This paper addresses the influences of device and circuit mismatches on paralleling the silicon carbide (SiC) MOSFETs. Comprehensive theoretical analysis and experimental validation from paralleled discrete devices to paralleled dies in multichip power modules are first presented. Then, the influence of circuit mismatch on paralleling SiC MOSFETs is investigated and experimentally evaluated for the first time. It is found that the mismatch of the switching loop stray inductance can also lead to on-state current unbalance with inductive output current, in addition to the on-state resistance of the device. It further reveals that circuit mismatches and a current coupling among the paralleled dies exist in a SiC MOSFET multichip power module, which is critical for the transient current distribution in the power module. Thus, a power module layout with an auxiliary source connection is developed to reduce such a coupling effect. Finally, simulations and experimental tests are carried out to validate the analysis and effectiveness of the developed layout.
7 July 2015 in IEEE Transactions on Industrial Electronics
More and more wind turbine manufacturers turn to using the full-scale power electronic converter due to the stricter grid code requirements to thoroughly decouple the generator from the grid connection. However, a commonly used type of this generator is still unclear, where the selections of the low-speed (LS; direct-drive) and medium-speed (MS; one-stage) permanent-magnet synchronous generators (PMSGs) are both promising solutions. This paper will assess and compare the reliability metrics for the machine-side converter (MSC) for those two configurations. First, a translation from the mission profile of the turbine to the current and voltage loading of each power semiconductor is achieved based on synchronous generator modeling. Afterward, a simplified approach to calculate the loss profile and the thermal profile is used to determine the most stressed power semiconductors in the converter. Finally, according to the lifetime power cycles, the lifespan can be calculated when operating in various wind classes. It is concluded that, although the LS PMSG is able to eliminate the gearbox, the lifespan of its MSC is lower than the one-stage MS generator.
18 March 2010 in IEEE Transactions on Industrial Electronics
Inverter systems that feed electrical power from fuel cells into the grid must convert the direct current of the fuel cell into the alternating current of the grid. In addition, these inverters have to adapt the different voltages of the fuel-cell system and the grid to each other. In this paper, different topologies of appropriate inverter systems in the medium power range of 20 kW and higher are presented briefly. The inverter operating behavior, power rating, and efficiency are compared. The power rating and efficiency are compared using an analytical calculation of the semiconductor losses. The study includes transformerless inverters as well as two-stage inverter systems with high-frequency transformers (dc/dc converter combined with an inverter). This paper compares converter systems using insulated-gate bipolar transistors (IGBTs), e.g., a boost converter in series with a voltage-source inverter (VSI), current-source inverter, and z-source inverter or converter systems using superjunction MOSFETs, such as voltage- and current-fed full-bridge converters or a boost converter with an autotransformer. The MOSFET-based dc/dc converters must be connected in series to a VSI with IGBTs to feed into the three-phase grid. The presented converters were tested in the laboratory. Some characteristics of their laboratory performance are shown.
Distributed power generation systems (DPGS) such as wind and solar become more and more widely spread. As a consequence grid operating companies demand system services. As part of the general fault ride through (FRT) requirements this paper deals with low voltage ride through (LVRT) capability of a three-phase-four-wire grid-tied solar inverter. The standard system will be described and necessary changes to the control such as positive (PS), negative (NS) and zero sequence (ZS) separation, a stable phase-locked-loop (PLL) as well as voltage support by means of reactive current as well as stress factors to the hardware will be identified.
1 June 2008 in 2008 IEEE Power Electronics Specialists Conference
In this paper the power losses of both Z-source- inverter (ZSI) and voltage-source inverter linked with a dc/dc boost inverter (VSI+BC) are compared. Therefore the circuits are simulated in time domain by means of numerical methods whereas the switching and conducting losses are calculated separately and in parallel. A main focus is on mathematical description of the current rms through the IGBTs of the ZSI to derive the conducting losses. It is shown that the losses of the diode in the dc link of the ZSI has a high percentage of the total power losses and cannot be neglected. Besides that the choice of the modulation method has a high influence on the power losses of the ZSI as well. Based on the calculated power losses the efficiency and the installed semiconductor power are investigated and compared for different input voltages. The results show that for an output voltage of 400V the VSI+BC topology outperforms the ZSI in almost all operating points. The reason is that for low input voltages the VSI+BC topology boosts the input voltage to limited 600 V and the VSI works with a modulation index of about 1 very efficiently. However the ZSI has to boost the input voltage to much higher values, since for high voltage gains long shoot- through-states are essential. However, long shoot-through- states are coming along with long zero-states which causes a buck behavior.
A physics-based analytical model for GaN high-electron-mobility transistors (HEMTs) with non-recessed- and recessed-gate structure is presented. Based on this model, the two-dimensional electron gas density (2DEG) and thereby the on-state resistance and breakdown voltage can be controlled by varying the barrier layer thickness and Al mole fraction in non-recessed depletion-mode GaN HEMTs. The analytical model indicates that the 2DEG charge density in the channel increases from 2.4 × 1012 cm−2 to 1.8 × 1013 cm−2 when increasing the Al mole fraction from x = 0.1 to 0.4 for an experimental non-recessed-gate GaN HEMT. In the recessed-gate GaN HEMT, in addition to these parameters, the recess height can also control the 2DEG to achieve high-performance power electronic devices. The model also calculates the critical recess height for which a normally-ON GaN switch becomes normally-OFF. This model shows good agreement with reported experimental results and promises to become a useful tool for advanced design of GaN HEMTS.