Nowadays, electronics has really flourished with the rapid progress in science and technology. Besides being light and efficient of the products, digitalization and faster speed are also required which make the problem of EMI, electromagnetic interference, become more serious. Newer power supplies using switching technology become smaller and smaller. However, they also produce EMI problems that need to be considered. In this thesis, we only address the conducted EMI. Generally, the noise is suppressed by adding a filter at the power-line end. Hence, the EMI filter plays an important role to the solution of conducted emission. The particular filter is related to the type of the conducted emission, including common-mode and differential-mode. As a result, it is a great help for an engineer to be able to separate the noise. There are several noise separators in the present days. We propose new separators, which are capable of detect common-mode and differential-mode separately. Our separators are low-cost, easily manufactured and are able to give exact noise readings without adjustment. Incorporating the known methods of filter design and combing programming measurement the filter design can be made automatic. One approach makes use of the spectrum analyzer, which is considered to be a relatively expensive instrument. We also propose another approach using oscilloscope. The time-domain measurement is performed and transformed into frequency domain using FFT. A realistic design example is shown to give satisfactory results.

This thesis presents the conducted electro-magnetic interference (EMI) prediction results for a continuous conduction mode (CCM) power factor correction (PFC) converter as well as the theoretical analysis for the noise generation and propagation mechanisms. In this thesis, multiple modeling and characterization techniques in the medium and high frequency ranges are developed for the circuit components that are important contributors to the EMI noise, so that a detailed simulation circuit for EMI prediction can be constructed. The conducted EMI noise prediction from the simulation circuit closely matches the measurement results obtained by a spectrum analyzer. Simulation time step and noise separator selection are two important issues for the noise simulation and measurement. These two issues are addressed and the solutions are proposed. The conducted EMI generation and propagation mechanisms are analyzed in a systematic way. Two loop models are proposed to explain the EMI noise behavior. The effects of the PFC inductor, the parasitic capacitance between the device and the heatsink, the rising/falling time of the MOSFET VDS voltage, and the input wires are studied to verify the validity of the loop models. / Master of Science

Power factor correction (PFC) converter is a species of switching mode power supply (SMPS) which is widely used in offline frond-end converter for the distributed power systems to reduce the grid harmonic distortion. With the fast development of information technology and multi-media systems, high frequency PFC power supplies for servers, desktops, laptops and flat-panel TVs, etc. are required for more efficient power delivery within limited spaces. Therefore the critical conduction mode (CRM) PFC converter has been becoming more and more popular for these information technology applications due to its advantages in inherent zero-voltage soft switching (ZVS) and negligible diode reverse recovery. With the emerging of the high voltage GaN devices, the goal of achieving soft switching for high frequency PFC converters is the top priority and the trend of adopting the CRM PFC converter is becoming clearer. However, there is the stringent electromagnetic interference (EMI) regulation worldwide. For the CRM PFC converter, there are several challenges on meeting the EMI standards. First, for the CRM PFC converter, the switching frequency is variable during the half line cycle and has very wide range dependent on the AC line RMS voltage and the load, which makes it unlike the traditional constant-frequency PFC converter and therefore the knowledge and experience of the EMI characteristics for the traditional constant-frequency PFC converter cannot be directly applied to the CRM PFC converter. Second, for the CRM PFC converter, the switching frequency is also dependent on the inductance of the boost inductor. It means the EMI spectrum of the CRM PFC converter is tightly related the boost inductor selection during the design of the PFC power stage. Therefore, unlike the traditional constant-frequency PFC converter, the selection of the boost inductor is also part of the EMI filter design process and EMI filter optimization should begin at the same time when the power stage design starts. Third, since the EMI filter optimization needs to begin before the proto-type of the CRM PFC converter is completed, the traditional EMI-measurement based EMI filter design will become much more complex and time-consuming if it is applied to the CRM PFC converter. Therefore, a new methodology must be developed to evaluate the EMI performance of the CRM PFC converter, help to simplify the process of the EMI filter design and achieve the EMI filter optimization. To overcome these challenges, a novel mathematical analysis method for variable frequency PFC converter is thus proposed in this dissertation. Based on the mathematical analysis, the quasi-peak EMI noise, which is specifically required in most EMI regulation standards, is investigated and accurately predicted for the first time. A complete approximate model is derived to predict the quasi-peak DM EMI noise for the CRM PFC converter. Experiments are carried out to verify the validity of the prediction. Based on the DM EMI noise prediction, worst case analysis is carried out and the worst DM EMI noise case for all the input line and load conditions can be found to avoid the overdesign of the EMI filter. Based on the discovered worst case, criteria to ease the DM EMI filter design procedure of the CRM boost PFC are given for different boost inductor selection. Optimized design procedure of the EMI filter for the front-end converter is then discussed. Experiments are carried out to verify the validity of the whole methodology. / Ph. D. / Power factor correction (PFC) converter is widely used in offline frond-end converter for the distributed power systems to reduce the grid harmonic distortion. With the fast development of information technology and multi-media systems, high frequency PFC power supplies for servers, desktops, laptops and flat-panel TVs, etc. are required for more efficient power delivery within limited spaces. Therefore the critical conduction mode (CRM) PFC converter has been becoming more and more popular for these information technology applications. However, there is the stringent electromagnetic interference (EMI) regulation worldwide. For the CRM PFC converter, there are many challenges on meeting the EMI standards. To overcome these challenges, a novel mathematical analysis method for variable frequency PFC converter is thus proposed in this dissertation. A complete approximate model is derived to predict the quasi-peak DM EMI noise for the CRM PFC converter. Experiments are carried out to verify the validity of the prediction. Based on the DM EMI noise prediction, worst case analysis is carried out and based on the discovered worst case, criteria to ease the DM EMI filter design procedure of the CRM boost PFC are given for different boost inductor selection. Optimized design procedure of the EMI filter for the front-end converter is then discussed. Experiments are carried out to verify the validity of the whole methodology.

Five chapters of this dissertation concentrate on the characterization and cancellation of high frequency parasitic parameters in EMI filters. One chapter addresses the interaction between the power interconnects and the parasitic parameters in EMI filters. The last chapter addresses the characterization, evaluation and design of noise separators. Both theoretical and experimental analyses are applied to each topic. This dissertation tries to explore several important issues related to EMI filters and noise separators. The author wishes to find some helpful approaches to benefit the understanding and design of EMI filters. The contributions of the dissertation can be summarized below: 1) Identification of mutual couplings and their effects on EMI filter performance 2) Extraction of mutual couplings using scattering parameters 3) Cancellation of mutual couplings to improve EMI filter performance 4) Cancellation of equivalent series inductance to improve capacitor performance 5) Analysis of mode transformations due to the imperfectly balanced parameters in EMI filters 6) Analysis of interaction between power interconnects and EMI filters on filter high-frequency performance 7) Modeling and design of high-performance noise separator for EMI diagnosis 8) Identification of the effects of parasitics in boost PFC inductor on DM noise Although all topics are supported by both theory and experiments, there may still be some mistakes in the dissertation. The author welcomes any advice and comments. Please send them via email to shuowang@ieee.org. Thanks / Ph. D.

With the development of power electronics, electromagnetic interference (EMI) and electromagnetic compatibility (EMC) issues have become more and more important for both power converter designers and customers. This dissertation studies EMI noise emission characterization and modeling in power converters. A modular-terminal-behavioral (MTB) equivalent EMI noise source modeling approach is proposed. This work is the first to systematically develop a 3-terminal EMI noise source model for a switching phase-leg device module. Each module is modeled as pairs of equivalent noise current sources and source impedances. Although the proposed MTB modeling approach applies the linear circuit theory to a semiconductor switching device, which exhibits nonlinear behavior during switching transients, the analysis and experiments show that the nonlinearity has negligible practical effect on the modeling methodology. The validation range of the modeling methodology has been analyzed. One of the differences between the proposed MTB model and the other state-of-the-art models is that the MTB model characterizes and predicts the CM and DM noise simultaneously. The inseparable high-frequency CM and DM noise characteristics contributed by the source impedance and propagation path are analyzed. A comprehensive evaluation of different EMI noise source modeling approaches according to the criteria of accuracy, feasibility and generality has been presented. Results show that the MTB modeling approach is more accurate, feasible and general than other approaches. The modular and terminal characteristics of the MTB noise source model are verified in two more complicated cases. One example is the application of the MTB equivalent source model in a half-bridge AC converter with variable switching conditions. Although the MTB model is derived under a certain operating condition, the models under different conditions can be combined together to predict the EMI noise for the converter with variable switching conditions. Another example is the application of the MTB equivalent source model in multi-phase-leg converters. The EMI noise of a full-bridge converter is predicted based on the MTB equivalent source model of one phase-leg module. The implementation procedures and results for both applications are verified by the experiment. The applicability of the MTB model in different type of converters is discussed. Based on the MTB model, EMI noise management is explored. The parametric study based on the MTB model is demonstrated by selecting DC-link decoupling capacitors for voltage source converter (VSC). The EMI effect of a decoupling capacitor for a device s safe operation is analyzed, and this analysis shows the terminal characteristics of the MTB model. Both the EMI and voltage overshoot are predicted by the MTB model. A completed converter-level EMI model can be derived based on the noise source model and propagation path model. This model makes it possible to optimize the EMI filter design and study the EMI noise interactions between converters in a power conversion system. / Ph. D.