Recent advancements in power electronics have drawn attention to high-frequency transformers (HFTs) and their role in boosting the efficiency of isolated DC-DC converters. These converters are pivotal for applications ranging from electric vehicles to renewable energy installations, where leakage inductance can significantly affect performance. A new study proposes a modelling technique aimed at improving the assessment of leakage inductance through innovative winding configurations, thereby enhancing transformer efficiency without necessitating additional materials or introducing increased copper losses.
Transformers used within bidirectional isolated DC-DC converters (BIDCs) require specific levels of leakage inductance to facilitate optimal power delivery and minimize losses associated with switching. The challenge, as outlined by the authors, is to fine-tune leakage inductance effectively, which has been traditionally reliant on cumbersome adjustments involving winding thickness or number of turns.
The core concept of this research revolves around modifying the height of the primary and secondary windings within the transformer’s core window. By introducing strategic adjustments to their vertical positioning, the proposed technique enhances the separation between specific winding turns, thereby increasing the leakage inductance. Notably, this method preserves the mean length turn (MLT), ensuring consistent copper loss throughout.
The research team conducted simulations and experimental validations using a three-phase high-frequency transformer (3P-HFT). The results were compelling, illustrating only slight variances (3.9%) and errors (4.53%) when comparing theoretical predictions with experimental outcomes.
The importance of the findings extends beyond academic curiosity; leakage inductance is directly linked to power transfer capabilities and zero-voltage switching (ZVS) characteristics, both of which are fundamental to energy efficiency. For example, excessive leakage inductance can hinder power transfer, complicate the control mechanism, and increase reactive power losses, whereas insufficient leakage can jeopardize ZVS operation, eleviating switching losses at high frequencies.
The innovative method introduced here stands out due to its capacity to efficiently control leakage inductance without sacrificing performance or incurring additional costs. Researchers argue this approach can easily be integrated within existing design frameworks, facilitating the rapid deployment of energy-efficient converter technologies.
Further developments include refining the mathematical models used for leakage inductance calculations. Researchers found the previously high computational costs associated with numerical methods—while accurate—were not practical for everyday application. This led to the formulation of new analytical methods grounded on the ideal geometry of HFTs and their operational parameters.
The study's experimental setup utilized advanced simulation tools, including ANSYS Maxwell, to visualize and predict leakage inductance variations corresponding to changes in winding arrangement. Following the parameters laid out, the team confirmed through rigorous tests and simulations how different winding configurations might influence overall transformer performance.
Future work aims to broaden the applicability of these findings across various types of power converters, especially as the demand for energy-efficient systems continues to rise. By proactively addressing leakage inductance issues through innovative winding strategies, this research paves the way for more reliable and efficient power electronics systems, with significant potential applications across multiple industries.
Comprehensively, this study exemplifies the intersection of theoretical modelling and practical application, showcasing how nuanced adjustments to established designs can yield prominent benefits. Moving forward, continuous exploration of these routes could lead to groundbreaking advancements within the burgeoning field of electronic energy conversion.