Jean-Pierre Keradec

Multiwinding transformers: a successive refinement method to characterize a general equivalent circuit

A lumped component equivalent circuit has been developed by our team to model the linear electrical behaviour of any two winding transformer. Its topology is independent of sizes and technology and a general method of characterization, based exclusively on external impedance measurements, has been presented. Unfortunately, when some key frequencies are out of the range of the measuring apparatus, many of its components remain unevaluated. Owing to a new understanding of magnetic and electrostatic coupling, a several step approach is presented. At each step a more accurate circuit proven to be effective in electronic simulation, is fully characterized. This method suits whatever the number of windings and experimental data will be presented for two and three winding transformers.

Equivalent circuits for transformers based on one-dimensional propagation: accounting for multilayer structure of windings and ferrite losses

In a recent paper, we linked the electromagnetic power propagation in a two-winding transformer to an electrical equivalent circuit. This study pointed out that an electrical equivalent circuit can be associated with each layer (conductor, insulator, ferrite) passed through, and that connecting all of these elementary circuits together leads to the equivalent circuit of the entire component. Propagation was assumed to be one-dimensional, and each winding was supposed to be made of a single layer of turns. In this paper, we first investigate the consequences of the multilayer structure of each winding, following the above approach. Skin and proximity effects are taken into account this way, and theoretical results are compared with experimental data. Then, we focus on the representation of the linear behavior of the ferrite core. On the basis of an experimental approach, it is shown that an accurate equivalent circuit must include a capacitor. This is consistent with the high permittivity of the material.

Digital oscilloscope measurements in high-frequency switching power electronics

The requirements for measuring instantaneous power of a switching power circuit using a digital storage oscilloscope are discussed. The measurements are difficult to perform and involve the use of high-speed, high-resolution measurement instrumentation. Careful consideration needs to be given to overload conditions. Averaging and advanced sampling techniques are necessary to obtain a satisfactory result. Performance enhancements of instruments that will allow improvement of the quality of the measurements remain desirable. An indication of measurement accuracy is obtained by comparing the mean dissipated power to the same quantity obtained by a calorimetric test (water flow heating). The mean dissipated power is the integral of the instantaneous power divided by the switching period. The two results agree within 5%. >

Volume optimization of a PFC flyback structure under electromagnetic compatibility, loss and temperature constraints

The aim of this paper is to present an analytical optimization approach of a flyback structure in PFC mode. Indeed, softwares like Saber, Pspice or Simplorer are effective means for the power electronics structure time-domain studies. However, if these structures have an AC input and a high switching frequency (various time scales), time-domain simulation becomes painful and expensive in memory and computing time. The study of EMC performances is also difficult because of the line impedance stabilizer network (LISN) time-constants which comes to penalize the time-domain simulation. In addition, in sizing and optimization process, results in short computing times are needed, so the time-domain simulation may be too time consuming. In this way, the paper proposes to carry out a compromise between the model accuracy and the tool rapidness and recommends the use of analytical models to optimize the passive element volume of a flyback structure by respecting EMC standards, by minimizing the whole losses dissipated in the structure (conduction and switching semiconductor losses, core and copper losses in the transformer) and by constraining the semiconductor junction, the winding and the magnetic circuit temperatures. Firstly, analytical models of the flyback structure for these various optimization aspects are developed, validated by numerical simulation or measures and integrated in an optimization process. Then, the optimization results are presented and validated thanks to a measurement workbench of the flyback structure.

Dielectric losses of capacitor and ferrite core in an LCT component

The lumped component equivalent circuit of an inductor capacitor transformer integrated component (LCT) is deduced from a set of impedance measurements. This circuit allows the whole electronic behavior of the component, including losses, to be accurately forecast by using standard electronic simulation software. A closer look at experimental data reveals that several physical phenomena, often neglected, have a strong impact on LCT electrical properties. Dielectric losses in the capacitor and permittivity related losses in ferrite core are among them. To sum up, this paper lists main phenomena to account for in future field computations, it supplies a suitable equivalent circuit that accounts for them and it shows how to extract all its element values from a set of impedances deductible from field computations.

Design and identification of an equivalent circuit for an LCT component: inventory and representation of losses

A new kind of passive component, the LCT (integrated inductor (L), capacitor (C) and transformer (T)) is becoming very fashionable in power electronics. Starting from our knowledge in transformer characterization, we recently elaborated an equivalent circuit for this component. This equivalent circuit is fully deducible from impedance measurements, no dismounting of the component is needed and no information about its internal design is required. This paper aims mainly to present the new equivalent circuit and the method leading to its identification using a commercial impedance analyzer. A LCT intended to a 300 W power supply working at 200 kHz has been built and characterized between 100 Hz and 40 MHz. Analyzing the frequency dependence of the real part of serial and parallel impedances we identified several kind of losses rarely taken into account. Among these losses are those related to the capacitor insulator and those due to the dielectric constant of the Mn-Zn core material. Equivalent circuits are supplied to represent every type of losses and their accuracy is checked.

Accounting for resistivity and permittivity in high frequency permeability measurements: application to MnZn ferrites

A careful look to methods used by manufacturers to measure the permeability of soft magnetic material let appears that these data are not reliable when induced currents flow in the magnetic material leading to a screening effect. This is obviously the case of bulk cast iron as soon as eddy currents flows inside. It is also applicable to MnZn ferrites because, beyond 100 kHz, they are gone through by both conduction currents due to their finite resistivity and displacement currents related to their very high permittivity. In both cases, flux density is not uniform on the core cross section. In this paper, we present an original method which allows, at every frequency, simultaneous evaluation of both complex permeability and permittivity (which accounts for resistivity). Completing the traditional inductance measurement by a capacitance one, enough experimental data are acquired, at every frequency, to deduce the two complex parameters. The material is assumed to behave linearly but, owing to special experimental care, no other hypothesis is needed. This method has proven to be reliable up to 10 MHz and, above some 100 kHz, obtained permeability shows significant differences with that measured traditionally. Contrary to traditional specifications, obtained parameters suit to design new core shapes intended to work in the 100 kHz-10 MHz range.

Modeling of passive electronic circuits with sensitivity analysis dedicated to the sizing by optimization

The paper deals with an approach that automates the computation of frequential characteristics of passive electronic circuit and their associated sensitivities according to all the components of the circuit. This method enables the designer to focus on the physical behavior of the circuit since the modeling and computing tasks are automatically performed without any computer science skills. It is useful to size circuits with many constraints by using optimization based on gradient algorithm.

Flyback converter surface minimization: Design procedure and formulas

This paper gives generic formulas to account for power components surface in the design of a flyback converter. Starting from the converter requirements, key design parameters are defined, and their influence on converter surface studied. It is shown that some parameters have opposite actions, thus some trade-off must be found. On the other hand, additional constraints must be taken into account, especially losses. These ones are also evaluated using analytic formulas, and included in the design procedure. After design, the converter has been built, and compared to previous version, surface reduction is obvious.

Characterization of material versus temperature using P(VDF-TrFE) SAW transducers

This work approaches the elastic properties of non-piezoelectric materials and reports their temperature dependence for a duraluminum substrate. The method is based on the propagation time measurement of ultrasonic Rayleigh waves and attenuation of electrical voltage between the emitter and receiver. Transducers made of P(VDF-TrFE) polymer film are used for Surface Acoustic Waves (SAW) generation and detection.