Optimizing Magnetic Component Design for High-Frequency Converters
Overview:
This article discusses the challenges of designing magnetic components for high-frequency converters and the advantages of using planar magnetic components.
Increasing the switching frequency decreases the size of the converter and the cost of the components, making a high-frequency (HF) converter a more attractive option.
The losses of conventional winding magnetic components, however, grow substantially as switching frequencies rise. Planar magnetic components, on the other hand, excel in HF applications due to their low profile and huge heat dissipation area.
How does a Conventional Magnetic Component impact the stability of a High-Frequency Converter?
In designing planar magnetic components, numerous related elements, including copper track architectures and winding patterns, will really have a significant impact on the characteristics of inductors and transformers. Furthermore, the skin and proximity effects grow more severe with an increase in switching frequency, and the self and mutual impedances make modeling more difficult, especially when parallel windings are included.
It is common practice to utilize non-analytical and time-consuming techniques like finite-element modeling (FEM) and experimental measurements when designing magnetic components. So, to aid the planar magnetic design, various effective methods are used. However, these methods often have complicated implementations and rely on a wide range of assumptions.
Designing Planar Magnetic Component
The modular layer model (MLM) is a framework for modeling planar magnetics, including impedances and current distribution. There are only two underlying assumptions in this model. This model can estimate impedances, losses, stored reactive energy, and current sharing between windings with high accuracy since it takes into account skin and proximity effects.
The essential idea of MLM is that the same method may be used to describe the various thicknesses of planar magnetic elements. It is ultimately possible to determine the impedance network of the magnetic components and the relationship between the two adjacent layers.
What are the factors that influence efficiency of high-frequency converters?
Planar magnetic components have many benefits, but they cannot be made without the inclusion of parasitic characteristics. The converter's efficiency increases with careful consideration of the three factors of leakage inductance, AC capacitance, and AC resistance.
Methods to Reduce Leakage Inductance
The leakage inductance is always present because there is always some fraction of the magnetic flux created by the primary side's ac current excitation that does not couple to the secondary winding. Power efficiency is decreased, switching losses are increased, and EMI is further exacerbated because of leakage inductance in hard-switching converters. Interleaving windings is a common technique used to lessen leakage inductance.
Methods to Increase Leakage Inductance
A greater leakage inductance is required in series-parallel resonance circuits and other rare circumstances. Adding a magnetic shunt and using fractional turns are two other ways to boost leakage inductance.
Magnetic Shunt Approach
In the first approach, ferrite polymer composites (FPC) are often placed between the primary and secondary windings to create a return channel for the magnetic flux with reduced reluctance. Because of this, the coupling coefficient between the windings is reduced and the leakage is increased. The findings demonstrate that the introduced FPC material increased the leakage inductance by a factor of four.
The thickness and permeability of the implanted magnetic shunts determine the amplitude of the leakage inductance. On the other hand, a thicker magnetic shunt or greater permeability both increase the leakage inductance and the eddy current effect, reducing the system's efficiency.
Fractional Turns Approach
Fractional turns are illustrated in Fig. 1. The illustration clearly demonstrates that the EI core's primary winding is located in the middle leg, while the secondary winding is located on the outside leg.
Fig. 1: Illustration of fractional turns Source: IEEE Open Journal of the Industrial Electronics Society
Due to its unique construction, this sort of core effectively divides the magnetic flux in half; however, only one half is connected to the secondary winding, resulting in a significant amount of leakage inductance. The ratio of core leg reluctance determines the leakage inductance.
Methods to Reduce Parasitic Capacitance
The leakage inductance can be decreased by using an interleaving structure; however, the parasitic capacitance will be raised. Leakage inductance and parasitic capacitance are opposing factors. The capacitances between the turns, the winding layers, and the windings and the core will all have a significant impact on the performance of the magnetic component. For instance, the interwinding capacitance, which causes leakage currents, makes the EMI issue much worse.
Four strategies for minimizing parasitic capacitance can be obtained by following these steps:
- Increasing the distance between windings and decreasing the overlapping surface area as much as possible;
- Decreasing the number of turns per layer and increasing the number of layers economically;
- Designing the structure so that as few intersections as possible occur between the primary and secondary windings;
- Arranging the winding connections reasonably to get the minimum amount of energy associated with the electric field.
A new interleaved structure was also proposed, as shown in Fig. 2, allowing for concurrently reducing leakage inductance and parasitic capacitance. As can be seen, both the primary and secondary windings use an interleaved structure.
However, in contrast to the conventional layout, all of the secondary winding's turns are contained within a single layer, and another layer is used for the outgoing line. The results of the simulation and calculations demonstrate the effectiveness of this approach. 
Fig. 2. A new interleved structure Source: IEEE Open Journal of the Industrial Electronics Society
Methods to Increase Parasitic Capacitance
Like leakage inductance, parasitic capacitance is predicted to increase in specific contexts, such as integrated LLC transformers and EMI filters. By sandwiching dielectric material with either high or low dielectric permittivity between two planar layers of windings, a capacitor can be produced to meet a wide range of capacitance needs.
This type of capacitor is already on the market, and it can be used in place of resonant capacitors, filters, decoupling capacitors, and snubber capacitors.
Methods to Reduce Parasitic Resistance
Furthermore, parasitic resistance is an essential parameter that has a direct impact on the features as well as the efficiency of power converters. The skin and proximity effects grow more severe as the switching frequency is increased. Because of the skin effect as well as the proximity effect, the current density is not distributed evenly across the cross section of the conductor, which results in a high winding resistance while operating at high frequencies.
It is easy to obtain the DC resistance; however, the AC resistance is a function of the DC resistance. To lower the DC winding resistance, a construction with a nonuniform-width winding can be implemented. Because the primary objective of this winding structure is to realize constant track resistance for each turn, the widths of the turns geometrically shift from the inner turn to the outer turn.
Within the context of this structure, the radius ratio of adjacent windings is the most significant element to consider. The radius ratio and the number of turns per layer are both taken into consideration concurrently in order to calculate the minimum DC resistance.
Summarizing the Key Points
- Designing magnetic components for high-frequency converters is challenging due to the impact of copper track architectures and winding patterns on their characteristics.
- Skin and proximity effects become more severe with an increase in switching frequency, making modeling self and mutual impedances difficult.
- The modular layer model (MLM) is a framework for modeling planar magnetics that can estimate impedances, losses, stored reactive energy, and current sharing between windings with high accuracy.
- Planar magnetic components excel in high-frequency applications due to their low profile and huge heat dissipation area.
- Using planar magnetic components can increase the efficiency of high-frequency converters and reduce their size and cost.
Reference
Wang, Yijie, Oscar Lucia, Zhe Zhang, Shanshan Gao, Yueshi Guan, and Dianguo Xu. “A Review of High Frequency Power Converters and Related Technologies.” IEEE Open Journal of the Industrial Electronics Society 1 (2020): 247–60.