New Magnetic Integration Technology Boosts EV Charger Performance
As the global push for sustainable transportation accelerates, the technological backbone of electric vehicles (EVs) is undergoing rapid transformation. Among the most critical components in this evolution is the on-board charger (OBC), the gateway through which energy flows from the grid into the vehicle’s battery. As consumer expectations rise for faster charging, longer range, and quieter, more reliable performance, the demand for higher efficiency, smaller size, and superior electromagnetic compatibility (EMC) in OBCs has never been greater. At the heart of this challenge lies a seemingly unassuming yet profoundly impactful component: the EMI filter inductor.
Traditionally, electromagnetic interference (EMI) filtering in three-phase four-wire (3P4W) OBC systems has relied on discrete or partially integrated magnetic components to suppress both common-mode (CM) and differential-mode (DM) noise. However, conventional designs—whether using “Y”-shaped, “cross”-shaped, or segmented magnetic structures—have long faced a fundamental limitation: their susceptibility to magnetic saturation under unbalanced load conditions. In real-world driving and charging scenarios, three-phase current imbalance is not an exception but a norm. When one phase draws more current than the others, it creates a DC bias in the magnetic core, leading to partial saturation of the differential-mode magnetic branches. This saturation not only reduces the inductance and impedance of the filter but also compromises the common-mode performance, ultimately degrading the entire EMI suppression capability of the system.
For engineers and designers in the EV power electronics field, this presents a persistent dilemma. How can one simultaneously achieve high power density, robust EMI performance, and resilience to real-world operating conditions? The answer, according to recent research, may lie in a novel magnetic integration architecture known as the “quasi-cross” differential-mode magnetic branch.
In a groundbreaking study published in the Journal of Power Supply, Haijun Yang and Zengyi Lu from Delta Electronics (Shanghai) Co., Ltd. have introduced and validated a new integrated DM-CM choke design specifically tailored for 3P4W OBC applications. Their work, titled Integration of DM and CM Chokes with Quasi-cross DM Magnetic Branches, represents a significant leap forward in magnetic component design, offering a solution that addresses the core weaknesses of existing technologies while enabling greater miniaturization and performance stability.
The innovation stems from a fundamental rethinking of how magnetic flux is managed within the inductor. Conventional approaches treat each phase—and the neutral line—individually, leading to isolated magnetic paths that are highly sensitive to imbalances. When, for example, Phase A carries more current than Phases B and C, the resulting magnetic flux in the corresponding DM branch increases disproportionately, pushing the core material closer to saturation. This not only diminishes the inductor’s effectiveness but also introduces nonlinearities that can generate additional harmonic noise, creating a feedback loop of performance degradation.
Yang and Lu’s “quasi-cross” architecture breaks away from this paradigm by reconfiguring the magnetic circuit based on paired current relationships rather than individual phase currents. Instead of summing the three-phase currents to determine the neutral current, their method groups adjacent phases—such as A and B—together and pairs them against the combination of C and N. This mathematical reorganization translates into a physical magnetic structure where the DM flux paths are shared and balanced in a way that inherently compensates for imbalances.
The result is a magnetic topology that behaves like a self-correcting system. When one phase current increases, the flux distribution across the quasi-cross structure redistributes in a manner that mitigates the peak flux density in any single branch. This “flux-correcting” effect is not merely theoretical; the researchers demonstrated through finite element analysis and hardware testing that the quasi-cross design reduces magnetic bias by up to 45% compared to traditional cross-shaped cores under a 25% current imbalance. More impressively, the overall anti-saturation capability of the inductor improves by approximately 80%, a figure that has profound implications for real-world reliability.
One of the most compelling aspects of the quasi-cross design is its versatility. Unlike previous integrated solutions that were often constrained by material limitations or complex manufacturing processes, this new architecture supports a wide range of magnetic materials, including ferrites, powder cores, silicon steel, and advanced nanocrystalline or amorphous alloys. This flexibility allows engineers to optimize the inductor for specific performance criteria—whether it’s maximizing inductance, minimizing core loss, or enhancing saturation resistance—without being locked into a single material system.
The research team explored several physical implementations of the quasi-cross concept, including stacked bar configurations, clamped block assemblies, and laminated sheet-stack structures. Each variant offers distinct advantages depending on the application. For instance, the stacked bar design is simple to manufacture and compatible with a broad spectrum of materials, making it ideal for rapid prototyping and low-volume production. The clamped block version, typically made from molded ferrite or powder cores, provides larger cross-sectional areas, which translate into higher inductance and better thermal performance. Meanwhile, the laminated sheet-stack approach—particularly when using high-permeability silicon steel—delivers superior flux handling and is well-suited for high-power applications where thermal management and saturation resistance are paramount.
In their experimental validation, Yang and Lu applied the quasi-cross silicon steel sheet-stack design to an 11kW dual-mode (single/three-phase) OBC platform. The results were striking. When compared to a conventional EMI filter using discrete or standard integrated chokes, the new design demonstrated a significant improvement in EMI suppression, particularly in the critical 150kHz frequency range. Measurements showed a noise reduction of 15 to 22 dB across all phases and the neutral line, a performance gain that directly contributes to compliance with stringent automotive EMC standards such as CISPR 25 Class 3.
This level of improvement is not just a technical achievement—it has tangible benefits for vehicle manufacturers. A more effective EMI filter means less electromagnetic noise radiating from the powertrain, which reduces the risk of interference with sensitive onboard electronics such as infotainment systems, advanced driver-assistance systems (ADAS), and vehicle-to-everything (V2X) communication modules. In an era where software-defined vehicles are becoming the norm, ensuring electromagnetic cleanliness is as important as mechanical reliability.
Moreover, the quasi-cross design contributes directly to the overarching goals of EV design: weight reduction, space savings, and cost efficiency. By integrating the DM and CM functions into a single, compact magnetic structure, the need for multiple discrete inductors is eliminated. In the 11kW OBC case study, this integration reduced the number of required magnetic components by four, significantly shrinking the footprint on the printed circuit board (PCB). In the tightly packed environment of an EV power module, where every cubic centimeter counts, such space savings can enable more compact powertrain layouts or free up room for additional functionality.
The manufacturing advantages are equally noteworthy. The quasi-cross architecture lends itself to modular, assembly-friendly designs that can be easily scaled and adapted. The clamped or stacked construction methods allow for automated assembly, reducing labor costs and improving consistency. Furthermore, because the design is less sensitive to material tolerances and current imbalances, it reduces the need for over-engineering—such as using oversized cores or derating components—which in turn lowers material costs and improves overall system efficiency.
Perhaps one of the most forward-looking aspects of this research is its scalability. While the current work focuses on 3P4W systems, the underlying principle of flux balancing through paired current relationships can be extended to more complex power architectures. The authors suggest that the quasi-cross concept could be applied to six-wire, eight-wire, or even higher-phase-count systems, provided that the sum of instantaneous currents remains zero—a condition that holds true in most polyphase power conversion circuits. This opens the door to next-generation EV charging systems, including bidirectional chargers, vehicle-to-grid (V2G) interfaces, and multi-port power electronics units that require sophisticated EMI management across multiple power paths.
From an industry perspective, the implications of this technology are significant. As automakers race to develop 800V architectures, ultra-fast charging capabilities, and integrated power electronics, the demand for smarter, more resilient magnetic components will only grow. Traditional inductor designs, constrained by their inherent limitations, may struggle to keep pace. The quasi-cross DM-CM choke, by contrast, offers a future-proof solution that aligns with the trajectory of EV innovation.
It is also worth noting that this advancement comes at a time when supply chain resilience and material efficiency are top priorities. By enabling the use of more cost-effective and widely available materials like silicon steel—without sacrificing performance—the quasi-cross design reduces reliance on rare-earth-based or specialized magnetic alloys. This not only lowers production costs but also enhances sustainability, a key consideration in the lifecycle analysis of EV components.
The work of Yang and Lu also underscores the importance of cross-disciplinary collaboration in advancing power electronics. Their approach combines deep theoretical insight with practical engineering know-how, bridging the gap between academic research and industrial application. By publishing their findings in a peer-reviewed journal and providing detailed experimental validation, they have set a high standard for transparency and reproducibility—key elements of trust in technical communication.
As the automotive industry continues its electrification journey, innovations like the quasi-cross integrated inductor will play a crucial role in shaping the next generation of EVs. They represent more than just incremental improvements; they are enablers of broader system-level advancements. By solving a long-standing challenge in EMI filtering, this technology helps pave the way for quieter, more efficient, and more reliable electric vehicles—vehicles that not only meet today’s standards but are ready for the demands of tomorrow.
In conclusion, the development of the quasi-cross DM-CM choke by Haijun Yang and Zengyi Lu of Delta Electronics is a testament to the power of innovative thinking in power electronics. It demonstrates that even in mature fields like magnetic component design, there is still room for breakthroughs that deliver real-world benefits. As this technology moves from the research lab into production vehicles, it has the potential to become a standard feature in high-performance OBCs, contributing to a cleaner, quieter, and more sustainable electric mobility future.
Haijun Yang, Zengyi Lu, Journal of Power Supply, DOI:10.13234/j.issn.2095-2805.2024.5.325