Characterization of inductors in partial saturation including thermal effects for DC/DC converters

Modern switching power electronic converters are widely used in many industrial and consumer applications, especially in step-up and step-down configurations. The optimization of the inductor in the converter design gives advantages in terms of power density increase and cost reduction; therefore, to properly design the converters, it is necessary to improve the knowledge of magnetic material behaviour as the current changes. The development of new models is becoming crucial in the design of modern power converters and there are some recent papers dealing with the use of inductors working in partial saturation [1-5]. This issue is of great interest for scientists and industry due to economic reasons tied both to the initial cost of the converter and to the reduced pay-back time of the equipment. Since available measurement systems are usually not well suited to test inductors in high current operating conditions, a dedicated system based on a DC/DC buck converter, in which its inductor represents the inductor under test and the load is a constant current sink, has been set up (Fig. 1). A virtual instrument that allows automatic inductance measurements has been implemented in LabVIEW® programming environment; the front panel of the instrument is shown in Fig. 2. These tests are performed on the same inductor as the current changes, in order to plot the inductance L versus the inductor current I at a fixed temperature. For the experimental measurement a test rig has been set up, the elements of which are listed in Table 1. Using the proposed measurement system, a family of isothermal inductance curves has been obtained for the inductor under test. The aggregate data relative to a 470μH Panasonic inductor (ELC18B471L) are shown in a 3-D surface in Fig. 3. For current lower than 2.5A the curves are similar and there is no significant temperature effect on the value of inductance. The same observation is valid for current higher than 4.5A; at this point, the inductance drops to 20% of the nominal value and the inductor is in deep saturation region. In the rolloff region, between 2.5A and 4.5A, it can be noted that for higher temperatures the inductance curves shift downwards, indicating a smaller inductance. As the inductor approaches the deep saturation region, the temperature minimizes his effect on the inductance variation. As expected, the operation with large currents signal highlights the variation of the inductance with temperature. These curves can be used in the design process for applications which involve inductors operating in moderate saturation region since it is possible to obtain a model based on polynomials including also the temperature effect. It has been underlined that the isothermal curves can be approximated by fifth-degree polynomials and that it can be assumed that the inductance depends linearly from temperature (Fig.4).