POW E R Jeo Be S(O Ee @ Bea Le GCURC UIT S ISSA BATARSEHEXECUTIVE EDITOR Bill Zobrist MARKETING MANAGER Katherine Hepbur. SENIOR PRODUCTION EDITOR Valerie A. Vangas SENIOR DESIGNER Dawn Sanley PRODUCTION MANAGEMENT SERVICES. Publication Services, “his book was st in 10/12 Times Roman by Publication Scrviees and printed end bound by Hamilton Printing. ‘The cover was printed by Phoenix Color. ‘This book i printed on aise paper. Copyright ©2004 John Wiley & Sons Inc. All righ reserved [No par of this publication may be reproduced, store ina etieval system or transmitted in any form or by any means, electonic, mechanical, photocopying, recording, ‘canning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior writen permission ofthe Publisher, oF uthorization through payment ofthe appropriate per-copy Zee wo the Copyright Clearance ‘Center, Inc. 222 Rosewood Drive, Dunvers, MA 01923, (508)7S0-8400, fax (508750 ‘4470. Requests tothe Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc, 111 River Suset, Hoboken, NJ.07030, 201)748- ‘601, fax (201)748-6008, E-Mail: PERMREQ@ WILE Y.COM, ‘Tocrder books ofr customer service please cll 1-800-CALL WILEY (225-594) cvasie: JKR. apguis Library of Congress Cataloging in Publication Data: “Author's mame entry. Brie ie entry: Subtitle if any xexxxx soar re eA NN {Series Statement) d Raver cicchrOnvic’ Notes wxxxxaxcaxccancxommmmnn 2. Pectipnie. Cureull S L.C. Subject Heading Joint Author Entry Title and Series added entries : Lc. Calne) Dewey Classification No. LC. Card No, ISBN 04471-12662-4 ISBN 0-471-45228.9 (WIE) Printed in the United States of America 987654321Preface Th recent years the field of power electronics has witnessed unprecedented research and teaching growth worldwide, emerging as a specialization in electrical enginecr- ing. This growth is due to expanding market demand for power electronic circuits for the energy conversion process. The need for power electronics engineers equipped with knowledge of new energy conversion technologies has never been greater. ‘Power Electronic Circuits is intended asa textbook to teach the subject of modem power electronics to senior undergraduate and first-year graduate electrical engincer~ ing students, Because of the breadth of the field of power electronics, teaching this suibject to undergraduate students is a challenge. This textbook is designed to intro- duce the basic concepts of power electronics to students and professionals interested in updating their knowledge of the subject. The objective of this textbook is to provide students with the ability to analyze and design power electronic circuits used in var ous industrial applications. ‘The prerequisites for this text are a fist course in circuit analysis techniques and a basic background in electronic circuits. Chapter 3 gives an overview of diode switch- ing circuits and basic analysis techniques that students will find useful in the remain- ing chapters. ‘Material Presentation Since the text is intended to be used in a three-credit-hour course in power electronics, topics such as power semiconductor devices, machine drives, and utility applications are not included. Because of limited lecture times, one course at the undergraduate level cannot adequately cover such topics and still present all power electronic cir- cuits used in energy conversion. This text contains sufficient material for a sin- tle-semester introduetory power electronics course, while giving the instructor flexibility in topic treatment and course design. The text is written in such a way as to equip students with the necessary back- ‘ground material in such topics as devices, switching circuit analysis techniques, con- ‘verter types, and methods of conversion in the first three chapters. The presentation of the material is new and has been recommended by many power electronics faculty. ‘The discussion begins by introducing high-frequency, nonisolated de-to-de converters in Chapter 4, followed by isolated de-to-de converters in Chapter 5. Resonant soft- switching converters are treated carly on in Chapter 6, The traditional diode and SCR converters and de-ac inverters are presented in the second part of the text, in Chapters 7, 8, and 9, respectively. Examples, Exercises, and Problems Unlike many existing texts, this text provides students with a large number of exam- ples, exercises, and problems, with detailed discussion of resonant and sofiswitching, de-to-de converters.vill Preface Examples are used to help students understand the material presented in the chap- ter, To drill students in applying the basic concepts and equations, and to help them: understand basic circuit operations, several exercises are given within cach chapter. ‘The text has more than 250 problems at different levels of complexity and difficulty. ‘These problems are intended not only to strengthen students’ understanding of the materials presented, but also to introduce many new concepts and circuits. To help imeet recent Accreditation Board for Engineering and Technology (ABET) require- ‘ments for design in the engineering curriculum, special emphasis is made on provid- ing students with opportunities 10 apply design techniques. Such problems are designated with the letter “D” next to the problem number, such as D5.32. Students should be aware that such problems are open-ended without unique solutions. A bibliography is included at the end of the text and a list of texthooks is given separately. About the Text Web-Based Course Materis Like the majority of textbooks, this book was developed from class notes the author prepared over the last eight years while teaching power electronics at the University of Central Florida. The author started teaching power electronics in 1991, when only a limited number of power electronic textbooks were available. Since then, a handful of additional textbooks have been published with very similar material coverage. Unlike ‘many existing texts, Power Electronic Circuits targets mainly senior undergraduate students majoring in electrical engineering. Ancillaries to this text are available on a dedicated Web site, waw.batarsch.org, estab- lished for both faculty and students to provide them with access to + Acomplete set of lecture notes + Sample quizzes + PSPICE- and Mathcad.-based simulation examples + A complete solutions manual + Transparency masters *+ Up-to-date text corrections and the opportunity to submit new corrections Even though the author's Web site is very useful in providing students and faculty «with teaching material, at the time of this writing it doesn’t offer an interactive platform. ‘The author believes that due to the nonlinear interdisciplinary nature of the field of ‘power electronics, computer simulations and computer tools must play an important role in delivering effective power electronics education. This is why any useful Web- ‘based education in the area of power electronics must include a platform that allows stu- dents to do design and simulation interactively online. To the author's knowledge, this has not been effectively achieved yet. However, one of the most useful interactive Web- based facilites in power electronics is the Interactive Power Electronics Seminar (iPES) developed by the Swiss Federal Institute of Technology, Zurich (ETH Zurich), hitp// sww.ipes.thz.ch/. The Java applets are designed to allow a degres‘of interactivity and ‘aniraation to aid in leaming the basics of power electronics typically taught in an intro- About the Text and its Nomenclature 16 2 Review of Switching Concepts and Power Semiconductor Devices 20 Troduetion 20 22 The Need for Switching in Power Electronic Circuits 21 2 Switching Characteristics 25 The Ideal Switeh 25 The Pructical Switch 25 25 Switching Funetions and Matrix Representation 30 Types of Switches 36 Available Semiconductor Switching Devices 37 Bipolar- and Unipolar-Based Devices 38 Thyristor-Based Devices $8 © Comparison of Switching Devices 67 27 Future Trends in Power Devices 67 © Soubber Circuits 68 Problems 69) & Switching Cireuits, Power Computations, and Component Concepts 75 Tnroduetion 75 Switching Diode Circuits 76 Switching Diode Circuits under de Excitation 76 Switching Diode Circuits with an ac Source 82 Switching SCR circuits 88 Basic Power and Harmonie Concepts 90 Average, Reactive, and Apparedt Powers 90) ‘Sinusoidal Waveforms 94 Nonsinusoidal Waveforms 98 34 Capacitor and Inductor Responses 106 Capacitor Transient Response 107 Capacitor Steady-State Response 110 Inductor Transient Response 111 Inductor Steady-State Response 112 Problems 114 4. Nonisolated Switeh-Mode de-de Converters 128 Introduction 128 4.1 Power Supply Applications Linear Regulators. 129, ‘Switched-Moxle Power Supplies 131 4.2 Continuous Conduction Mode 132 ‘The Buck Converter 136 ‘The Boost Converter 147 ‘The Buck-Boost Converter 154 Fourth-Order Converters 159 Bipolar Ousput Voltage Converters 171 129 4.3 Discontinuous Conduction Mode 173 ‘The Buck Converter 173 ‘The Boost Converter 179. ‘The Buck-Boost Converter 182 44 The Effect of Converter Nonidealities 186 Inductor Resistance 187 ‘Transistor and Diode Voltage Drop 191 ‘Switch Resistance 192 48 Switch Utilization Factor 199 Problems 201 Isolated Switch-Mode de-to-de Converters 209 Introduction 209 5.1. Transformer Circuit Configurations 210 3.2 Buck-Derived Isolated Converters 213 Single-Ended Forward Converter 216 HalfBridge Converter 224 Full-Bridge Converter 226 Push-Pull Converter 227 5.3 Boost-Derived Isolated Converters 232 ‘Single-Ended Flyback Converter 232 Half-Bridge Converter. 237 Full-Bridge Converter 238xiv Contents 54 Other Isolated Converters 242 Isolated Cuk Converter 242 Weinberg Converter 244 5.5 Multi-Output Converters 251 Problems 252 6. Soft-Switching de-de Converters 262 6.1 Types of de-de Converters 263 6.2 Classification of Soft-Switching Resonant Converters 265 63 Advantages and Disadvantages of ZCS. andZVS 265 64 Zero-Current Switching Topologies The Resonant Switch 65 Zero-Voltage Switching Topologies Resonant Switch Arrangements ‘Steady-State Analyses of Quasi Resonant Converters 292 66 Generalized Analysis for ZCS 303, ‘The Generalized Switching Cell ‘The Generalized Transformation Table 305 Basic Operation of the ZCS QRC Cell 306 Basie Operation of the ZVS ORC Cell 311 6.41 Zeco-Voltage and Zero-Current Transition Convenes 319 Switching Transition 320 The Boost ZVT PWM Converter Problems 325 267 290 2m 304 321 7. Uncontrolled Diode Rectifier Circuits Introduction 332 7.1. Single-Phase Rectifier Circuits Resistive Load 334 Inductive Load 336 Capacitive Load 344 Voltage Source in the de Side 7.2. The Effect of ae-Side Inductance Half-Wave Rectifier with Inductive Load 348 Half-Wave Rectifier with Capacitive Load 352 Full-Wave Rectifier with Inductive Load 356 73 Three-Phase Rectifier Circuits 360 Three-Phase Half-Wave Rectifier 360 Three-Phase Full-Wave Rectifier 365, 74 a¢-Side Inductance in Three-Phase Rectifier Circuits 368 332 334 348 348 Half-Wave Rectifiers _ 368 Full-Wave Bridge Rectifiers 371 Problems 374 8. Phase-Controlled Converters Introduction 382 8.1 Basic Phase Control Concepts 8.2 Half Wave Controlled Rectifiers Resistive Load 385 Inductive Load 393 8.3 Full-Wave Phase-Controlled Rectifiers. 391 Resistive Load 391 Inductive Load 393 8.4 Effect of ac-Side Indueunee Half-Wave Circuits 404 Full-Wave Bridge Cireuits 85 Three-Phase Controlled Converters Talf:Wave Converters 409 Full-Wave Converters 411 Problems 417 402 9. de-ac Inverters 422 “Totroduction 422 9.1 Basic Bleck Diagram of do- Inverter 382 383 385 407 409 423 92. Basic Half-Bridge Inverter Circuits 426 Resistive Load 426 Inductive-Resistive Load 9.3 Full-Bridge Inverters 437 Approximate Analysis 443 Generalized Analysis 44 9.4 Harmonic Reduction 451 9.5 Pulse-Width Modulation 457 Equal-Pulse (Uniform) PWM Sinusoidal PWM. 465 9.6 Three-Phase Inverters 478 9.7 Curreni-Souree Inverters 485 Problems 487 a9 459 Appendix A. Introduction to Magnetic Cireuits 494 Introduction 494 ‘AJL Types of Magnetic Materials ‘A2 Magnetic Fields 497 Toroidal Structure 498 A3 Magnetic Circuits $00 ‘Ad The Magnetizing Curve AS Inductors 509 A6 Transformers $10 Ideal Transformers Nonideal Transformers ‘Transformer Equations Leakage Inductances 495 506 su 513 516 317Magnetic Losses Core Material and Types A7 Three-Phase Transformers Appendix B. Fourier Series for Common Contents xv Bl Square Waveforms $31 B2 Sinusoidal Waveforms 539 Appendix C. Useful Functions 549 C.1 Trigonometri Identities 549 C2 Some Laplace Transformations 550 C3 Derivatives and Intograls 550 C4 Definite Integrals 552Chapter 1 Introduction INTRODUCTION 1.1 WHAT IS POWER ELECTRONICS? 1.2. THE HISTORY OF POWER ELECTRONICS 1.3 THE NEED FOR POWER CONVERSION 14 POWER ELECTRONIC SYSTEMS 1.4.1 Classification of Power Converter Circuits 14.2 Power Semiconductor Devices 143 Converter Modeling and Control 1.5 APPLICATIONS OF POWER ELECTRONICS 16 FUTURE TRENDS 1.7 ABOUT THE TEXT AND ITS NOMENCLATURE |NTRODUCTION This chapter is intended to give the reader an overview of the field of power electron ics and its applications. Basic block diagrams will be provided for a power electronics system and its major functions. Different types of power electronic circuits used to achieve power conversion will be presented. 1 WHAT IS POWER ELECTRONICS? To date, there is no widely accepted definition that clearly and specifically delimits the field of power electronics. In fact, many experts in the academic and industrial ‘communities feel that the name itself does not do justice to the field, which is applica~ tions oriented and multidisciplinary in nature, and which also encompasses many sub- areas in cleetrical engineering." Because of the multidisciplinary nature of the field of power electronics, experts must have a commanding knowledge of several electrical engineering subjects, such as electronic devices, electronic circuits, signal processing, ‘magnetism, electrical machines, control, and power. In a very broad sense, power clectronic circuits perform the task of processing one form of energy supplied by a source into a different form required at the load side. Henee, power electronics ean be closely identified with the following suibdiscipline areas of electrical engineering: "Many schools today offer power electronics under either the “power” or the “electron limited number of schools presenti separately "area, while a2 Chapter 1 Introduction 1 Gamenson Teasley Few Seniconactr Tehacogy | | + cheat teny pave dees ' 1 Comer esney Pcie i 12 Setting ato + Dine cre 1 1 stlpmeaning + Rraecton ee ' {entation Eosers | 1+ Magnetic components | + Rotating machines Power Control Technology TC control packages ‘Microelectonie iets Microprocessor eres Digital and analog cletone circuits ‘Conte theory wasnt and stability issues Digital nd information signal processing Silation Figure 1.1 Power electronics encompasses three technologies: power semi= ‘conductor, power conversion, and power control. electronics, power, and control. Here, electronics deals with the semiconductor devices and circuits used in signal processing to implement the control functions, power deals with both statie and rotating equipment that uses electric power, and con! deals with, the steady-state stability of the closed-loop system during the power conversion pro- cess. Hence, the subject of power electronics deals specifically with the application of electronic semiconductor devices and circuits in the conversion and control of electric power. In summary, power electronics is a technology that brings together three fun- damental technologies: power semiconductor technology, power conversion technology, and power control technology, as illustrated in Fig. 1. ‘A final observation is that in power electronic circuits there exist two types of switching devices: one type in the power stage that handles high power up to hun- dreds of gigawatts (which represents the muscle of the system) and another type in the eedbuck control circuit that handles low power up to hundreds of milliwatts, repre- senting the brain or intelligence of the system, Hence, today’s power clectronic cit- ‘cuits are essentially digital electronic circuits whose switching elements manipulate power ftom milliwatts to gizawatts. Asa result, one may conclude thatthe task of power electronics is to convert and control power using low-power switching devices that process power that is at much higher levels (a hundred times as great, or even greater). For example, a six-pulse SCR inverter with a power rating of a few kilowatts can pro- cess and control megawatts of power. Recent Growtl Power Electronics ‘The field of power electronics has recently experienced unprecedented growth in terms of research and educational activities, Its applications have been steadily and rapidly expanded to cover many sectors of our society. This growth is due to several factors; paramount among them is the technological advancement by the semiconduc- tor device industry, which has led to the introduction of very fast high-power capabil- ities and highly integrated power semiconductor devices. Other factors inelude (1) the revolutionary advances made in the microelectronics field that have led to the devel-1.2. The History ofPower Electronics 3 ‘opment of very efficient integrated circuits (ICs) used for the generation of control signals for processing and control purposes, (2) the ever-increasing demand. for smaller-size and lighter-weight power electronic systems, and (3) the expanding mar- kot demand for new power electronic applications that require variable-speed motor drives, regulated power supplies, robotics, and uninterruptible power supplies. This ‘increasing reliance on power electronic systems has made it mandatory that all such systems have radiated and conducted electromagnetic interference (EMI) limited ‘within regulated ranges. The industry's interest in developing power systems with low harmonic content and with an improved power fuctor will continue to place the field of power electronics at the top of the research priority lis. 42 THE HISTORY OF POWER ELECTRONICS Before a review of the history of power electronics in the past century, it might be use- fal to cover the history of the development of ac and de electricity in the last two de- cades of the nineteenth century. The inventions of the 1880s resulted in the present ‘worldwide use of the ac electric power system, providing the energy form that must be processed for any power electronics application. The History of de and ac Electricity in the Late Nineteenth Century It was decided in the middle of the nineteenth century that electrical energy was the most practical and economic form of energy for human use. Electricity was recog- nized as an excellent form of energy in terms of generation, transmission, and distri- bution. However, not long afterward, a heated debate began among scientists on ‘whether the future of transmitting and distributing electricity to industries and homes, would be based on alternating current (ac) or direct current (dc). George Westing- house and Nikola Tesla (1856-1943) represented the ae camp, and Thomas Edison (1847-1931) represented the de camp. After more than 15 years of intellectual debate, supported by new inventions and developmental and experimental studies, the ac advocates won; consequently, the entire world today uses an ac-based power distribu- tion system? ‘Thomas Edison was a self-educated inventor wito was awarded 1033 patents over «30-year period. He is best known for the invention of the phonograph and the incan- descent lamp, which was invented in 1879 after many years of repeated experiments. In 1878, he formulated the concept of a centrally located power station from which power can be distributed to surrounding areas. On September 4, 1882, using de gener ators (at that time called dynamos) driven by steam engines, he opened Pearl Strect Station in New York City to supply electricity to 59 customers in a one-square-mile area, It was the first de-based power station in the world, with a total power load of only 30 kW, It was the beginning of the electric utility industry, which grew at a remarkable rate. In. 1884, Frank Sprague produced a practical de motor for Fdison’s systems, This invention, coupled with the development of three-wire 220 VDC power, enabled Edison to distribute de electrical power to larger areas and supply heavier loads and consequently more customers. Edison thereby prompted the adoption of Tesla and Eiison worked together fora shor tae, and soon developed hatred for one another, reulting in Tesla opening his own business, believing in uc transmission systems. In 1912, both were nominated lor ‘the Nobel Prize in physics, Because of the feud between them, Tesla declared that he had nothing to do ‘with Edison, and the prize was given toa thied party!4° Chapter 1 Introduction de-based power distribution systems. As transmission distances and load demands in- creased, Edison’s de systems ran into trouble. The de distribution lines suffered very high power losses because of the high. voltage and current that existed simultaneously. This severely limited the transmission distance and resulted in highly inefficient sys- ‘tems. In order to sustain the power level, Edison had to build a de power station every 20 km! This was costly and very impractical. However, he didn’t give up on the de transmission idea and insisted that these problems could be overcome. George Westinghouse and Nikola Tesla didn’t hesitate to develop ae-based power Aistribution systems, despite Edison’s plans to continue to construct de transmission systems in New York. In 1885, Westinghouse took a major step in developing ac sys- tems when he bought the American patents of L. Gaulard and J. D. Gibbs of Paris for ac systems. Westinghouse challenged the de transmission system and went ahead with developing an ac system, ‘A major step in supporting ac systems oceured in 1885, when William Stanley, an early associate of George Westinghouse, developed a commercially practical trans- former, allowing the possible distribution of ac-based electricity. Using transformers, it ‘was possible to transmit high-level voltages with a very low-level current, resulting in a very low voltage drap (low power dissipation) in the transmission line. In the winter of 1886, Stanley installed the first experimental ac distributed system in Great Bur- rington, Massachusetts, supplying power to 150 lamps in the covered area. In 1889, the first single-phase distributed power system was operational in the United States be- tween Oregon City and Portland, covering a 21 km distance with 4 kVA of power. ‘The second major event that boosted the potential of ac systems took place on May 16, 1888, when Tesla presented a paper at the annual meeting of the American Institute of Electrical Engineers, discussing two-phase induction and synchronous motors. Ba: sically, he showed that itis more practical and more efficient to use polyphase systems to distribute power. The firs three-phase ac transmission power system was installed in ‘Germany in 1891; itwas rated at 12 kV and transmitted over a distance of 179 km. Two years later (1893), the first three-phase power transmission system in the United States ‘Was installed in California, rated at 2.3 kV and covering a distance of 12 km. Also in 1893, a two-phase distributed system was demonstrated at the Colombian Exposition in Chicago. The apparent advantages of ac, especially the three-phase systems, over the de system led to the gradual replacement of de by ac systems. Today, the transmission of electricity is done almost entirely by means of ac. However, de transmission of electric [power is used in some locations in Europe and is rarely used in the United States. Since the late nineteenth century, economic studies have shown that ac transmission is much more cost-effective, resuiting in its worldwide acceptance. ‘The History of de and ac Electricity in the Late Twentieth Century Over the last 25 years, the technological advancement by the semiconductor device industry, the revolutionary advances made in the microelectronics field, and the ever- increasing demand for smaller and lighter-weight power systems for space, industrial, and residential applications has led to renewed interest in using de transmission sys- tems, Many experts believe that because of technological advances, it is now possible to develop de transmission electric power systems economically and efficiently. To- day's systems for conversion from ac to de and back to ac can be produced using very fast, high-power, and highly integrated semiconductor devices. What we can achieve using today’s technology was unimaginable only 10 years ago. This is why many power electronies researchers believe that the old debate between de and ac camps is coming back under a new set of technological ruts.1.3. ‘The Need for Power Conversion § In the mid-1890s, ac was declared a winner over de, and in the mid-1990s, the de promoters, mostly power electronics experts, had another shot at the ae camp! History repeats itself! The twenty-first century might very well be friendlier to de transmis- sion system advocates! Who will win the next century is still to be seen! The History of Modern Power Electronics Many agree that the history of power electronics began in 1900, when the glass-bulb rmereury-are rectifiers were introduced, signaling the beginning of the age of vacuum tube electronics, also called glass tube-based industrial electronics. This period con- tinued until 1947, when the germanium transistor was invented at Bell Telephone Laboratory by Bardeen, Bratain, and Shockley, signaling the end of the age of vacuum tubes and the beginning of the age of transistor electronics, During the 1930s and 1940s several new power electronic circuits (then known as industrial electronics) were introduced, including the metal-tank rectifier, the grid-controlled vacuum-tube rectifier, the thyratron motor, and gas‘vapor tube switching devices such as hot-eathode ‘thyratrons, ignatrons, and phanotrons. In the 1940s and early 1950s, solid-state mag- netic amplifiers, using saturable reactors, were introduced. The modem era of power electronics began in 1958, when the General Electric ‘Company introduced a commercial thyristor, two years after it was invented by Bell ‘Telephone Laboratory. Soon all industrial applications that were based on mercury-are rectifiers and power magnetic amplifiers were replaced by silicon-contfolled rect fers (SCRs). In les than 20-years after commercial SCRs were introduced, significant improvements in semiconductor fabrication technology and physical operation were made, and many different types of power semiconductor devices appeared. The growth in power electronics was made possible with the microelectronic revolution of the 1970s, and 1980s, in which the low-power IC control chips provided the brain and the intelligence to control the high-power semiconductor devices. Moreover, the intro- duction of microprocessors made it possible to apply modern control theory to power electronics. In the last 20 years, the growth in power electronics applications has bbeen remarkable because of the introduction of very fast and high-power switching devices, coupled with the utilization of state-of-the-art control algorithms. Today power electronics is a mature technology. The future direction of the era of power electronics is hard to predict, but it is certain that as long as humans seek to improve the quality of life, produce a cleaner environment, and implement energy-saving ‘measures, the demand for clean energy will continue to grow. This implies that power electronics must be used to address the tremendous changes in the way we generate, transmit, and distribute electricity as we cross the bridge into the new cen- tury. For a more detailed discussion of the modem history of power electronics, see the paper by D. Wyke. 13 THE NEED FOR POWER CONVERSION With the invention of a practical transformer by Stanley in 1885 and of polyphase ac systems by Tesla in 1891, the advantages of low-frequency ac over de were compel- ling to power engineers. The basis of utility power system generation, transmission, and distribution since the beginning of this century has been ac at a fixed frequency of cither 50 or 60 Hz. The most outstanding advantage of ac over de is the maintenance of high voltage over long transmission lines and the simplicity of designing distribu- tion networks. However, the nature of the electricity being distributed is totally differ- ent from the nature of the energy required by the electrical load.6 Chapter 1 Introduction ‘At the consumer end many applications may need de or ac power at line, higher, lower, or variable frequencies. Therefore, it is necessary to convert the available ac systems to de with precise control. Furthermore, in some cases the ‘generated power is from de sources such as batteries, fuel cells, or photovoltaic, and in other cases the available power is generated as variable-frequency ac from sources such as wind or gas turbines. The need for this power conversion, ac-to-dc, became more acute with the invention of vacuum tubes, transistors, ICs, and com- puters. Moreover, modern electric conversion goes beyond ac-to-de conversion, as we shall shortly discuss. Tn the late 1880s, power conversion from ac to de was done by using ac mo- tors along with de generators in series (motor-generator set), The motor-generator arrangement was used in de and with 50/60 Hz motors and generators. The difficulties of using the electromechanical conversion system include large weight and size, noisy ‘operation, servicing and maintenance problems, short lifetime, low efficiency, limited range of conversion, and slow recovery time, To avoid the problems of electrome- chanical conversion systems, industrial engineers tumed to linear electronics in the Tate 1960s, where power semiconductor devices are operated in theit linear (active) region, To obtain electrical isolation, input line-frequency transformers were used, resulting in bulky, heavy power converter systems. Furthermore, with power devices operating in the linear region, the overall efficiency of the system is low. Compared with electromechanical systems and linear electronic systems, power electronics has many advantages, including (1) high energy conversion efficiency, (2) highly inte- grated power electronic systems, (3) reduced EMI and electronic pollution, (4) higher reliability, (5) use of environmentally clean voltage sources such as photovoltaic and fuel cells to generate electric power, (6) the integration of electrical and mechanical systems, and (7) maximum adaplability and controllability. In short, all forms of electrical power conversion will be needed as long as con- ‘sumers live in homes and use light, heat, electronic devices, and equipment and inter- face with industry. 14 POWER ELECTRONIC SYSTEMS ‘Most power electronic systems consist of two major modules: (1) The power stage (forward circuit) and (2) the control stage (feedback circuit). The power stage handles the power transfer from the input to the output, and the feedback cireuit controls the amount of power transferred to the output. A generalized block diagram of a power electronic system with n sources and m loads is given in Fig. 1.2 where, ee es Hest) aeons % Pe —e-o{ Conc) fil 4 Ail eee al or Figure 1.2 Simplified block diagram for a power electronic system,1.4 Power Electronic Systems 7 XjpX3, 5% Are input signals (voltage, current, or angular frequency). Yin Jor + «J Are Output signals (voltages, currents, or angular frequency). ‘Pa(t) 8 the total instantaneous input power in watts. ‘Pou(® is the total instantaneous output power in watts, « fiofin +++ sfedte feedback signals: voltages or currents in an electrical system, or ‘angular speed or angular position in a mechanical system. Efficiency, 7, is defined as follows: = Po x 100% Figure 1.3 shows a more detailed description of a block diagram for a power elec- ‘ronie input system with electrical and mechanical output loads. The main function of ‘a power electronic circuit is to process energy from a given source to a required load, In many applications, the conversion process concludes with mechanical motion. Classification of Power Converter Circuits ‘The function of the power converter stage is to perform the actual power conversion and processing of the energy from the input to the output by incorporating & matrix of power ‘witching devices. The control of the output power is caried out through eontrol signals applied to these sWitching devices. Broadly speaking, power conversion refers 10 the power electronic circuit that changes one of the following: voltage form (ac or de), voltage level (magnitude), voltage frequency (line or otherwise), voltage waveshape (sinusoidal or nonsinusoidal, such as square, triangle, or sawtooth), or voltage phase (single- or three- hase). Moy soe Monty coud ponte sotap ig saegalde wseuale tes) luge ets Fam ft presage) PMS RBEAHE A | pomauge ——Ooput ‘technical ™~ Teiont | Powecekcwonic |” Fineng & | _,J esi ren | stn seat siren Meche Sieh com shiver cut Ue Imerscebeween Press eedback enue and power signals nd ‘Smut desi nl Figure 1.3 Detailed block diagram of a power electronic system.8 Chapter 1 Introduction Broadly speaking, there are four conversion circuits that are used in the majority of today’s power electronic cireuits: . acto-ae b. ac-to-de ¢, de-to-ae d. de-to-de In terms of the functional description, modem power electronic systems perform ‘one or more of the following conversion functions: 1. Rectification (ac-to-de) 2, Inversion (de-to-ac) 3. Cycloconversion (ac-to-ac, different frequencies) or ac controllers (ac-to-a¢, same frequency) 4, Conversion (de-to-de) Rectification (ac-to-de) ‘The term rectification refers to the power circuit whose function is to alter the ae chi acteristic ofthe line electric power to produce a “rectified” ac power at the load site that contains the de value. Figure 1.4(a) and (b) shows the block diagram representation of. an ac-to-de converter and its typical input and output waveforms, respectively. To ‘smooth out the output voltage by removing the unwanted ac component, an additional “filtering” circuit is added at the output side. Depending on the switch implementa tions, these converters are further divided into two types, diode converter circuits (uncontrolled ) and thyristor converters (phase-controlted); each type can have either single-phase or three-phase input voltages. Both types are extensively used in various off-line applications. Rectification circuits will be discussed in Chapters 7 and 8. The topologies that perform the rectification function include half-wave, full-wave (full- bridge), semi-bridge, and transformer-coupled center-tapped. From the beginning of the industrial electronics era, the ac-to-de line commutation converter class, utilizing thy- © © Figure 14 ac-de rectification. (a) Simplified block diagram representation. (b) Example of ac-de conversion.14 Power Electronic Systems 9 ristors, has been the most popular among power electronics converters because oftheir sim- plicity in design, efficiency, and higher current and voltage ratings. Inversion (de-to-ac) ‘The term inversion is used in power electronic circuits for the function that alters the de sourée (.g.,a battery) with no ac components into an “inverted” ac power at the load that has no de components, as shown in Fig. 1.5(a). Example input and output wave- forms are shown in Fig. 1.5(). The ac output can have an adjustable magnitude and frequency. Additional filtering is normally used to extract the fundamental component of v,(0 at © = 27/T. Generally speaking, do-to-ac inverters are classified as voltage- fed and current-fed inverter types as discussed in Chapter 9. In the last few years, reso- nant-link technology that has been successful in the design of PWM power supplies ‘has been applied to the design of de-to-ac inverters, producing ac outputs at variable voltages and variable frequencies. A resonant-link de-to-ac inverter is a two-stage con- version circuit that takes the de voltage and changes it to high ac resonant voltage, which in tum is changed to a variable low-frequency output, as shown in Fig, 1.5(c). Generally speaking, since cascaded systems involve a two-stage conversion, power processing passes through more than one switching device and therefore increases con- duction losses. However, in some cases, by using intermediate stages, itis possible to insert an electrical isolation transformer, in other cases, cascaded stages produce high- frequency resonant waveforms that could result in the soft switching of the power de- vices, which in tum could reduce the overall switching losses of the cascaded system. The ac-to-de rectification and de-to-ac inversion represent the broadest functions of power electronic cireuits, Veet @ o o Figure 1.5 de-to-ac inversion. (a) Simplified block diagram representation. (6) Example of de-ac inversion, (c Block diagram representation with high frequency inversion,10 Chapter 1 Introduetion ot @ » ac, de, a, © Figure 1.6 Cycloconversion. (a) One-stage ac-to-ac cycloconyersion. (b) Example of ac-to-ac conversion waveforms. (c) Two-stage ac-to-ac eyeloconversion. Cycloconversion or Voltage Controllers (ac-to-ae) “The term cyeloconversion is used for power electronic circuits that convert the ac input power atone frequency to an ac output power ata different frequency using one-stage con- version, as shown in Fig. 1.6. However, two-stage conversion is also possible, as shown in Fig. 1.6(c),i.,a¢-to-de and then de-to-ac, resulting in wat is called a de-link converter. An ac controller is a power electronic circuit that alters the rms ae input atthe same frequency. Conversion (de-to-de) De-to- Fog and Vy 72 Vay then the instantaneous power, 260) = LoNy Cam be given as follows: os tg StS = tos Vaal Talal = tan fe He 7 cigetan gare 2.4(c) shows a plot ofthe instantaneous power where the maximum power during turn- ox and turn-ofF is Vrlog/4. ©) The total average dissipated power is given by "plo de Yetta det ; ee Ba tate ¢ rf) ME=T,) dt] Evaluation of this integral gives Pare ‘he First expression represents the total switching loss, and the second expression represents ‘Se total conduction loss aver one switching cycle. We notice that as the frequency increases, the erage power increases Tinearly. Also, the power dissipation increases with an increase inthe ‘cused conduction current and the reverse blocking voltage.30 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices (@) The maximum power occurs atthe time when the first derivative of p(0) during switehins is set to 210, i.e, wpe dt ‘Solving this equation for ta,,, we obtain the following values at tum-on and tum-off, respective Repeat Example 2.2 forthe switching waveforms shown in Fig. E2.1. ANSWER © Pam = SCR af) @ Pres ™ Votan Vea f toa 5 i Boia P teuetnd fT Teta 1 Figure E2.1. Waveforms for Exercise 2.1 Find the efficiency of the circui shown in Fig. 24a) with fg = 1005, fg in Fig, 2.1(¢) assuming the switching charactristies for S are = 1500s, 7, = 1M. log = 0, Veg = 0,and D = 1.0 ANSWER 80% 2.3. SWITCHING FUNCTIONS AND MATRIX REPRESENTATION Since switches perform the duties of conversion, rectification, inversion, regulation, ané 0 on, itis possible to use the block diagram of Fig: 2.5 as a useful representation of ‘many power electronic circuits. This system has 1 inputs and m outputs that can be either voltages or currents. There are n x m switches, where each of the m input lines2.3. Switching Functions and Matrix Representation 31 “— A A A. a wt At SF Z u— é ie aod wo 4h A aged een i is Figure 2.5 Switching matrix representation. veld be connected to any of the m outputs, resulting in what is known as a switching searrx, with the control of the switches described by a switching finetion. For illstra- ‘son purposes, Fig. 2.6(a) and (b) shows the switching matrix representation for the single-phase full-bridge and the three-phase full-bridge, respectively. The switching. Sanction is a mathematical model for the switching matrix, describing the operation of ‘Se switches in the matrix. The literature is full of different techniques and technologies Se the generation of switching functions. The switching function approach provides @ compact matrix representation for the power converter and serves as a convenient tool ‘Se modeling all kinds of power conversion circuits. Due to the fact that no resistors are secluded in the structure, there is no power dissipation. In practical systems, additional ommuy storage elements are present to facilitate energy transfer; however, losses are sssociated with these components and the parts of the switching devices. If we assume sécal storage elements and ideal switches, then it is conceivable to achieve 100% effi- Sency in the switch matrix arrangement shown in Fig, 2.5. In order to process power sutiectionally, switches must be able to block voltages of cither polarity and conduct current in either direction. One of the most challenging problems in designing and ana- jpzng the system is the design and implementation of the switching network within the storage elements. Theoretically, since we assume ideal switches and because there are no energy ssorage elements, the instantaneous input power must be equal to the instantaneous sexpat power. Also, there are no restrictions on the form and frequency of the sources.32 Chapter 2 Review of Switching Conceps and Power Semiconductor Devices sith on Sa Sa SPM @ Figure 2.6 Examples of power electronic cit its. (a) Single-phase. However, only one variable at the source terminal can be fixed (either voltage or cu rent) and the corresponding terminal variable is determined by the switching function For example, if 1. Yi. «+ Yq Fepresent fixed voltage sources, their corresponding cur Tents ig fy +++» Ate determined by the switching, as will be illustrated in Example 2.3. Similarly ifthe output variables are represented by fixed output current sources ings ++ dang theit corresponding terminal voltages are also determined by the switchine functions. The reverse is also true for both the input and output terminals. Even thoug® the switching matrix and its function determine the type of power conversion in a gives power-processing circuit, the detailed implementation and terminal characteristics of the source and load sides are also a major part of the power conversion circuit. Normally the ‘energy source is represented by an ideal voltage source that supplies a constant voltage over a wide range of currents. Similarly, the ideal current source can provide a constazt current over a wide range of voltages. We will be using both types of energy sources throughout the book. As for the load side, the power conversion circuit must be designed to provide a stable and fixed ousput that can be represented by either a curres: source or a voltage source. Ifthe output is to be a current source, the load is connected in series with an inductor; fora voltage source output, a capacitor is used. ‘Two important design issues need to be addressed in designing a power electronic switching circuit: (1) the “hardware,” or physical implementation of the semiconductor switching matrix, and (2) the “software,” or logical implementation that guarantees the operation of the switching matrix. The hardware implementation of the switching matrix is restricted by Kirchhof’s voltage law (KVL) and Kirchhoff’s current law (KCL). These circuit laws must be observed at all times. KVL states that the algebraic sum of the volt ‘age drops around a closed loop must be zero, Hence, there shoild not be a switching s= uence that will allow two unequal voltage sources to be connected in parallel or allow = short circuit across a voltage source (see the reference by Philip Krein). For example, avoid establishing a short circuit across v in Fig. .6(@), S, and Sy or S, and S, are no a lowed to close simultaneously. Similarly, S, and S, of Fig. 2.6(6) cannot be closed simults- ‘neously in order that two unequal voltage sources, y, and v,, are not connected in parallel2.3. Switching Functions and Matrix Representation 33, vee | 52 | | Se 4 cnimtne | yt sy | 5 sa ap | mei A at are E A ehh let oF ® Figure 2.6 (continued) Examples of power electronic circuits. (b) Three-phase. Because KCL guarantees that the algebraic sum of the currents entering a node is s=r0, no switching sequence should allow two unequal current sources to be con- ssccted in series. For example, to avoid establishing an open cireuit in series with a cazrent source in Fig. 2.6(6), KCL dictates that neither 5, S;, and S, not S,,S,, and S, se opened simultaneously. For more discussion about switching matrices and theit sssociated switching fimetions, see the reference by Peter Wood. Consider the single-switeh, single-input power-prooessing circuit given in Fig, 2.7. Assume the source voltage, 1(0, i triangular wavefonm with a peak voltage ¥, and frequency f= 1/7, as ssoown in Fig, 2716), Assume the switch is ideal and initially of, and its control works in such @ sexy that it togaes every time v0) crosses zero, Use VY, = 12 V, R= 10 Mand 7 = 1 ms. fa) Sketch the waveforms fori, and v, (>) Calculate the average and rms values forthe output voltage34 Chapter2 _ Review of Switching Concepts and Power Semiconductor Devices aa " cy @ 120 : t ; T © Figure 2.7 (a) Circuit and (b) waveforms for Example 2.3. (©) Caleulate the average input power, average output power, and efficiency. (@) Repeat parts (a)-(c) by assuming T = 1 us. © Repeat parts (a)-(d) by assuming the switch has 1 V voltage drop when closed. SOLUTION (@) The output voltage and the source current waveform are shown in Fig. 2.7(0) (b) ‘The average output voltage is given by23. Switching Functions and Matrix Representation 35) ‘The rmsis given by = ftoa= (Fey, Jape Bere sr,)a) y, = gry (©) The average input power is calculated from Pam Lticomrae=H(4)r7 (Meera feet Morssy,) a) Pee 8 Sune pee and the average outpt power, P, oat = =r i,(tv.s (a= Bar4aw Taco = Pag = 106 (sr en ei par (© Taegan en men Fame bfta= 1) (eras farsa, —1)a) apaee Teta wth igi (ee SrA] (a {etna = Sees, =) je) = See I can be shown thatthe average input power is v,-1) jae Pa Hinde “een 4¥, aa, 1/-4¥, 2 es b1+3y,\1( Ales 37, - er, -1)ar (27S sy, a 1+37,-1)dl] 7 25a) _ ee RR ahr and the average ouput power, Pau = F{il0vsOae = Hide Cgt re) re preg reear,-1) a) “ae — 1p +t) sw resulting inefficiency of 85/2.1 x 100% = 88.2%. aul Py 10036 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices Repeat Example 2.3 for (a) v, = 12sinef and (6) v, a squarewave with peak-to-peak voltage ‘equal to +12 V, ANSWER (a) -382 V, 6 V, 3.6 W, 3.6, 100% same; ~3.32 V, 5.37 ¥, 3.22 W, 2.89 W, 89.7% (b) 6 V, 85 V, 7.2 W, 100%; same; ~5.5 V, 7.8 V, 6.6 W, 6.05 W, 91.7% 2.4 TYPES OF SWITCHES To implement a given switching function, the ideal switches in the switching matrix ‘must be realized by practical power devices. Functionally speaking, any switch must have the ability to conduct current and/or the ability to block voltage by means of con- ‘tol signals. In the on state, the current conduction state is the task under consider- ation, and in the off state, the voltage blocking state is what we are considering. Practical switches have limitations in their conduction current and in their voltage blocking. Since the switch current can flow in the forward, reverse, or both directions and the voltage can be blocked in the forward, reverse, and both directions, there are nine different combinations of current carrying and voltage blocking directions. Four ‘of these combinations are duplicates of another four, ie. forward current carrying and reverse voltage blocking is the same as reverse current carrying and forward voltage blocking. As a result, switches are classified into five general types of restricted switches, as shown in Table 2.1 (for more details see the reference by Philip Krein). 1. Forward current carrying and reverse voltage blocking. This is an uncontrolled device with unidirectional current flow. Uncontrolled device means that turn-on and turn-off are not controlled by an external control signal but rather by the power circuit itself: An example of this type of switch isthe diode since it carries the cur- Table 2.1. Types of Semiconductor Switches, Their Control Features, and Their Possible Switch Implementations Type Current flow Voltageblocking Switch implementation 1 Forward Reverse Diode Z Forward Forward ‘Tramistor 3 Forward Bidirectional scr 4 Bidirectional Forward mie with yb Siode Bidirectional Bidirectional2.5 Available Semiconductor Switching Devices 37 rent only in the forward direction when the anode-cathode voltage is positive, The diode is an uncontrolled device since no external control signal can be applied to initiate carrying the forward current orto initiate blocking the reverse voltage. The controlling voltage is derived from either the source or the load or both, 2. Forward current carrying and forward voltage blocking, This is a controlled de- vice with unidirectional current flow. Such a switch should be able to carry the current in the forward direction and block voltage in the forward direction. Of ‘course, the diode is unable to block a forward voltage across it, Another type of switeh is needed with an external control signal that allows the device to decide ‘Whether the forward current is to flow or not even when a forward voltage is ap- plied. This is the same as implying that the switch performs a current conduction delay function, An example of such a switch is the transistor, which is able to block voltage in the forward direction when the base (gate) current is absent. Forward current carrying and bidirectional voltage blocking. This switch can block current flow in both directions (.c., supports forward and reverse volt- age blocking) but carries the current only in the forward direction, An example is the silicon-controlled rectifier (SCR), to be studied later in this chapter. 4, Bidirectional current carrying and forward voltage blocking. This switch is sii lar to type 2, except the current can flow bidirectionally. Hence, the implementa- tion is 2 transistor with a diode connected as shown in Table 2.1 to allow reverse ccurrent flow. The diode is known asa flyback or body diode since it picks the eur rent in the'teverse direction. This switch can carry the current in the forward di- rection through the transistor and in the reverse ditection through the flyback: diode. The base (gate) signal is used to allow the switch to determine whether to carry the current in the forward direction or to be inthe voltage blocking state; be- cause of the presence ofthe diode, the switch is unable to block reverse voltage, 5. Bidirectional current carrying and bidirectional voltage blocking. This switch is the most general power electronic switch and is similar to type 3 with additional characteristics that allow it to support current flow in both rections. Unlike type 4, where a flyback diode is added to allow reverse cur- rent flow, here the forward and reverse current flow must be controlled. An ‘example of this type is the triee (iriode ac), which is simply two SCRs con- nected in parallel and in opposite directions, as shown in Table 2. 3, 2.5 AVAILABLE SEMICONDUCTOR SWITCHING DEVICES In this section, the emphasis will be on the i-v switching characteristics of devices and their corresponding power ratings and possible applications. Selecting the most appropri ate device for a given application is not an easy task, requiring knowledge about the de~ -vice’s characteristics and unique features, innovation, and engineering design experience. Unlike low-power (signal) devices, power devices are more complicated in structure, diver design, and operational i-v characteristics, This knowledge is very important for ‘enabling power electronics engincers to design circuits that will make these devices elose to ideal. In this section, we will briefly discuss two broad families of power devices: Bipolar and Unipolar Devices 1. Power diodes 2. Bipolar junction transistors 3, Insulated gate bipolar transistors (IGBTS) 4, Metal oxide semiconductor field-effect transistors (MOSFETs)38 Chapter 2. Review of Switching Concepts and Power Semiconductor Devices ‘Thyristor-Based Devices 1. Silicon-controlled rectifiers (SCRs) 2. Gate turn-off (GTO) thyristors 3. Triode ac switches (triacs) 4. Static induction transistors (SITs) and thyristors (SITHs) 8. MOS-controlled thyristors (MCTS) 2.5.1 Bipolar and Unipolar Devices The Power Diode ‘The power diode is @ two-terminal device composed of a pn junction and whose turn- on state cannot be controlled (uncontrolled switch). The diode turn-on and turn-off is decided by the extemal circuitry: A positive voltage imposed across it will tur it on and a negative current through it turns it off. ‘The symbol and the practical and ideal i-v characteristic curves of the power diode are shown in Fig. 2.8(a),(b), and (c), respectively. In the conduction state, the forward voltage drop, Fr, is typically | V ot less. The diode current increases exponentially with the voltage across i; ie, small increase in Vp produces a large increase in Jp (see Prob- lem 2.5) In the reverse-bias region, the device is in the off state and only a reverse satu- ration current, /,, exists in the diode (also known as leakage current). The breakdown voltage, Vg, is the maximum inverse voltage the diode is capable of blocking. Vag is a diode-rated parameter with values up to a few kilovolts, and in normal operation the reverse voltage should not reach Vga. Zener diodes are special diodes in which the break- down voltage is approximately 6-12 V, controlled by the doping process. In power circuits, power diodes have two important features: 1, Power-handling capabilities, including forward current carrying and reverse voltage blocking 2, Reverse recovery time (t,) at turn-off ‘The parameter 1, is very significant because the speed of turning off the diode could be large enough to affect the operation of the circuit. At turn-on, the delay time is nor- mally insignificant compared to the transient time in power electronic circuits. Broadly speaking, two types of power diodes are available: 1. The bipolar diode, which is based on the pn semiconductor junction. Depend- ing on the applications, a bipolar diode can be either the standard line-frequency @ o @ Figure 2.8 Power diode, (a) Circuit symbol. (b) Practical and (c) ideal switching characteristics.25 Available Semiconductor Switching Devices 39 type or the fast-recovery high-frequency type, with its ‘varying between 50 ns and 50 is, Typical voltage drop is 0.7-1.3 V, with reverse voltage block- ing of 3 KV and forward current of 3.5 KA. 2. The Schottky diode, which is based on the metal-semiconductor junction. It has a lower forward voltage drop than the bipolar (about 0.5 V or less). Unlike the bipolar diode, whose current conduction depends on the minority and majority carriers, the Schottky diode current depends mainly on the majority carriers. Finally, the /- characteristics of the Schottky diode are similar to the -v charac- torictiow of the bipolar diodo, Unlike bipolar diodes, Schattky diodes generate fewer excess minority carriers than majority carriers; hence its current is prima- fily generated duc to the drift of majority carriers. Due to its large-leakage cir- cuit, it is normally used in low-voltage, high-current de power supplies To study the reverse recovery characteristics of the diode, we consider the circuit of Fig. 2.9(a), which has a typical diode current waveform during turn-off as shown in. Fig. 2.9(6). For simplicity we assume the switch is ideal. Such a cireuit arrangement is normally eneountered in switch-mode de-to-de converters with the switch replaced by either a BJT or a MOSFET. Initially we assume the diode is conducting with forward current Ip. At = fg, the switch is turned on, forcing the diode to turn off due to the de input voltage Vg. The turn-on characteristics of the diode are simpler to deal with since turn-on only involves charging the diode depletion capacitor. The diode’s for- ‘ward conduction begins when its depletion capacitor has been charged. At turn-on, the diode voltage drop is larger than the normal forward drop during, conduction. This transient voltage exists due to the large value of diode resistance at turn-on. This is why during the diode’s turn-on time the power dissipation is much larger than when it is in the steady conduction state. The diode turn-off characteristics are mote complex since significant stored charges exist in the body of the pm junction and at the junetion. As shown in Fig. 2.9(b), during turn-off, the diode current linearly decreases from its forward value, fy, at = fy {0 zero at ¢ = f, and then continues to go negative until it reaches a negative peak value at 1 = fa, known as the reverse recovery current, fgg at ‘which the current starts to rise exponentially to zero at ¢ = f. The time intervals can be broken down as follows: Between fy and 1, the diode current is positive and the diode forward voltage is small. Hence, we assume the rate of change of diode current is con- stant and is determined by the total circuit inductance in series with the diode. In our ex- ample, since the switch is ideal, dip/dt and Iya are limited by the diode’s and leads parasitic inductances, At r = f,, the current becomes zero and the diode should begin ‘turing off by supporting reverse voltage. But because of the excess minority carriers in «a © Figure 2.9 Typical diode switching characteristics. (a) Switching cireuit with S closed at ¢ = f(b) Diode current.40. Chapter 2 Review of Switching Concepts and Power Semiconductor Devices the pn junction that need to be removed before the diode’s reverse voltage begins to rise, the diode remains in the conduction state for longer time, i.., until ¢ = f ‘The delay from 1, to f is due to the minority carriers in the depletion region, whereas the delay from f, to f is caused by the charge stored in the bulk of the semi- conductor material. Ats =, all charge carriers are removed, causing the device to be fully switched off. The time it takes from the moment the diode current becomes neg- ative until it becomes zero is the reverse recovery time, fas shown in Fig. 2.9(b). Between t and f,, the junction behaves like a capacitor whose voltage goes from zero to the reverse voltage via a charging current in this interval. The total charge carriers that cause a negative diode current flow when itis turned off constitute the reverse recovery charge, Q,, and can be expressed in terms of fea and f, The time between f, and 1, may be very short compared t0 fq, resulting in high di /dt, The ratio between (t5 ~ f) and ty is a parameter that defines what is known as diode snappiness. The smaller this ratio, the quicker the diode recavers its reverse blocking voltage, resulting in what is known as a fast-recovery ot hard-recovery diode. Meanwhile, a diode with a high ratio of (t, ~ f,) to ¢, takes a relatively long time to bring its forward current to zero from its negative peak value, These diodes are known as sofi-recovery diodes. The fast-recovery diodes have high di/dt and normally experi- ‘ence oscillation at turn-off. Standard or general-purpose diodes have soft recovery time ‘and are used in low-speed applications where the frequency is less than a few kHz. In general, an attempt to reduce either ¢, OF Jag Will result in an increase in the other. The forward recovery voltage limits the efficiency because of device stresses and higher switching losses, and i, limits the frequency of operation, Power electron- ies engineers should keep in mind that the transient voltage at turn-on and the tran- sient current at turn-off might affect the external circuitry and cause unwanted stresses. External snubber circuits are added to suppress these transient values. Consider the switching circuit shown in Fig. 2.10 by modeling the circuit parasitic inductance as a lumped discrete value, L,. Assume the switch was open for a long time before being turned ‘on at = fy. Assume the same diode switching characteristics of Fig. 2.9(b), except that its a fast-recovery diode with, ~ f= 0. Derive the expressions for yx and the peak switch current in terms of the diode reverse recovery time. SOLUTION While the diode is in the conduction state, its forward current is Jy. When the switch is ‘closed at ¢ = fy, the diode voltage remains zero and its current is given by and i, for 12 fis given by Vi LO= Tt) Wsise faa iy Figure 2.10 Diode switching circuit with parasitic inductor.25 Available Semiconductor Switching Devices 41 ‘Ae 4 the diode curent becomes zero and j becomes Hence, the interval, ~ i given by ‘Since t5 ~ f = O, then 4 ~ r=, and Ian is given by At this point the diode is tured off and the peek inductor current is higher than the load eur- rent ly, Since the load is highly inductive, its value cannot inerease suddenly by the amount Valy/L, without creating high reverse voltage across the diode. As a result, a snubber circuit must be added across the diode to dissipate excess stored energy in the inductor. ‘Another important point is that when £, becomes very small, a very large reverse recovery current occurs that could damage the diode and cause large switching losses, ‘The peak value of the reverse current, ~Zaa, is @ very important parameter and it can be less than, equal to, or lager than the forward current J, depending on the extemal circuitry connected tothe diode, and the diode parasitic inductance. The fast-recovery diodes have low recovery time, normaly less than 50 ns, and are used in applications such as high-frequency de de converters, where the speed of recovery is critica ‘The Bipolar Junction Transistors (BJT) ‘The schematic symbol and iv characteristics for the bipolar junction transistors (BIT) are shown in Fig. 2.11(@), (6), and (c), respectively. Itis a two-junction, three- terminal device with the minority carriers being the main conducting charges. The switching speed of the BUT is much faster than that of thyristor-type devices. A major drawback is the second breakdown problem? Unlike the SCR, the BJT is turned on by constantly applying a base signal. Power BJTs have two different properties from the low-power BIT and logic transistor: large blocking voltage in the off state and high forward current-carrying capabilities in the on state. BIT power ratings reach up to 1200 V and 500 A. These high rating values suggest that the power BIT’s driving circuits are more complicated. Because the BJT is a current-driven device, the langer the base current, the smaller Pa.si and the deeper the transistor is driven into saturation. In saturation, the collector-emitier voltage is almost constant and the collector current is determined largely by the external circuit to the switch. It is sometimes useful to define what is known as an overdrive factor, which gives a measure to how deep in saturation the transistor is. For example, if the transistor is at the edge of saturation with given base current Ip, then with an overdrive factor of 10 the base current becomes 107, and the transistor becomes deeper in saturation. Normally, te first breakdown vohage refers tothe avalanche breakdown eaused by the increas in the reverse bias voltage, which ean-be nondestructive. The second breakdown voltage isa destructive phenomenon caused by locelized overheating spots in the devi. Bea i defined asthe ratio /c/J, when the transistor is operating inthe saturation mode.42 Chapter2 Review of Switching Concepts and Power Semiconductor Devices ic fe if Onstate vee on siue © o Active region Thereasing ae ceonrent © Figure 211 BIT switching characteristics. (a) npn transistor. (b) deal i-v ‘characteristics. (c) Practical i-v characteristics. Since the base thickness is inversely proportional to the current gain i, Darlington connected BJT pairs have been developed in which the collectors of two devices are joined and the base of the first is connected to the emitter of the second, as shown in Fig. 2.12. This arrangement results in an overall gain that approximately equals the prod- ‘uct of the individual fs of the two transistors, Transistor Q, serves as an auxiliary tran- sistor, which provides the base current necessary to tum on Q3. Because there is @ high ccurrent gain, a smaller base current to Q, is needed to drive the power Darlington pair. Darlington power transistors are widely used in UPSs and various ae and de motor drives up to hundreds of kilowatts and tens of kilohertz. A modem Darlington pair has ratings up t0 1.2 kV with current up to 800 A and operating frequency up to several kilohertz. Triple Darlingtons are also available, in which the current gain becomes propor- ‘ional to the product of the three individual currents gains of the transistors. To tum. off the Darlington switch, all base currents must become zero, resulting in slower switching speed compared to a single transistor. Also, the overall collector-emitter saturation voltage, Fax i8 higher than for a single transistor, as will be illustrated in Exercise 25.25 Available Semiconductor Switching Devices 43 Figure 2.12 Darlingion-connected BIT, ‘There are three regions of operation: saturation, active, and cutoff, As a power switch, the BJT must operate either in the saturation region (on state) or in the cutoff region (off state). The third state is when the transistor is in the linear region and is used as an analog amplifier. To investigate the turn-on and turn-off processes, we consider a simple inverter circuit shown in Fig, 2,13(a) with its switching waveforms as shown in Fig. 2.13(6). ‘The voltage v, is the base driving voltage with positive polarity, ¥), to push positive current into the base, Ig, = (V; ~ Vag)/Rp, and a negative polarity, V, to quickly dis- charge the base current, Jy =—(V2 + Vag)/Rp- At time 1 = iq, % is applied with positive de voltage, +V). Because if takes time to charge the internal depletion capac~ itor to turn the junction on at Ve = 0.7 V, a delay time, ty, elapses before the collec- tor current starts flowing. Afier the junction is tured on, the collector starts flowing exponentially through R» and the emitter-base junction capacitor. During this period, the minority carriers are being stored in the transistor base region. The collector cur~ reaches its maximum saturated value, /.q, determined by Vo Vee mt R ‘The time it takes for the collector current to rise from 10% to 90% of its maxi- mum value, J, is called the rise time. For simplicity, Fig. 2.13(6) shows the rise time from Jc = 0 t0 Ay. The total switching on time is given by ty, = ty + 4,. To tum off the transistor, a negative (or zero) base voltage is normally applied, resulting in a base ‘current J,» being pulled out of the base as shown in Fig. 2.13(b). The collector current does not start decreasing until sometime later after the stored saturation charge in the ‘base has been removed. This time is called the storage time, 1; it is normally longer than the delay time, f, and usually determines the limiting range of the switching speed, If the base voltage is not negative (ie, in the absence of /y.), the entire base ‘current must be removed through the process of recombination To turn on the BJT, a lange current must be pushed to the base. This base current ‘must be large enough to safurate the transistor. In the saturation region, both base- ‘emitter and base-collector junctions are forward biased. This is why the BJT is known as a current-driven device. When it is operated in the saturation region, lp > Ic/B, where B is the de current gain, In this region a new dc current gain B is defined to indicate the depth of the transistor saturation, The saturation collector-emitter voltage is given as Vep.u 8nd Pye is defined as log te Paes =F44 Chapter 2 _ Review of Switching Concepts and Power Semiconductor Devices Re « Figure 2.13 Switching characteristics for the BIT. (a) Circuit. (6) Switching ‘waveforms. where Jc and Jy are the collector and base currents in saturation, respectively, and Beret < B. The smaller Byyas the deeper the transistor is driven into saturation. Typ- ically, Bjacea CaN be as low as 1. Ideally, Vcp.., = 0, but in practice this value varies between 0.1 and 0.6 V, depending on how deep in saturation the device is driven. The new ratio of collector to emitter current is much smaller than the case when the tran- sistor is operated in the active mode, At the edge of saturation, Braes ~ B. ‘The total power dissipation in the transistor is obtained by adding the input power supplied by the collector current and the input power supplied by the base current; hence, the total power dissipation is defined as follows: Pass = Verlot Vaele arcs | Consider the transistor cireuit shown in Fig. E2.4, Assume the transistor is operating in the sat- uration region with Veg gy = 0.5 V, Pap = 0.75 V, Vo = 0.7 V, and D = 0.5, Sketch ip. vor. and i. Determine the overall efficiency ofthe circuit (neglect the power supplied through the base),25 Available Semiconductor Switching Devices 45 iB mh Figure £24 ANSWER 1) = 80.6% ‘To determine the voltage, curent, and power operational limits, normally a plot of i=» characteristics is given, as shown in Fig. 2.14. It gives the region in which the transistor can ‘operate within its limits, the region is known as the safe operation area (SOA), Itrepresents the permissible range of current, voltage, and power of the device in operation. The locus of ‘switch voltage versus switch current during tum-on and turn-off must le within the SOA. a (a) Show thatthe curent gin ofthe triple Daligton transistors shown in Fig. E25 is given by fe. $= 6.B:B: is (b) Assume transistor Q, has collector-emitter saturation voltage Vrs ay Show that Vers ux fOF ‘tansistor Q, is given by Veesan = Veer * Vax Assume identical Pg forthe three transistors. Max power ena) Second breakdown iit Voltage timit Sceimax Figure 2.14 Safe Operation Atea (SOA) for a BIT.46 Chapter2 Review of Switching Concepts and Power Semiconductor Devices Figure E25 Consider the simple BJT switch shown in Fig E26. Determine Brgae for Ry = 1 KO, Ry = WOK Ry = 20 KA. Use Vyg = 0.7 V, Very = 0.3 V and assume an ideal diode. i On cn Figure £2.6 ANSWER 1.25, 125,25 Initially, the BJT was developed to be used in linear audio output amplifiers. ‘Soon BJT devices were used in switch-mode and high-frequency converters for acro- space applications to reduce the size and weight of magnetic components and filter capacitors. In applications where self-turn-off devices are needed, such as de choppers and inverters, BJTs quickly replaced thyristors. ‘The Power MOSFET In this section, an overview of power MOSFET semiconductor switehing devices will be given. A detailed discussion of the physical structure, fabrication, and physical be- havior of the device and its packaging is beyond the scope of this chapter. The empha- sis here will be on the device's regions of operation and its terminal #-v switching, characteristics. Unlike the bipolar junction transistor, the metal oxide semiconductor field-effect transistor (MOSFET) device belongs to the unipolar device family, since it uses only the majority carriers in conduction. The development of metal oxide semiconductor technology for microeleeuonie circults opened the way for the power MOSFET de~ vviee in 1975, Selecting the most appropriate device for a given application is not an ‘easy task, requiring knowledge about the device characteristics, and unique features, as well as innovation and engineering design experience. Unlike low-power (sig- nal) devices, power devices are more complicated in structure, driver design, and ‘operational i-y characteristics. This knowledge is very important in enabling a power electronics engineer to design circuits that will make these devices close to ideal.2.5 Available Semiconductor Switching Devices 47 A 8 source(s) - @ o > D oa a Figure 2.18. MOSFET device symbols. (@)n-channel enhancement-mode (6) channel enhancement-mode. " S — (c) n-channel depletion-mode. © (@ (A prchanneldepletion-mose ‘The device symbols for p- and n-channel enhancement and depletion types are shown in Fig. 2.15. Figure 2.16 shows the i-v characteristics for the n-channel enhancement- type MOSFET. It is the fastest power switching device, with a switching frequency of ‘more than 1 MHz, a voltage power rating up to 600 V, and a current rating as high as 40 A. MOSFET regions of operations will be studied shortly. Teiode Saturation gion (sear region) active esion) Gute (@) KA vos Source (8) «@ © Figure 2.16 (a) n-channel enhancement-mode MOSFET and (6) its i; vs. vps characteristics.48 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices Gate Source Figure 2.17 (a) Vertical cross-sectional view of a power MOSFET. (®) Simplified representation. MOSFET Structure Unlike the lateral-channel MOSET devices used in many IC technologies, in which the gate, source, and drain terminals are located on the same surface of the silicon wafer, power MOSFETs use a vertical channel struc- ture to increase the device's power rating. In the vertical channel structure, the source and drain are on opposite sides of the silicon wafer. Figure 2,17(a) shows a ver- tical cross-sectional view of a power MOSFET. Figure 2.17(b) shows a simplified representation. There are several discrete types of the vertical-structure power MOSFET available commercially today, such as the V-MOSFET, U-MOSFET, D-MOSFET, and $-MOSFET. The pn junction between the p-base region (also referred to as the body or bulk region) and the n-drift region provides the forward voltage blocking capabilities. The source metal contact is connected directly to the p-base region through a break in the n* source region to allow for a fixed potential to the p-base region during normal device operation. When the gate and source terminals are set at the same potential (cs = 0), no channel is, established in the p-base region (i.e., the channel region remains unmodulated). The lower doping in the n-drift region is needed to achieve higher drain voltage blocking capabilities. For the drain-source current, /p, to flow, a conductive path must be established between the n* and n- regions through the p-base diffusion region,25 Available Semiconductor Switching Devices 49 On-State Resistance When the MOSFET is in the on state (triode region), the chan- nel of the device behaves like a constant resistance, Rign, that is linearly propor- tional to the change between vp and jp, as given by the following relation: vps Rosen) = Fi, ‘The total conduetion (on-state) power loss for a given MOSFET with forward current Ip and on-resistance Rpsom is given by Pon iss = IDR ps(on) The value of Rpga) Can be significant and varies between tens of milliohms and a few ohms for low-voltage and high-voltage MOSFET, respectively. The on-state resis~ tance is an important data shect parameter, since it determines the forward voltage drop across the device and its total power losses. Unlike the current-controlled bipolar device, which requires base current to allow the current to flow in the collector, the power MOSFET is a voltage-controlted unipolar device and requires only a small amount of input (gate) current. As a result, it requires less drive power than the BJT. However, itis a nontatching current like that of the BJT; i.e, a gate-source voltage must be maintained. Moreover, since only majority carriers contribute to the current flow, MOSFETs surpass all other devices in switching speed, with speeds exceeding a few megahertz. Comparing the BJT and the MOSFET, the BJT has higher power-handling capabilities and lower switching speed, while the MOSFET device has lower power-handling capabilities and relatively fast switching speed. The MOSFET device has a higher on-state resistance than the bipolar transistor. Another dif- ference is that the BJT parameters are more sensitive to junction temperature compared to the MOSFET parameters. Unlike the BJT, MOSFET devices don’t suffer from sec- ond breakdown voltages, and sharing current in parallel devices is possible. Internal Body Diode The modern power MOSFET has an intemal diode called a body diode connected between the source and the drain, as shown in Fig. 2.18(a). This diode provides a reverse direction for the drain current, allowing a bidirectional switch ‘implementation. Even thongh the MOSFET’s body diode has adequate current and switching speed ratings, in some power electronic applications that require the use of ultra-fast diodes, an external fast-recovery diode is added in an anti-parallel fashion, with the body diode blocked by a slow-tecovery diode, as shown in Fig, 2.18(). Internal Capacitors Another important parameter that affects the MOSFET’s switching behavior is the parasitic capacitances between the device's three terminals, namely, the gateto-souree (C,,), gate-t-drain (C,,), and drain-o-soutce (C.,) capacitances, shown in Fig, 2.19(a). The values ofthese capacitances are nonlinear and a function of the device’s structure, geometry, and bias voltages. During turn-on, capacitors C,, and C,, must be charged through the gate; hence, the design of the gate control circuit must take into con sideration the variation in these capacitances. The largest variation occurs in the gate-to- <érain eapacitance as the drain-to-gate voltage varies. The MOSFET parasitic capacitances are given in terms of the device's data sheet parameters Coos Cys atl Cru 25 follows, Cga = Crs igs = Crue Case Cay = Coss — CrsChapter 2 Review of Switching Concepts and Power Semiconductor Devices batt 2 : —" Boa dade : pe eee ca ; oe) «@ Figure 2.18 (a) MOSFET internal body diode. (6) Implementation of a fast body diode. oo fa [fie ate o Figure 2.19 (a) Equivalent MOSFET representation including {junction capacitances. (b) Representation ofthis physical location, where Coa: = small-signal reverse transfer capacitance ‘Gi = small-signal input capacitance with the drain and source terminals shorted Coes = small-signal output capacitance with the gate and source terminals shorted. ‘The MOSFET capacitances C,., C,y, and C,, are nonlinear and are a function of the de bias voltage. The variations in Cg and C,,, are significant as the drain-to-souree and gate-to-source voltages cross zero, respectively. The objective of the drive circuit is to25 Available Semiconductor Switching Devices $1 charge and discharge the gate-to-source and gate-to-dtain parasitic capacitances to turn the device on and off, respectively In power electronics, the aim is to use power switching devices to operate at higher and higher frequencies. Hence, the size and weight associated with the output transformer, inductors, and filter capacitors will decrease. As a result, MOSFETs are used extensively in power supply designs that require high switching frequencies, including switching and resonant-mode power supplies and brushless de motor drives. Because of the device's large conduction losses, its power rating is Limited to a few kilowatts. Because of its many advantages over BJT devices, modern MOSFET devices have received high market acceptance. Regions of Operation Most MOSFET devices used in power electronics applica- tions are of the n-channel, enhancement type, like that shown in Fig, 2.16(a). For the MOSFET to carry drain current, a channel between the drain and the source must be created. This occurs when the gate-to-source voltage exceeds the device threshold voltage, Vy. For vgs > Vi» the device can be either in the triode region, which is also called “constant resistance” region, or in the saturation region, depending on the value of vy Fora given vgs, with a small vine (ips <_ vgs ~ Vay) the device operates in the triode region (saturation region in the BJT), and with a latge vps (ns > vos ~ Vays the device enters the saturation region (active region in the BIT). For vgs < Vy, the device tums off, with the drain current almost equal to zero. Under both regions of op- eration, the gate current is almost zero. This is why the MOSFET is known as.a voltage- driven device and, therefore, requires a simple gate control circuit The characteristic curves in Fig. 2.16(b) show that there are three distinct regions of operation, labeled as triode region, saturation region, and cutoff region. When used as.a switching device, only the triode and cutoff regions are used; when it is used as an amplifier, the MOSFET must operate in the saturation region, which corresponds to the active region in the BIT. The device operates in the cutoff region (offstate) when vcs < Vy. resulting in no induced channel, In order to operate the MOSFET in either the triode or saturation region, a channel must first be induced. This can be accomplished by applying a gate- to-source voltage that exceeds Vp, ie. oss Once the channel is induced, the MOSFET can operate either in the triode region (when ‘the channel is continuous with no pinch-off, resulting in the drain current being propor- tional to the channel resistance) or in the saturation region (the channel pinches off, re- sulting in constant J,). The gate-to-drain bias voltage (yqp) determines whether the induced channel undergoes pinch-off or not. This is subject to the following restrictions. For the triode mode of operation, we have Yan > Van and for the saturation region of operation, we have ven < Vim Pinch-off occurs when vay ~ Vin In terms of vps, the preceding inequalities may be expressed as follows: 1. For the triode region of operation, vps Vin 3. For the cutoff region of operation Vos < Vin It can be shown that the drain current, ip, can be mathematically approximated as follows: ip = K[2(ves—Vn)¥vs Vis] (triode region) ip = Mvos- Van? (saturation region) 1 kalae, CaF) where by, = electron mobility Co, = oxide capacitance per unit area, L = length of the channel W = width of the channel ‘Typical values for these parameters are given in the PSPICE model discussed later. At the boundary between the saturation (active) and triode regions, we have ¥ps= Yos— Van resulting in the following equation for ip: ip=av} n= kvas ‘The input transfer characteristic curve for ip ¥S. vgs When the device is operating in the saturation region is shown in Fig. 2.20. ‘The large-signal equivalent circuit model for an n-channel enhancement-type MOSFET operating in the saturation mode is shown in Fig. 2.21. The drain current is represented by a current source as a function of Vp, and vgs. If we assume the channel is pinched off, the drain-source current will no longer be constant but rather will depend on the value of vp as shown in Fig, 2.22. The increased ‘o Ym “Gs Figure 2.20 Input transfer characteristics for a MOSFET device operating in the saturation region.25. Available Semiconductor Switching Devices 53 i» fs kvasYa? s Figure 2.21 Large-signal equivalent circuit model value of vps results in reduced channel length, resulting in a phenomenon known as channel length modulation. If the vps~ip lines are extended as shown in Fig. 2.22, they all intercept the vn; axis ata single point labeled -1/A, where A is a positive con- stant MOSFET parameter, The term (1 + Avpg) is added to the jp equation to account for the increase in ip due to the channel length mostulation. jp is thus given by ip = K(vgs~ Vn). + Avps) (saturation region) From the definition of 7, given, itis easy to show that the MOSFET output resis tance can be expressed as follows: 7 Ras Vin) If we assume the MOSFET is operating under small-signal conditions, ie., the variation in vgs on the i, VS. vas characteristic curve is in the neighborhood of the de operating point Oat J, and Vs, as shown in Fig. 2.23, the /p current source can be represented as the product of the slope g., and vgs. as shown in Fig, 2.24. Increasing Figure 2.22. MOSFET characteristic curve including output resistance.54 Chopter 2 Review of Switching Concepts and Power Semiconductor Devices Figure 2.23 Linearized i, vs. vgs curve with operating de point (Q). oo . D s Figure 2.24 Small-signal equivalent circuit including MOSFET output resistance. Input Capacitance Because the MOSFET is « majority-carrier transport device, it is inherently capable of high-frequency operation, Still, the MOSFET has two limitations: 1. High input gate capacitances 2. ‘Transient/delay due to carrier transport through the drift region {As stated earlier, the input capacitance consists of two components: the yate-to-source and gate-to-drain capacitances. The input capacitances can be expressed in terms of the device junction capacitances by applying the Miller theorem to Fig. 2.25(a). Using the Miller theorem, the total input capacitance, C,,, between the gate and source is given by Cin = Cpe t 1 Bho) Cga ‘The frequency response of the MOSFET circuit is limited by the charging and discharging times of C,,, Miller effect is inherent in any feedback transistor circuit with resistive load that exhibits a feedback capacitance from the input to the out- put, The objective is to reduce the feedback gate-to-drain zesistance. The output capacitance between the drain and source, C,,, does not affect the turn-on and turn-off MOSFET switching characteristics. Figure 2.26 shows how Cy, and C,, vary under increased drain-source voltage, Vps. {In power electronics applications, power MOSFETs are operated at high frequen- cies in order to reduce the size of the magnetic components. To reduce the switching2.5 Avsilable Semiconductor Switching Devices 55 (+ EnraCad © Figure 2.25 (a) Small-signal equivalent circuit including parasitic ‘capacitances, (6) Applying the Miller theorem, Capacicance & Cea = ‘Vottage Figure 2.26 Variation of Cy and C,, 38a function of vp. losses, power MOSFETs are maintained in either the on state (conduction state) or off state (Forward blocking state). Safe Operation Area ‘The safe operation area (SOA) of a device provides the cur- rent and voltage limits the device must handle to avoid destructive failure. The typical SOA for a MOSFET deviee is shown in Fig. 2.27. The maximum current limit while the device is on is determined by the maximum power dissipation. Paisson = Losions® oso) As the drain-source voltage starts increasing, the device starts leaving the on state and enters the saturation (linear) region: During the transition time the device exhibits,56 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices ot lo aap rm Second breakdown init SoA, Voltage limit Yosmax ~s Figure 2.27 Safe operation area for a MOSFET. large voltage and current simultaneously. At higher drain-source voltage values that approach the avalanche breakdown, it is observed that a power MOSFET suffers from a second breakdown phenomenon. The second breakdown oceurs when the MOSFET is in the blocking state (off), and a further increase in vg will cause a sudden drop in the blocking voltage. The source of this phenomenon in MOSFETS is the presence of a parasitic n-type bipolar transistor, as shown in Fig. 2.28. The inherent presence of the body diode in the MOSFET structure makes the device attractive to applications in which bidirectional current flow is needed in the power switches. Temperature Effect Today's commercial MOSFET devices have excellent response for high operating temperatures. The effect of temperature is more prominent on the ‘on-state resistance, as shown in Fig. 2.29. As the on-state resistance increases, the conduc- ole Ey ‘Temperature Figure 2.29 The on-state resistance asa function of temperature.25 Available Semiconductor Switching Devices $7 oe 5 ‘Pap tu =I a4 4 ae ea : : i Figure 2.30 (a) IGBT equivalent citeut, (6) simplified equivalent circuit, and (c) symbol, tiom losses also increase. This large Vpyoq limits the use of the MOSFET in high- voltage applications. The use of silicon carbide instead of silicon has reduced Vp manifold. As the device technology keeps improving in terms of switch speeds and power- handling capabilities, it is expected that the MOSFET will continue to replace the BJT. in all types of power electronic systems. ‘The Insulated Gate Bipolar Transistor (IGBT) ‘The detailed equivalent circuit model, the simplified two-transistor circuit model, and the schematic symbol for the insulated gate bipolar transistors (IGBT) are shown in Fig. 2.30(a), (b) and (c), respectively. Its -v characteristic is similar to the MOSFET device and is not shown here. Since the IGBT architecture consists of a MOSFET and a BJT as shown in Fig. 2.30(, itis clear that the IGBT hes the high in- put impedance of the MOSFET along with the high current gain and small on-state conduc- tion voltage of the BIT. The device was commercially introduced in 1983 and combines the advantages of MOSFETs, BJTs, and thyristor devices: the high cur- rent density allowed in BIT devices and the low-power gate drive needed in MOSFET devices. ‘The device is turned off by zero gate voltage, which removes the conducting channel, However, a negative base cannot turn off the pnp transistor current. As a result the turn-off time is higher in the IGBT than in the bipolar transistor. However, like the GTO (to be discussed shortly), the IGBT has a tail current at turn-off due to the recombination of carriers from the base region. At turn-on, & positive gate voltage is applied with respect to the emitter of the npn transistor, creating an n-channel in the MOS device that causes the pnp transistor to start conducting, Tis input capacitance is significantly smaller than that of the MOSFET device, and the device does not exhibit the second-breakdown phenomenon. It is faster than the BJT and can operate up to 20 kHz. in medium-power applications. Currently, itis available at ratings as high as 1.2 kV and 400 A. The improvement in its fabrication is promising, and it is expected that it will replace the BIT in the majority of power electronics applications.58° Chapler 2 Review of Switching Concepts and Power Semiconductor Devices 2.5.2. Thyristor-Based Devices The generic term thyristor refers to the family of power semiconductor devices made of three pm junctions (four layers of pnpn) that can be latched into the on state through an external gate signal that causes a regeneration mechanism in the device. In this section, we will discuss four main members of the thyristor fam- ily that are currently used in power electronic circuits: The silicon-controlled rectifier (SCR), gate turn-off thyristor (GTO), triode ac switch (triac), static in- duction transistor (SIT), static induction thyristor (SITH), and MOS-controlled thyristor (MCT). ‘The Silicon-Controlled Rectifier ‘The silicon-controlled rectifier (SCR) is the oldest power controllable device utilized in power electronic circuits, introduced in 1958. Unlike the diode, the SCR can block voltages bidirectionally and carry current unidirectionally. Until the 1970s, when power transistors were presented, the conventional thyristor had been used exten- sively in various industrial applications. The SCR is a three-terminal device com- posed of a four-semiconductor pa junction, Unlike the diode, the SCR has a third terminal called the “gate” used for control purposes. ‘The symbol and i-v characteristics for the SCR are shown in Fig, 2.31(a) and (b), respectively. The ideal switching characteristic curves are shown in Fig. 2.31(c), where vy and i, are the voltage across the anode-cathode terminals and the current through the anode, respectively. ‘The latching current is always less than the minimum trigger current specified in the device’s data sheet. The holding current is the minimum forward current the SCR can carry in the absence of a gate drive. The forward breakover voltage, Vyo, is the voltage actoss the anode-cathodte terminal that causes the SCR to turn on without the application of a gate current. Reverse avalanche (breakdown) occurs when v,x is neg- atively large. ‘The normal operation of the SCR occurs when its gate is used to control the tum- on process by injecting a gate current ig; to allow the forward current t0 flow; v4y is Positive and can be turned off by applying a negative v.. across it. It must be noted that once the SCR is turned on, the gate signal can be removed, For this reason, this device is also known as a latch device. The gate current must be applied for a very short time and normally can go up to 100 mA. Once the SCR is tured on, it has @ 0.5-2 V forward voltage. ‘The physical structure of the SCR consists of three pm junctions, as shown in Fig. 2.32. The different doping levels shown are used to help sustain a large block voltage and speed the breakdown process. Under no external bias voltage, the ma- jority carriers diffuse across the junctions and recombine with the minority carriers, resulting in zero net current. The Off State Gencrally speaking, thyristor turn-off can be carried out by reversing the anode-cathode voltage (i.e., through natural ac commutation), of it can be turned off through forced commutation by switching a previously negatively charged capaci- tor across the SCR or by the insertion of a series impedance to reduce the forward cur- rent below the device's holding current. In order to turn off the SCR, v 4x must be negative (vx < 0); of course, the trig- sgering gate signal is immaterial (the presence of i, only increases electron movement scross the junction J,). Under this condition, junctions J, and J, are reverse biased and2.5 Available Semiconductor Switching Devices $9 “Anode (A) fal + m Ye Forward blocking reson ibe aso fas bgt fn ig et fe 0 va » in og «formas cae Reverse volage | 7 *ean7ying (on) blocking Forward volage blocking (ff) © Figure 2.31 SCR switching characteristics. (a) Symbol. (b) -v characteristics. (c) Ideal ing characteristics, Anode at a be Cue ot » A 1s ig Cathode Figure 2.32 Simplified physical structure ofthe SCR: no biasing.60 Chapter 2. Review of Switching Concepts and Power Semiconductor Devices Jjumetion J is forward biased. Like two reverse-bias diodes, the only currents that flow ‘through the device are the leakage currents, /, and f,s, of junetions J, and J3, respec tively. Notice that the larger the depletion region driven to the low doping level of 7, ‘compared to pas shown in Fig. 2.33(a), the larger the reverse voltage and the wider the depletion regions of J; and J become. Junction J is forward biased with forward ‘current, Za, equal to (J,) = Zs). The junction J breaks down at lower voltages than J, due to different doping levels, and the large reverse blocking voltage is sustained by junotion J. ‘Typical turn-off times for the SCR range from a few microseconds to 100 ps for low and high voltage ratings, respectively. The On State To tum on the SCR in the conduction state, a positive anode-cathode voltage must be applied (v- > 0) and a gate current must be injected to initiate the on- state regeneration process. For v4 > 0 and ig = 0, junctions J, and J are in the forward-bias states and J, is in the reverse-bias state. Since n, is less doped than the p) region, the depletion region grows mostly into the m, region, as shown in Fig. 2.3305). Without a trigger current (ic = 0), the junction J, is in the forward-blocking con- ition, and the only current that flows in the device is the small leakage current, J», through Jp since itis reverse biased. The forward current remains small until the criti- cal forward breakover voltage, go, is exceeded. At this point, the potential energy of the barrier at junction J, increases, accelerating the electron-hole pair generation until avalanche breakdown occurs, resulting in the thyristor being switched rapidly into the conduction state. This trigger mechanism should be avoided unless specified as safe by the manufacturer, Gate triggering is achieved when a small pulse current, ig, is injected at the gate, introducing an avalanche condition across J, and forward currents, Ic, and Jp, that flow through J; and J, respectively. In order to explain the thyristor gate firing mechanism, normally the SCR is replaced by the model of two intercon- nected complementary pnp and npn transistors Q, and Q>, respectively, as shown in Fig. 2.34. ‘We consider only the forward blocking and forward conducting states, since in the reverse blocking state the only current that flows is the leakage current of Q, and Q,. Depletion region “a a ud h ds Figure 2.33 Depletion layer (@) under reverse bias (74 < Onc (8) under forward2.5 Available Semiconductor Switching Devices 61 » fay a fa o cor A Pe « ’ ie Figure 234 Two-ttansistor analogy Bg Imaal forthe SCR (ep justion Fr representation. (6) Transisior cireuit @ o representation, ‘To derive the terminal current relation for the SCR, we tur to the large-signal model of the transistor, called the Ebers-Moll (EM) model, which is used for both transistor june- tions, the npn and the pnp. Figure 2.35(a) and (6) shows the EM models for the npn and _pnp transistors, respectively. Here ayy* and aj denote the forward a, which is close to unity, and ag, and ay, denote the reverse a which is normally very small (0.02). For simplicity we will assume a, and ap are zero. The currents Ip gc and Ip cy represent the diode leakage currents for the base-collector diode of Q, and the collector-base iode of @, when both transistors are in the active mode. These currents are also known as the collector-to-base saturation current while the emitter is open-circuit, Jcgon, for the ‘npn transistor, and base-to-collector current, Jycoi, forthe pnp transistor. The direction of /egoa and Ico, are opposite to the direction of the leakage currents as shown in Fig. 2.35(a) and (4). Recall that, in the active region, the base-emitter junction is for- ward biased and the base-collector junction is reverse biased. First, let us assume there's no gate triggering (/g — 0). Replacing the equivalent EM models of Fig. 2.35(a) and (6) into Fig. 2.35(c), we obtain the two-transistor equivalent circuit model. With simple algebraic manipulation, we can show that ig =i, = oot tacor RNA T= (a+) With ig =0, the only current that will flow is the leakage current (a and a are small). Normally a + a; <€ 1 to keep it off. Ifa, +a, = 1, the SCR will enter a sustained breakdown, with the anode current limited only by the external circuitry. To avoid entering the breakdown region, a gate signal is injected, resulting in the following forward current: fe * Toxo * Incon 2, = Ae tae ater 2.1 T= (ay +a) a) “Forward a (az) isthe same a the single transistors er when operating in the active region, given by a = B/D), where Bis the transistor eureat gain, B = 1c/Ty62 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices a G y eninge tecoif V | nce <> anines D> entoca=0 binwle feria % a ® ® Figure 2.35 Complete Ebers-Moll model for (a) the npn transistor and (6) the pmp transistor. (c) Equivalent circuit model for Fig. 2.34(b). ig + Tonos + acon a Fa) 2) For the generation process to start, we design the SCR for a + a, = 1 ‘Show that for the generation process to start, the following relation must be satisfied: BB SI2,5 Available Semiconductor Switching Devices 63 When the thyristor was invented, all the schemes for force-commutating mercury- arc rectifiers of the 1930s soon became thyristor-based circuits with expanded applications to include ac drives and UPSs. However, because of their cost and low efficiency, thyristor circuits did not penetrate the adjustable-speed-drive application area. Today's applications range from single phase-controlled rectifier circuits to static var compen- sation in utility systems. Because of its limited frequency of operation, the application of the thyristor has reached saturation. ‘The Gate Turn-off Thyristor (GTO) ‘The schematic symbol and the practical and ideal switching i-v characteristics for the gate tumn-off thyristor (GTO) are shown in Fig. 2.36(a), (b), and (c), respectively. The device is as old as the SCR and was introduced commercially in 1962. Like the SCR, it can be turned on with a positive gate signal, but unlike the SCR, applying a negative gate signal, as shown in Fig, 2.36(6), turns off the GTO. Once the GTO is turned on or Off, the gate signal can be removed. The device has a higher on-state voltage than the SCR at comparable currents. The GTO is normally an aff device; it has a very poor turn-off current gain and it exhibits a second-breakdown problem at turn-off @ ‘Tiszer offby negative gate sgoal o On sate bo Rovese blocking sate | orate YAK © Figure 2.36 GTO switching characteristics. (a) Symbol. (b)i+v characteristics (6) Ideal switching characteristics.64 Chapter 2 _ Review of Switching Concepts and Power Semiconductor Devices Because of its high switching power dissipation, the GTO's frequency of opera- tion is limited to less than 1 kHz, and modern GTO devices are rated at 4.5 kV and at currents as high as 3 kA. The GTO is used in high current and voltage applications, such as voltage-fed inverters and induction heating resonant converters. ‘The Triode AC Switch (Triac) Like the GTO, the triode ac (triac) switching device was introduced immediately after the SCR. In fact, the triac is nothing but a pair of SCRs connected in reverse- parallel on one integrated chip, as shown in Fig. 2.37. It is also known as a bidi- rectional SCR. The triac’s equivalent circuit and the circuit schematic symbol are shown in Fig. 2.37(a) and (b), respectively. The device can be triggered in the positive and negative half-cycle of the ac voltage source by applying a positive or a negative gate signal, respectively. Today’s triac ratings are up to $00 V at 40 A. ‘The é-v characteristics and the ideal switching characteristics are shown in Fig, 237(c) and (d). Use of the triac is considerably limited duc to its low rates of rise of volt- age and current, Applications include light dimming, heating control, and various home appliances. ‘The Diac ally, we should mention another power device known as the diac, which is essen- tially a gateless triae constructed to break down at low forward and reverse voltages. ‘The diac is mainly used as a triggering device for the triac. Static Induction Transistors and Thyristors In 1987 a device known as the static induction transistor (SIT) was introduced. One year later, the static induction thyristor (SITH) was introduced. The symbols for the SIT and SITH are shown in Fig. 2.38(a) and (6), respectively. ‘The SIT is @ high-power and high-frequency device. The device is almost identi- Ito the JFET, but with its special gate construction it has a lower channel resistance compared to the JFET. ‘The SIT and SITH are normally on devices and have no reverse voltage blocking capabilities. The SITH device turns off in the same way as the GTO, by applying a negative gate current, but it has a higher conduction drop than the GTO. Finally, both devices are majority-carrier devices with positive temperature coefficients, allowing. device paralleling. Among the SIT’s major applications are audio and VHF/UHF amplifiers, micro- waves, AM/EM transmitters, induction heating, and high-voltage, low-current power supplies, It has a large forward voltage drop compared to the MOSFET; hence, itis not normally used in power electronic converter applications. The applications of the SITH include static var compensators and induction heating. The MOS-Controlled Thyristor = The simplified equivalent circuit model and the schematic symbol fora p-type MOS- controlled thyristor (MCT) are shown in Fig. 2.39(a) and (6), respectively. Its ideal i-v switching characteristic is similar to that of the GTO, as shown in Fig, 2.39(c). The device was commercially introduced in 1988. Like the GTO device, it has a high tum- off current gain,2.5 Avtlable Semiconductor Switching Devices 65 Aanose) 4 My eR, sc, 2 = eux) is $m (cathode) @ © onsite Revere caren ‘eying re 2.37 Triae switching characteristics, (a) Equivalent representation wsing two SCRs. (6) Symbol. (c) iv characteristics. (@) Ideal switching characteristics, ‘The p-MCT is turned on by applying a negative gate voltage (less than -5 V) with respect to the cathode, turning on the p-FET and turning off the n-FET, initiating the regenerative mechanism in the SCR connected npn and pnp transistors. Similarly, applying a positive gate signal with respect to the cathode initiates the turn-off. The a-MCT has the same device structure, except that the p-FET and n-FET are66 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices Diy Ano cq oe G (gate) ; Kcaade) Stone) Figure 238 (a) SIT symbol. () ITH @ © symbol. Alar) 4 tase) Fe Ere * i 29 igo —| 6 oe rer ah ° z -K (cathode) fh @ ® Figure 2.39 MCT switching characteristics (a) Equivalent circuit (®)p-MCT symbol. e) Ideal switching characteristics. (f)n-MCT symbol. © interchanged; hence, a positive and a negative gate signal tums the m“MCT on and off, respectively. The schematic symbol for the n-MCT is shown in Fig, 2.3%(d). ‘The MCT’s current and voltage ratings exeeed 1 kV and 100 A and are continu- ously being improved. The device can be easily connected in series and in parallel ‘combinations to boost power rating.2.7 Future Trends in Power Devices 67 This device is serious competition for the IGBT. It has the same frequency of ‘operation as the IGBT but with a smaller voltage drop and a higher operating temper- ature, Intensive efforts are under way to introduce a new improvement in the device, ‘and itis expected to reveive wider acceptance in medium: and high-power applivations, Other Power Devices Other devices of the thyristor family include the reverse-conducting thyristor (RCT), ‘whieh is nothing but a built-in anti-parallel body diode connected across the SCR to allow current to flow in the opposite direction, and the light-activated SCR (LASCR), ‘which is used in high-voltage and high-current applications such as HVDC systems. Their power ratings go up to hundreds of kilovolts and hundreds of kiloamperes, and they provide complete electrical isolation between the power and control circuits 25 COMPARISON OF POWER DEVICES Depending on the applications, the power range processed in power electronics is very ‘wide—from hundreds of milliwatts o hundreds of megawatts, Therefore, itis very diffi- cult to find a single switching device type to cover all power electronics applications. ‘Today's available power devices have tremendous power and frequency rating ranges as well as diversity. Their forward current ratings range from a few amperes to few kilo- amperes, theit blocking voltage rating ranges from a few volts to a few kilovolts, and their switching frequency ranges from a few hundred hertz to a few megahertz, as illus- trated in Table 2.2. This table gives only relative comparison between available power semiconductor devices because there is no straightforward technique that gives a rank- ing for these devices. Devices are still being developed very rapidly with higher current, voltage, and switching frequency ratings. Figure 2.40 shows a plot of frequency versus, power, illustrating these rating ranges for various available power devices, 27 FUTURE TRENDS IN POWER DEVICES It is expected that improvements in power-handling capabilities and increases in the frequency of operation of power devices will continue to drive the research and devel- opments in semiconductor technology. From power MOSFETs to power MOS-IGBTS ‘Table 22 Comparison of Power Semiconductor Deviees Deview ‘Yearmade Rated Rated Rated Rated Forward ‘ype available vole current’ frequency _—power_—_voltage ‘Thyristor 1987 SRV 3SKASO0HZ —100sMW «15-28 (scr) Trine 1958 kv 100A -S00Hz —100sKW 15-2 GTO 1962 4SKV SKA 2kHz lsMW 3-4 Bur 1960; L.2KV «800A LOK IMW 153V (Dartington) MOSFET 1976 S00V. SOA MHz 1ookW 34 IGpt 1983 12kV 400A. 20KkHz 108k 3-4 SIT 1987 4KV 600A LOKHz_10skW 10-20. sITH 1975 4KV 600A «10K 1skW 0 2-4V. MCT 1988 3RV2kV 20-100 0sMW 1-2.68 Chapter? Review of Switching Concepts and Power Semiconductor Devices 108 wer g omy 5 0? 2 er? ae Avtoqeons 10? ‘increases, power aoe 10! 10° 10° 10' 10? 10310 10810807108 f Prey Figure 2.40 Frequency versus power rating ranges for various power devices. to power MOS-controlled thyristors, the power rating has consistently increased by a factor of 5 from one type to another. Major research activities will focus on obtaining new device structures based on the MOS-BIT technology integration to rapidly increase power ratings. It is expected that the power MOS-BIT technology will capture more than 90% of the total power transistor market. The continuing development of power semiconductor technology has resulted in power systems with driver circuits, logic and control, device protection, and switching devices designed and fabricated on a single chip. Such power IC modules are called “smart power” devices. For example, some of today’s power supplies are available as ICs for use in low-power applications. There is no doubt that the development of smart power devices will continue in the near future, addressing more power electronics applications. 2.8 SNUBBER CIRCUITS To relieve switches from oversiress during switching, switching aid circuits, known as snubber circuits, are normally added to the power switching device. The objectives of smubber circuits may be summarized as (1) reducing the switching power losses in the main power device in the power electronic circuit, (2) avoiding second breakdowns, and (3) controlling the device's dv'/dt or di/ dl in order to avoid latching in pnpn de- vices. There are a wide range of turn-on and turn-off snubber circuits available in today’s power electronic circuits. These include dissipative and nondissipative pas- sive snubber circuits, and nondissipative active snubbet circuits. In dissipative snubber circuits a capacitor is used to slow the device’s voltage rise during turn-off, or an inductor to slow the device’s current rise during turn-on. Figure 2.41(a) and (5) shows popular turn-off and turn-on snubber circuits, respectively. In Fig. 2.41(a), a capacitor is used to reduce the voltage rise dy,,/dt across the switch during turn-off In Fig, 2.41(b),a snubber inductor, L, is used to slow down the rise of the inductor currentProblems 69 Ber @ Figure 2.41 Passive snubber circuits: (a) turn-off and (b) turn-on suber circuits ‘With mon snd uot subbers Figure 2.42 ig, Versus vi, switching loci. diz,/dt (the inductor current equals the switch current, iz). Figure 2.42 shows the switching loci for a practical switch (transistor) with and without snubber circuits. For a detailed discussion on all types of snubber circuits and their design methods, refer to the references at the end of the textbook. PROBLEMS ‘aie! Switch Characteristics (©) Determine the average output voltage. Consider the switching circuit shown in (€) Determine the average output power delivered Se P21 with a resistive load. Assume the 10 the load. sonich is ideal and operating at « duty ratio cf (@) Determine the average output power supplied by the de source ich the waveforms fr iy 8d Ye (@) Determine the efficiency ofthe circuit70 Chapier 2 Review of Switching Concepts and Power Semiconductor Devices a Pat a le on [on 100 %S5a 0 cal Figure P2.1 igure P2.3 22 Consider the two-switch circuit given in Fig, P2.2. Assume ideal §, and S with the shown switehing sequence, Sketch lyin Yowis yar ADE ‘ya: Determine (a) the average output voltage, (b) the average power delivered to the load, and (c) the efficiency. 23 The circuit shown in Fig. P23 is known as @ ccurrent-driven full-bridge inverter. Assume 5), 5. Sp and S; are ideal, with their switching wave- forms as shown, Sketch the waveforms £0F fy, Vy Yuts fons You 40d Fog and derive the expression for the average output voltage. Nonideal Switching Characteristics 14 Use Fig. £21 and by assuming that the switching waveform has the following parameters, 50 V, Jog = 10 JA, Vag =2 Vs, 200s, fe = 10005 determine the equation for the instantaneous power and find Py. 25 Repeat Problem 2.1 by assuming the switch has 2 0.2 0 on-state resistance and a 2 V forward Arop during conduction. 2.6 Repeat Problem 2.3 by assuming each switch has 0.2 9 on-state resistance and a 2 V forward voltage drop during conduction 2.7 Consider the circuit of Fig. P2.3 by assum- ing each switch has « forward voltage drop, Fs and an on-state resistance 7yy, Derive the ex- pression for the circuit's efficiency in terms of Tog, R, 8, Vn, and ray. Determine 8 for maximum efficiency. 28 Determine the conduction and switching aver- age power dissipation for Example 2.2 by using 1S, far = 8 ns, F, = 150 ps, Vag = 150, Fog = 0.7 V, ad log = 15 A.Figure P2.10 |» Repeat Problem 2.8 for Fig. E2.1 with y = £28 Consider @ power switching device whose “si and voltage waveforms are as shown in <= P20 Pe | sewer and sketch it, fe Determine the average power di se Calculate part (6) for 4, = 1200ns, = 180 ns, ns, f =100 kHz, Voq = 150 V, and _=15A se Repeat parts (a)-(c) by assuming Veg =2 V sat ag = 10 MA. 228 Use the linear approximated switching cur- == snd voltage given in Fig. 2: ~c voltage and zero leakage (reverse) current. st from 10% t0 90% of Ig, and fy Is 10% to eof Vag 3 The switching cuent nd voltage wave- ae of Fig. 23 ot errata ati cally flows: lye fl) Ya Ngee Sse 7 represents the time constant, which is a sxcson of the on-state resistance of the device cad &s capacitance. Assume negligible and Voy sow that the switching power dissipation at von is given by Felten Pan Be cane fgg 2 Tad yg 7 255 Assume the switching current and voltage of 1 at the turn-off interval (0 <1 $ 1, + fg) Problems 71 are represented as follows: Yaa) = Vor fg = Toye! Derive the expression for the average switching power dissipation during the off time. Assume thn st 2.14 The diode i-v characteristic curve in the for- ward region can be mathematically represented by ipelew"rr reverse saturation current thermal voltage, equal to 25 mV at 20°C ‘n= empirical constant whose value de- pends on the semiconductor material and the physical construction of the device (normally between 1 and 2) (@) Show that a decade change in the forward di- ‘ode current results in a 2.3n¥ change in the for- ‘ward voltage, mathematically expressed as follows: VV = 2300 ig (b) Consider the circuit of Fig, P2.14, Assume the diode has Ip = 5 A at Vp = 1Vwithn = 1.5. De- temine and vp for()R = 10.0 and i) R = 5.0. io, Figure P2472 Chapter 2 General Device Siitching Problems 218 Consider Fig. P2.15, which shows an ap- proximated reverse recovery turn-off characteris- tie for @ power diode. Show that the following relation can express the total reverse recovery chante, One 1, Mat, , aa glotitge = 3alog where di /di and dis /dt are the slopes of the code current during fan, respectively. Qn » 4h (fa teat ole Figure P2.15 2.16 Consider the switching circuit with the series snubber R,-C, network connected across the diode as shown in Fig. P2.16. Assume the diode switch- ing characteristic curve is given by Fig. P2,16(4) and the switch is ideal @ Figure P2.16 Review of Switching Concepts and Power Semiconductor Devices (a) Derive the expression for all the branch cur- rents and vp, and sketch them (b) Derive the expression for the eapacitor C, that is needed. 2.17 One way to speed the tum-off time of the Darlington connection is to include a diode be- twoen the bases of the two transistors as shown in Fig. P2.17. Discuss how the turn-off speed is im- proved by adding the diode D. Sa py “LT Figure P2. 2.18 Consider the transistor switching circuit and its switching waveforms in Fig. P2.18(a) and (b), respeetively. (@) Sketch the waveforms for ip and vp. (b)- Calculate the average power dissipated in the transistor, Use f, 10 KHz, ty @ Figure P2.18(a)Sf Problems 73 Figure P2.18(5) 2.29 Figure P2.19 shows @ BIT switching circuit wen as the Baker clamp, whose objective is to fimat how deep in saturation the transistor is pennit- sett go. Show that if we assume yp = Vaz, then the caninum Veg is clamped to Vz Ifthe collector is ammnected toa 15 V power supply through a collector sssssor, Rc, and the base is connected toa 5 V power supply through a base resistor, Ry, design for Vi, Rey V2 ~ Figure P2.19 and R so that the transistor is diven into saturation with Fora: = 18 and Berns = 10. 2.20 Derive the SCR forward cument relation ‘given in Eqs, 2.1) and (2.2) 221 Consider the half-bridge inverter circuits With the unidirectional MOSFET switching a re- sistive load given in Fig, P2.21. Use & = 10.0, Sy on | om | on ov (7 55 2 - a on | on | of «ov ost! i Vierameer oy 60a, o sr T r -o0v 60 Figure P2.2174 Chapter 2 Review of Switching Concepts and Power Semiconductor Devices (a) Determine the average power delivered 1 the 2.23 Consider the Darlington transistor pair give= load when S, and S; ae assumed ideal in Fig, P2.23. Assume transistor 0, is driven int () Repost part (a) for Rosa) = 0.2 0 for cach srt with Very =03 Vand By gas = 0 MOSFET. Calculate But 86 Versa Use ary = Va (©) Find the effcioney ofthe cireuit in pat (dy, OT Veand 7y = 5. 2.22 Calculate nw in Fig. P222(a) and (b) for Ry = Vidand Ry = 1OKO) Assume Vy, = 0.75 V a Verge = OV. wv Figure P22Chapter 3 Switching Circuits, Power Computations, and Component Concepts 3 32 33 34 INTRODUCTION SWITCHING DIODE CIRCUITS 3.1.1 Switching Diode Citeuits under de Excitation 3.1.2. Switching Diode Circuits with an ae Source ‘SWITCHING SCR CIRCUITS BASIC POWER AND HARMONIC CONCEPTS 33.1 Average, Reactive, and Apparent Powers 33.2 Sinusoidal Waveforms 3.32.1 Instantaneous and Average Powers 33.22 Power Factor 33.3 Nonsinusoidal Waveforms 3.33.1 Fourier Analysis 3.33.2 Line Current Harmonies 33.33 ‘Total Harmonie Distortion 3.334 Power Factor CAPACITOR AND INDUCTOR RESPONSES 34.1 Cepacitor Transient Response 342 Copacitor Steady-State Response 3.43 Inductor Transient Response 3.44 Inductor Steady-State Response PROBLEMS ENTRODUCTION In this chapter, several general concepts will be discussed to provide the necessary back- ground material in power electronics. The review material will cover switching diode 816 BA 3d Chapter 3 Switching Circuits, Power Computations, and Component Concepts circuits, power computation, harmonic analysis, and component concepts. Magnetic cir- cuits, and inductor and transformer components will be covered in Appendix A. SWITCHING DIODE CIRCUITS Switching power electronic circuits process power more efficiently than linear power elec- tronic cireuits. This is why power electronic design engineers must understand the analysis ‘of power switching circuits. In this section, we will discuss the analysis of switching cir~ ‘cuits that include diodes, SCRs, and ideal switches under both de and ac excitation. Diode circuits are nonlinear and their analysis is normally not as straightforward as the analysis of linear circuits. By adding switches to diode circuits, additional non- linearity is introduced, making the analysis even more complex. These circuits are en- countered frequently in power electronics, such as in diode and SCR rectifier circuits, and in pulse-width-modulation (PWM) and resonant converters. To simplify the anal- ‘yses, we will mostly assume that the diodes, SCRs, and switches are ideal. ‘Switching Diode Circuits under de Excitation We begin by analyzing diode resonant networks under de excitation, which are often en- countered in resonant-type de-de and de-ae power electronic circuits. It will be shown later that resonance is used in power electronics to achieve several important functions, including filtering and soft switching. Let us first consider a series RLC circuit as shown in Fig, 3.1. Assume the switch is closed at ¢ = 0 and v,(0-) and i, (0) are the capaci- tor and inductor initial values, respectively, just before the switch is closed at t = “The analysis ofthis circuit is straightforward and is given here as a review. Using KVL around the loop yields the following integral-differential equation in terms of the resonant current i,() oe + Rit of i, dt = Veg de By taking the first derivative of this equation, we obtain a second-order differential equation: Pi, Rdi, 1, Fe Res ay =0 6) ‘Since the excitation is a de source, there exists only a transient (natural) solution,’ ‘whose roots are obtained from the characteristic equation given by R srbsede 32a) Ge ate et 120 c ve ze Fig 2 Ser nit REC iit ith de source. "Also known asthe homogeneous soluion3.1 Switehing Diode Circuits 77 ‘The roots of Eq. (3.2a) are given by (2 oa Depending on the different values of these roots, there exist three different well- understood general solutions for the transient responses. Case 1 (equal real roots). When the roots are teal and equal, we obtain what is called acritically damped circuit, which occurs when the following eireuit condition is met: a= ‘The roots of Eq. (3.2) then become Under this condition, the general solution for the inductor current is given by i) = (Ay + Age 2 63) and the capacitor voltage is given by i= Yar ttt * 3 G4) = ~Ry)-4, 8 Jorn resold) 4B ‘The constants 4, and 4, are obtained from the given initial conditions. Case IT (unequal real roots). Under this condition, the circuit is known as over damped and oceuts when RY 1 (%) > 70 ees aces aes ,(t) = Aye + Aye“ G5) and the capacitor voltage is given by v(t) = Vgg— (Ls, - R)Aye" ~(Lsy~ R)Aze™ G6) where, (37a) 7) Case II! (complex pair of roots). Under this case, the circuit is known as under- damped and ccc when RY 1 (i) fae R eae ri = ft = Ire where @ is known as the resonant frequency, ais called the damping factor, and «, is known as the damped resonant frequency. The ratio a/ is defined as the damp- ing ratio, 6: teal fut @ 2JE7C Itean be shown that the constants 4, and 4; fora given initial capacitor voltage v,(0}, and an initial inductor current, é,(0), for the series resonant REC circuit of Fig. 3.1 can be found using Table 3.1. For the critically damped case, we have 6 = 1, and fora purely eapacitive-induc- tive circuit (R =0), we have 5 = 0. In the latter case, the response is purely oscilla- (ory. Such a response is encountered frequently in de-de soft-switching power electronic circuits. Another parameter that is normally given in the RLC circuit is the quality factor, Op, which is defined as ‘Table 3.1 Three Possible Cases and Their Response Constants for the Series Resonant RLC Circuit Circuit type As Ar Case I: ritcally damped io) ideo ee Eo) Case I: overdamped Paar) , sif(0) Va MOY sain) Ts, - 52) 4-82 Us~ 82) 1-2 (Case Ik: underdamped i) Yu-"40°)_ R106) ot fae = [email protected]= i Ge Ge3.1 Switehing Diode Circuits 79 G9) The higher Qo, the more oscillatory the current response becomes. Another parameter of particular interest in power electronics is what is commonly referred to as the circuit characteristic impedance, Zy, defined as a= fe G.19) TER in the REC circuit represents the load, then Qp is known as the normalized load and is given by 0-2 Gitta) To give the same measure of oscillation in the series case for parallel resonant REC circuits, the normalized load is defined as, R 2 @.11) Qo a ‘The higher Q,, the more oscillatory the voltage response becomes. Normalized loads will be studied in de-de resonant converters in Chapter 6. Notice that Q, is defined the same as the quality factor of the resonant circuit. However, one should distinguish between the Qj of the resonant circuit in which R represents the losses in the resonant circuit that can be ignored, and the case when R represents the load, where Qy becomes a normalized load, i.e., a design parameter, Consider the circuit of Fig 3.2(a) with = 200 02, L=2 mH, C= 0.01 yaF and V4, = 20 V. Derive the expressions for 1, (2) and v,(2) for £ > 0 after the switch is closed. Assume the inital inductor current and the initial capacitor voltage are ze, i, i(0°) = Oand v(0-) = 0. SOLUTION Here we have R/b= 100X109 rad/s, and 1/LC= 50% 10% rad?is?. Since (R/2L) < 1/LG, the circuit is underdamped; hence the roots ofthe characteristic equation are 5; = (50x 109/218 x 103) mais, and s, = (50 x 105—j218 x 10°) rads ‘The general solution forthe inductor current is given by g(t) = €-50'(4 cos 218 x 1034) + Agsin(218 x 103) A 6.12) The eonstants 4, and A, can be obtained from Table 3.1. However, fr illustration purposes ‘we will show how to find 4, and 4; Since i, (0°) = i,(0-) = 0, then A, = condition. Applying KVL to the circuit, we obtain To solve for 42, we use the capacitor inital diy Vag t LE + ve RI =O80 Chapter 3 Switching Circuits, Power Computations, and Component Concepts Caps 4p rus) Figure3.2 (a) Circuit for *#8) Example 3.1 and (b) its ccurrent and voltage » ‘waveforms. 0°, we obtain the first derivative at ¢ = 0°: iO") _ Fae ¥0" Evaluating this equation at LO) Ya.) 19 ang soe ek ee een BO seman ited eta He Solving for 4 , we obtain 4, = 0.046. ‘The inductor current and capacitor voltage expressions for # > 0 are given by E(t) = 0.046e-500"' sin (218 x 1032) A i Ryde 1 Ri LG ve) ‘Substituting for 1,(0) the capacitor voltage is given by ¥,(2) = 20-6 $*101(4.6sin(218 x 10%r) ~20.056c0s(218 x 10°1)) V ‘The diode switches off when i,(0)=0, which oceurs when 218% 10% =a, or 1 = n/(218% 10") = 144 yas, at which the capacitor voltage equals 29.76 V. The plots for and v, ae given in Fig. 3.20).3.1. Switching Diode Circuits 81 EXAMPLES? Repeat Example 3.1 by replacing the diode across 2 as shown in Fig, 3.3(a), Again, assume {,(0") = 0 and »,(0") = 0. SOLUTION At ¢= 0° the diode begins conducting, because the inductor current just after # > 0 is, positive (assuming an ideal diode). The equivalent circuit for > Ois shown in Fg, 3.3(6). This circuit has a damping ratio of zero, resulting in a purely sinusoidal response with () given by i,(0)= 4 coseyt + Aysinoyst where oy = 2.23 x 105 rads, Since the intial inductor current is zero, the constant A, = 0 and, from Table 3.1, 4; =45 x 10 = 0,045; then the inductor current and capacitor voltage become (0) = A5sinaayt mA v, (0) = 20(1 ~ cose) V ‘The inductor and capacitor waveforms are shown in Fig. 33(0). ‘At 1 = 17/4 the diode becomes reverse biased since i, starts becoming negative. At this ‘time the equivalent circuit model changes to a series REC eircuit with a new capacitor initial tgp hietaet ie ee cy yen 10 c ¢ ae nS yo % nS Yo @ w Me ZOD @ Figure 333 (a) RC diode circuit for Example 3.2. (6) Equivalent cireuit for 0 <1 n/a (c) The inductor and capacitor waveforms. (4) Equivalent circuit for F2 7a.82 Chapter 3 Switching Cireuits, Power Computations, and Component Concepts value, v, (7/4) = 40 V, and i,(7/4) = 0. Figure 3.3(d) shows the equivalent circuit = £> n/a. The solution for i, and v. is similar to that in Example 3.1 except that when i, be ‘comes zero again, the diode will conduct and the equivalent cizcuit mode of Fig. 3.3(b) aga= becomes valid. This process continues until the diode current decreases to zero, (REROIBED Tepet Example wi he doe don vend. Stee he cpa volgen in framers te 0se 1/a ANSWER — For 051 < /ep iy(¢) = 0.046€°°*""'sin(2.18 x 10°) A v(t) = 20~ (4.68in2.18 x 10°r+ 20.1 sin2.18 x 10°r)e Solve for i,(#) and v,(t) after the switch is closed at = 0 in Fig, E32. This cireuitis fre quently encountered in de-de resonant converters. Assume the initial values are (0°) = 0 an 1, (0°) = L.5f,. (The circuit that established this initial inductor current isnot shown.) Figure E3.2 Circuit for Exercise 3.2. ANSWER i,(t) = 1.5I,c0s04¢ A, v(t) = 15Leysinugt V, 04 = YI7EC for ayt St 1,0) =-1.SIy ¥.(0) = 0 for ant > il conditions are v,(0") = 0 V and #,(0-) = 0.51, A ANSWER 4,(0) = 051,cosayt A,v.(0) = 0.5I,Leysinayt V, 0 = Y17EE for ayt 05 ly.¥,(0) =0 fot yt > a Give switch implementations (unidirectional or bidirectional) for Exercises 3.2 and 3.1.2. Switching Diode Circuits with an ac Source Large classes of switching converter circuits use an ac source excitation rather than de source. The analysis of diode switching circuits with ac sources is carried out in ‘two steps: first obtain the transient response (also known as the natural response) by setting the ac source to zero, and then obtain the steady-state response (also known as3.1 Switching Diode Circuits 83 cee chi acne 4 2 Quer i a @ Iipere BLC chal ac xt (te deni clad Saunas muon the forced response) by converting the circuit to the phasor domain. The final solution is the sum of the natural and the forced responses. Figure 3.4(a) shows an REC circuit with an ac source, v(t) = V, sin ax V. Assume the switch is turned on at = 0, and we wish to solve for the inductor current. For > 0, KVL yields the following differential-integral equation in terms off, (¢): Vy sinaw + is + al. i(0 dt + Ri() =0 Talking the first derivative of this equation, we obtain @i, edi, 1 Vo SER + Hic) = tease G.13) ‘The right-hand side of Eq. (3.13) represents the forced excitation. First we obtain the natural-response component of the complete solution by setting the forced function (source function v, and its derivatives) to zero, to yield i de ‘The matural response of i, is the same as the response for the de-source RLC circuit shown in Table 3.1 for the overdamped case; ic, the transient response, ng i given by Fiat) = Aye + Age G15) The steady-state response component is obtained by transferring the circuit to the phasor domain as shown in Fig. 3.4(6). The steady-state response, [is easily obtained: v LO ViL-6 pain al 6.19 Root i In the time domain, i is given by diy te Vy 14) ae tnt) =0 G.14) where ‘iO = pp sin(on- 0) G17)84 Chapter3 Switching Cireuits, Power Computations, and Component Concepts ‘The total response is obtained by adding Eqs. (3.15) and (3.17), to yield 5,2) = Us ant) * ir gl) G.18) i: psin(or- 8) ‘The constants 4, and 4, are obtained from thei and capacitor voltage at ¢ = 0°. DOMPESS Consider the ac diode cireuit shown in Fig. 3.5(a) with a de source in the load side representing cither a charged battery or a back electromotive force (enxf) to excite the armature circuit in a de motor. Sketch the waveforms for i,, vp, and »,, Assume v, = 100sin3771, What isthe average value of v,? = Apel + Ane! ial condition of the induetor current SOLUTION The diode will tum on when vp > 0, which occurs at = 1, when v(t) = 60 V, 0 100sin3771, = 60 V resulting in) = 1.7 ms. For 0-<1<, the diode is off, and for the interval 1, <1 < 1, the diode is on, where f, = 7/21) = 6.63 ms. During f, $$ tthe output voltage, v,, equals v, and the output current is given by = Va oi = 12Ssin3771- 7.5 & For all other times, namely 0St < f, and f$1< Tj, = andy, = Vy, = 60V. ‘The average ovtput voltage is obtained from the following equation: 100 Fame =4[ [2 ¥ sinoe d+ Vq072+21)] cate eS Te d+ .- +34 ms) | = 67. reahaal | oss: access m+ me] 6778 V ‘The voltage and current waveforms are shown in Fig, 3.5(6). Repeat Example 3.3 by including a freewheeling diode across the load as shown in Fig. F3.5 a + 80 % Dy % . Oe. Figure E3.5 Circuit for Exervise 3.5, ANSWER 67.79.Switching Diode Circuits 8S at ccurent waveforms. (se diode eeu for (6) The vlige and Example 33. 2 z z Repeat Exercise 35 by reversing the plait ofthe 60 V de source. ANSWER 318386 Chapter 3. Switching Circuits, Power Computations, and Component Concepts Consider the circuit shown in Fig. 3.6(a) with @ de soure inthe load side, Assume ideal di- ‘odes, zero intial inductor current, and, < V, Find the expressions for ip iy and, for 0 <1 <27, where Tis the period of v, given by T =2n/a Assume) = ¥-4in at ith V, = OV, @ = 377 rads, 7 = 1667 ms SOLUTION Att = 0 the switch closes and v, remains off and so does D,. Atl =f), ¥, = Vig, which forces D, to turn on and D, to tum off, resulting in the equiva- lent cireut of mode I for t > r, shown in Fig. 3.6(0). f, can be determined by the following equation: switched into the circuit, However, when ¥; S Vag. D. af Ye) _ in| | = 0.67 ms ‘The differential equation for mode 1 is given by tk aE R=30 L=10mH @ Dy Dy Yas Msc Dy Lb ~@) D2 6: ® © 11206Tms TR n= Figure 3.6 (a) Circuit for Example 3.4. (b) Mode 1: 0 <+ < 1/2, (Mode 2: 7/2 <1 < F-(d) Plotofi,(#) for <1 < 273.1 Switching Diode Circuits 87 "The transient response of (0) is eines Fingal) = Ly where 7 = L/R = 3.33 ms, and the steady-state response is given by inal!) = anton o- * 20.75sin(3771—51.5°)-8.33 A where [Z| = JRE (WL)! = 4.82 0 and 6= tan” (wL/R) = 515° The over response is given by baa Ye 1) = he! sion 8) G19) ‘The constant J is obtsined by setting /, = Oat ¢ = 1, in Eq. (3.19), which yields ote 1 = Fanon 958 =20824 The general solution for (0 is given by 5) = 24.5704 20.75sin(3T7E-~S1S°)-833 OS1» V,,sin(neot + 8,,) = ‘The fundamental current and voltage components are iy (t) = Tysin(wr+ 6) (3.450! vy() = V,,sin(wt+ 8.) (3.456 where J,, and V,, are the peak values of the current and voltage fundamental compo- nents, respectively. The rms values of i,(#) and v,(#) can be easily obtained using Bqs. (3.43) ane G4): Bam = +[aoe F > hy Gas 1 =a +(2) (#2 m= (4) (8 He Bie + aom + aso Similety, ie tw PR+) peed’ Hane(i)-( ba = VE + VA pos + Veg ++ VEmg = 152, ‘These equations are obtained since the integration of the product of two different fre- quency components over Tis zero. ‘The instantaneous power is given by PO = iC), Ian +TasS° Sino 0,,)* Ve Sly sin(nt~ 8.) Bis : a + (Siasnoer- wf y V,q8in (ment — tn) Evaluating the average value of p(t) in Eq. (3.48) over the fundamental frequency, shows that the second and thied terms in Eq. (3.48) are zeros and the fourth term is simply the sum of expressions similar to Eq, (3.46) but evaluated at each harmonic component as shown below: Pave = Hag ¥de + Letras cts COS (Oy1 — 8:1) + Erase s2smas C0882 hs) + Gas ict 5 bsml am 89,3.3. Basie Power and Harmonic Concepts 103 where 0, = 6,,~ O,( = 1,2, . nth voltage and current harmonics. It is clear from the above equation that true de power can be obtained only if both the line current and voltage have de components, The second term represents the average power at the source terminal obtained from the rms value of the hat- ‘monic components. which represents the phase shift between the 33.33 Total Harmonic Distortion Since the wanted portion of the distorted waveform i,(t) is the fundamental compo- nent, the difference between the “desired” rms value of f(0) and the “wanted” value is appropriately called the distorted portion of (1), defined as igaan = ()~ 6100 = > ion (3.504) co Yea =~ Ve) = > Vas 3.508) ‘The relative measure of the distortion is defined through an index called the sonal harmonic distortion (THD), which is the ratio of the rms value of the distorted wave form and the rms value of the fundamental component. The THD expression for the current and voltage are given in Eqs. (3.514) and (3.516), respectively, assuming no de components. Fosse = flan * Tso * Bare + Tie Tele et a (@GSla) +(e lees, Jerome, 22m * Vis ere + Vise + Vans (3.518) = a = (lem) (Yom - kez) “(e)* In terms of the rms of the original waveform, Eqs. (3.51a) and (3.518) may be re- written as (3.524) (3.525) ey itis also common to refer to THD as a percentage.104 Chapter 3. Switching Cireuits, Power Computations, and Component Concepts “Edenesse a. Calculate THD, for (0) given in Example 3.6 and THD, for v, of Exercise 3.11. ANSWER 121%, 100% 3.3.3.4. Power Factor The equations to calculate the power factor for distorted waveforms are more complex compared to the sinusoidal case discussed earlier. Applying the definition of thx power factor given in Eq. (3.37) to the distorted current and voltage waveforms Eqs. (3.43) and (3.44) and the average power given in Eq, (3.49) (with zero de compo- nents), pf may be expressed as See: !enmsV mss 008 Oy, son) Vins Bs ‘This expression for pf can be significantly simplified if we assume the line voltage is pure sinusoidal and distortion is limited toi); thus it can be shown that pfcan be expressed a= pf= Ets cns 6, @ss where 0 isthe phase angle between the voltage v,(?) and the fundamental compones= ‘fi,(). This assumption i valid in many power electronics applications, The line v= age is normally undistorted, and the line curent is what gets dstored, i, ¥,(0 =V,sineot G.550 i,(1) = distorted (nonsinusoidal) @.586) ‘The current is expressed in terms of the Fourier series as follows: 1,0) =hsin(or+ )+ hsinQort 0,)+-+J,sin(nor+ 6) (3.56 resulting in the average power given by Pe Hf vad 0 “4 * (i,sineot)( {sin t+ 8,) + sin(2ext+ 63) +--+ J,sin(nar + 6,)) de 0 0 | | = e054, Pave = Vs sd rms C08 9} if Vt sina sin(oo + eat]3.3. Basic Power and Harmonic Concepts 105 Hence, the power factor is given by Tysnms¥ ems 0086 Power factor = /amst sean 008 0s Jats ¢056, 57) ‘The expression /, ms/ lym i8 caused by the distortion of the line current and is ap- propriately called the distortion power factor, kau. The term cos is caused by the dis- placement angle between the line voltage and the fundamental current component and is. ‘commonly known as the displacement power factor, Kygy- Hence, the power factor in power electronics is more useful if it is represented as a product of ka and fa? pf = kackaisp G58) where Kaigy = 608 8, Kise = To) rme/ Laon In terms of kyo it can be shown that the current THD can be expressed as I Zz ia THD, G59) Caleulate Aggy gay: THD,, and THD, for the wavefoems shown in Fig. 3.16 (MENTION Since the voltage is purely sinusoidal, THD, = 0%. ‘To obtain THD,, first we obtain the rms value of i(#) and its fundamental component. 1 sft, ‘ndars fc) aur] Boos = sof idee =8 ‘The fundamental component of) is given by Kysiner + cosa where °- 9) sina det [275 "-tesinor dor] TY 7 o Te Figure 3.16 Waveforms for Example 3.7106 Chapter3._ Switching Circuits, Power Computations, and Component Concepts If we let or = at + 8, the integral becomes i i Af [Ftesinceor— Adar [?*Tsin(wr~ das] 1, [- cos(ar— 6) + cos(-#) + cos(2a— 6) + eos(-6)] 41, bcos 8 Similarly, 1%; = (4,/7)sin 0. Therefore, the peak fundamental component is given by 4 fy = AT = Hence, the fundamental component of i() is given by 1,0) = 1, sin wr + 6) ‘and the ms of J, is given by CCaleulate THD, and pf forthe curent waveform shown in Fig. 3.15, by using upto the fifth harmonic ANSWER 117%, 0.64 CCaleulate THD, and pf forthe current waveform shown in Fig. E3.14, ANSWER 136%, 11 34 CAPACITOR AND INDUCTOR RESPONSES ‘The transient and steady-state values of the capacitor and inductor voltage and current are well understood by undergraduate electrical engineering students. However, © brief review of such responses might be useful to some readers at this point.3.4 Capacitor and Inductor Responses 107 40 Figure £3.14 Waveform for Exercise 3.14, - Capacitor Transient Response Consider the RC circuit of Fig, 3.17(a) with a de excitation and an ideal switch. Assume the switch is open for t < tp and at t = fy, the switch is closed. The capac- itor voltage for ¢ < fis equal to the de source, Vi. For f > fg the time-domain capacitor current is given in terms of its voltage, dv, c& 3.60) @ o Vie “|e To teirem vol | | vaits took | { Figure 3.17 (a) RC circuit i with de excitation and an + ideal switch, (6) The current * change through the o ‘capacitor.108 Chapter 3. Switehing Cirenits, Power Computations, and Component Concepts ‘Substituting for i, «ral sotution: (0—v.(#))/R and solving for v,(t), we obtain the following ge= yelt) = vest Wes Ved tO" G6 where Ye. = final value at ¢ = 9° Vo = initial value at +7= circuit time constant, with = (R, ILR,)C ‘Since the capacitor voltage does not change instantaneously, we have Vell) = elt.) From the circuit diagram, at =e the capacitor becomes fully discharges through R, with time constant RC. The capacitor’s final value is given by Veg = Velo) = 0 ‘The final expressions for v.(Q and i,(#) for ¢ > fy are velt) = Vgge tet (B.622 Gye cee ‘ 1A) = Hes) 3.628) AO) Re « ‘The plot is shown in Fig. 3.17(6). The steady state is reached at approximately 5 ‘ALE = fo. the capacitor curent is zero, but at =, it suddenly becomes —F./. ‘This brings us to another important statement about the capacitor current: The current throws? the capacitor can change instantaneously in svatching circulis, as shown in Fig. 3.17(). Consider the circuit of Fig, 3.18(a), where the switch has been closed for ¢< 0. Aty = 0, he switch is opened. Sketch the capacitor vollage and current waveforms for 1 > 0. Assume the enact sally uncharged @ iu Sar x eee Oe. o i ae Figure 3. (a) Circuit for Example 3.8. (8) Capacitor current and yoltage waveforms.3.4 Capacitor and Inductor Responses 109 i=in Ie D "ik ° 4 RCH ay van) @ 0 HT é o Figure 3.19 (a) Series combination of capacitor and diode. (b) Capacitor current and voltage waveforms. “TION ‘The capacitor voltage equation is given by Therefore, ¥() is volt) ‘The capacitor current and voltage waveforms are shown in Fig, 3.18(b). Since the capacitor voltage does not change instantaneously, the switch inthis example is not allowed to close again unless a way is found by which the switch voltage is prevented from sppearing across the capacitor (.¢, voltae diversion), Ths ean be accomplished by adding a ( the time-domain inductor voltage is given by diy aoe 3.65) Substituting for v, = RU ~i,(0), where R = R, || Ry, and solving for i,t), we ob- tain the following general solution for (0): WOH hap ag h, peer” G66) where iy.p= final value at t = 0° iy, = initial value at ¢= fy. 7 = circuit time constant, with t= L/R ‘Since the inductor current does not change instantaneously, we have Lied = Tilt) G.67) From the circuit, itis clear that at 1 ~oe, i,( comes zero. The final expression for i(t) is given by fa and its voltage be- y(t) = Ty - ry and the inductor voltage is given by VAC) = RT vt Again at ¢ = ty, Vglig-) = 0, and at 1 = fp-, the inductor voltage is veltos) = Rl We conclude that the voltage across the inductor can change instantaneously under pitching action, as shown by the inductor current and voltage waveforms in Fig. 3.23, “@ en Figure 3.22 Inductor switehing circuit.112 Chapter 3 Switching Circuits, Power Computations, and Component Concepts ‘0 he 0 Luin) rales eal ye, + Figure 3.23 Inductor cure % : and voltage waveforms. Consider the switching circuit given in Fig. F3.16 with an ideal diode. Assume the initial 0. 1F the diode has a 5 9 forward resistance, what is ; = = 05 pis, 10 4s, and 100 ys. At what time does the inductor becomes discharged? wie ines. yee Figure £3.16 Switching circuit for fe Exercise 3.16. ANSWER 50 pA, 1 mA, 0.95 mA, 10 ms Notice in the preceding exercise that when the switch is opened again at ¢ = 10 ms the inductor current gets trapped and the diode tums on and diverts the current from the switch branch to the diode branch. This is known as current commrdation and will be com cred extensively in Chapters 5 and 6. In practice, trapped energy in the inductor would dis sipate through the inductor and diode resistances. To improve the converter’s efficiency the trapped energy is normally allowed to be reconnected by retuming it to the source through a feedback circuit along withthe use of transformers, to be illustrated in Chapter 5 3.4.4 Inductor Steady-State Response Let us reconsider the circuit of Fig. 3.22 except that the switch tums on and off repea:- edly according to the waveform shown in Fig. 3.24(a).3.4 Capacitor and Inductor Responses 113 4 Swick on | om —P ae aT eer © uo SIO Io Ss aft GHNTEG ® “ a 5 ‘ is Figure 3.24 (a) Switch action and inductor {)eurtent and (c) voltage © waveforms for Fig. 3.22. Afier a few switching cycles, the capacitor reaches steady state, at which its value at the beginning of each switching cycle is the same. Mathematically, we express the steady-state condition as i (aT + to) = i,t 1)T + to) G68) This condition makes it possible to show that the average value of the steady-state inductor voltage is zero: Vine Hf em (0 diy Le re e ibd = Eli (n+ YT) i(n7 (3.69) =0 The steady-state current and voltage waveforms are shown in Fig. 3.24(b) and (c), respectively. The zero average inductor voltage is illustrated by the equal negative and positive shaded areas for v,(0) shown in Fig. 3.24(c).114 Chapter 3 Switching Circuits, Power Computations, andl Component Concepts PROBLEMS In all problems assume ideal diodes unless stated 3.3 Consider the switching diode circuits otherwise Fig, P3.3 for ¢ > 0 when the switch is turned c= 3.1 Derive the expressions for (i) and vi) inthe Derive expressions for i,(#), i,(0), and v0). A ‘switching circuits shown in Fig, P3.1, Assume all ini SU iO) =1.0/,, and v(0) =¥,. Suge: Pan vacpaiget ig ire edhe enh RE al Ee cy BAe R200 C= O00 pF, and L=2 mH, and Peditectional sketch i, andy, 34 Consider the transistor switching ciruits 3.2 Assuming the switch in Fig P3.2is opened at Fig. P34. At = 0 the transistor is tamed on by 120" dite the expressions fori, and 4, si@Ralt0 the base. Derive and sketch the wa forms for ig, i, vq andl vp. Assume v,(0) = -20 and sketch them for ¢ > 0. Assume 4,(0) = J, and ‘o 5 rd ee : os and (0) =0. Compare the t= circuits. @ i © Figure P31 i= Ig 12 (0)= Va Figure P32@ Figure P3.4 Consider the diode circuit OF Fig. P3.5 with 5; opened and S, closed at ¢ = 0. Assume seal conditions are given as v,,(0) =0 and ~0.5fy Derive the expressions for v3 and snd sketch them, Use J, =2 A, C, =C: = o to L a8 zo 10 7 i c ud Assume the switch in Fig. P36 is opened at % Derive the expressions and sketch the seems for Ve Yea fi, and ip. Assume ¥.:(0) = m= Oand i(0) = 0. Ceasider the capacitor current shown in Fig. * Sor a given switch-mode power supply with ‘2F,and assume v0) = 100'V. ‘Sesch v, showing the peak values and times. Seesmic the ripple voltage across the capacitor, Consider the SCR-diode switching circuit ‘= Fig. P3.8 with the switch closing at = 0. w ae 45s 3a 1 ra 2 Figure P3.6 ‘This cits vollage commutation cet known a an impulse- O's given by (et WE T Discuss the drawback of such an arrangement 3.9 For the SCR to tum offin the circuit of Fig. P38, the capacitor voltage must first be charged to a large negative value. This is done through an ex- ternal circuit consisting of a diode and an induc- tance as shown in Fig. P39. The purpose of Dy Figure P3.7116 Chapter 3 Wie Figure P3.9 snd Lis that whem the switch is tured off, the ca- pacitor voltage returns to its original negative value, Vo. Assume Sand SCR are open for along time witha capocitor initial value equa 10+ ¥%, At 1 = Othe SCR is triggered. Sketch the waveforms {0 ip ion aces Yo #3 ¥,, Compare this circuit with the iret given in Problem 3.8. 3.10 Consider the SCR circuit shown in Fig 3.10, Because the SCR turns aff naturally due to the fact that its eurrent becomes zero, the circuit is ‘knowin asa self-commutated circuit. Atr = 0, the SCR is tuned on by applying jg. Derive the ex- pressions for v,(), (0, and the SCR commutation time. Let L=0.1 mH, C=47 pk, and 4, = 120 V. Assume the initial capacitor vollage is () zero and (i) -¥ (where |] < Fa) sce fb } + Yee cz Figure P3.10 3.11 Design for L and C in Problem 3.10 so that the SCR commutation time is 100 jus and the peak ‘capacitor voltage does not exceed 200 V. Switching Cireuits, Power Computations, and Component Concepts 3:12 Consider a source terminal connected 2 power electronic cireait whose current and vols ‘waveforms are shown in Fig. P3.12, wit (a) Calculate the average and rms source curr and source voltage values. (b) Calculate the average input power. 3.13 Consider the circuit of Fig. P3.13, where». a train of pulses as shown. Assume RC 2 T'so Sar ¥, is assumed constant and equal to 75 V. Skew the steady-state inductor current waveform, 3.14 The switeh in Fig. P3.14 is elosed at 1 = ‘Assume ideal diodes and all inductor initial cont tions are zero. Obtain the expressions forthe in tor current and sketch them for 0:<#<27. Assam: 1(0 is sinusoidal with «9 = 377 rads and ¥, = bil Veet = Lm, R = 19, and ¥,, =20V. 3.15 Consider a power source with its termini voltage and current given as ¥,(0) = 100+ 80sin(or — 100°) 4 10c0s(2ar+ 120°) + 25sin3ex V 1,() = 12+ 10sin(wr +25°) | + Ssin(2av~30°) + 2¢0s30" A where a is the fundamental angular fequens Calenlat: (@) The rms value of f(0) and v0) (©) The averaye input power supplied bythe so (©) The ms values of the fundamental comp ents of i anv) (@ THD, snd THD, (0) Kiss Kays and PE 346 Consider the phasor circuit of Fig. P3.0¢ with Z, = 202-360, Vy = 80202 V. (a) Determine the cizeuit’s real, spparent, und = active powers and the input power factor (©) Determine the type and value of the loss needed tobe connected etween «and a achieve a unity power factor. 3.17 Prove the following intograls r= Jf sin(an sin mena = J cos(nan)cos(maryar nem ve S& Ns Problems 117 Power letonic 1 rts s rn ToT ce where @ = 2n/T and n= 1,2,...,0,m = 1, aha 3.18 Sketch the steady-state #, and v, wave forms by assuming the switch in Fig. P3.18 is repeatedly opened and closed as shown, Use T= 10ms, 3.19 Show thatthe Fourier series forthe half2wave rwatifier ouput voltage of Fig. P3.19 is given by oon Fast ar P33 F, [fsin(nmepcos(menyde 3.20 Show that the Fourier series for the output voltage waveform of the full-wave rectifier of Fig, P3.20/is given by ie orig ad 3.21. By considering only the first four terms of the Fourier series given in Problem 3.19, verify the answers given in Exercise 3.11118 Chapter 3 Switching Circuits, Power Computations, and Component Coneepts @ » 7) “ a o Figure P3.14 10 R=020 4 > b= tm f +4 Switch waveform % 4 on on oT f Figure P3.16 Figure P3.18Figure P3.20 3.22 Derive the Fourier coefficient equations 2, sad b, given in Eqs. (3.41). 3.23 The phase current waveform in a three phase full-wave rectifier SCR circuit under Sighly induetive load is shown in Fig, P3.23 Show that the harmonies of () are given by the lowing expression: 087. 7 Seale isan) 1 3 with no triple harmonics in X?) and nm = 6k£1 f= 0,1,2,-..). Show that tbe rs magnitude of ‘Sem harmonics 61,/ — eee Figure P3.23 3.24 Show that the Fourier series for the phase carent of the six-pulse SCR converter shown in 3.24 is given by Problems 119 Stal i@ 7 ae A 4 elope | let Figure F324 3.25 (a) Figure P3.25(a) shows a typical ac line current waveform in a single-phase, full-wave SCR-controlled rectifier circuit under resistance load and sinusoidal ac source. Show that the fun- damental line current component is given by iC) =Lsin(eo+ 9) cos2a—1 Talat ainda MIF 2Cr— a) 2G a)sin2a— costae L, Da (b) Find i,,(0) for the half-wave SCR-controlled line current shown in Fig, P3.25(0). 3.26 Show that the high-harmonics Fourier coef ficients for Fig. P3.26 are given by, = tan 4, a,=0 1 at sin2ay+S? [: mie re) 2 ~ 2 sina +n) + aswel) What value of h will achieve 95% input power factor assuming the waveform represents a line ‘current with a sinusoidal line voltage 4.27 Figure P3.27 shows the ac line current for a full-wave, single-phase SCR-controlled convertor under highly inductive load. Show that the a, co- efficients are given by120 Chapter 3 & 4 io io Figure P3.26 Switching Cireuits, Power Computations, and Component Concepts 3.28 Show that the Fourier components of i(#) of Fig, P3.28 are given by 10- Ze S comasinnar 40 E sale 2x[_ peas Figure P3.28 3.29 Determine the Fourier components for the waveforms given in Fig. P3.29. General Problems 3.30 In the circuit of Fig. P3.30, the switch is ‘opened at ¢ = 0 after it has been closed for along time. Derive the expressions for (9), ipf0), and ¥(@) for t > 0, and plot them. 3.31 Show thatthe load current after the switch is closed at ¢ = 0 in the diode cireut in Fig. P3.31 is given by io=0 10 = Bante" ede 451 osren, i) =0 rT Assume the inductor is not initially charged, and ‘here 4 isthe time when the diode turns on and is given by ox, = sin-!(V/V,), and tis the time at which the diode turns off before the next cycle starts at ¢ = 7. Find the expression fort. 3.32 The circuit of Fig. P3.32(a) is known as a single-phase bridge inverter whose purpose is to convert the de input voltage, V4, to an ac output voltage, v,. Ifthe switching sequence of 5, S,,S;, and S, is done in such a way that the output volt age is shown in Fig. P3.32(6):Problems 121 Figure P3.29 | = 2 Eas Show that the Fourier series for v, is given by 7 sino + Vysin3a1 + VysinSar+ o ae +¥,sinnot122 Chapter 3 Switching Circuits, Power Computations, and Component Concepts Figure P3.33 (b) Use the above results to find the Fourier series for i,(0. 3.33 Consider the two-phase switching commuta- tion circuit shown in Fig. P3.33. Assume 5; has ‘been on for a long time prior to uxt = 150°, when S, 4s tured on (a) Sketch the waveforms for iQ) and {g(0, and (b) determine the time during which both 1D, and D wore on, Assume v{@) = 100 sin 377¢ and, ysl) = 100 sin(77¢ — 120°) V. Use L, = 1 mi and R, = 2.0. 3.34 The circuit shown in Fig. P3.34 is known as ' forced commutation circuit whose voltage wave~ Figure P3.34 form over one switching period, [,, is shown for 1> 0. Derive the expression for ¥,(0) over one switching cycle, The conduction states of SCR, and SCR, are shown on the waveform. 3.35 Figure P3.35 shows a disc-triae switching Circuit used in heater controller, motor speed vari- sation, and light dimmer applications. Design for R and R,, so that the triae triggers at 30° and 2: during the positive and negative cycles, respeo tively. Assume v,(9 = 110sin 27760 V and te die breakover voltage is 24 V. R x 2 Wise Diae c Figure P3.38 3.36 Assume the switch is turned on at ¢ = 0 ant ‘the capacitor i initially discharged inthe circuit ot Fig. 13.36, Derive the expression for »,(i) ant sketch it for 0-< 1 < 40 ms, where (9 = 1 sin2aS01, 8 = 1002 and Figure P3.36 3.37 Figure P337 shows a self-osilating LC circu that allows the SCR to turn of raturlly wathout using ‘an additional auxiliery SCR. This eieuit is known ss = series resonant turn-off circuit. (a) Derive the e» pression for v,(0) and i,(0, and (b) determine the power rating of the SCR. Assume the SCR is fe ‘wiggered at ¢ = 0 and then repeatedly every T= = rms, Assume ¥{0-) = 30°V and (0°) = 0. 3.38 The circuit given in Fig. P3.38 represen ‘one possible implementation of a family of de-c= converters known as soft-switching converters.Wwe assume the switch is tumed on at 7 = 0, ‘Sat the capacitor voltage for r > is given by Yer = V+ Hag ¥ooseayt | | | | Figure P3.39 Problems 123 where @, = (C51 and C,, = C, +C, () In order to tum S back on while the capacitor voltage across C, is zero, the voltage across it must be allowed to reach zero again during the off time of ‘the switch. Show that for zero-voltage switching to ‘occur, the following condition must be met; Veg wee 3.39 Figure P3.39 shows output and input wave forms in a cycloconverter-type power electroni cireuit, where the output frequency is one-half of the input voltage frequency. Find the Fourier series representation for such waveforms. 3.40 Figure P3.40 shows a typical inductor cur- rent waveform in a switched-mode power supply when operating at the boundary of continuous- and discontinuous-mode operations. (a) Determine the Fourier components for i(t). (b) Calculate its average and rms values, (©) Calculate THD, 3.41. Repeat Problem 3.40 for the inductor current shown in Fig. P3.41, which represents a discontin- ‘uous conduction mode of operation ina switched ‘mode power supply. 3.42 Figure P3.42 shows a half-wave rectifier wave~ ‘form with a sine-squared pulse represented mathemat- ically for 0 < 1 < T/2>by the following equation: (0) = Isnt124 Chapter 3 Toot omar Figure P3.41 £50) ora a T ava Figure P3.42 (@) Determine the Fourier coefficients. (b) Calculate the ms and THD, values. 3.43 Consider a sinusoidal line current as shown in Fig. P3.43 that can be represented mathematically for 0 < 1 < Ty the following equation: (= 1, sintat ‘where » is an odd integer. Assuming thet the line voltage is given by (0) = V, sin wt V, derive the ‘expression forthe power factor and THD, 3.44 The simplified equivalent circuit of a de-to-de ‘converter known as a buck converter is shown in Fig, P3.44(a). Assume the switch is tuned on and ‘off according to the waveform shown in Fig. 3.44(b), Sketch the steady-state waveforms for us fs if, and derive the expression for the av= erage output power. 345 Figure P3.45 shows eight different switched- ‘mode topologies, of which only one is valid. Iden- tify this topology and state which circuit law (KVL or KCL) is violated for each of the other topologies. ‘Switching Circuits, Power Computations, and Component Concepts Figure P3.43 @ ° DT Tr ® a+r Figure P3.44 3.46 Sketch the waveforms for ipy pas nd vy and find the average output voltage for the cireut of Fig, P3.46(a), where v3, (2) and v,0) are as shown inFig. P3.46(0). 3.47 Repeat Problem 3.46 by replacing Dy and D- with two switches 5, and 5, a8 shown in Fig 347. Assume that 5; is conducting only when |vsa(O| > ¥,/3 and $;'s the complement of 5) ‘3481 Consider the full-bridge rectifier of Fig. P3.48 with a current source i() = J sin o€ and a con- stant output voltage, V, Sketch v, and i, and find tne average output power. 3.49 Consider the full-bridge rect P3.49 with a voltage souree »,(0 4 constant output current, /,, Sketch i, and v, and find the average output power,