With the goal of promoting the technology of power electronics, CPSS TPEA is expected to be a favorable platform to strengthen information exchange in this area through publishing and disseminating research findings worldwide. Authors are cordially invited to submit your papers through the website.

The objective of this journal is to finish the 1st round review within 6 weeks from submission.

CPSS TPEA Vol.11 No.3 (September 30, 2026)

Regular Papers

Investigation of GaN HEMT Degradation in High Temperature Power Electronics Using Application Level Reliability Test

M. A. C. N., L. Hu, M. J. Kim, and B. Akin

Abstract Extreme environmental and high-power density applications, such as those in oil and gas industry, rely on power electronic circuits to drive sensors and control systems. These circuits are often subjected to harsh thermal conditions and must demonstrate high reliability to ensure uninterrupted operation. Gallium nitride (GaN) based semiconductor switches offer several advantages over conventional silicon (Si) devices, including a higher breakdown electric field and superior electron mobility, enabling more efficient and compact power electronics. However, as a relatively new technology with immature epitaxial processes, GaN reliability remains a critical concern, particularly in applications where unexpected downtime can lead to substantial economic losses. In high temperature environments, this challenge is further amplified by the frequent need for devices to operate beyond their datasheet-specified limits. In this work we explore the reliability of different GaN power semiconductor technologies in these conditions using realistic test procedure that is very closer to an actual converter operation. Devices were stressed for extended durations of up to 1000 h with their static parameters continuously monitored. The collected data were analyzed to identify failure modes and their underlying root causes, providing insights into the robustness of GaN devices under extreme condition.

High Power Density High Current Density LLC-DCX Converter for Datacenter Application

Y. Cao, B. Wang, Z. Zhu, M. Xu, and D. W. Zhang

Abstract This paper presents a high-efficiency LLC resonant bus converter with a 10:1 step-down ratio, tailored for 48 V datacenter applications. To address the challenges of power density and current density, a compact design was achieved by employing a single-PCB planar transformer and a high-density integration scheme, significantly reducing the converter's footprint compared to conventional approaches. Furthermore, a primary winding connection method for current sharing and a copper-plated secondary termination design were proposed. The implemented prototype demonstrates excellent performance: with a 48 V input, it delivers a maximum output current of 85 A and a peak efficiency of 98.8%; under 60 V input conditions, the converter achieves a current density of 0.29 A/mm² and a power density of 3548 W/inch³.

A Wideband Modeling Method for Magnetic Components Considering Strong/Weak Near-Field Coupling Characteristics

X. Huang and K. Fu

Abstract Magnetic components in switching power supplies generate severe leakage electromagnetic fields, which not only affect the filtering performance of EMI filters but also may disrupt the normal operation of circuits. This paper first analyzes the mechanism of near-field leakage generated by magnetic components. It investigates the effect of frequency-dependent permeability on the self-parameters of these components and further explores how the self-parameters influence mutual parameters. By integrating both effects on near-field coupling, a wideband near-field coupling model for magnetic components is established, which accounts for the transmission characteristics of both magnetic and electric field coupling. Insertion loss is measured using a network analyzer and converted into Z-parameters to extract the values of equivalent inductance, equivalent capacitance, and equivalent resistance in the self-parameter module and mutual-parameter module. The near-field transmission characteristics of magnetic components under both strong and weak coupling scenarios are tested to verify the accuracy of the proposed wideband near-field coupling model, which exhibits good conformity in the frequency range from 100 kHz to 60 MHz. The results indicate that magnetic field coupling and electric field coupling dominate the near-field transmission characteristics of magnetic components in different frequency bands.

A Current Measurement Offset Error Elimination Method Based on Signal Injection for Interior Permanent Magnet Synchronous Motor Drives

J. Zhao, H. Wang, S. Wan, R. Liu, and H. Zou

Abstract Accurate information of the measured phase current is of high importance for interior permanent magnet synchronous motor (IPMSM) drives. However, under the long-term effects of electrical and thermal stress, the current sensors may suffer from the current measurement offset errors (CMOEs), which causes obvious speed fluctuations and torque ripples. To address this, a current measurement offset error elimination (CMOEE) method based on the signal injection is proposed in this paper, which enables to maintain the performance of the IPMSM drives. In the proposed CMOEE method, the effects of the CMOEs on the performance of the IPMSM drives are analyzed. On the basis of this, the details of the proposed CMOEE method are provided. Then, considering that the response of the injected signal is extracted by the band-stop filter, obvious amplitude attenuation and phase deviation are introduced into the proposed CMOEE method. With this, the performance of the proposed CMOEE method may be degraded. Accordingly, a simple yet effective compensation method is adopted. Consequently, the amplitude attenuation and phase deviation are effectively compensated, ensuring the performance of the proposed CMOEE method. Finally, experimental tests are carried out to verify the effectiveness of the proposed CMOEE method under different cases.

Research on Wide-Range Core Loss Measurement Model and Method Based on DC Power Method

Q. Chen, W. Qiu, J. Li, Q. He, X. Guo, and W. Chen

Abstract To address the engineering bottleneck of frequent recalibration in core loss measurement using the DC power method that caused by the narrow prediction range and low accuracy of existing inverter loss models, this paper proposes a novel wide-range calibration model. The model comprehensively incorporates multi-modal 

loss mechanisms, including hard- and soft-switching, as well as the dependence on duty cycle. By imposing physical constraints on the fitting parameters, a general model applicable to both asymmetric and symmetric pulse width modulation (PWM) excitations is established. Experimental results demonstrate that over a wide measurement range of 20–100 kHz switching frequency, 0.2–0.5 duty cycle, and 1.01–43.01 V input voltage, the proposed model requires only a single calibration to achieve an average relative error of 1.80%. Compared with existing methods, the maximum relative error is reduced by at least 16.27% and up to 59.07%.

Near-Field Coupling Analysis and Modeling of the Input Plug on EMI Performance in Power Delivery Adapter

Q. Chen, Z. Qian, Y. Lan, Q. He, and W. Chen

Abstract With the continuous enhancement of power density in power delivery (PD) adapters, devices are positioned in close proximity to each other. Consequently, the proximity raises the concern for electromagnetic interference (EMI) due to magnetic near-field coupling effects, making EMI mitigation a priority in PD adapters. This paper addresses potential near-field couplings within a PD adapter and introduces a differential-mode (DM) EMI circuit model that accounts for near-field coupling effects. The magnetic field coupling between high-frequency (HF) power printed circuit board (PCB) loops and the input plug loop is identified as the pivotal factor affecting the differential-mode EMI noise. By electromagnetic field theory, a mathematical model is established 

for near-field coupling between the power PCB loops and the input plug loop. The model calculations and simulation examine the impact of power PCB loop dimension parameters. Ultimately, this study proposes three primary PCB layout design approaches for both single-sided and double-sided PCBs to mitigate near-field 

coupling effects. Empirical results validate the accuracy and applicability of the theoretical analysis

An Improved Two-Dimensional Calculation Method for PCB Winding Loss of Planar Transformers

J. Li, B. Xie, W. Chen, and Q. Chen

Abstract Accurate winding-loss calculation is critical for the design of printed circuit board (PCB) planar transformers, especially for considering the edge effects of the PCB winding. In conventional winding-loss modeling, the winding is uniformly divided into a limited number of segments due to computational constraints. However, this approach fails to accurately represent the nonuniform current distribution within the winding under high-frequency operating conditions. To address this issue, an improved winding segmentation algorithm is proposed, based on an iterative procedure. Each PCB turn of the winding is modeled as multiple parallel-connected conductors with nonuniform dimensions, and the segmentation size is adaptively determined according to the actual current distribution. An 8-turn planar transformer is used as a sample, and the winding-loss calculation results are validated against those obtained from two-dimensional (2-D) finite element method (FEM). The results indicate that the proposed algorithm reduces the maximum winding-loss calculation error from 2.9% to 0.22% compared with conventional uniform segmentation methods, in which each turn is divided into eight segments for calculation. The experiment results show that the winding AC resistance is lower than 8% up to 1 MHz, which verifies that the proposed method is correct and flexible.

A Review of Supercapacitor Balancing Circuits

Y. Fan, B. Pang, C. Lin, M. Fu, and H. Yang

Abstract The adoption of the supercapacitor technology in electrified transportation systems and renewable energy systems is accelerating. To enhance the safety and economy of supercapacitor energy storage systems, balancing circuits are typically required. This paper reviews the existing works on supercapacitor balancing. Basic balancing circuits based on resistors, capacitors, non-isolated and isolated converters, and resonant converters are covered. Advanced balancing circuits including modular circuits and reconfigurable circuits are also examined. As for the balancing circuit control strategy, distributed control, observer-based control, fuzzy control, and model predictive control are investigated. Furthermore, this paper proposes to conduct a holistic investigation of the existing and emerging topologies and control strategies of the balancing circuits to transform the empirical works on supercapacitor balancing to a general framework.

Modeling of Magnetic Core Losses From Global to Local Scales Based on Multimodal Neural Networks 

C. Wang, S. Sang, S. Qu, D. Yang, and Y. Wei

Abstract This paper presents a comprehensive framework for modeling magnetic core losses, validated through a systematic multi-scale approach. The framework is built on two pillars: a suite of physically-grounded data augmentation and training optimization techniques, and a multimodal neural network that synergizes CNNs and Transformers. Data augmentation techniques—including random period shifts, waveform flipping, noise injection, and parameter perturbations—simulate real-world waveform irregularities, enabling direct applicability to arbitrary-phase and non-periodic inputs. Complementary training strategies ensure stable learning. The core network architecture dynamically fuses scalar parameters and sequential waveforms via attention mechanisms. Extensive validation uses both an in-house experimental platform and the public MagNet database, extending beyond common ferrites to include nanocrystalline and iron powder cores, with data for some materials reaching magnetic saturation. The framework generalizes consistently across these materials, with average prediction errors below 3.4% for most, and handles non-integer periodic waveforms like 3C94 with about 2.44% error. Compared to existing methods, the core contribution lies in the holistic framework integrating expanded experimental data, enhancement techniques tailored to real-world variability, and rigorous multi-material validation, offering a flexible and directly applicable solution for predicting core losses in power electronic devices.

An Improved Sliding Mode Control Approach for the Enhancement of Wind Energy Conversion Systems Based on PMSG 

P. Kumar and Y. Kashyap

Abstract Recently, researchers have increasingly focused on enhancing the quality of grid-supplied power while efficiently controlling wind energy conversion systems (WECS) under variable wind conditions. This study presents an improved control method to mitigate the chattering phenomenon often encountered in traditional approaches. By integrating maximum power point tracking (MPPT) with pitch angle control, the efficiency and 

performance of WECS utilizing permanent magnet synchronous generators (PMSG) are significantly improved. The primary objective is to regulate the generator and grid-side converter to precisely track the wind speed reference determined by the MPPT algorithm. The proposed sliding mode control (SMC) employs a unique, smooth, and continuous switching mechanism to address the chattering issue. Its stability is rigorously analyzed using Lyapunov’s stability function. The control framework is straightforward to implement, and simulation results demonstrate its superior performance, achieving a total harmonic distortion (THD) below 2%, accurate current injection, and a near unity power factor. The MATLAB/Simulink based system is also validated through real-time hardware-in-the-loop (HIL) testing using an OPAL-RT simulator, confirming the model’s effectiveness. Finally, the proposed method is benchmarked against existing schemes, highlighting its competitive advantages.

A Universal EV Charging System With Extended Output Voltage Range for Future Electric Mobility 

M. S. H. Naidu and J. R. Louis

Abstract The rapid diversification of electric vehicles and energy storage systems has created a need for universal battery chargers capable of supporting wide and evolving voltage standards. Present charging infrastructure is designed for fixed or narrow voltage ranges, leading to limited interoperability and reduced future readiness as battery voltages continue to increase. In this paper, a universal battery charger with a wide output voltage range of 100–900 V is proposed to address these challenges and enable future-proof charging solutions. The architecture integrates a bridge less power factor correction (PFC) converter, a full-bridge active power decoupling (APD) stage, and an LLC resonant converter with a voltage doubler circuit. Compared to conventional PFC topologies, the bridgeless PFC reduces conduction losses and ensures near-unity power factor with sinusoidal input current. To mitigate low-frequency second-harmonic power ripple without bulky electrolytic capacitors, the APD stage employs compact film capacitors, improving reliability, thermal performance, and lifetime. The LLC resonant converter, combined with a relay-controlled voltage doubler, enables flexible high-voltage operation while maintaining high efficiency. A 3.3 kW hardware prototype is implemented and experimentally validated, achieving a peak efficiency of 96.5%, demonstrating the feasibility of the proposed universal charger for next-generation wide-voltage battery charging applications.

An Improved Double-2D Electromagnetic Simulation Methodology for Winding Loss Analysis of High-Frequency Magnetic Components 

K. Fu, J. Ye, J. Tu, Y. Lan, and W. Chen

Abstract Precise and efficient winding loss calculation is critical for optimizing high-frequency magnetic components. While three-dimensional (3D) finite element method (FEM) provides accuracy, its computational demands limit practical applications. Conventional two-dimensional (2D) FEM, though computationally efficient, fails to balance calculation accuracy and efficiency. This paper develops a double-2D electromagnetic simulation methodology achieving 3D-equivalent accuracy through magnetic potential equivalence in the air gap location. The proposed double-2D approach can be applied to common core structures such as EE, EC, and others: (1) Synthesis modeling for symmetric structures. (2) Magnetic potential compensation for complex asymmetric structures. A deviation of less than 5% from the measured results is achieved over the frequency range of 10 kHz to 1 MHz, and a 130-fold acceleration was achieved, resolving the accuracy–efficiency trade-off in 3D FEM simulations. The proposed double-2D simulation technique provides practical significance for high-frequency magnetic device design, particularly in GaN-based power delivery.

High Voltage Gain Bidirectional DC-DC Converters for Supercapacitor Assisted Electric Vehicles: A Review

Ankit Kumar Singh; Anjanee Kumar Mishra; Krishna Kumar Gupta; Yam P. Siwakoti

LLC and CLLC resonant converters based DC transformers (DCXs): Characteristics, issues, and solutions

Yuqi Wei; Quanming Luo; Homer Alan Mantooth

A Review of Magnetic Core Materials, Core Loss Modeling and Measurements in High-Power High-Frequency Transformers

Zenong Li;Weijian Han;Zhen Xin;Qing Liu;Jianliang Chen;Poh Chiang Loh

A review of SiC power module packaging: Layout, material system and integration

Cai Chen;Fang Luo;Yong Kang

Critical Review of Vehicle-to-Everything (V2X) Topologies: Communication, Power Flow Characteristics, Challenges, and Opportunities

Gaurav Kumar; Suresh Mikkili

Overview of Voltage Regulator Modules in 48 V Bus-Based Data Center Power Systems

Jiawei Liang; Liang Wang; Minfan Fu; Junrui Liang; Haoyu Wang

A novel model predictive current control with reduced computational burden based on discrete space vector modulation for PMSM drives

Jun Sun;Yong Yang;Jiefeng Hu;Xinan Zhang;Xinghe Li;Jose Rodriguez

A SiC-Based Liquid-Cooled Electric Vehicle Traction Inverter Operating at High Ambient Temperature

Chi Zhang;Srdjan Srdic;Srdjan Lukic;Keyao Sun;Jun Wang;Rolando Burgos

Review of GaN totem-pole bridgeless PFC

Qingyun Huang;Alex Q. Huang

A new high gain DC-DC converter with model-predictive-control based MPPT technique for photovoltaic systems