AlGaN/GaN FinFET: A Comparative Study

A number of the circuit and design decisions that provide a reduction in overvoltage and dynamic power losses in power switches of multilevel voltage inverters with clamping diodes is considered. It is shown that the integral or unpackaged version is the most effective method for a reduction in the stray inductance in the multilevel circuit, which is the general cause for occurrence of dynamic overvoltage. To reduce dynamic losses of power, a method for resonant commutation on fronts, which provides commutation of switches in the multilevel circuit at zero voltage (SZV) and zero current (SZC), is proposed. The resonant switching method as applied to the multilevel circuit simultaneously solves a problem of dynamic overvoltage by means of limitation for the rate of voltage rise in the SZV mode or the mode of preliminary fault of energy stored in the inductance of the switching circuit in the SZC mode.

The article describes a method for determining static and dynamic power losses in power transistors, taking into account the influence of the module temperature. The results of a three-dimensional approximation of the energy characteristics of an IGBT module with a reverse diode of the type 2MBI150U2A-060, namely, the voltage drop between the collector and the emitter depending on the collector current and temperature, as well as the dependence of the turn-on energy, the turn-off energy and the return energy on the flowing current and temperature. The above expressions make it possible, by means of computer modeling in the Matlab / Simulink software environment, to determine the static and dynamic losses in the power switches of semiconductor converters of various topologies with different control systems, taking into account the influence of the temperature regime.

High-performance turbomachinery relies on tilting pad journal bearings (TPJBs) for high-speed operation while maintaining acceptable stability margins. However, TPJBs can produce large drag power losses that increase with machine size and surface speed. In a conventional configuration, the unloaded pads will drag on the surface of the shaft while, presumably, having a minimum impact on the bearing dynamic performance. The effects of removing unloaded pads from a TPJB have been studied before, but research has yet to describe the force coefficients of this bearing type. The present work aims to evaluate the effect on dynamic forced performance and power loss of a 4-pad TPJB configured as load on pad (LOP) and having the unloaded pad, opposite to the loaded pad, replaced by a fixed insert with a large clearance. The replacement insert has a significantly larger clearance (∼14.4 × nominal clearance), effectively making this a 3-pad bearing. The original 4-pad bearing was previously tested, hence allowing for an accurate comparison before and after the pad exchange. The experiments include operation at both 6 and 12 krpm (shaft surface speeds of 32 m/s and 64 m/s, respectively) with unit loads equaling 345 kPa, 862 kPa, and 1379 kPa. The bearing is supplied with ISO VG 46 oil (inlet T = 50 °C) and flow equaling 50%, 100%, and 150% of a nominal flow set for the original 4-pad bearing. For the lowest flow (50%) and a unit load of 345 kPa, the modified bearing shows a substantial decrease in drag power loss, ∼24% at 6 krpm and ∼19% at 12 krpm. For the largest supplied flow, however, the drag power loss increases, 15% at most. Higher churning losses within the 3-pad bearing internal cavities produce the power loss increase. On the other hand, the power reduction is always positive and with higher margins when comparing power losses derived from oil temperature rise measurements. The 3-pad bearing shows signs of lubricant starvation due to the large volume created by replacing the unloaded pad with the large clearance fixed pad. Incipient subsynchronous vibrations (SSVs) appeared at 100% nominal flow, at both 6 and 12 krpm and a unit load of 1379 kPa. SSV hash is a common occurrence in bearings supplied with a low flow rate and low loads. During tests with the lowest flow rate, the modified bearing produced very small stiffness (∼75% reduction) and damping (∼57% reduction) coefficients along the unloaded direction, most likely due to lubricant starvation.

Using a pulse density modulation (PDM) to regulate a load current of a resonant voltage-source inverter causes fluctuation of a current amplitude. The paper proposes a software phase-locked loop (SPLL) of control system in a resonant voltage inverter with PDM for induction heating equipment. The proposed SPLL changes a dead-time between control pulses of inverter transistors, taking into account current amplitude changes at the inverter output under PDM to provide a phase shift between its output voltage and current, which ensures lower dynamic power loss in these transistors. The paper presents results of experimental investigation of the proposed SPLL.

A comparison of different power electron switches for power electron converters is presented. The estimation method of steady and dynamic power loss in the switches is also described.

This study proposes an integrated eco‐driving controller, in which speed profile and power split optimisation are simultaneously treated. A speed profile controller is designed based on a multi‐phase dynamic programming (DP) algorithm, which is utilised to optimise speed trajectories with the detailed model of the traction power system, vehicle dynamics system, and signal constraints at intersections. The optimised speed and demand power profiles are designed as the reference of the power split optimal controller. A DP algorithm is presented to determine voltage trajectories of supercapacitors with power plant voltage limitations and dynamic power losses. The performance of the proposed integrated method is evaluated by numerical simulation based on an actual route of Chengdu Tramway Line 2. The energy reduction can be achieved by the speed profiles controller, compared to the eco‐cruising benchmark strategies. Energy efficiency can be further improved by the integrated eco‐driving controller. Moreover, the influences of green signal time windows and trip time on the shape of the optimised solution are investigated. For a given journey, the trip time with minimum energy consumption can be obtained. The integrated eco‐driving control method can be implemented by driver assistant systems or automatic train operation systems.

We have improved the performance of integrated antennas on Si for possible application in wireless communications and wireless interconnects. For practical VLSI integration, we have reduced the antenna size and optimized the proton implantation to a low energy of /spl sim/4 MeV with a depth of /spl sim/175 /spl mu/m. To avoid any possible contamination, the ion implantation is applied after device fabrication. Excellent performance such as very low RF power loss up to 50 GHz, record high 103 GHz antenna resonance, and sharp 5 GHz bandwidth have been achieved.

The low frequency power loss of two 1200V planar gate field-stop insulated gate bipolar transistor (IGBT) devices with different power rating is measured with double pulse test method. The switching loss and saturation loss with various collector current values are analyzed. The results show that power loss is strongly correlated with the device area and collector current. Device temperature is 6.6K lower for large area device when operating with collector current at 10A. The saturation power loss and turn-off trailing loss are discussed as well. Finally, the methods for power loss reduction are discussed from the standpoint of device structural design, layout optimization and manufacturing process.

The Elrod algorithm was implemented to study the effects of axial lubricant supply groove parameters, i.e., groove length, width, and location, on the performance of a one-axial groove journal bearing, taking into consideration thermal effects. The results and presented recommendations will assist bearing designers in determining the groove’s preferred angular location and main dimensions to achieve a solution in terms of low total power loss and high minimum film thickness. It is suggested that with regard to low total power loss and high minimum film thickness, for a lighter loaded bearing, the axial groove should be located in the angular range of 100–140 deg to the load line with a groove length ratio Lg/L over 0.6 or the angular range of 300–350 deg with an Lg/L over 0.75. For a heavier loaded bearing, it is suggested that the groove be placed at the angular range of 80–115 deg with an Lg/L over 0.6.

In this research project we provide one quick design method for a high current density power distribution board design and improvement on a high end server system. It is based on system power source to sink the current path for design PCB (Printed Circuit Board) copper shape dimension and thickness, and add some power or ground via for stitching each power or ground plane on a multi-layer PCB design to reduce DCR (Direct Current Resistance), and also considers airflow for cooling the power distribution board to make a lower temperature increase with lower power loss and a higher power efficiency with performance per watt.

Designing for low power can drive you crazy. What is commonly referred to as low-power design actually comprises two different, but related activities. Power minimisation seeks to reduce average and instantaneous power consumption. This process will consider different modes of operation, such as standby or sleep mode, as well as dynamic and leakage power losses. Power integrity management on the other hand takes a look at the delivery of power to various parts of the chip. This process aims to minimise any negative power effects on design timing and functionality. To optimise power effects, a system should be properly architectured from word go and successively refined as the design proceeds. Design software can be used at all stages of chip design to detect and address critical issues. Indeed, the latest EDA tools can support high level register-transfer-level (RTL), optimise power, timing, and noise characteristics at the physical level, and also verify cell-based electrical characteristics for verification and sign-off.

Development of a new soft ferrite core for power applications

Differential power processing (DPP) is a power electronics system architecture that configures dc–dc converters in parallel with the PV string to improve its power yield. The parallel nature of the DPP architecture brings a number of benefits, such as low converter power rating and low power losses, all of which make DPP especially suitable for submodule-level maximum power point tracking (MPPT). Meanwhile, microinverters typically perform only module-level MPPT and do not address power losses due to uncompensated submodule mismatch. In this paper, we introduce DPP converters into a microinverter system to improve its energy capture by recovering power losses due to submodule mismatch. The control method to interface DPP converters with a microinverter for submodule MPPT is presented and the tradeoff between tracking accuracy and control overhead is analyzed. A small-footprint DPP converter is designed. A digitally assisted windowed sensing technique is also implemented to address the challenge of precise current sensing. Altogether, the proposed solution seamlessly integrates DPP into the existing microinverter design. To demonstrate the effectiveness of the proposed solution, a hardware prototype has been built and tested with an off-the-shelf commercial microinverter. The improvement in energy capture with DPP converters has been experimentally verified.

A novel laterally and micro-electro-thermally actuated RF MEMS switch is presented in this paper. Despite many RF MEMS switches requiring continuous actuation voltage to hold the switch at ON-state, the proposed switch remains ON without any DC power. The mechanical design of the switch is in such a way that, the contact force of the switch is in mN range with zero DC power consumption. The switch is modeled and its operation is comprehensively analyzed. The fabrication process is completely compatible with standard MetalMUMPs process. The switch requires actuation voltages of 0.5 and 0.9 V for U-shaped and V-shaped actuators, respectively. The switch is normally OFF with isolation of ý 20 dB up to 100 GHz. The insertion loss of the switch at ON-state is better than ý 0.1 dB up to 40 GHz and the return loss is below ý 10 dB up to 100 GHz. The power consumption of the switch is confined to the transition time of the switch between OFF and ON states. Based on analytical and numerical simulations, the contact resistance of the switch is 0.028 ý. The proposed switch in this paper is dedicated for applications in mobile front-ends especially antenna switch networks where the signal loss and power consumption are the main constraints.

Wide bandgap (WBG) semiconductors including gallium nitride (GaN) and silicon carbide (SiC) offer significant performance improvement compared with conventional silicon power devices. The quasi-Z-source cascaded multilevel inverter (qZS-CMI) provides many advantages over the conventional CMI while applied in photovoltaic (PV) systems. In this paper, two solutions are proposed and compared to the design goal of a high efficiency and low-cost qZS-CMI based 1 MW/11 kV PV system. The first solution is based on 650 V GaN enhancement mode high-electron-mobility transistors (E-HEMT) and 650 V SiC Schottky diodes. The second solution uses 1200 V SiC power modules and 1200 V SiC Schottky diodes. The power losses and costs of the two candidate designs are compared in details. It is concluded that the first solution shows lower power losses and costs per quasi-Z-source inverter (qZSI) module. However, due to the low voltage rating of GaN E-HEMTs, more qZSI modules are needed to achieve the overall 11 kV inverter rating. Therefore, the second solution shows lower total power loss and cost in the medium-voltage, MW-scale qZS-CMI PV system.