Power Electronics

Ferrite Material Improves High Frequency Magnetics

New processes yield core material with low power losses at 500 kHz and high magnetic saturation flux density.

Development and continued success of switchmode power supplies is challenging the ferrite industry to produce new high-quality ferrite cores capable of operating at increasingly higher frequencies [1,2]. For this reason, it's important to reduce the power losses of MnZn-ferrites used as transformers cores.

The electromagnetic characteristics of MnZn-ferrite are not only dependent on the composition of the main elements, but also on its microstructure [3]. Efforts have been made to improve their loss characteristics by controlling the size of the grains and the distribution of small amounts of additives in the grain-boundary region [4].

You can divide core loss into three components: hysteresis loss (Ph), eddy-current loss (Pe) and residual loss (Pr). The proportions of these components in the total loss can vary widely, according to measurement conditions such as frequency and magnetic flux-density.

At low frequencies, Ph is dominant. To reduce these losses, it's important to form a uniform microstructure free from lattice defects and pores. At high frequencies the proportion of Pe increases, but you can reduce it by increasing core resistance. Decreasing Pe involves grain boundaries with a high electrical resistance and use of a ceramic microstructure with small grains. Therefore, the choice of raw materials and technological parameters influence the power losses.

When considering MnZn-ferrites, we can identify two extreme cases regarding the eddy current in the magnetic core by applying the Brick-Wall Model [5]. In the first case, when the magnetic grains are isolated and the eddy current in this hypothetical case is localized inside the grains, the core behaves as an assembly of individual magnetic grains in which each grain contributes to the eddy-current loss. A different dependence between the total loss and the microstructure holds when the grain boundaries are permeable to the eddy current. According to this model, the bulk material can be approximated by a group of small cubes of size D, separated by high-resistance layers of thickness and resistance. Using this model [5], the Pe, which prevails in the frequency range above 500 kHz, can be effectively suppressed by decreasing the average grain size and increasing the grain-boundary resistance and width.

The saturation, Bs of a material mainly depends on its ferrite composition. The maximum saturation for MnZn ferrites is in the region 55 mol% Fe2O3, 35 mol% MnO and 10 mol% ZnO. You can enhance saturation by increasing the density of the sintered material. Concerning the initial permeability, a high μi and low core losses at high frequencies are in contradiction with regard to the microstructure because the small grains lead to an internal shearing and so a lower initial permeability. This means the composition has to be chosen as a compromise between the saturation, Bs, the initial permeability, μi, and SMP (minimum of power loss). There are many contradictory requirements for low-core-loss ferrites. Achieving the lowest core losses possible requires compromise and novel processing techniques.

The development process must ensure achieving the required core characteristics. In high-frequency power ferrites, the minimization of the Pe is the most important factor in determining the optimum core characteristics. When developing a process for the production of ferrites, the problem should not be approached by trying to optimize each magnetic property. Because of the interdependency of the magnetic properties, improving one property may lead to the degradation of several others. The way to approach the problem is to find the compromise that best fits the design requirements. When designing the process, you must consider the reproducibility and cost.

Two main factors governing the reproducibility of a process are control of the processing variables and the chemical homogeneity. Switching to more sophisticated equipment (DTA, TGA, X-ray) reduces the number of uncontrollable processing variables. You can achieve chemical homogeneity mainly through good-quality starting powders and intensive mixing procedures.

Fig. 1, on page 48, shows the power loss of the standard high-frequency power transformer-material, 35G, compared with the new material, 55G. By introducing the design principles described above, the power loss could be decreased from 180 mW/cm3 to 120 mW/cm3 at 400 kHz, 50 mT, 100°C. The Table summarizes the sintered density, the initial permeability and the saturation of 35G and the new 55G. The upgraded version of 55G increases the density by 0.1 g/cm3 and the saturation from 300 mT to 390 mT (at 1200 A/m) at 120°C.

Initial permeability varies with temperature, as shown in Fig. 2, on page 52, for the standard Iskra 35G material and the new 55G in the range of -25°C to +235°C.

The distinction between usable flux density and saturation flux density is critical for magnetic materials such as MnZn ferrites that exhibit hard saturation. Hard saturation is a sharp decrease of the magnetic permeability as the excitation results in flux-density values located in the nonlinear region of the hysteresis loop [8,9].

Usable flux density is defined as the calculated flux density at the point of minimum cross-sectional area (Amin) in the magnetic core that corresponds to the dc bias condition. A dc current (IDC) equivalent to the peak current is typically used to measure the inductance for a specific peak-current condition and the maximum peak-flux density in the core (BPK_MAX) is:

L=Inductance at 0 Adc
I=Specified test current
Amin=Minimum cross-sectional area
Np=Number of turns

Fig. 3, on page 52, compares the inductance of air-gaped toroid cores using the standard Iskra 35G material and the new 55G, as a function of BPK_MAX at 120°C.

The new 55G material is appropriate for operation at 500 kHz; compared with the standard material, 35G, it decreases power loss by 40%. We achieved this mainly by a microstructure with smaller grains and a higher electric resistance of the grain boundaries. At the same time, we increased saturation by choosing an appropriate composition and an enhanced density. With these properties, 55G, the upgraded version of 35G, is suitable for dc-dc applications.


  1. Goldman, “Modern Ferrite Technology,” Van Nostrand Reinhold 1990, N.Y.

  2. C.R. Hendrics, W.R. Amarakoon, “Processing of Manganese Zinc Ferrites for High-Frequency Switchmode Power Supplies,” C. Buliten, Vol. 70, No. 5, 1991.

  3. H. Tsunekawa, A. Nakata, T. Kamijo, K. Okutani, R.K. Mishra and G. Thomas, “Microstructure and Properties of Commercial Grade Manganese Zinc Ferrites,” IEEE Trans. Mag. 15 1855-57 (1979).

  4. A. Žnidaršič, M. Limpel, M. Drofenik, “Effect of Dopants on the Magnetic Properties of MnZn Ferrites for High Frequency Power Supplies,” IEEE Trans. Mag. 31(2), 950-953 (1995).

  5. M. Drofenik, A. Žnidaršič and I. Zajc, “Highly Resistive Grain Boundaries in Doped MnZn Ferrites for High Frequency Power Supplies,” J. Appl. Phys. 82(1), 333-340 (1997).

  6. A. Žnidaršič, M. Drofenik, “Influence of Oxygen Partial Pressure During Sintering on the Power Loss of MnZn Ferrites,” IEEE Trans. Mag. 32(3), 1941-1945 (1996)

  7. A. Žnidaršič, M. Drofenik, “High Resistivity Boundaries in CaO-Doped MnZn-ferrites for High-Frequency Power Application,” J. Am.Cer. Soc. 82(2) 359-65(1999)

  8. J.Wrba, S. Plutzer, “Recent Progress in the Development of MnZn Ferrites for Applications at High Frequencies in Electronics,” 7th Conference of the European Ceramic Society, 1247-1250 (2000).

  9. G.E. Schaller “Excerpt on usable flux density from IEEE paper on review of power magazine,” Internal Paper, March 13, 2000.

For more information on this article, CIRCLE 334 on Reader Service Card

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.