Evaluation of Input Current in the Critical Mode Boost PFC Converter for Distributed Power Systems* Jindong Zhang, Jianwen Shao, Peng Xu and Fred C. Lee Milan M. JovanoviC Center for Power Electronics Systems Electrical and Computer Engineering Department Virginia Polytechnic Institute and State University Delta Product Corp. Power Electronics Laboratory P.O. Box 12173,5101 Davis Dr. Blacksburg, VA 2406 1-0 1 1 1 ABSTRACT This paper presents the analysis and evaluation of the input current in the critical mode single-phase boost power-factor-correction (PFC) converter. First, the switching-frequency ripple of the input current is derived and the differential-mode EM1 filters of the critical and continuous mode PFC rectifiers are calculated and compared. Next, the benefits and challenges of interleaving the critical mode PFC rectifiers are discussed. Finally, the low-frequency input current distortions are analyzed with respect to the switching-frequency limits of the critical mode boost PFC rectifier. I. INTRODUCTION The front-end of distributed power systems (DPS) used in telecom and high-end server applications requires low input current distortions, high power density, good efficiency with universal line input, and low cost. Generally, to meet the low-input-current-distortion requirement, the continuous-conduction-mode (CCM) boost power-factor-correction (PFC) rectifier is the preferred topology for the input side of a DPS front end because the differential mode electromagnetic interference (DM-EMI) is minimized due to its low input current ripple. However, because of the hgh output voltage of the boost PFC rectifier, a high-voltage fast-recovery boost diode is required, which generates significant reverse- recovery losses when switched under hard-switching conditions. These reverse-recovery losses limit the converter efficiency, especially at low line and full load. One technique to eliminate the boost diode reverse- recovery losses is to operate the boost rectifier at the boundary between the continuous and discontinuous conduction mode (DCM). This lund of boundary operation is also referred to as the critical mode of operation. Figure l(a) shows the boost PFC circuit, whereas Fig. l(b) shows the critical mode inductor current waveform. In the circuit in Fig. l(a), the switch S is turned on immediately after the inductor current reaches zero, i.e., at the moment the rectifier current is zero, which eliminates the diode reverse-recovery losses. Moreover, with proper control timing and component choices, the boost switch can be Research Triangle Park, NC 27709 turned on with zero-voltage-switching (ZVS) to further improve the converter efficiency. However, further studies need to address two major issues in the critical mode PFC rectifier. The first issue is existence of a high input current ripple. It requires a large DM-EM1 filter with increased converter size and cost. To reduce the input current ripple and the EM1 filter size, interleaving techniques for the critical mode PFC rectifiers have been proposed [ 1],[2]. However, the interleaving the two variable-frequency boost converters requires relatively complex synchronization circuit. The second issue of the critical mode PFC is its wide switching-frequency range, which increases the switching losses. To limit switching loss, the maximum switching-frequency is usually limited. This frequency limit causes low-frequency distortion of the line current. Fig. 1 Critical mode boost PFC rectifier and its input current: (a) circuit diagram (with EM1 filter); (b) input current and boost inductor current waveforms. ms paper presents an in-depth study of these two issues of the critical-mode PFC rectifier. The first part of this paper analyzes the input current ripple spectrums and compares the DM-EM1 filter sizes of the CCM mode and critical mode PFC converters. Different interleaving schemes are discussed and the benefits and limitations of interleaving are also presented. The second part of this paper analyzes the dependence of the line current harmonics on the switching-frequency limit in two critical mode PFC rectifiers. Potential problems are exposed and possible solutions are proposed. This work was supported in part by a fellowship from Delta Product Corp. This work made use of ERC Shared Facilities supported by the U.S. National Science Foundation under Award Number EEC-9731677 0-7803-6618-2/01/$10.00 0 2001 IEEE 13011. EVALUATION OF THE SWITCHING nUEQUENCY CURRENT RIPPLE 2.1 Boost inductor ripple current spectrum analysis To achieve unity input power factor, the switch S in the critical mode boost rectifier shown in Fig.1 should be dnven with a variable switching cycle, constant on-time gate signal. The peak inductor current iLB(peak) always has a magnitude twice as great as that of the instantaneous sinusoidal input current ih. Due to its high inductor current ripple, the critical mode boost rectifier requires a much larger DM-EM1 filter than does the CCM boost rectifier. In order to evaluate the EM1 filters, the boost inductor current spectrum should be derived. This section presents the analysis of the CCM and critical mode inductor current in the frequency domain. A) Critical-mode inductor current spectrum analysis An analysis of the ripple current harmonics of the DCM boost rectifier was presented in [3] and [4]. Generally, the same analysis can be extended to the critical-mode boost rectifier. Ton Toff t TK TK+I t (b) Fig. 2 Mathematical function of critical inductor current: (a) one switching cycle current function f(t); (b) current iLB(t) is the sum of switching cycle fk(t). To calculate the critical mode inductor current spectrum in the frequency domain, the first step is to represent mathematically the high-frequency inductor current ripple of Fig. l(b). In Fig. 2(a), in the kth switching cycle, the inductor current can be regarded as a triangle function fK(t), which has a peak value of IK and two time variables To, and Toff. Equation (1) shows function fK(t). In order to obtain the current frequency spectrum, the next step is to apply the Laplace Transformation to the triangle function fK(t). Equation 2 shows the Laplace Transformation function fK(t). Figure 2(b)
View Full Document
Unlocking...