Accurate calculation of losses incurred in the power electronics converter is particularly essential in selection of devices and optimal thermal design. This paper deals with the loss analysis in space-vector modulated voltage source inverters, which is accomplished by meticulous formulation of currents flowing through the devices in the converter. Inverter dead-time is integral for safe operation of devices. Introduction of dead-time leads to undesired effects, pronounced effect of dead-time is seen in phase current. Hence, in estimation of losses in the converter the effect of dead-time must be taken into consideration. This paper presents a method to find the actual phase current. Device currents are then devised using the actual current and modified modulating signal. These equations facilitate for quick calculation of conduction losses in the converter at different load conditions. The effectiveness of the derived expressions is validated experimentally using 415 V, 5 kW SiC MOSFET based voltage source inverter setup.
The increase in hydrogen production to support the energy transition in different sectors, such as the steel industry, leads to the utilization of large scale electrolyzers. These electrolyzers have the ability to become a fundamental tool for grid stability providing grid services, especially frequency regulation, for power grids with a high share of renewable energy sources. Alkaline electrolyzers (AELs) have low cost and long lifetime, but their slow dynamics make them unsuitable for fast frequency regulation, especially in case of contingencies. Proton Exchange Membrane electrolyzers (PEMELs) have fast dynamic response to provide grid services, but they have higher costs. This paper proposes a dynamic power allocation control strategy for hybrid electrolyzer systems to provide frequency regulation with reduced cost, making use of advantages of AELs and PEMELs. Simulations and experiments are conducted to verify the proposed control strategy.
In order to enhance the saliency-based position sensorless drives with a single current sensor (SD-SCS), this paper proposes a current derivative-based sensorless control using a tri-active vector pulse width modulation (TAVPWM). In the SD-SCS, the performance of the position estimation is deteriorated due to the injected voltage at the current reconstruction dead zone (CRDZ) and the unaligned sampling with a single current sensor. To improve the performance of the SD-SCS, the TAVPWM scheme is adopted to eliminate the CRDZ in the low-modulation region. Furthermore, the phase current derivatives, which are accurately reconstructed at each active vector of the TAVPWM, are utilized to minimize the position estimation error resulting from the unaligned current sampling. Owing to the proposed method, the switching frequency has been increased up to 16 kHz, which eliminates the audible noise virtually in the SD-SCS. Lastly, the enhanced dynamic performance of the SD-SCS is verified by various experiments.
Data center power consumption has been increasing remarkably in the last decades, mainly due to the massive adoption of cloud computing. Due to the vast amount of power consumed, server rack architecture has switched from 12 V to (48 V – 60 V). The commonly used power delivery system employs two stages of DC-DC conversion, cascading an unregulated first-stage converter and a regulated one. The 4-to-1 switched tank converter (STC) is one of the main topologies used as a first stage thanks to its very high efficiency and power density. However, a limitation of the STC is its large inrush current during the startup phase. To avoid this, usually the converter is preceded by a DC-DC converter, a hot-swap controller or an eFuse. Moreover, ensuring the zero current switching condition in both resonant tanks is not straightforward in presence of a mismatch between the resonant frequencies. In this paper, a novel control technique that avoids a large inrush current at startup without the usage of an auxiliary converter and a strategy to minimize the mismatch in the resonant frequencies are proposed for the STC. Experimental results for a 600 W prototype show the validity of these approaches.
Press-pack power module gradually replaces traditional power module in high power applications such as High Voltage Direct Current and Flexible AC Transmission Systems. There are many semiconductor chips connected in parallel inside press-pack power module. The uneven distribution of clamping force and temperature will cause the imbalance current distribution of semiconductor chips, which may lead to the overcurrent of some chips. Since press-pack power module is usually used in high power converters, online current monitoring should be applied to provide fault diagnosis. However, the accurate chip-level current sensing inside press-pack power module is difficult due to the size and bandwidth of sensors. To solve this problem, a compact hybrid current sensor for chip-level online sensing, which consists of anisotropic magnetoresistance (AMR) and compact planar Rogowski coil is proposed in this paper. AMR sensor is used to sense low-frequency and DC current, and the high-frequency part is measured by Rogowski coil. The parameters of current signal processing circuit are designed by calculation and simulation. The hybrid sensor senses the current of a single-chip press-pack IGBT in buck converter and verified by a commercial current probe.
This paper proposes a novel current controller for three-phase grid-connected inverters under unbalanced conditions, which is based on the series virtual impedance algorithm. Unlike the conventional methods, the transformation of unbalanced current to positive and negative sequence components is not required. Thus, the current controllers are only two, considering the controller is applied in the d-q synchronous reference frame (SRF). Here, the current controller regards unbalance as disturbance, so current reference and disturbance suppression responses can be separated. By doing separation, the controller has two degrees of freedom, meaning the suppression of disturbance does not affect the reference response. Simulation and experiments have been done to justify the effectiveness of the proposed method, which can suppress the current unbalance factor from 23.33% to 0.13 % and current THD from 8.9 % to 3 %.
This paper presents the degradation testing and failure mechanisms analysis of metalized film capacitors used for AC filtering in MW power converters. Based on more than 2,800 hours of accelerated testing under accelerated AC voltage, temperature, and AC current, various electro-thermal parameter data are recorded. The results reveal that capacitance values have negligible reduction until the testing samples catastrophically fail. The capacitor hot spot temperature and case temperature are measured along the testing, which are increasing. The observations provide a new perspective on the possible failure mechanisms and condition monitoring of film capacitors in AC filtering applications.
This paper presents a Maximum Torque per Ampere (MTPA) control algorithm for an interior permanent magnet synchronous motor (IPMSM) drive considering the PM flux linkage change due to permanent magnet (PM) temperature variation. Reluctance torque and PM torque are generated from an IPMSM and parameter variations affect development of torque, efficiency and reliability of IPMSM drives. In this paper, a PM flux linkage are estimated in real time via a Gopinath style stator flux linkage observer and a torque error correction factor is calculated from the estimated PM flux linkage. A 2-dimensional (2D) MTPA look-up table (LUT) is developed to achieve the MTPA trajectory reflecting PM flux linkage variation for compensating torque error occurred by parameter variation. The proposed IPMSM control algorithm is verified through simulations and experiments.
Detecting power-device State-Of-Health (SOH) during converter operation can enhance system reliability by predicting imminent failure scenarios. While various aging indicators for GaN devices have been demonstrated in the literature, few are practically measurable in an active converter. This paper demonstrates that the large-signal device output capacitance $(C_{\text{oss}})$ is a reliable indicator of short-circuit (SC) aging, and proposes an insitu measurement technique to capture its value by leveraging the operational waveforms of soft-switching converters. Experimental results show a 5% decrease in the large-signal $C_{\text{oss}}$ after 5000 SC cycles, proving the usefulness of this parameter as an SOH indicator. The in-situ measurement technique is demonstrated in a synchronous buck converter operating in discontinuous conduction mode, successfully capturing the SC-aging-induced change in $C_{\text{oss}}$. The presented results and proposed measurement technique pave the way for system-level monitoring of power-device SOH and self-calibrating operation.
Digital controls are generally characterized by significant phase delays due to the analog-to-digital conversion process, sampling time, algorithm computation time, and the digital pulse-width modulator's architecture. Usually, the delay introduced by the latter has a more significant impact than previous ones, especially when approaching the switching frequency. The multi-sampling operation is largely adopted to reduce this delay. Recently proposed, double-sampling asymmetrical dual-edge (ADE) carrier-based digital pulse-width modulators (DPWM) have proven to operate with zero phase delay. Therefore, the multi-sampling architecture might provide even better results. Unfortunately, its documented operating point dependence does not allow such modulators to be used effectively. This article examines an improved multi-sampling ADE-DPWM architecture where the dependency on the operating point is significantly reduced. This manuscript also includes an accurate small-signal transfer function model. The proposed architecture and the developed small-signal model are validated in simulation and experimentally. In addition, experimental tests on a multi-loop voltage-controlled single-phase voltage-source inverter revealed the advantages of the proposed architecture over the more traditional DPWM one based on a trailing-triangle edge carrier.
High voltage (HV) batteries, from 400 V up to 800 V, are a recent breakthrough in the automotive field. These batteries along with the required power converter are able to provide higher efficient systems with enhanced power density, at the same time faster charging speeds can be achieved. Depending on the automobile manufacturer and the electric vehicle type, multiple DC buses of HV and LV are employed simultaneously, which demands several conversion stages. For this purpose, isolated multiport DC/DC converters are a promising solution to adapt these different HV and LV levels while allowing a reduced number of conversion stages and high efficiency. This is only achievable because a multiwinding transformer (MWT) is used to couple the multiple cells and hence compose the overall multiport DC/DC converter. Therefore, in this paper, a GaN-based multiport resonant converter with a wide input voltage range is investigated to interconnect multiple HV and LV DC buses. To reduce the voltage and current effort of the GaN devices, the approach employs an input-series and output-parallel interconnection (ISOP). Further, as a novel approach, only one resonant tank on the input side is used to reduce parameter deviations among the multiple ports. Finally, the investigated topology is validated in simulation as well as by means of a hardware demonstrator with a peak efficiency of 94.23 %.
In this paper, a novel single-stage single-phase microinverter with a beat frequency modulation (BFM) and fully soft-switching operation is presented for photovoltaic (PV) applications. To convert the low-level DC voltage of the PV panel to the desired AC voltage of the grid, a DC-AC converter is required to amplify the input DC voltage and also make a sinusoidal AC voltage at the output. The proposed microinverter is configured based on a resonant LLC converter to provide both amplification and isolation. A new frequency modulation is applied to the converter to provide a pure sinusoidal waveform at the output of the microinverter. The soft-switching conditions are provided for all switches of the converter due to the utilization of the resonant elements. The rectifier and unfolder at the output stage of the microinverter are merged to utilize the minimum number of semiconductors. A new control strategy is employed to provide the output current regulation based on a single-phase DQ current controller. The proposed beat frequency modulation and topology derivation of the microinverter are discussed in this paper. To validate the theoretical analysis, a 250W prototype is implemented and experimental results are represented.
Litz wires are widely used in wireless charging stations of light-duty electric vehicles. In this work, several alternatives to the traditional copper litz wires are evaluated. Specifically, aluminum litz wires, copper-clad aluminum litz wires and copper tubes are considered as alternatives to copper litz wires for the mentioned application. Potential benefits of the considered alternatives can include weight and cost savings. However, these advantages could not be enough if the resistance of the magnetic coupler is deteriorated. For this reason, a model based on the combination of analytical formulas and finite element simulations for calculating the resistance of each cable is proposed. The model is verified with resistance measurements with several prototypes. The results point to some conclusions about the suitability of the frequency range of use.
The unbalance loss distribution of the three-level (3-L) T-Type inverter increases the cost and reduces the overall reliability. A novel loss-balance modulation scheme is proposed for 3-L T-Type inverter in this paper. With the proposed modulation scheme, the switching loss and conduction loss of all switches in the 3-L T-Type inverter can be balanced respectively regardless of the power factor and modulation index. This method is independent of the power factor and the modulation index, and only the on-resistance of devices are required. With the proposed method, the total loss of the inverter can also be reduced, and the neutral point potential can be balanced naturally. The proposed modulation scheme is validated in a 200 kW SiC-based 3-L T-Type inverter.
The four-switch buck-boost (FSBB) topology is often used in combination with other isolated converters to extend the voltage range capability of the overall structure. In such applications, the duty-cycles of the two legs of the FSBB are independently controlled, and a phase-shift is introduced in order to shape the inductor current ripple and thus achieve zero voltage switching. This paper proposes a non-linear average model, and the corresponding linearized small-signal model, for the FSBB operated in the described way. The derived average and small-signal models are shown to be in excellent agreement with simulation and are also validated by measurements on an experimental prototype.
In this paper, a novel method for modeling and simulation of large-scale energy storage systems (ESS) is provided. Specifically, the model is developed for large-scale series connected supercapacitors (SCs) intended for power electronic applications. This method is especially useful for high voltage applications where a large number of series connected energy storage units (ESUs) are required. The proposed solution reduces a multi-node string of series connected SCs–together with their corresponding voltage balancing circuit–to a single unit with two electrical nodes. The proposed model is connected to the dc link of a three phase grid-connected modular multilevel converter (MMC). In this system, the effectiveness of the proposed model and the proposed voltage balancing scheme is demonstrated for a string comprising ten thousand series connected ESUs. The efficacy of the proposed model and the balancing algorithm is proven by simulations in the MATLAB/Simulink environment.
This paper introduces the packaging and characterization of a novel die-integrated PCB SiC MOSFET half-bridge module. Due to the use of standard PCB technology, the proposed module is manufacturing-friendly and very cost-effective. Two 1.2 kV SiC MOSFET dies are fully embedded, resulting in over 95% size reduction compared to the state-of-the-art half-bridge module with the same voltage rating. The weight of the power module is reduced by 90% to only 0.06 oz. The power loop inductance is optimized and minimized to 2.3 nH while the gate loop inductance drops to 3.8 nH. Copper filling via PCB is further utilized to reduce thermal resistance. The thermal and electric performance of the developed module is presented in the paper. The insulation capability is also verified to be two times higher than the rated voltage.
Increased disturbances in the AC power system are caused by the rapid charging of high-capacity battery packs, hence increasing the requirement for efficient, low-distortion smart chargers. Therefore, a battery charger that can react to the battery state and adjust its charging process accordingly is essential. In this context, the application of the LLC resonant converter employed in battery chargers is considered in this paper. A hybrid control strategy that integrates both the pulse frequency modulation (PFM) and the phase shift modulation (PSM) is proposed, which can handle a wide voltage range of the battery. This control method includes non-overlapping (NOVL) mode of operation. In this study, the input voltage of the LLC converter is assumed to be fluctuated. When the control state changes due to the fluctuation of the input voltage in the NOVL mode, a surge current is observed in the output current. This surge causes heat generation and deterioration of the battery. Therefore, an alternative overlapping (OVL) mode of operation is further proposed as a new control state change method to curtail the surge current in the charging process. Finally, the obtained experimental results confirmed the surge current reduction are introduced, which also demonstrated the effectiveness of OVL mode over NOVL mode control strategy.
In this paper, a high boost factor single phase switched-capacitor five-level boost inverter (1P-SC5LBI) has been presented. In this topology, one more capacitor and diode have been added to the conventional single-phase H-bridge following DC/DC boost converter. By this way, the introduced inverter can enhance the boost factor to twice of that in conventional inverters. Beside conventional three voltage levels of H-bridge inverter, two extra voltage levels have been introduced to improve quality of output voltage. Simulation and experimental results based on 150-W laboratory prototype have been presented to verify the operating of proposed inverter.
