Newly available Medium-Voltage (MV) Silicon-Carbide (SiC) transistors are setting new limits for the design space of MV converters. Unprecedented blocking voltages, higher switching frequencies, higher commutation speeds, lower losses, and high temperature operation can be reached, which, however, create new challenges for the electrical insulation. In particular, the dielectric losses can become significant for MV converters operated at higher switching frequencies. Moreover, the evaluation of the dielectric losses is a key element for assessing the insulation stress and for insulation diagnostics. Therefore, this paper analyzes the modeling and the computation of dielectric losses with PWM voltages. After a review of the dielectric loss mechanisms occurring in polymeric insulation materials, scalable analytical expressions are proposed for the losses produced by PWM voltages. Afterwards, the dielectric losses of a Medium-Frequency (MF) transformer, employed in a 25 kW MV DC-DC converter (7 kV to 400 V) operated at 48 kHz, are analyzed and measured in detail. With epoxy resin, the insulation losses represent a significant share, i.e. up to 17 %, of the total transformer losses. As shown, the transformer performance can be significantly improved with a silicone elastomer insulation. Finally, design guidelines are provided for the selection of insulation materials.

Solid-state transformers (SSTs) are employing compact high-power medium-voltage (MV) medium-frequency transformers (MFTs) and enable the power transfer between galvanically isolated DC systems. Considering a typical SST isolation stage, i.e., a 166 kW unregulated series-resonant DC–DC converter acting as a DC transformer (DCX) with equal input and output MV DC voltages of 7 kV (1:1-DCX), we derive component- and system-level performance limits of air-cooled realizations with either an air-core transformer (ACT) or with a magnetic-core transformer (MCT). We describe the design of two fully rated MFT prototypes in detail and provide a comprehensive experimental characterization to validate the derived performance limits, including the dielectric losses and the analysis of stray magnetic fields. The realized ACT and MCT prototypes achieve measured efficiencies of 99.5% and 99.7% at gravimetric power densities of 16.5 and 6.7 kW/kg, respectively. Considering 10 kV SiC MOSFETs, calculated system-level efficiencies of the ACT-based and MCT-based 1:1-DCX result in 99.0% and 99.2% at the nominal operating point, with similar part-load efficiency characteristics. This article concludes with an application-oriented qualitative evaluation of the two concepts.

The galvanic isolation of solid-state transformers (SSTs) is typically realized with a medium-frequency (MF) magnetic-core transformer (MCT). Previous demonstrations indicate that achieving highly power-dense and lightweight MCTs imposes several challenges on the design because of stringent requirements related to insulation and cooling. This work investigates the achievable efficiency of an optimized 166 kW/7 kV air-core transformer (ACT), which is a core part of a dc transformer (DCX), i.e., an unregulated dc–dc SST with a voltage scaling defined by the transformer turns ratio. The ACT features relatively low complexity of the construction, comparably high coupling values, and high efficiency. Modeling, optimization, and construction of the realized ACT are explained and guidelines regarding insulation, cooling, and shielding of the magnetic stray flux are discussed in detail. Furthermore, the prototype is experimentally validated to demonstrate its full functionality. In the investigated DCX, which is based on a series resonant converter (SRC) topology, the realized ACT is found to achieve a full-load efficiency of 99.5% and an unprecedented gravimetric power density of 16.5 kW/kg. With the use of 10 kV SiC MOSFETs, the complete DCX is estimated to reach an efficiency of 99% at 166 kW output power.

This paper studies the loss-optimal design of a power inductor employed in a 2kW, 400V input DC-DC converter, for a given magnetic core (E55/28/21, N87), and with a focus on optimal choices of key design parameters, i.e., switching frequency, f, relative peak-to-peak current ripple, r, and number of turns, N. The investigation is based on analytical equations and, subsequently, a more detailed semi-numerical model. Surprisingly, the evaluation of the losses calculated in the f-r domain reveals that nearly minimum inductor losses are obtained for a current ripple that is inversely proportional to the frequency, i.e., for a constant inductance, within a wide frequency range, from 200kHz to 1MHz. Furthermore, the investigation reveals a decrease of the losses for increasing frequencies up to 375kHz. The detailed analysis related to these results enables the compilation of a simple two-equations guide for the design of an inductor that achieves close to minimum losses. In a next step, interesting trade-offs are identified based on a study of the design space diversity, e.g., with respect to low cost and increased partial-load efficiency. The findings of this work are experimentally verified, i.e., the losses of three different inductors are measured with an accurate calorimetric method and at four different frequencies, ranging from 150kHz to 700kHz.

An accurate and fast transient calorimetric ferrite core-loss measurement method is proposed in this paper. In contrast to electrical measurements, the accuracy of the calorimetric approach is largely independent of the magnetic excitation and operating frequency. However, accurate values of the thermal capacitance and the temperature of the Core Under Test (CUT) are required. Accurate measurement of the specific heat capacity of the core material can be achieved with a Differential Scanning Calorimeter (DSC) or by using the CUT as a DC electric conductor and measuring its thermal response for known Joule heating. Accurate temperature measurements can be realized with NTC temperature sensors. A thorough uncertainty analysis of the presented method is conducted by identifying the impact of each source of uncertainty in the course of a sensitivity analysis. The results are verified by means of FEM simulations and experiments. Furthermore, a step-by-step description of the workflow for preparing and conducting the experiments is provided. The proposed method is tested experimentally and compared to a state-of-the-art electrical loss measurement method for MnZn N87 and N49 ferrite cores of EPCOS-TDK. In addition, it is used to measure the loss-map of the NiZn ferrite material 67 from Fair-Rite for very high frequencies up to 50MHz.

The endurance of polymeric insulation foil is investigated under a mixed medium-voltage stress (DC + medium-frequency rectangular pulse) by means of accelerated lifetime testing. A dedicated setup is used that allows us to selectively eliminate the known risk factors for premature insulation failure under medium-frequency pulse voltage stress: partial discharges (PDs) during pulse transitions, excessive dielectric heating, and systemic overvoltages. The obtained results on polyethylenterephtalat (PET) insulation foil suggest that the adequate consideration of these factors is sufficient for eliminating the adverse effects of the pulse modulation under the investigated conditions. Indeed, if all mentioned risk factors are eliminated, the time to failure observed under a pure DC stress is shorter than with a superimposed pulse (keeping the same peak voltage). There is then no indication of an additional detrimental “per pulse” degradation process (i.e., the time to failure is not dependent on pulse frequency). In contrast, when repetitive PDs are present, the lifetime under combined DC + rectangular pulse stress strongly decreases with increasing pulse switching frequency. PD erosion of the foil is quantified by means of confocal microscopy, and the applicability of the streamer criterion for predicting PD inception is discussed.