GaN Transistors for Efficient Power Conversion E-Book

Alex Lidow

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 582

Veröffentlichungsjahr: 2019

Beschreibung

An up-to-date, practical guide on upgrading from silicon to GaN, and how to use GaN transistors in power conversion systems design This updated, third edition of a popular book on GaN transistors for efficient power conversion has been substantially expanded to keep students and practicing power conversion engineers ahead of the learning curve in GaN technology advancements. Acknowledging that GaN transistors are not one-to-one replacements for the current MOSFET technology, this book serves as a practical guide for understanding basic GaN transistor construction, characteristics, and applications. Included are discussions on the fundamental physics of these power semiconductors, layout, and other circuit design considerations, as well as specific application examples demonstrating design techniques when employing GaN devices. GaN Transistors for Efficient Power Conversion, 3rd Edition brings key updates to the chapters of Driving GaN Transistors; Modeling, Simulation, and Measurement of GaN Transistors; DC-DC Power Conversion; Envelope Tracking; and Highly Resonant Wireless Energy Transfer. It also offers new chapters on Thermal Management, Multilevel Converters, and Lidar, and revises many others throughout. * Written by leaders in the power semiconductor field and industry pioneers in GaN power transistor technology and applications * Updated with 35% new material, including three new chapters on Thermal Management, Multilevel Converters, Wireless Power, and Lidar * Features practical guidance on formulating specific circuit designs when constructing power conversion systems using GaN transistors * A valuable resource for professional engineers, systems designers, and electrical engineering students who need to fully understand the state-of-the-art GaN Transistors for Efficient Power Conversion, 3rd Edition is an essential learning tool and reference guide that enables power conversion engineers to design energy-efficient, smaller, and more cost-effective products using GaN transistors.

Leseprobe

GaN Transistors for Efficient Power Conversion

Third Edition

Alex Lidow

Efficient Power Conversion Corporation (EPC)USA

Michael de Rooij

Efficient Power Conversion Corporation (EPC)USA

Johan Strydom

Kilby LabsTexas InstrumentsUSA

David Reusch

VPT, Inc.USA

John Glaser

Efficient Power Conversion Corporation (EPC)USA

This edition first published 2020© 2020 John Wiley & Sons Ltd

Edition History1e 2012 Power Conversion Publications, 2e 2015 Wiley

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Alex Lidow, Michael de Rooij, Johan Strydom, David Reusch and John Glaser to be identified as the authors of this work has been asserted in accordance with law.

Registered OfficesJohn Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USAJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial OfficeThe Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of WarrantyIn view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication DataNames: Lidow, Alex, author. | de Rooij, Michael, author. | Strydom, Johan, author. | Reusch, David, author. | Glaser, John (Electrical engineer), author.Title: GaN transistors for efficient power conversion / Alex Lidow, Ph.D., Efficient Power Conversion Corporation (EPC), USA, Michael de Rooij, Ph.D., Efficient Power Conversion Corporation (EPC), USA, Johan Strydom, Ph.D., Kilby Labs, Texas Instruments, USA, David Reusch, Ph.D., VPT, Inc., USA, John Glaser, Ph.D., Efficient Power Conversion Corporation (EPC), USA.Description: 3rd edition. | Hoboken, NJ : John Wiley & Sons, Inc., 2020. | Includes bibliographical references and index. |Identifiers: LCCN 2019015122 (print) | LCCN 2019017751 (ebook) | ISBN 9781119594376 (Adobe PDF) | ISBN 9781119594420 (ePub) | ISBN 9781119594147 (hardback)Subjects: LCSH: Field‐effect transistors–Materials. | Power transistors–Materials. | Gallium nitride.Classification: LCC TK7871.95 (ebook) | LCC TK7871.95 .G355 2020 (print) | DDC 621.3815/284–dc23LC record available at https://lccn.loc.gov/2019015122

Cover Design: WileyCover Image: Courtesy of Efficient Power Conversion Corporation

In memory of Eric Lidow, the original power conversion pioneer.

Foreword

It is well established that the CMOS inverter and DRAM are the two basic building blocks of digital signal processing. Decades of improving inverter switching speed and memory density under Moore's Law has unearthed numerous applications that were previously unimaginable. Power processing is built upon two similar functional building blocks: power switches and energy storage devices, such as the inductor and capacitor. The push for higher switching frequencies has always been a major catalyst for performance improvement and size reduction.

Since its introduction in the mid‐1970s, the power MOSFET, with its greater switching speed, has replaced the bipolar transistor. To date, the power MOSFET has been perfected up to its theoretical limit. Device switching losses can be reduced further with the help of soft‐switching techniques. However, its gate‐drive loss is still excessive, limiting the switching frequency to the low hundreds of kilohertz in most applications.

The recent introduction of GaN, with much improved figures of merit, opens the door for operating frequencies well into the megahertz range. A number of design examples are illustrated in this book and other literatures, citing impressive power density improvements by a factor of 5 or 10. However, I believe the potential contribution of GaN goes beyond the simple measures of efficiency and power density. GaN has the potential to have a profound impact on our design practice, including a possible paradigm shift.

Power electronics is interdisciplinary. The essential constituencies of a power electronics system include switches, energy storage devices, circuit topology, system packaging, electromagnetic interactions, thermal management, EMC/EMI, and manufacturing considerations. When the switching frequency is low, these various constituencies are loosely coupled. Current design practices address these issues in piecemeal fashion. When a system is designed for a much higher frequency, the components are arranged in close proximity to minimize undesirable parasitics. This invariably leads to unwanted electromagnetic coupling and thermal interaction.

This increasing intricacy between components and circuits requires a more holistic approach, concurrently taking into account all electrical, mechanical, electromagnetic, and thermal considerations. Furthermore, all operations should be executed correctly, both spatially and temporally. These challenges would prompt circuit designers to pursue a more integrated approach. For power electronics, integration will take place at the functional level or the subsystem level whenever feasible and practical. These integrated modules will serve as the basic building blocks of further system integration. In this manner, customization can be achieved using standardized building blocks, in much the same way as digital electronics systems. With the economy of scale in manufacturing, this will bring significant cost reduction in power electronics equipment and unearth numerous new applications previously precluded due to high cost.

GaN will create fertile ground for research and technology innovations for years to come. Dr. Alex Lidow mentions in this book that it took 30 years for power MOSFET to reach its current state of maturity. While GaN is still in an early stage of development, a few technical challenges require immediate attention. These issues are recognized by the authors and addressed in the book.

High d

v

/d

t

and high d

i

/d

t

render most of the commercially available gate drive circuits unsuitable for GaN devices, especially for the high‐side switch.

Chapter 3

offers many important insights into the design of the gate drive circuit.

Device packaging and circuit layout are critical. The unwanted effects of parasitics should be contained. Soft‐switching techniques can be very useful for this purpose. A number of important issues related to packaging and layout are addressed in detail in

Chapters 4

to

6

.

High‐frequency magnetic design is critical. The choice of suitable magnetics materials becomes rather limited when the switching frequency goes beyond 2–3 MHz. Additionally, more creative high‐frequency magnetics design practice should be explored. Several recent publications suggest design practices that defy the conventional wisdom and practice, yielding interesting results.

The impact of high frequency to EMI/EMC has yet to be explored.

Dr. Alex Lidow is a well‐respected leader in the field. Alex has always been in the forefront of technology and a trendsetter. While serving as the CEO of International Rectifier, he initiated GaN development in the early 2000s. He also led the team in developing the first integrated DrMOS and DirectFET®, which are now commonly used in powering the new generation of microprocessors and many other applications.

This book is a gift to power electronics engineers. It offers a comprehensive view, from device physics, characteristics, and modeling to device and circuit layout considerations and gate drive design, with design considerations for both hard switching and soft switching. Additionally, it further illustrates the utilization of GaN in a wide range of emerging applications.

It is very gratifying to note that three of the five authors of this book are from CPES, joining with Dr. Lidow in the effort to develop this new generation of wide‐band‐gap power switches – presumably a game‐changing device with a scale of impact yet to be defined.

Dr. Fred C. LeeDirector, Center for Power Electronics Systems (CPES)University Distinguished Professor, Virginia Tech

Acknowledgments

The authors wish to acknowledge the many exceptional contributions toward the content of this book from our colleagues.

Dr. Edward Jones contributed to Chapters 3, 6, and 7 with significant content and excellent insights into these as well as several other chapters. Dr. Yuanzhe Zhang is the resident expert on envelope tracking and created much of the work behind Chapter 14 as well as numerous insights and suggestions on other chapters. Dr. Suvankar Biswas did work supporting Chapters 5, 10, and 11. Steve Colino was the driving force behind our class D audio in Chapter 12, while giving many suggestions and insights into almost every chapter. Mohamed Ahmed, in addition to being a star Ph.D. candidate at Virginia Tech, was a brilliant intern for EPC during the summer of 2018 and did much of the experimental work behind the LLC converters in Chapter 10.

We would also like to acknowledge Jianjun (Joe) Cao, Robert Beach, Alana Nakata, Guang Yuan Zhao, Yanping Ma, Robert Strittmatter, and Seshadri Kolluri for providing much of the technical foundation behind GaN transistors and integrated circuits.

A special thank you is due to Joe Engle who, in addition to reviewing and editing all corners of this work, put all the logistics together to make it happen. Sometimes this logistics meant long continuous hours of editing, coupled with amazing diplomacy working with a wide spectrum of personalities. Jenny Somers, the lead graphic artist on this work, as well as many other GaN‐related papers and application notes, deserves a medal of honor as well as an honorary degree in GaN for her creative, and extremely accurate, projection of scientific data into documentary communications.

A note of gratitude to the editors and staff at Wiley who were instrumental in undertaking a diligent review of the text and shepherding the book through the production process.

Finally, we would like to thank Archie Huang and Sue Lin for believing in GaN from the beginning. Their vision and support will change the semiconductor industry forever.

Efficient Power Conversion CorporationJanuary 2019

Alex LidowMichael de RooijJohan StrydomDavid ReuschJohn Glaser

1GaN Technology Overview

1.1 Silicon Power Metal Oxide Silicon Field Effect Transistors 1976–2010

For over four decades, power management efficiency and cost have improved steadily as innovations in power metal oxide silicon field effect transistor (MOSFET) structures, technology, and circuit topologies have kept pace with the growing need for electrical power in our daily lives. In the new millennium, however, the rate of improvement has slowed as the silicon power MOSFET asymptotically approaches its theoretical bounds.

Power MOSFETs first appeared in 1976 as alternatives to bipolar transistors. These majority‐carrier devices were faster, more rugged, and had higher current gain than their minority‐carrier counterparts (for a discussion of basic semiconductor physics, a good reference is [1]). As a result, switching power conversion became a commercial reality. Among the earliest high‐volume consumers of power MOSFETs were AC–DC switching power supplies for early desktop computers, followed by variable‐speed motor drives, fluorescent lights, DC–DC converters, and thousands of other applications that populate our daily lives.

One of the first power MOSFETs was the IRF100 from International Rectifier Corporation, introduced in November 1978. It boasted a 100 V drain‐source breakdown voltage and a 0.1 Ω on‐resistance (RDS(on)), the benchmark of the era. With a die size over 40 mm2 and with a $34 price tag, this product was not destined to supplant the venerable bipolar transistor immediately. Since then, several manufacturers have developed many generations of power MOSFETs. Benchmarks have been set, and subsequently surpassed, each year for 40‐plus years. As of the date of this writing, the 100 V benchmark arguably is held by Infineon with the BSZ096N10LS5. In comparison with the IRF100 MOSFET's resistivity figure of merit (4 Ω mm2), the BSZ096N10LS5 has a figure of merit of 0.060 Ω mm2. That is almost at the theoretical limit for a silicon device [2].

There are still improvements to be made in power MOSFETs. For example, super‐junction devices and IGBTs have achieved conductivity improvements beyond the theoretical limits of a simple vertical, majority‐carrier MOSFET. These innovations may still continue for quite some time and certainly will be able to leverage the low‐cost structure of the power MOSFET and the know‐how of a well‐educated base of designers who, after many years, have learned to squeeze every ounce of performance out of their power conversion circuits and systems.

1.2 The Gallium Nitride Journey Begins

Gallium nitride (GaN) is called a wide bandgap (WBG) semiconductor due to the relatively large bonding energy of the atomic components in its crystal structure (silicon carbide [SiC] is the other most common WBG semiconductor). GaN HEMT (High Electron Mobility Transistors) devices first appeared in about 2004 with depletion‐mode radio frequency (RF) transistors made by Eudyna Corporation in Japan. Using GaN‐on‐SiC substrates, Eudyna successfully produced transistors designed for the RF market [3]. The HEMT structure was based on the phenomenon first described in 1975 by Mimura et al. [4] and in 1994 by Khan et al. [5], which demonstrated the unusually high electron mobility described as a two‐dimensional electron gas (2DEG) near the interface between an AlGaN and GaN heterostructure interface. Adapting this phenomenon to GaN grown on SiC, Eudyna was able to produce benchmark power gain in the multigigahertz frequency range. In 2005, Nitronex Corporation introduced the first depletion‐mode RF HEMT device made with GaN grown on silicon wafers using their SIGANTIC® technology.

GaN RF transistors have continued to make inroads in RF applications as several other companies have entered the market. Acceptance outside of this application, however, has been limited by device cost as well as the inconvenience of depletion‐mode operation (normally conducting and requires a negative voltage on the gate to turn the device off).

In June 2009, Efficient Power Conversion Corporation (EPC) introduced the first enhancement‐mode GaN on silicon (eGaN®) field effect transistors (FETs) designed specifically as power MOSFET replacements (since eGaN FETs do not require a negative voltage to be turned off). At the outset, these products were produced in high volume at low cost by using standard silicon manufacturing technology and facilities. Since then, Matsushita, Transphorm, GaN Systems, ON Semiconductor, Panasonic, TSMC, Navitas, and Infineon, among others, have announced their intention to manufacture GaN transistors for the power conversion market.

The basic requirements for semiconductors used in power conversion are efficiency, reliability, controllability, and cost effectiveness. Without these attributes, a new device structure would not be economically viable. There have been many new structures and materials considered as a successor to silicon; some have been economic successes, others have seen limited or niche acceptance. In the next section, we will look at the comparison between silicon, SiC, and GaN as platform candidates to dominate the next‐generation of power transistors.

1.3 GaN and SiC Compared with Silicon

Silicon has been a dominant material for power management since the late 1950s. The advantages silicon had over earlier semiconductors, such as germanium or selenium, could be expressed in four key categories:

Silicon enabled new applications not possible with earlier materials.

Silicon proved more reliable.

Silicon was easier to use in many ways.

Silicon devices cost less.

All of these advantages stemmed from the basic physical properties of silicon combined with a huge investment in manufacturing infrastructure and engineering. Let us look at some of those basic properties and compare them with other successor candidates. Table 1.1 identifies five key electrical properties of three semiconductor materials contending for the power management market.

Table 1.1 Material properties of GaN, 4H‐SiC, and Si.

One way of translating these basic crystal parameters into a comparison of device performance is to calculate the best theoretical performance achievable for each of the three candidates. For power devices, there are many characteristics that matter in the variety of power conversion systems available today. Five of the most important are: conduction efficiency (on‐resistance), breakdown voltage, size, switching efficiency, and cost.

In the next section, the first four of the material characteristics in Table 1.1 will be reviewed, leading to the conclusion that both SiC [6] and GaN are capable of producing devices with superior on‐resistance, breakdown voltage, and a smaller‐sized transistor compared to Si. In Chapter 2, how these material characteristics translate into superior switching efficiency for a GaN transistor will be explored and in Chapter 17, how a GaN transistor can also be produced at a lower cost than a silicon MOSFET of equivalent performance will be addressed.

1.3.1 Bandgap (Eg)

The bandgap of a semiconductor is related to the strength of the chemical bonds between the atoms in the lattice. These stronger bonds mean that it is harder for an electron to jump from one site to the next. Among the many consequences are lower intrinsic leakage currents and higher operating temperatures for higher bandgap semiconductors. Based on the data in Table 1.1, GaN and SiC both have higher bandgaps than silicon.

1.3.2 Critical Field (Ecrit)

The stronger chemical bonds that cause the wider bandgap also result in a higher critical electric field needed to initiate impact ionization, which results in avalanche breakdown. The voltage at which a device breaks down can be approximated with the formula:

(1.1)

The breakdown voltage of a device (VBR) is therefore proportional to the width of the drift region (wdrift). In the case of SiC and GaN, the drift region can be 10 times smaller than in silicon for the same breakdown voltage. In order to support this electric field, there need to be carriers in the drift region that are depleted away at the point where the device reaches the critical field. This is where there is a huge gain in devices with high critical fields. The number of electrons (assuming an N‐type semiconductor) between two terminals can be calculated using Poisson's equation:

(1.2)

In this equation q is the charge of the electron (1.6 · 10−19 C), ND is the total number of electrons in the volume, ɛo is the permittivity of a vacuum measured in Farads per meter (8.854 · 10−12 F/m), and ɛr is the relative permittivity of the crystal compared to a vacuum. In its simplest form under DC conditions, permittivity is the dielectric constant of the crystal.

Referring to Eq. (1.2), it can be seen that if the critical field of the crystal is 10 times higher, from Eq. (1.1), the electrical terminals can be 10 times closer together. Therefore, the number of electrons, ND, in the drift region can be 100 times greater, but only have one‐tenth the distance to travel. This is the basis for the ability of GaN and SiC to outperform silicon in power conversion.

1.3.3 On‐Resistance (RDS(on))

The theoretical on‐resistance of a one square millimeter majority‐carrier device (measured in ohms [Ω · mm2]) is therefore

(1.3)

where μn is the mobility of electrons. Combining Eqs. (1.1) to (1.3) produces the following relationship between breakdown voltage and on‐resistance:

(1.4)

This equation can now be plotted as shown in Figure 1.1 for Si, SiC, and GaN. This plot is for an ideal structure. Real semiconductors are not always ideal structures and, therefore, it is always a challenge to achieve the theoretical limit. In the case of silicon MOSFETs, it took 30 years.

Figure 1.1 Theoretical on‐resistance for a one square millimeter device versus blocking voltage capability for Si, SiC, and GaN based power devices.

1.3.4 The Two‐Dimensional Electron Gas (2DEG)

The natural structure of crystalline GaN, a hexagonal structure named “wurtzite,” is shown in Figure 1.2a and the 4H‐SiC structure is shown in Figure 1.2b. Because both structures are very chemically stable, they are mechanically robust and can withstand high temperatures without decomposition. The wurtzite crystal structure gives GaN piezoelectric properties that lead to its ability to achieve very high conductivity compared with either silicon or SiC.

Figure 1.2 (a) Schematic of wurtzite GaN. (b) Schematic of 4H‐SiC.

Piezoelectricity in GaN is predominantly caused by the displacement of charged elements in the crystal lattice. If the lattice is subjected to strain, the deformation will cause a miniscule shift in the atoms in the lattice that generate an electric field – the higher the strain, the greater the electric field. By growing a thin layer of AlGaN on top of a GaN crystal, a strain is created at the interface that induces a compensating 2DEG, as shown schematically in Figure 1.3 [7–9]. This 2DEG is used to efficiently conduct electrons when an electric field is applied across, as shown in Figure 1.4.

Figure 1.3 Simplified cross section of a GaN/AlGaN heterostructure showing the formation of a 2DEG due to the strain‐induced polarization at the interface between the two materials.

Figure 1.4 By applying a voltage to the 2DEG an electric current is induced in the crystal.

This 2DEG is highly conductive, in part due to the confinement of the electrons to a very small region at the interface. This confinement increases the mobility of electrons from about 1000 cm2/V·s in unstrained GaN to between 1500 and 2000 cm2/V s in the 2DEG region. The high concentration of electrons with very high mobility is the basis for the HEMT, the primary subject of this book.

1.4 The Basic GaN Transistor Structure

The basic depletion‐mode GaN transistor structure is shown in Figure 1.5. As with any power FET, there are gate, source, and drain electrodes. The source and drain electrodes pierce through the top AlGaN layer to form an ohmic contact with the underlying 2DEG. This creates a short circuit between the source and the drain unless the 2DEG “pool” of electrons is depleted and the semi‐insulating GaN crystal can block the flow of current. In order to deplete the 2DEG, a gate electrode is placed on top of the AlGaN layer. When a negative voltage relative to both drain and source electrodes is applied to the gate, the electrons in the 2DEG are depleted out of the device. This type of transistor is called a depletion‐mode, or d‐mode, HEMT.