Analysis And Optimization Of EMI Characteristics In PSR Based Flyback Constant Power Supply

Flyback switching power supply based on primary side return (PSR) has a broad prospect in products which pay a strict requirement on volume and cost. The conducted electromagnetic interference(EMI)would have some effects on power supply itself and other devices. Many countries have developed appropriate EMI standards for peremptory norms of flyback converters. Therefore, how to predict and optimize the conducted EMI of PSR flyback converters effectively becomes the main content of this thesis.First of all, the research status, mechanism, test method of conducted EMI and the mechanism of PSR flyback converter were introduced in this thesis. Then, the equivalent circuits of different mode and common mode were built and several related formulas were induced. After that, the characteristics of transformer, power MOSFET, power diode and passive components in system were analyzed and the high-frequency modules were built. At the same time, the module integrated circuit was described by the software Saber and parasitic parameters were extracted by the software Q3D.Thus, the conducted EMI module of flyback constant power supply based on PSR was obtained. On the basis of the module, the simulation was done in Saber and followed by the verification through actual testing. At last, several methods for optimization of conducted EMI were put forward from three aspects:different mode path, common mode path and interference source. These methods includes adding a Y capacitor, selecting filter capacitors with low-ESR and adding a snubber circuit for secondary diode. And these methods were verified through simulation and actual testing.The simulation and testing results demonstrated that the simulation module built above could predict the actual situation accurately. The peak value error of conducted EMI between simulation and testing is less than 5dBμV. After optimization, the testing result of EMI peak values around 6MHz and 22MHz frequency band have improved more than 5dBμV compared with the result before optimization. The simulation and testing results have verified the correctness of theoretical analysis and suppression methods.