to patients by a highly specialized nurse using a time-consuming and complex testing paradigm. Automated odor delivery devices, such as olfactometer systems, exist but are expensive, bulky and typically lab based, making them difficult to use for “on the ground” odor delivery. We have developed a portable, affordable, odor delivery device that can deliver 24 odors through individual channels with high temporal precision and without cross-contamination. The device allows for the fast, flexible sequencing of odors via digital control using a mobile application and has been experimentally validated in the lab, as well as tested on patients. The novel design provides several unique advantages for investigating olfactory perception and offers the possibility that users can one day self-administer smell tests in a range of settings, including at home, allowing smell healthcare services to evolve and become part of a routine practice and self-care culture.









Citations (64)


... However, high dV/dt also causes EMI noise that can interfere with nearby circuitry, such as microcontrollers, memories, or measurement equipment, indirectly impacting the reliability of SiC IGBTs. It also makes power converter design more complex in various ways as was explained in [32]. Therefore, controlling the dV/dt on the device level will reduce the time and effort required for converter development, improve its ruggedness, and optimize costs. ...

Reference:

Silicon Carbide n-IGBTs: Structure Optimization for Ruggedness Enhancement
10kV+ Rated SiC n-IGBTs: Novel Collector-Side Design Approach Breaking the Trade-Off between dV/dt and Device Efficiency
Key Engineering Materials

Key Engineering Materials

... These techniques include self-aligned channel [2,3], self-aligned source contact [4], specific doping on JFET region [5], and current spreading layer [6]. On the design front, cell topology and pitch reduction play vital roles in further improving the performance of VDMOSFETs [7][8][9][10][11][12][13][14][15][16]. ...

Clamped and Unclamped Inductive Switching of 3.3 kV 4H-SiC MOSFETs with 3D Cellular Layouts

... To achieve a lower R on,sp and V on , a thinner drift layer and the use of lower work-function metal (like TiN, Mo, Ti) can be employed. In addition, a super-junction, a novel structural concept with a challenging fabrication process, can be considered, which further relaxes the stress relationship between V on and reverse leakage current, as extensively reported in SiC, [24][25][26] and sparingly reported in GaN, 27) β-Ga 2 O 3 . 28) ...

First Demonstration of Vertical Superjunction Diode in GaN
  • Citing Conference Paper
  • December 2022

... The heat transfer from a heated structure to the surrounding gas will be higher for gases with a higher thermal conductivity. Thermal conductivity-based gas sensors have advantages, including eliminating catalysts and adsorbents [5,6]. Moreover, thermal conductivity gas sensors have been found in many applications, including coal mines, natural gas monitoring (for methane sensing), and light gases such as hydrogen [7]. ...

Micromachined Thermal Gas Sensors—A Review

Sensors

... This may lead to the inapplicability of the gate lifetime obtained from DC tests to device operations in power converters. Recently, a circuit method is proposed to evaluate the gate lifetime under the inductive load switching [225], the condition of which is not accessible by DC or pulse I-V methods. As the parasiticinduced gate overshoot has a resonance nature, this method features a resonance-like gate ringing with the pulse width down to 20 ns and an inductive switching concurrently in the drain-source loop. ...

Dynamic Gate Breakdown of p-Gate GaN HEMTs in Inductive Power Switching
  • Citing Article
  • January 2022

IEEE Electron Device Letters

... During the fast voltage rising phase, the voltage can increase as fast as several hundred kV/μs, which can cause long-term reliability issues. Our previous work [29], explained in detail that positive and negative voltage spikes are induced into the gate electrode due to this high dV/dt which can either cause unintentional turn-on of the IGBT or stress the gate oxide. This issue is particularly severe in bridge-leg configurations ( Fig. 6(a)), such as DC/DC converters or inverters, where positive and negative gate voltage spikes can be produced in the low-side switch during the switching transient of the high-side switch and vice versa ( Fig. 6(b)) leading to short-circuiting of the DC bus. ...

Influence of Emitter Side Design on the Unintentional Turn-on of 10kV+ SiC n-IGBTs
  • Citing Conference Paper
  • October 2022

... In the plateau region, while the drain voltage (V DS ) slowly decreases, the freewheeling Schottky diode begins to form a depletion region to support 400 V. As a result, the displacement currents from the Schottky diode are added to the drain current, leading to an instantaneous increase in the MOSFET's drain current, as shown in Fig. 8 [22]. ...

Gate Ringing in Superjunction MOSFETs with a Parasitic Capacitance in the Load Inductor
  • Citing Article
  • November 2022

Power Electronic Devices and Components

... Furthermore, 10 400-V BV was obtained on HEMTs with an L GD of 125 µm by means of a thick poly-AlN passivation layer [21]. Furthermore, pioneering multidimensional device architectures were proposed to balance V BD and R ON [22]. ...

Multidimensional device architectures for efficient power electronics
  • Citing Article
  • November 2022

Nature Electronics

... An initial attempt to present the short circuit behaviour of a the conventional structure of a 10 kV-rated IGBT was done in [38] and showed that the device can fail due to parasitic thyristor latching, caused by a positive feedback mechanism that increases temperature and current density, and high leakage current flowing inside the IGBT during the blocking state after the short-circuit has been detected and the device turned off. The latter mechanism can cause the IGBT to fail several > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < hundred microseconds after it has been turned off, a behaviour that is similar to what has been observed in silicon IGBTs [39] and SiC MOSFETS [40]. Figure 13 shows that the same failure mechanisms exist in the 27 kV validated IGBT during a short circuit event under hard switching at a DC voltage of 14 kV, gate to emitter voltage of 15 V and stray inductance of 100 nH. ...

Short-Circuit Performance Investigation of 10kV+ Rated SiC n-IGBT
  • Citing Conference Paper
  • September 2022

... The parallel connection of a sense transistor (sense-FET, also called sense-HEMT or sense-GaN for GaN HEMTs [30]) does not require the critical loop to be cut, and can be also monolithically integrated with the GaN IC. GaN-based sense-FET ICs have been already demonstrated monolithically [31][32][33][34][35][36][37] and also by external wiring of several GaN devices [30,38] for single discrete transistors and used in half-bridges. This work focuses only on current-mirror sensing; other current sensing methods such as shunt-sensor [39][40][41][42][43] or hall-sensor [44] integration, as well as related magnetic flux concentrators [45] or magnetic sensors [46], were also demonstrated in GaN technologies. ...

ICeGaNTM technology: the easy-to-use and self-protected GaN power IC
  • Citing Article
  • November 2022

Power Electronic Devices and Components