Optimal Area-Product Model (OAPM) based Non-Iterative Analytical Design Methodology for Litz-wired High-Frequency Gapped-Transformer (LHFGT) in LLC Converters

2020 
In dc-dc LLC resonant converters, the Litz-wired high-frequency gapped-transformer (LHFGT) plays an important role, as it provides the necessary isolation, the required voltage-conversion, and the desired magnetizing inductance (L m ) for efficient converter operation. Since the LHFGT makes a significant contribution to the overall converter weight and size, so the converter designers must rely on complicated and advanced optimization techniques, with a large number of iterations, for its design. This dependence is due to the shortcomings in the conventional analytical modeling techniques for optimal size-selection of the core and winding in the LHFGT. Hence, this manuscript proposes an optimal area-product (A prod ) model (OAPM)-based non-iterative LHFGT design methodology that maximizes the efficiency and power density, minimizes the losses and volume, integrates the L m , and maintains the temperature-rise within limits. The method takes into consideration the LLC circuit parameters, the Litz-wire strand-radius (rs), the core material and geometrical parameters, the excitation-waveform shape, the stored energy due to coreairgap, and the peak flux-density (B pk ) inside the core. The accuracy improvement is attained through the proposal of accurate A prod -based core-geometry features estimation (ACGFE) models and by keeping in view the interdependency between the transformer-design parameters (TDPs). The optimized design is obtained in a single iteration, based on the proposed OAPM, the proposed optimal r s selection model, and the proposed optimal TDPs' models. The proposed design routine is validated through the analytical and experimental results of a prototype LHFGT for a 200W-110kHz-400VDC/12VDC LLC resonant converter.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    2
    Citations
    NaN
    KQI
    []