No abstract

No abstract

SUMMARY Because of the improved performance of power devices, the volume of the ac filter inductors used in high‐frequency PWM inverters has been reduced. However, the temperature rise in the filter inductor due to this miniaturization has become more pronounced. Therefore, we have proposed an iron loss calculation method for the ac filter inductor. However, the accuracy of the value calculated via the loss map method cannot be verified, because the iron loss arising during each switching period cannot be measured with conventional power measuring instruments. In order to resolve this problem, we developed an inductor loss analyzer (ILA), which allows precise measurement of the iron loss in the inductor during each switching period. The accuracy of the calculation of iron loss in the filter inductor by the loss map method was verified with the ILA. We found that the value calculated by the loss map method differed slightly from the value measured with the ILA. However, these differences can be reduced if we take into account the accurate flux density calculation and the effect of the duty ratio of PWM pulses on the loss. Finally, we verified that the loss map method can provide accurate iron loss calculations.

No abstract

SUMMARY The inductor losses in a three‐phase ac filter inductor used in a three‐phase pulse‐width modulation (PWM) inverter are evaluated. First, a three‐phase inductor is designed to obtain the same value of inductance for each phase. Then, based on the design, a three‐phase inductor that uses two magnetic materials is proposed. The conversion efficiency of a 1 kVA three‐phase PWM inverter that uses the conventional and proposed ac filter inductors is simulated. Simulation results show that conversion efficiency improves. Finally, the conversion efficiency of an actual three‐phase 1 kVA PWM inverter that uses the conventional and proposed ac filter inductors is measured. In the experiment, the conversion efficiency obtained for the case of the proposed inductor improves by approximately 1% at low power load as compared to the conventional inductor. Furthermore, the calculated inductor losses are in good agreement with measured losses. Improvement in efficiency is verified trough simulations and experiments.