Sequential offline-online-offline measurement approach for high-frequency LCLC resonant converters in the TWTA applications

• Main Authors: Zhao, Bin, Zhang, Xin, Zhang, Zhe
• Other Authors: School of Electrical and Electronic Engineering
• Format: Article
• Language: English
• Published: 2022
• Subjects: Engineering::Electrical and electronic engineering; Resonant Converters; High Frequency
• Online Access: https://hdl.handle.net/10356/155314
• Institution: Nanyang Technological University

The high-frequency LCLC resonant converter is one of the important parts of the two-stage power supply in the space travelling-wave tube amplifier application. Usually, the high-frequency LCLC resonant converter utilizes open-loop control and simultaneously sets its switching frequency and duty cycle to the required values to guarantee low cost and high power efficiency. However, the required switching frequency and duty cycle are determined by the transformer parasitic parameters. Therefore, how to measure the real transformer parasitic parameters under the real working conditions precisely becomes very important to the high-frequency LCLC resonant converter. The conventional way to measure the transformer parasitics is to employ an offline impedance analyzer. However, the transformer parasitics under real operating conditions may deviate from the offline measured results. The online measurement methods can obtain the real values when the high-frequency LCLC resonant converter is working, but these online approaches also mean additional cost and complex implementation requirement. To solve the above-mentioned problems, a sequential offline-online-offline (SO3) measurement method is proposed in this paper. With this SO3 measurement approach, all the real transformer parasitics can be easily obtained in a low cost and simple implementation way, which combines the advantages of both traditional offline and online measurement methods while removing their corresponding shortcomings. The proposed method is validated by the experiments.