electronics in motion and conversion july 2018 · 2019-07-03 · electronics in motion and...
TRANSCRIPT
Electronics in Motion and Conversion July 2018
ISSN: 1863-5598 ZKZ 6471707-18
WELCOME TO THE HOUSE OF COMPETENCE
PRODUCTIONENGINEERING GvA SOLUTIONS DISTRIBUTION
ENGINEERING
POWER IS IN OUR NATURE!
Whether small or large development tasks: speed is no witchcraft but the result of perfectly
fine-tuned processes in interplay with highest competence - our engineering power for your
success.
Quick design-to-product
Time and cost savings through synergy effects
Extensive experience gives you an edge
State-of-the-art technology
Simulation and in-house tests ensure reliability
GvA Leistungselektronik GmbH
Boehringer Straße 10 - 12
D-68307 Mannheim
Phone +49 (0) 621/7 89 92-0
www.gva-leistungselektronik.de
GvA_Anzeige Engineering_151215_engl_RZ.indd 1 15.12.15 12:16
— Rectifier diodes Exceeding nominal and surge current capabilities.
ABB Semiconductors’ high-power rectifier diodes are the first choice in many de-manding applications in industry and traction. We offer two families of high-power rectifier diodes: Standard and avalanche diodes, both featuring reverse repetitive voltage up to 6000 V and junction temperature from up to 190 °C. abb.com/semiconductors
www.bodospower.com July 2018 Bodo´s Power Systems® 1
CONTENT
Viewpoint .......................................................................................... 4 What A Hot PCIM Europe!
Events ................................................................................................ 4
News ............................................................................................. 6-13
Product of the Month ..................................................................... 14 Military-Grade Laptop PSU is 90% Smaller and Lighter, 80% Lower Cost
Guest Editorial ........................................................................... 16-19 International Symposium on Power Semiconductor Devices and ICs—ISPSD 2018 By Gary M. Dolny, Bodo’s Power Systems, [email protected]
Guest Editorial ........................................................................... 20-21 Book Review: “Electric Powertrain” By Dr. John M. Miller, Industry Consultant
Guest Editorial ................................................................................ 22 Secrets of ‘Fluid-Like’ Heat Flow in Solid Semiconductor at Nanoscale By Henning Wriedt, US-Correspondent Bodo’s Power Systems
Cover Story ................................................................................ 24-28 Automotive Traction Inverter Utilizing SiC Technology By Aly Mashaly and Masaharu Nakanishi, ROHM Semiconductor
Wide Band Gap .......................................................................... 30-34 x-GaN 2.0: The Destination FiT10 is Demonstrably Exceeded By Francois Perraud, Teamleader Power & Automotive Solutions, Panasonic Industry Europe GmbH
Power Supply ............................................................................. 36-37 USB Type-C: Consumer Convenience Comes with Design Challenges By Robert Heinzelmann, Toshiba Electronics Europe GmbH
Technology ................................................................................. 38-41 Gate Drive Optocouplers for GaN Power Devices By Robinson Law, Applications Engineer and Chun Keong Tee, Product Manager, Broadcom Inc.
Wide Band Gap .......................................................................... 42-45 Inevitability of Near Chip-Scale SMD Packaging for Power GaN & SiC By Courtney R. Furnival, Semiconductor Packaging Solutions, Inc.
Power Supply ............................................................................. 46-53 Preventing Start-Up Issues Due to Output Inrush in Switching Converters By Fil Paulo Balat, Jefferson Eco, and James Macasaet, Analog Devices
Power Conversion ..................................................................... 54-57 99% PFC Efficiency – with Silicon, at Low Cost! By Garry Tomlins and Dr. Trong Tue Vu PHd., ICERGi Ltd.Technology
Design and Simulation .............................................................. 58-60 Modular Real-Time Development Platform for Power Electronic Systems By Julian Endres, Fabian Bayer, Andreas Linke and Ansgar Ackva, TTZ-EMO - University of Applied Sciences Würzburg-Schweinfurt
New Products............................................................................. 61-64
CONTENT
Bodo´s Power Systems® July 2018 www.bodospower.com2
The Gallery
Allinclusive.
© e
iSos
#MAGICPOWERMODULES
MagI³C Power Modules are easy-to-use DC/DC converters with integrated regulator IC,
power inductor and capacitors. Design and layout reviews as well as support with EMI fi lter
design are offered as a service for all customers. Datasheets contain detailed specifi cations
and application information.
For more information please visit:
www.we-online.com/powermodules
Simple design-in process Design and layout support EMI filter design for EN55022
class B compliance Evaluation Boards for all products
LGA-6 LGA-16 QFN TO263 SIP-3 SIP-4 SIP-7
Bodo´s Power Systems® July 2018 www.bodospower.com4
CONTENTVIEWPOINT
For the first time since 2005, PCIM Europe was held in June. OK, for perfectionists, PCIM 2006 ended on June 1st. This one brought nice weather - and the opportunity to sit outside until late in the evening to review the day at the show and the ton of impres-sions. Bodo was still able to enjoy Spargel from the menu. And as hot as the weather was when entering the halls, so also were the “Hot” contents: 506 exhibitors from all around the world, presenting to over 11,000 visitors. Things are getting so large; it was nice to find most of the booths at the same place as last year.
PCIM is clearly committed to be an important event in Europe, with large players as well as smaller companies seeing the show as a must to attend. And they are right, proving loyalty and solidarity to an event being held since 1979. I will bet it stays the same next year for its 30th anniversary.
And while Wide Band Gap material product applications are still at a very early stage of development, it was interesting to see some big players trying to get a foot in both SiC and GaN. New teams, facilities and laboratories, and strategic cooperation between established companies from “the other technology”, seem to me like logical steps towards servicing customers by offer-ing the whole package – certainly a sign of some maturity. It’ll be very interesting to see how these developments work out - see the News section in this issue to illustrate what I’m referring to.
But while Wide Band Gap surely was the dominating topic, a lot of other interesting things were presented. For example, using Artificial Intelligence via IC’s, to eliminate switching losses in Power Conversion, is a really interesting new approach, as present-ed by a Silicon Valley based company. Again, more information can be found in this issue.
And not to forget Bodo’s traditional Podium’s Discussion! Every single chair was taken, with many interested people making the ef-fort of standing to catch the latest news from SiC and GaN experts. A big thank you to all the presenters for taking the time to partici-pate in the podium!
Don’t worry if you missed it, we will have most of the presentations published in up-coming issues.
Bodo’s magazine is delivered by postal service to all places in the world. It is the only magazine that spreads technical informa-tion on power electronics globally. We have EETech as a partner to serve North America more efficiently. If you are using any kind of tablet or smart phone, you will find all of our content on the website www.eepower.com. If you speak the language, or just want to have a look, don’t miss our Chinese version: www.bodospowerchina.com
My Green Power Tip for July: Use a fan instead of ramping up your air conditioner. A ceiling fan can make a room feel noticeably cooler and uses 10 percent of the energy of a central air conditioner.
Best Regards
What a Hot PCIM Europe!
SEMICON West 2018 San Francisco CA, USA, July 10-12
www.semiconwest.org
Thermal Management 2018 Denver CO, USA, August 8-9 www.thermalconference.com
IEEE-PEMC 2018 Budapest, Hungary, August 26-30
www.ieee-pemc2018.org
ECSCRM 2018 Birmingham, UK, September 2-6
www.ecscrm.org
SEMICON Taiwan 2018 Taipei, Taiwan, September 5-7
www.semicontaiwan.org
EPE ECCE 2018 Riga Latvia September 17-21
www.epe2018.com
Events
A Media Katzbek 17a D-24235 Laboe, Germany Phone: +49 4343 42 17 90 Fax: +49 4343 42 17 89 [email protected] www.bodospower.com
Publishing Editor Bodo Arlt, Dipl.-Ing.
[email protected] Editor
Holger Moscheik Phone + 49 4343 428 5017 [email protected] Editor China Min Xu Phone: +86 156 18860853 [email protected] Support June Hulme
Phone: +44 1270 872315 [email protected]
US Support Cody Miller Phone +1 208 429 6533 [email protected]
Creative Direction & Production Repro Studio Peschke
[email protected] Subscription to qualified readersBodo´s Power Systems is available for the following subscription charges:Annual charge (12 issues)
is 150 € world wideSingle issue is 18 €[email protected]
circulation
print run 24 000
Printing by: Brühlsche Universitätsdruckerei GmbH
& Co KG; 35396 Gießen, GermanyA Media and Bodos Power Systems
assume and hereby disclaim any liability to any person for any loss or damage by errors or omissions in the material contained herein regardless of whether such errors result from negligence accident or any other cause whatsoever.
YOUR SOLAR LIFEGUARD
LDSR series
At the heart of power electronics
www.lem.com
New closed-loop current transducers, based on a custom Hall Effect LEM ASIC, measure the leakage current up to 2 KHz frequency. Used in transformerless photovoltaic (PV) inverters for the residential market, LDSR measures AC & DC fault currents and ensures the safety of people around the installation.LDSR offers a competitive price, low dimensions and complies with all regulatory standards. LDSR is also an excellent alternative to expensive fluxgate solutions due to its small footprint and simple construction.
• 300 mA nominal current• PCB mounting• Small dimensions & light weight (25g)• -40 to +105°C operating temperature• Single or three phase configuration
Bodo´s Power Systems® July 2018 www.bodospower.com6
CONTENTNEWS
To support its accelerating growth in the defense and aerospace markets, Efficient Power Conversion Corporation (EPC) is proud to announce the appointment of Spirit Electronics as a distribution partner focusing on these key market segments. Spirit Electron-ics, in operation since 1979 and located in Phoenix, Arizona and Irvine, California, supplies products and services to the Department of Defense, aerospace, and telecommunication industries. “Spirit
Electronics’ knowledge of the market, along with their extensive his-tory and proven successful track record for working with defense and aerospace customers make them a perfect partner for representing EPC’s industry-leading, off-the-shelf eGaN FET and IC product line,” commented Nick Cataldo, EPC vice president of sales and marketing.Marti McCurdy, CEO of Spirit Electronics, noted that, “Our new part-nership with EPC is an exciting addition to our portfolio of products and will allow us to bring the superior performance of eGaN power transistors and ICs to defense and aerospace customers, so they can design leading edge power system solutions.”
www.epc-co.com
www.spiritelectronics.com
Efficient Power Conversion Announces Spirit Electronics as Distribution Partner
LEM has launched a new, modern, intuitive and responsive website with a focus on providing quick access to product documentation. When asked about the changes to the site, Stéphane Rollier, Product & MarCom Manager at LEM said “The previous website was not adapted to handle the needs of today’s engineer, the new site has been simplified to make it easier for the user to find the information they are looking for. A fantastic technical query and easy to navigate menu are just some of the improvements that have been made.” The website is currently available in English, Russian and Japanese with
Chinese launching in June 2018.LEM invite you to explore the new website with your computer, smart-phone or tablet. Alongside the recent launch of LEM’s Smart Grid dedicated platform – www.lemcity.com- LEM is taking every step to make the customer experience as easy and stress free as possible.
www.lem.com
New Website to Enable Users to Search Product Information More Efficiently
Husum, Germany is to become the center for capacitor development in the Mersen Group: The worldwide expert for electrical power and advanced materials has signed a contract to take over the capacitor manufacturer FTCAP. The merger of the two enterprises will result in substantial growth potentials for both parties. FTCAP will oper-ate as an independent subsidiary – which will ensure the continued existence of a successful brand. “FTCAP has undergone extensive growth in recent years and has significantly developed its market
position,” according to FTCAP Managing Directors Nora Reimers and André Tausche. “That is one of the reasons why FTCAP is an attractive addition to the Mersen Group, which so far does not include a capacitor manufacturer.” The transaction will enable the group to expand its current product spectrum to include capacitors, which are of central importance for the development of efficient and high-performance electronic systems. The merger will result in significant growth opportunities for both enterprises – for FTCAP due to syner-gies in development, for example, but also due to a more extensive sales network. The competence and experience of FTCAP are of the utmost importance for Mersen – which is why the acquisition goes hand in hand with a clear commitment to the Husum location and its employees. “Husum will become the center for capacitor development in the Mersen Group,” Managing Directors Reimers and Tausche explain. “In recent years we have implemented diverse joint projects, which allowed us to define technological priorities.”
www.ftcap.de
www.mersen.com
Joining Forces for a Strong Future
ROHM´s partnership in Formula E.
A short guide to the Official Technology Partnership with Venturi Formula E team in season 3.
SMALLER STRONGER FASTER
The Formula E Venturi team has adapted the latest range of ROHM inverters derived from full SiC module technology in its electric-powered racing cars. ROHM has enabled the broad implementation of e-mobility by delivering the next generation of power semiconductor-based SiC modules. It produces these in-house using a vertically integrated manufacturing system, thus guaranteeing high quality and consistent supply to the market.
www.rohm.com/eu
SiC technology allows the chip to be reduced in size, leading to a SMALLER inverter in terms of dimensions and weight.
SMALLER SiC increases thermal efficiency and power density for a STRONGER performance.
STRONGER FASTER
SiC helps vehicles to cross the finish line FASTER and supports fast-charging solutions.
Bodo´s Power Systems® July 2018 www.bodospower.com8
CONTENTNEWS
A fully automated chip factory for manufacturing 300-millimeter thin wafers will be constructed at the Villach location in Austria along-side the existing production facility. Austria’s Chancellor Sebastian Kurz, Dr. Reinhard Ploss, Chief Executive Officer of Infineon, and Dr. Sabine Herlitschka, Chief Executive Officer of Infineon Austria, presented the project in Vienna. Investments totaling around €1.6 bil-lion are planned over six years. Some 400 new jobs, especially highly qualified ones, will be created by the new, highly efficient factory. Construction is scheduled to start in the first half of 2019 and produc-tion is expected to commence at the start of 2021. The additional sales potential of the new factory, given full capacity utilization, is put at around €1.8 billion a year.“Global demand for power semiconductors is soaring. As the market and technology leader, Infineon is particularly sought-after by custom-ers and is even growing more strongly than the market,” said Dr. Reinhard Ploss, Chief Executive Officer of Infineon. “Growth is under-pinned by global megatrends such as climate change, demographic change and increasing digitization. Electric vehicles, connected and
battery-powered devices, data centers or power generation from renewable sources require efficient and reliable power semiconduc-tors. We recognized that trend early on and so are rapidly expanding production capacities for 300-millimeter technology at our Dresden location. The new facility at Villach will help us cater for the growing demand that our customers anticipate, and continue on our path to success in the coming decade.”
www.infineon.com
Infineon Prepares for Long-Term Growth
The company is proud to announce that two technologists have earned their designation as SMTA-Certified Process Engineers. Joining a host of certified Indium Corpora-tion engineers are: Kimberly Flanagan, Technical Support Engineer, Eastern USA and Canada; and David Strabel, Techni-cal Support Engineer, Midwest USA. SMTA Certification is a unique program, sponsored by the Surface Mount Technol-ogy Association (SMTA), which recognizes and certifies competence across the entire SMT assembly process at an engineer-ing level. This certification is one of the electronics assembly industry’s most respected validations of process excellence.Flanagan provides technical support and guidance related to process steps, equipment, techniques, and materials to customers. In addi-tion, she delivers technical training to staff and industry partners.
Flanagan was introduced to Indium Corporation through the com-pany’s summer college internship program, working in the Quality Department. She remained with Indium Corporation as a part-time Quality Engineering Technician while she finished her bachelor’s degree in Physics from Le Moyne College in December 2016.
Strabel provides expert technical assis-tance and guidance to Indium Corporation’s current and potential customers to resolve soldering process-related challenges and enable increased levels of performance. Strabel joined Indium Corporation in 2017, and has more than eight years of experi-ence in SMT/PCBA process engineering. At his previous employer, Eastek International, he received the company’s most prestigious recognition–the President’s Award.
www.indium.com
Indium Corporation Engineers Earn SMTA Certification
Faraday Semi, based in California, USA, is a fast-growing semi-conductor company with innovations in systems solutions focused on power management through DC-DC power conversion. Faraday Semi solutions incorporate high performance semiconductor in ad-vanced packaging technologies such as semiconductor embedded in substrate (SESUB) and advanced electronic components to achieve unique system integration in a smaller size and lower profile by 3D assembly. This integration allows TDK to deliver higher efficiency and ease of use at a lower total system cost to what is currently available today. This results in the world’s smallest class of high power density Point of Load solutions (product brand: μPOL™). The continued pop-ularization of “big data” usage is driving the rapidly increasing global demand for high density and highly efficient data center and reliable data storage, and these miniature devices provide high-performance solutions that meet the challenging requirements of modern digital ICs such as ASICs, FPGAs and CPUs. Their emphasis on high efficiency and thermal management makes these devices suitable for distribut-ed power management systems including ICT and other applications.This will enable TDK to deliver high-efficiency, high-performance,
super-compact solutions catering to the evolving trend of miniatur-ized power solutions and their accompanying semiconductors. TDK is acquiring Faraday Semi in anticipation that these technologies will help to improve systems efficiency and reduce system costs for customers around the world, in ICT as well as in the industrial devices and automotive markets.
www.global.tdk.com
www.faradaysemi.com
TDK Acquires Faraday Semi LLC
a
b
c
d
ef
g
h
Please read right-to-left, in respect to Akira Toriyama and Manga artists around the world. Just follow the numbers.
650V – 1200V SUIJIN AUTOMOTIVE IGBT MODULESwww: pdd.hitachi.eu/ev tel: +44 (0)1628 585151 email: [email protected]
Bodo´s Power Systems® July 2018 www.bodospower.com10
CONTENTNEWS
Microsemi Corporation announced it won Light Reading’s 2018 Leading Lights Award in the “Outstanding Components Vendor” category. Now in its 14th year, Leading Lights is one of the communications industry’s leading awards programs, recognizing top companies and executives for their outstanding achievements in next-generation communications technology, ap-plications, services, strategies and innovations. “Microsemi is honored to be recognized by Light Reading editors and the analyst team from Heavy Reading as the outstanding components vendor for 2018,” said Babak Samimi, vice presi-dent and business unit manager at Microsemi. “As finalists of last year’s award, we are excited to win the
prestigious recognition this year, as it demonstrates our team’s ongoing achievements and validates our commitment to
the industry and our customers.” Microsemi received the top honor for its unique innovation, industry-
leading product portfolio, performance and ability to set industry trends. The judges highlighted the newest member of Microsemi’s award-winning DIGI OTN processors franchise, the DIGI-G5, which supports new 5G optimized architecture enabling the stringent synchronization, latency and network slicing requirements being placed on
optical networks to support 5G deployments.
www.microsemi.com
“Outstanding Components Vendor” Award
The ST Voltage Regulator Finder app helps engineers, purchasers, students, or makers find and buy STMicroelectronics’ voltage regula-tors, precision voltage references, or DC/DC converters, quickly and
easily from a smartphone or tablet. Users can search by part number, parameter, or product family to identify suitable components and view key features in the app. Parametric searches are easily filtered, using checkboxes to choose options such as automotive or industrial quali-fication, and sliders to limit parameters like input or output voltage, output current, or reference voltage. Other searchable parameters include package style, temperature range, and marketing status. Shortlisted components can be saved in My Favorites for convenient access later. The app works offline, so users can identify the parts they need even if no Internet connection is available. When con-nected, datasheets for the selected parts are just a tap away. There is a Sample and Buy option that gives live inventory at ST distributors, and Order Now goes directly to online purchasing from any chosen supplier. There are also handy links to find ST sales offices and dis-tributors worldwide for support or further information.
www.st.com/st-vreg-finder
Mobile App Simplifies Sourcing of Voltage Regulators, Con-verters, and References
Pre-Switch, Inc. introduced Pre-Flex™ Technology, an embedded AI (Artificial Intelligence) controller IC (Integrated Circuit) which effectively eliminates transistor switching losses in hard-switched archi-tectures for virtually any power converter topol-ogy. Pre-Flex dramatically reduces cost, size and weight, while increasing efficiency and reliability of power converters in the server, solar, wind, EV,
traction, and motor drive markets. Pre-Flex technology learns and adapts ‘in-system’, on a cycle-by-cycle basis to reliably force resonant soft-switch across changing loads, input voltages, temperatures, and manufacturing tolerances. The technology has proven efficiencies greater than those of five-level topologies at a fraction of the cost and complexity. Pre-Flex also significantly lowers EMI (Electromagnetic interference) and dramatically decreases dV/dt for any switch type.
Pre-Flex enables Silicon carbide-like performance for low-cost IGBTs and allows SiC- and GaN-based topologies to switch up to 20x faster than they do today. Built into the architecture are innovative cycle-by-cycle safety features and communications that were never possible before. Pre-Flex has switched 900V Wolfspeed SiC MOSFETs up to 1MHz, and 650V Infineon IGBTs at over 100 kHz with unprecedented efficiencies. The technology presently scales from 1 KW to above gigawatts. “We have removed the biggest barrier to advancing the power conversion industry,” said Bruce Renouard, chief executive officer of Pre-Switch, Inc. “Our technology enables simple, reliable, and cost effective forced resonant soft-switching on any topology.”
www.pre-switch.com
Cross-Platform Soft-Switching Technology Using Artificial Intelligence
MORE CHOICESGREATER SECURITY
www.vincotech.com//MiniSKiip
New MiniSKiiP®PIM and MiniSKiiP®PACK with IGBT M7 FOR REAL CHIP-LEVEL MULTI SOURCINGVincotech has extended its range of MiniSKiiP® products with the new IGBT M7 Chip. This new IGBT M7 chip delivers the dual benefi ts of superior performance and multiple sourcing to minimize customers‘ supply chain risks.
Discover the new PIM (CIB) and sixpack confi gurations with power ranges extending up to 200 A.
CHOICESCHOICESCHOICESCHOICESCHOICESCHOICESCHOICESGREATER GREATER GREATER GREATER GREATER GREATER GREATER GREATER GREATER SECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITY
Vincotech has extended its range of MiniSKiiP® products with the This new IGBT M7 chip delivers the dual benefi ts of
superior performance and multiple sourcing to minimize customers‘
SECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITYSECURITY
Main benefi ts/ Multiple sourcing for enhanced supply chain security
/ Lower power losses and improved effi ciency
/ Extended power range for easily scalable inverter designs
/ Pre-applied, high performance thermal paste (HPTP) for superior thermal performance
Power Loss Benchmarking MiniSKiiP® PACK 1: Pout~6.5KW, Fsw~4KHz
Pow
er L
oss
[W]
90
80
70
60
50
40
30
20
10
0
K210-F40 with IGBT4 K210-F70 with M7
– 10 %
VIN_ADs_06_2018_MiniSKiiP_210x297.indd 1 15.06.18 07:10
Bodo´s Power Systems® July 2018 www.bodospower.com12
CONTENTNEWS
ROHM and GaN Systems announced their collaboration in the GaN (gallium nitride) Power Semiconductor business, with the goal of con-tributing to the continuing evolution of power electronics.This strategic partnership leverages GaN Systems’ industry leading capabilities in power GaN transistors along with ROHM’s compre-hensive footprint in semiconductor and considerable resources in the design and manufacture of electronic components. The companies have agreed to jointly develop form-, fit-, and function-compatible products using GaN semiconductor dies in both GaN Systems’ GaNPX™ packaging and ROHM’s traditional power semiconductor packaging. GaN Systems and ROHM customers will now have the advantage of having two possible sources for package-compatible GaN power switches, presenting the widest selection of dual-sourced
GaN devices. Customers will also benefit from greater access to GaN products and resources globally, especially in Asia, one of the fastest growing market for GaN. In addition, GaN Systems and ROHM will work together on GaN semiconductor research and development activities to propose ground-breaking solutions for the industrial, auto-motive, and consumer electronics fields. And to contribute to greater energy savings and increased power densities in the power electron-ics market, both companies will continue to collaborate to expand their line-up of GaN products and broaden the range of choices.
www.rohm.com/eu
www.gansystems.com
Joining Forces for GaN Power Semiconductors
VisIC Technologies has closed $10 million in a Series D round of financing lead by a private investor. The electrification of automotive vehicles has been growing at an unprecedented pace in the recent past and will continue to grow for the foreseeable future. GaN power devices get the maximum performance out of high power, high voltage power conversion systems inside hybrid and electric vehicles. The improved size, weight, efficiency and heat manage-ment of the on-board charger and the DC/DC converter, designed with GaN power devices, all contribute to faster charging and longer driving range. High performance power supplies for telecom systems and datacenters are using GaN power devices to reach new levels of density and efficiency, bringing down the electricity costs of the
operators significantly. “GaN technology opens a new space in power electronics - from shifting the performance envelope up to the point of new topologies development. We are delighted to see VisIC offering specifically rugged GaN devices with negligible fast transient dynamic RDSon.”, said Ivan Feno, Principal Power Design Engineer from Bel Power Solutions. “The insulated thermal pad is another welcome fea-ture enabling the increase of the power stage reliability and density. Ultimately, 1200V rated GaN devices might be an attractive alterna-tive in the 1200V segment dominated by SiC technology today.”
www.visic-tech.com
$10M Raised to Speed Up Market Adoption
The 64th annual IEEE International Elec-tron Devices Meeting (IEDM), to be held at the Hilton San Francisco Union Square hotel December 1-5, 2018, has issued a Call for Papers seeking the world’s best original work in all areas of microelectron-ics research and development.The paper submission deadline this year is Wednesday, August 1, 2018. Authors
are asked to submit four-page camera-ready papers. Accepted papers will be published as-is in the proceedings. A limited number of late-news papers will be accepted. Authors are asked to submit
late-news papers announcing only the most recent and noteworthy developments. The late-news submission deadline is September 10, 2018. At IEDM each year, the world’s best scientists and engineers in the field of microelectronics gather to participate in a technical pro-gram consisting of more than 220 presentations, along with a variety of panels, special sessions, Short Courses, a supplier exhibit, IEEE/EDS award presentations and other events highlighting leading work in more areas of the field than any other conference.
www.ieee-iedm.org
International Electron Devices Meeting 2018 Call for Papers
HiTech Korea is a premier electronics assembly polymer and epoxy-based materials supplier, with its primary operational facility in Korea. “We are excited to announce that HiTech is joining the Alpha family,” said Rick Ertmann, President of Alpha Assembly Solutions, part of the MacDermid Performance Solutions group of businesses. “We believe that HiTech’s talent and resources will augment our expertise to develop new and improved products and capabilities. With this trans-action, Alpha is well-situated to capitalize on many emerging trends in the electronics industry. As the lines between the assembly tech-nologies in electronics continues to merge, we see this acquisition as one of many steps that we are taking to position the business for continued success as we work with customers to solve their assembly
challenges.” This ac-quisition is aligned with Alpha and MacDermid Performance Solutions strategic goals and priorities, namely to drive growth through the delivery of innovative, value added solutions and services into the electronics assembly industry. HiTech’s leading products and development initiatives focus on high growth segments, such as underfills, encapsulants, low-temperature adhesives and ultraviolet adhesives.
www.AlphaAssembly.com
Alpha Assembly Solutions Acquires HiTech Korea CO., LTD
www.bodospower.com July 2018 Bodo´s Power Systems® 13
CONTENTNEWS
Researchers at the Fraunhofer Institute for Solar Energy Systems ISE developed and successfully put into operation an inverter for direct feed-in to the 10 kV medium-voltage grid. The inverter contains high-voltage silicon carbide (SiC) transistors which allow for coupling to the medium voltage grid without requiring an additional transformer. The three-phase inverter can be used to regulate reactive power as well as to actively filter undesirable harmonics in the electricity grid. Thus, it actively contributes to the stabilization of future power grids with a large share of renewables. According to the current state of technol-ogy, power electronics are coupled to the electricity grid mainly in the low voltage range. For grid stabilization, power converters, so-called STATCOMs (Static Synchronous Compensators), are used to supply continuous inductive or capacitive reactive power. The coupling to the medium voltage grid is affected by means of a 50 Hz transformer. The newly developed inverter from Fraunhofer ISE can now feed directly into the medium voltage grid without a transformer, due to the use of high voltage transistors made of silicon carbide (SiC). First compo-nent prototypes with a blocking voltage of 15 kV were used for this purpose. www.ise.fraunhofer.de
High-Voltage Silicon Carbide Inverter Enables Stabilization of Medium-Voltage Grids
Three billion people around the world live in severe energy poverty, including 1.1 billion who live completely off grid. Providing affordable energy access to them can dramatically impact their living standard, health, education, productivity and ability to be a part of modern society.
Many programs and initiatives have been doing stellar work in tackling energy poverty, but much remains to be done. Solving energy poverty using mature and proven 20th century solutions, an obvious path
forward, could result in an additional 3.8 gigatons per year of carbone-missions an environmental catastrophe! New solutions that can scale are clearly needed! IEEE Empower a Billion Lives (EBL) is a recurring global competition organized by the IEEE Power Electronics Society, to crowdsource regionally relevant innovation to accelerate deployment ofenergy access solutions in the affected areas. It is anticipated that fast moving 21st century technologies with rapidly declining prices can allow a holistic approach to the design of energy solutions to ad-dress the needs of families and communities living completely off grid or suffering from a severe lack of energy access. Effective solutions should be economically viable today, and should be able to provide continuing value to the families and communities as they improve their lives.A primary focus for EBL is to help develop new energy access solu-tions with reduced technology and market risk. Another is to prove out new business models that show impact and scale can be achieved with solutions that are economically viable and environmentally sustainable.
www.empowerabillionlives.org
Empowering a Billion Lives with Power Electronics
Transphorm Inc. announced that Seasonic Electronics Co. uses Transphorm’s high voltage GaN FETs in its new 1600 Watt bridge-less totem-pole power factor correction (PFC) platform. The 1600T is the power supply manufacturer’s highest performing PFC platform to date at greater than 99 percent efficiency. Notably, the introduction of GaN delivers a two percent efficiency increase and 20 percent power density increase over Seasonic’s previous Silicon-based platform. The 1600T platform will be scaled and deployed in various catalog products targeting the Charger (e-scooters, industrial, etc.), Gaming, Server, and PC power markets. “When researching semiconductor technologies that would enable us to reach world-leading efficiency levels, gallium nitride stood out as an attractive alternative to Silicon,”
said Paul Lin, Director of Research and Development, Seasonic. “We knew the bridgeless totem-pole PFC was the topology we would use in our first high voltage GaN power platform. So, we needed power semiconductors capable of successfully capitalizing on that topology. What’s more, we wanted a GaN solution that could be backed by our standard warranty. We ultimately opted for Transphorm’s FETs within the 1600T given their proven performance and reliability that allowed us to meet those requirements.”
www.transphormusa.com
www.seasonic.com
GaN Platform for Broad Power Supply Portfolio
Bodo´s Power Systems® July 2018 www.bodospower.com14
CONTENT
Military-Grade Laptop PSU is 90% Smaller and Lighter,
80% Lower CostUK company, On-Systems, launches revolutionary disruptive technology; delivers new
technology platform for harsh-environment PSUs
PRODUCT OF THE MONTH
On-Systems, a young British company that specializes in innovative power supply design for harsh environments, launched Pebble, a new COTS, rugged, EMI-silent laptop power supply for defence, avionics and mobile applications that is a fraction of the size, weight and cost of current designs. Pebble is the first in a series of harsh-environment PSUs from the company which combine a new topology, custom-designed FET drivers and fast switching speed to deliver industry-changing benefits.
The new topology combines three stages into one, reducing I²R losses and increasing efficiency. On-Systems has worked with a leading power semiconductor company to design new very fast FET drivers, allowing the Pebble to switch at up to 40 times typical switch-ing frequencies. This high speed reduces the size of the EMC filtering components required, such that the laptop PSU meets all defence EMC performance standards, including MIL-STD 1275A-E, MIL-STD 704A-F, DEF-STAN 59-411 and MIL-STD 461 E-F EMC, in a package
measuring just 115 x 52 x 36mm and weighing only 180g. Pebble is 96% efficient, operates over universal AC input range and wide DC input range, features short circuit, overload and overvoltage protec-tion, and carries a four year warranty.
Explains On-Systems’ CCO Mike Harvey: “Traditional laptop power supplies that must be electrically-‘silent’ for use in secure, harsh environments are about the same size as a laptop and weigh 2-3kg, because they need so much filtering and heatsinking. Pebble, by con-trast, retains the familiar look and feel of a commercial power adaptor, yet it is fully certified for defence use. More, we can use the same technology in other applications – notably in aircraft fuel pump sys-tems – where the space and weigh saving will be of especial interest.”
www.on-systems.co.uk
Broadcom OptocouplersA Superior Technology for High Voltage Protection!
Optocouplers are the only isolation devices that meet or exceed the IEC 60747-5-5 International Safety Standard for insulation and isolation.
IEC 60747-5-5 Certifi ed
Broadcom, the pulse logo, connecting everything are among the trademarks of Broadcom. Copyright © 2017 Broadcom. All Rights Reserved. The term “Broadcom” refers to Broadcom Limited and/or its subsidiaries. For more information, please visit broadcom.com.
Stringent evaluation tests show Broadcom optocouplers deliver outstanding performance on essential safety and deliver exceptional High Voltage protection for your equipment. Alternative isolation technologies such as magnetic or capacitive isolators do not deliver anywhere near the high voltage insulation protection or noise isolation capabilities that optocouplers deliver.
IPD-IEC-Superior Technology 2-190.5mmx266.7mm_051817.indd 1 5/22/17 10:37 AM
Bodo´s Power Systems® July 2018 www.bodospower.com16
CONTENT
The 30th Annual International Symposium on Power Semiconductor Devices and Integrated Circuits, (ISPSD) was held from May 13-17, 2018 in Chicago, Illinois, USA. ISPSD, which is sponsored by the IEEE Electron Devices Society (EDS), and technically co-sponsored by the IEEE Power Electronics Society (PELS), Industrial Applications Society (IAS), and Institute of Electrical Engineers of Japan (IEEJ) is considered the world’s premier forum for scientific and technical discussions in all areas of power semiconductor devices, power inte-grated circuits and packaging.
This year’s conference brought together over 450 attendees from around the world as well as many exhibitors from fields related to power semiconductors including test equipment, design software and wafer foundries. Each year the ISPSD host site rotates between Europe, Asia, North America, and a developing country. This year’s venue was the historic Palmer House Hilton Hotel located in the well-known loop area of downtown Chicago within walking distance of numerous shops, museums and other points of interest.
2018 marked the 30th anniversary of the ISPSD. Conference Gen-eral Chairman, Professor John Shen of Illinois Institute of Technology stated that since the first meeting in 1988 in Tokyo, Japan, ISPSD has been instrumental in the growth of the global power semiconductor industry. He noted that more than 1600 papers have been presented over the past three decades and that most of the breakthrough power device technologies were first reported at ISPSD before they were introduced commercially. These breakthroughs enabled the global power semiconductor industry to grow into a $30 billion market, facili-
tating applications such as solar power, wind power, electric vehicles, lighting and industrial drives. As part of the anniversary celebration, 32 distinguished members of the power device community were inaugurated into a newly established ISPSD Hall of Fame during the Wednesday 30th Anniversary Celebration Banquet.
The conference technical program consisted of 17 technical sessions devoted to both silicon and wide-bandgap discrete power devices, power IC’s, packaging technologies and novel device structures. The sessions were made up of 50 oral and 79 poster presentations selected from 245 abstracts submitted from 23 different countries.
The conference began on Sunday, May 13 with a full day technical short course taught by various experts in the field. The short course was designed to educate both students and working professionals on topics of current interest as well as introducing emerging tech-nologies. This year’s short course topics included Advanced Silicon Technologies, Loss Mechanisms in Silicon and Wide-bandgap Power Devices, Silicon Carbide Device Design and Applications, Vertical Power Electronics Based on GaN, AlGaN/GaN Device Reliability, and Multi-Chip Power Module Design.
The technical sessions opened on Monday May 14 with four invited plenary presentations. The first by M. A. Shibib of Vishay Siliconix, L. Lorenz of ECPE, and H. Ohashi of NEPRC was titled “ISPSD-A 30 Year Journey Advancing Power Semiconductor Technology” [1]. The paper reviewed the history of ISPSD and highlighted the most impor-tant contributions that the conference has made to the power semi-conductor field. A second, by L. Spaziani of GaN Systems was titled “Silicon, GaN, and SiC—There is Room for All, An Application Space Overview of Device Considerations” [2]. The presentation consid-ered the general systems priorities such as power density, efficiency and cost for several key applications areas for power semiconductor devices and then weighed these against the properties of the three major materials technologies. While the wide-bandgap technologies are making steady progress the need for lower cost, wafer availabil-ity and proven reliability were identified as concerns. N. Machida of Sumco addressed “Si Wafer Technology for Power Devices: A Review and Future Directions” in the third presentation [3]. He stated that the future direction is in the increased utilization of the magnetic field applied Czochralski (MCZ) crystal growing method as well expanding the production of 300 mm diameter wafers for power applications. The final plenary presentation was given by Bert De Colvenaer, ECSEL JU and addressed “The Future of Power Semiconductors: An EU Perspective” [4] He noted that Europe and the ECSEL JU are
GUEST EDITORIAL
International Symposium on Power Semiconductor Devices
and ICs—ISPSD 2018ISPSD brings together technical professionals in all areas of power
semiconductor devices and power integrated circuits By Gary M. Dolny, Bodo’s Power Systems, [email protected]
Newly inducted members of the ISPSD Hall of Fame
Digital Power Designs Made EasierProducts, Tools, Software and Reference Designs
Microchip’s digital power design suite includes the Digital Compensation Design Tool (DCDT), MPLAB® Code Configurator (MCC), Microchip compensator libraries and design examples.
These four components of the digital power design suite provide the tools and required guidance for developing complete digital power designs. Once the initial simulation model of your design is ready, the DCDT can be used to analyze the design and the feedback transfer function, and to generate compensator coefficients. Device initialization code can be generated with the help of MCC; and the final firmware can be created with some help from the code examples and the code generated from MCC and the DCDT.
Key Features Digital Compensation Design Tool to analyze your design Libraries and design examples to jump start your development Feature-rich dsPIC33EP “GS” family of DSCs
The Microchip name and logo, the Microchip logo and MPLAB are registered trademarks of Microchip Technology Incorporated in the U.S.A. and other countries. All other trademarks are the property of their registered owners. © 2017 Microchip Technology Inc. All rights reserved. DS00002536A. MEC2199Eng12/17
www.microchip.com/DDSMCU16
Bodo´s Power Systems® July 2018 www.bodospower.com18
CONTENT
investing in power semiconductors as a key enabling technology for achieving EU goals for increased renewable energy usage. He stated that success hinges on regulation and standards, technology avail-ability, reliability, seamless integration, and acceptance by the users.
A number of conference presentations highlighted the steady progress being made in silicon carbide power MOSFET technology. Chowdhury et. al. from Monolith Semiconductor presented a 1200V SiC MOSFET with an improved tradeoff between on-resistance and reverse bias gate oxide electric field [5]. The improved tradeoff was obtained by optimizing the JFET doping profile and unit cell design. These MOSFETs showed a specific on-resistance of 3.5 mΩ-cm² at room temperature, increasing to 5.9 mΩ-cm² at 175°C. The device also exhibited excellent High Temperature Reverse Bias (HTRB) reliability as shown by no failures after stressing at 1440V, 175°C for 1000 hours.
Han et. al. from North Carolina State University presented a novel 1.2 kV Buffer Gate SiC-NMOS structure [6]. In this device the gate polysilicon is removed from above the JFET region, similar to a conven-tional split-gate structure but in addition, the buried p-layer is extended beyond the gate edge to further minimize gate to drain charge. The devices were formed with both accumulation and inversion mode channels. Experimental data confirmed the accumulation mode Buried-Gate MOSFET has reduced figures of merit Ron×Cgd and Ron×Qgd compared with both the conventional accumulation mode SiC-MOSFET and split-gate SiC-MOSFET, due to the significantly improved Cgd and Qgd. The improvement was in the range of 2.6x to 4x.
Yen et. al. of Hestia Power Corp. demonstrated a monolithic SiC junction barrier controlled Schottky diode (JBS) integrated MOSFET (JMOS) using an identical process flow as a standard SiC double implanted MOSFET (DMOS) without area penalty [7]. To incorporate a good Schottky contact into the MOSFET cell, the Schottky open-ings were formed together with the openings to the gate, after the ohmic contacts were formed in the source(body) openings on the n+/p+ regions. A standard Ti/TiN/AlCu metal stack was used to form gate contacts for the gate electrode and Schottky contacts and ohmic contacts. The device provides similar on-resistance and drain-source breakdown voltage with the same chip size as the standard double-implanted MOSFET (DMOS). The reverse recovery charge at 650V and 1200V at 150°C were 22% and 53% lower than the correspond-ing DMOS while the peak reverse recovery current of the 650V and 1200V JMOS were 26% and 40% lower than the corresponding DMOS. The device was demonstrated to be reliable via forward cur-rent stress, surge current test, and 1000 hours HTRB.
The threshold voltage hysteresis effect in SiC MOSFETS received considerable attention. This effect is observed as a negative shift in the Id-Vg transfer characteristics that becomes more pronounced as the off-state gate voltage becomes more negative. The effect is attrib-uted to interface states in the bandgap which can become positively charged when a negative gate voltage is applied but are neutralized under positive gate bias. The positive charges are in a direction to enhance channel inversion, thus the onset of drain current occurs at a lower value of Vgs. Peters et. al. [8] addressed this phenomenon, and concluded that the effect is fully reversible, harmless for turn-on and almost irrelevant for turn-off. They also performed Bias Tempera-ture Instability tests and showed that drift is predictable and within data sheet limits. Unger et.al. [9] of Technische Universität Dortmund, performed an application-oriented investigation of the impact of nega-tive off-state gate-source voltages on 600V class SiC MOSFETs. They observed that as the off-state gate voltage becomes more negative,
a more pronounced drain current overshoot immediately after turn-on is observed. The overshoot decays over time as the channel current neutralizes the traps. The phenomenon especially affects the accu-rate determination of the threshold voltage and transfer characteristics but can also lead to a reduced short-circuit withstand time depending on the circuit operation.
SiC JBS rectifiers received significant interest despite the fact that commercial products have been available for several years. Three presentations addressed issues related to surge current reliability. Van Brunt et. al. [10] from Wolfspeed presented a study of surge current failure mechanisms in 4H-SiC JBS rectifiers. They noted that the transition between the normal Schottky operating mode, and the bipolar mode, in which the p-n junction forward biases, is a function of the dynamic heating that occurs during the surge environment and not just the electrical bias conditions. For non-repetitive transients the energy-to-fail at 10 µs is over 10 times larger than the energy which causes a failure at 10 ms due to heat diffusion into the package. Palanisamy et. al. [11] of Technische Universität Chemnitz and Infi-neon studied repetitive surge currents in state of the art 650V, 1200V and 1700V SiC MPS diodes operating in the bipolar regime with 10 µs pulses typically 20% lower in magnitude than the destructive limit for single-event pulses. They observed that most devices could withstand a large number of repetitive pulses, typically over 1000. Observed failure mechanisms included a change in the Schottky bar-rier or power metal, chip cracks at the middle of the die, partial bond wire lift-off and front side metallization delamination. Xu et. al. [12] of Zhejiang University studied the surge capability of 1.2 kV SiC JBS di-odes in which the p+ region is formed with a 500°C implantation step. Their surge current experiments verified that an improved capability is achieved in the JBS diodes with high-temperature implantation.
Advances in GaN technology were highlighted in three dedicated con-ference sessions. Posthuma et. al. [13] of IMEC and On Semicon-ductor described an industry ready 200 mm GaN-on-Si technology. The process features 650 V rated enhancement mode p-GaN gate HEMTS that were fabricated with an Au-free process. Source and drain ohmic contacts are realized by recess etching, cleaning, Ti/TiN/Al based metallization and low-temperature ohmic anneal. The thresh-old voltage was 2.8 V with an off-state leakage of less than 1µA/mm at room temperature. Large area power devices with gate width up to 36 mm were demonstrated. For the optimum field plate design no key device parameters including dynamic Ron shifted by more than 10% after 1008 hours of HTRB testing at 80% of rated voltage. Tajalli [14] et. al. of Univ. of Padova, ON Semiconductor, and CMST IMEC/Ghent University discussed the use of proton implantation for the control of dynamic Ron in AlGaN/GaN transistors. They attributed the improve-ment to a small increase in leakage current in the unintentionally-doped GaN layer which facilitates charge de-trapping from the buffer layer. Proton irradiation at fluences of 1.5 x 1014 cm-² with energy of 3 MeV were shown to completely suppress dynamic Ron over the entire voltage range and up to 150°C without significant modification to other device parameters. The tested devices had gate width of 200 mm and were designed for 650 V operation.
Riccio et. al. [15] of University of Naples and Cambridge University analyzed the short-circuit robustness of new generation p-GaN HEMTs. Their tests were performed on commercial devices with rated breakdown voltage of 650 V and on-state resistances of 200 mΩ and 50 mΩ. They observed a large increase in gate current, from a 40 µA typical value at room temperature to up to 9 mA during short circuit testing. They attributed this increase to a fast temperature rise within the device during the short-circuit pulse. This current imposes
GUEST EDITORIAL
www.bodospower.com July 2018 Bodo´s Power Systems® 19
CONTENTGUEST EDITORIAL
a voltage drop across the gate resistance which results in a reduction of the effective Vgs at the gate terminal. The combination of mobility reduction as a result of self-heating along with the effective decrease of gate drive voltage results in a drastic reduction of the drain current, thus enhancing the short-circuit capability of the device. Efthymiou et. al. [16] of Cambridge University and Vishay Siliconix investigated the effect of layout on the switching of enhancement mode, 650 V, 15 A, p-GaN gate HEMTs. Three layouts with variation in the placement of the contact pads for the different terminals (drain, source, gate) and the metallization tracks which connect the fingers to the pads were studied. The different designs were found to exhibit varying degrees of susceptibility to oscillatory behavior during high dI/dT switching. A design with peripheral drain pads and an interior source contact exhibited the highest immunity to oscillations, but at the penalty of increased on-state resistance. The oscillatory behavior was found to be related to the level of source inductance and the unbalances of these inductances within the die.
Smart-power gate driver integrated circuits for use with GaN power devices were discussed in several presentations. Tang et. al. [17] of the Hong Kong University of Science and Technology and Taiwan Semiconductor Manufacturing Company demonstrated a 650-V enhancement-mode GaN power switch with a monolithically integrat-ed gate driver fabricated on a commercial GaN-on-Si power device platform. The low-voltage logic and control circuits are fabricated in enhancement/depletion HEMTS and require no additional process steps. This monolithic integration of the gate driver circuit minimizes parasitic inductance in the gate loop and thus suppresses ringing and alleviates overshoot. This enables high switching speed in the GaN power transistor enabling reduced switching losses. Yu et. al. [18] of the University of Toronto presented an integrated smart gate driver IC for GaN power transistors. The chip featured a segmented output stage topology, programmable sense-FET, current sensing circuits and an on-chip stacked-based CPU for flexible digital control. The circuit was fabricated on a commercial 0.18 µm BCD technology and was targeted for driving depletion mode GaN HEMTS in a cascode configuration. The circuit dynamically adjusts the gate driving strength to achieve slope control during switching, monitors the load current using a built-in sense-FET, provides peak-current regulation and actively adapts the driving pattern in less than 1 µs to optimally match the load conditions.
The full conference proceedings will be available to IEEE Members through the IEEE Explore website.
ISPSD 2019 will be held from May 19-23, 2019 in Shanghai, China, one of the country’s most dynamic, vibrant, and diverse cities. This will be the first time in its 31-year history the conference will be held in mainland China. Abstract submission deadline in November 12, 2018. Topics of interest include, but are not limited to High Voltage Power Devices, Low Voltage Power Devices, SiC, GaN, and other Wide-Bandgap Power Devices, Power ICs, and Module and Packaging Technology. More information is available at the conference website www.ispsd2019.com.
References:[1] A. Shibib, L. Lorenz, H. Ohashi, “ISPSD: 30 year journey in
advancing power semiconductor technology”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 1-7.
[2] L. Spaziani, L. Lu, “Silicon, GaN and SiC: there’s room for all”, in Proceedings of 30th ISPSD, Chicago Ill., USA, May, 2018, pp 8-11.
[3] N. Machida, “Si wafer technology for power devices-A review and future directions”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 12-14.
[4] B. De Colvenaer, “The Future of Power Semiconductors: an EU Perspective”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 15-23.
[5] S. Chowdhury, K. Matocha, B. Powell, G. Shieh, S. Banerjee, “Next generation 1200V, 3.5mΩ.cm2 SiC planar gate MOSFET with excellent HTRB reliability”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 427-430.
[6] K. Han, B. J. Baliga, W. Sung, “Accumulation channel vs. inver-sion channel 1.2 kV rated 4H-SiC buffered-gate (BG) MOSFETs: analysis and experimental results”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 371-374.
[7] C.-T. Yen, F.-J. Hsu, C.-C. Hung, C.-Y. Lee, L.-S. Lee, Y.-F. Li , K.-T. Chu, “Avalanche ruggedness and reverse-bias reliability of SiC MOSFET with integrated junction barrier controlled Schottky rectifier”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 56-59.
[8] D. Peters, T. Aichinger, T. Basler,G. Rescher, K. Puschkarsky, H. Reisinger, “Investigation of threshold voltage stability of SiC MOSFETs”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 40-43.
[9] C. Unger, M. Pfost, “Influence of the off-state gate-source voltage on the transient drain current response of SiC MOSFETs”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 48-51.
[10] E. Van Brunt, T. Barbieri, A. Barkley, J. Solovey, J. Richmond, B. Hull, “Surge current failure mechanisms in 4H-SiC JBS rectifiers”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 415-418.
[11] S. Palanisamy,, J. Kowalsky, J. Lutza, T. Basler, R. Rupp, J. M. Fallah, “Repetitive surge current test of SiC MPS diode with load in bipolar regime”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 367-370.
[12] H. Xu, J. Sun, J. Cui, J. Wu, H. Wang, S. Yang, N. Ren, K.Sheng, “Surge capability of 1.2kV SiC diodes with high-temperature implantation”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 419-422.
[13] N.E. Posthuma, S. You, S. Stoffels, D. Wellekens, H.Liang, M. Zhao, B. De Jaeger, K. Geens, N. Ronchi, S. Decoutere, P. Moens, A. Banerjee, H. Ziad, M. Tack, “An industry-ready 200 mm p-GaN E-mode GaN-on-Si power technology”, in Proceed-ings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 284-287.
[14] A.Tajalli, A. Stockman, M. Meneghini, S. Mouhoubi, A. Banerjee, S. Gerardin, M. Bagatin, A. Paccagnella,E. Zanoni, M. Tack, B. Bakeroot, P. Moensand G. Meneghesso, “Dynamic-Ron control via proton irradiation in AlGaN/GaN transistors”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 92-95.
[15] M. Riccio, G. Romano, L. Maresca, G. Breglio, A. Irace, G. Longobardi “Short circuit robustness analysis of new generation enhancement-mode p-GaN power HEMTs”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 104-107.
[16] L. Efthymiou, G. Camuso,G. Longobardi . F. Udrea, T. Chien, M. Chen, A. Shibib, K. Terrill “ Effect of device layout on the switch-ing of enhancement mode GaN HEMTs”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 220-223.
[17] G. Tang., M.-H. Kwan, Z. Zhang., J. He, Ji. Lei, R.-Y. Su, F.-W. Yao, Y.-M. Lin, J.-L. Yu, Th. Yang, C-H Chern, T Tsai, H. C. Tuan, Alexander Kalnitsky, K. J. Chen, “High-Speed, High-Reliability GaN Power Device with integrated gate driver”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 76-79.
[18] J. Yu, W. J. Zhang, A. Shorten, R. Li, W. T. Ng, “A smart gate driver IC for GaN power transistors”, in Proceedings of 30th ISPSD, Chicago, Ill., USA, May, 2018, pp 84-87.
www.ispsd2019.com
Bodo´s Power Systems® July 2018 www.bodospower.com20
CONTENT
My over-arching comment, having thoroughly read this amazing book, is that I cannot improve on the authors own assessment, but I will offer an independent critique below. The book was years in prepara-tion and reflects the authors’ own wealth of industrial and academic experience. First, the authors’ prologue.
This is primarily an engineering textbook covering the automotive powertrain, energy storage and energy conversion, power electronics and electrical machines. A significant additional focus is placed on the engineering design, the energy for transportation, and the related environmental motivations and impacts. The book is both an inte-grated and holistic university teaching textbook, from undergraduate to introductory postgraduate levels, and an educational and research reference for our industry colleagues.
This textbook is an educational tool for practicing engineers and others, such as transportation policy planners and regulators. This material is also written to be of interest to the general scientific reader, who may have little or no interest in the power electronics and machines. The general reader can read the introduction to each of the chapters and move to the next as soon as the material goes beyond the reader’s interest.
While the technical topics are diverse, the technical content and teaching paths are narrowed and focused by concentrating on the existing and emerging technologies for the technology evolution underway in the automotive industry. The focus is on cars, trucks and buses. However, the amazing electric vehicle Mars rovers are also considered, especially the Opportunity, a rover which finished a mara-thon on Mars in 2015 after 11 years (and which featured in the movie and book The Martian).
The textbook results from almost three decades of collaboration between Dr. John Hayes of University College Cork, Ireland, and Dr. Abas Goodarzi of US Hybrid in California. John and Abas previously worked together on the General Motors’ electric vehicle, the GM EV1, the first production electric car of the modern era.
It has been almost three decades since General Motors (GM) displayed the Impact concept car at the Los Angeles Auto Show in January 1990. The Impact inspired the design of the GM EV1 and for California to subsequently issue its zero emissions mandate. The
EV1 became available for leasing in 1996 and demonstrated to the public that electric cars were feasible. It would require very consider-able education for the electric car to go mainstream, and in fact, it still does.
The automobile is a complex machine that has been continuously re-fined ever since Henry Ford rolled out the first Model T in 1908 as an affordable mass-produced vehicle. By the 1990’s the motoring public
Book Review: “Electric Powertrain”
Energy Systems, Power Electronics and Drives for Hybrid, Electric and Fuel Cell Vehicles
Empowers engineering professionals and students with the knowledge and skills required to engineer electric vehicle powertrain architectures, energy storage systems, power electronics converters and electric drives.
By Dr. John M. Miller, Industry Consultant
GUEST EDITORIAL
www.bodospower.com July 2018 Bodo´s Power Systems® 21
CONTENT
was accustomed to the automobile’s refined propulsion and highly tuned ride and handling. Replacing the internal combustion engine (ICE) and gasoline tank with an electric traction drive and relatively massive battery pack constituted a quantum jump in engineering for manufacturers to just stay on par.
Dr. Hayes and Dr. Goodarzi have focused their considerable talent and experience on teaching the inner workings of the electric car. Readers, whether engineers, students, or the interested public, will find this book a treasure trove of knowledge on modern automotive technology. The book is divided into four parts in 16 chapters, start-ing with Part I Vehicles and Energy Sources. This section is actually a standalone, concise description of what it means to say a more highly electrified automobile. Here the reader is taken on a path that traverses our present dependence on fossil fuels and CO2 emissions of the ICE into the world of fuel cells and battery energy storage as the electric fuel for the traction motor. What is most refreshing in all the chapters of this book are the worked examples, plus the exercises that delve more deeply into each topic. But not just the fact that the examples are relevant to each section but that specific examples and problems get further refined as the reader moves into each new sec-tion of the book.
Part I of the book has set the stage that hybrid, fuel cell, and battery electric vehicles must match or exceed the expectations of modern drivers. Part II Electric Machines, is a painstaking deep dive into the four main players in electric traction motors: brushed dc, induction (asynchronous), surface permanent magnet, and interior permanent magnet types. Engineers familiar with electric machines will find these chapters an ideal refresher, students will find the topics fascinating, and the informed lay person will garner a deeper appreciation of how that stored electric fuel gets converted into mechanical energy.
Part III Power Electronics, is a more esoteric topic as many readers wishing a deeper understanding of the electric car will view these chapters as really getting “under the hood”. This is the world of power converters, inverters, and how they are controlled. Most will recognize these 5 chapters as a real strong suite of the authors because of their ability to describe with such clarity the dc transformer and ac synthe-sizer. The reader will be led by an expert hand through switch mode power conversion and be shown the intricacies of the complementary behavior of active (transistor) and passive (diode) switching devices. Examples again will reinforce their knowledge of not only computing average and rms quantities of signals but of realistic power dissipation and efficiency computations. All extremely beneficial to the practicing design engineer and fundamental to understanding by the student and inquiring reader. My favorite chapter was on battery charging because of the need to not only understand the role of power factor correction (PFC) to meet utility grid power quality regulations but also to minimize ripple exposure of the battery. Some will need to dust off their control theory as the examples in chapter 15 get into tuning the inner current/torque loop and what can be a finicky outer voltage/speed loop for the electric vehicle drivetrain.
Part IV Basics, is another standalone section but this time in a single chapter on electromagnetism and energy conversion topics that are in fact fundamental; basic to understanding not only electric machines but the electrified automobile in general. My recommendation is that readers may find some topics in parts II and III more understandable by consulting chapter 16 as needed. Others may actually wish to read over chapter 16 after Part I and before getting into electric traction drives.
My overall impression is that this book was well worth the investment, and one which I’ll proudly display on my shelve (and also available as PDF). For those wishing to dig even deeper, or for appreciation of sources, the authors provide excellent references after each chapter. In conclusion, what a great book!
John M. Miller ([email protected]) is owner and founder of J-N-J Miller Design Services PLLC, which was established in 2002 to provide profes-sional consulting to industry. Dr. Miller has over 42 years of experience in electrical engineering across various industries that include automotive, aerospace, white goods, and electri-cal practice. He joined Momentum Dynamics Technical Advisory Board in 2014 as senior scientist working on
wireless power transfer for heavy-duty vehicles. In 2018 he was invited to Ashok Leyland’s Advisory Panel and he is an industry consultant on electric traction drive and energy storage systems for an electric vehicle manufacturer. His previous work experience includes Distinguished R&D Scientist at Oak Ridge National Labo-ratory (ORNL) where he held positions as Director of the Power Electronics and Electric Power Systems Research Center and Program Manager of the DOE Vehicular Technologies subprogram APEEM. Previously, Dr. Miller held various senior management and engineering positions at Maxwell Technologies, Ford Motor Company, and Texas Instruments. He has published several books related to wireless charging, ultracapacitor applications (translated to Chinese 2015), propulsion systems for hybrid vehicles (translat-ed to Chinese 2016), automotive power electronics, and vehicular electric power systems. He holds a B.S.E.E. from the University of Arkansas-Fayetteville, a M.S.E.E. from Southern Methodist University, Dallas, TX, and a Ph.D. from Michigan State University, East Lansing, MI. He is also an Adjunct Professor in ECE at the University of Texas at Dallas (UTD). Dr. Miller is a Life Fellow of the IEEE, a Fellow of the SAE, and a registered professional engineer in Michigan (1980) and in Texas (2014). Dr. Miller holds 60 U.S. patents.
www.wiley.com/WileyCDA/WileyTitle/productCd-1119063647.html
www.wiley.com
GUEST EDITORIAL
Bodo´s Power Systems® July 2018 www.bodospower.com22
CONTENT
The new findings have potentially important implications for “thermal crosstalk,” in which multiple heat sources next to each other impact the overall temperature of the system, hindering performance. The researchers used a technique called full-field thermoreflectance ther-mal imaging to directly visualize temperature changes produced by ultra-small heat sources, gold strips formed on top of the semiconduc-tor indium gallium arsenide.
The research concerns the crucial role of phonons, quantum-mechan-ical objects, or “quasiparticles,” that describe how vibrations travel through a material’s crystal structure. The phonons are said to be “heat carriers” in solid materials.
“This is the first time such hydrodynamic effects are indirectly ob-served for heat propagation in a solid,” said Ali Shakouri, Purdue Uni-versity’s Director of the Birck Nanotechnology Center and a professor of electrical and computer engineering.
“While structures called vortices are common in fluid flows such as water or air, this is the first time we’ve seen that they can be pres-ent inside solids for phonon flow in the typical semiconductor indium gallium arsenide, which is used in high-speed transistors and lasers.” Findings are detailed in a research paper appearing in “Nature Com-munications”.
“The observed thermal crosstalk reduction has important implications in the design of nanoscale electronic and optoelectronic devices,” said Purdue postdoctoral research associate Amirkoushyar Ziabari, the paper’s lead author. “As the size of electronic and optoelectronic
devices are getting smaller, there are more and more devices being packed into a smaller area, so the thermal crosstalk between these devices becomes important. Knowing the accurate thermal behavior in the neighborhood and a few microns from heat sources would help design better state-of-the-art devices in terms of performance, speed, thermal reliability, and so on.”
The researchers found that the reduced thermal crosstalk is caused by vortices generated near the edge of the heat sources. “This is similar to the vortices that are observed at the edge of an obstacle placed inside of a current of air or water, such as behind an airplane wing,” Shakouri said.
The governing law of heat conduction, known as the Fourier Law or the heat-diffusion equation, does not accurately predict thermal transport for devices on the nanoscale. Because the Fourier diffusion equation doesn’t explain the heat transport at those scales, this trans-port regime is called non-diffusive.
“As the size of electronic and optoelectronic devices are getting smaller, it is important to consider this non-diffusive behavior for design and optimization of such small devices,” Ziabari said. “These new measurements show that at nanoscales, heat propagation has interesting ‘fluid-like’ behavior.”
Conventional methods do not account for vortices of heat transport found at the nanoscale. “Vorticity only becomes important when the characteristic source dimension is comparable to the hydrodynamic length scale of about 150 nanometers,” he said. The Fourier theory substantially overestimates the experimentally observed temperature a short distance away from the heater lines.
“The surprising effect was that the temperature decays much faster than what Fourier theory predicted,” Shakouri said. “Within a distance of 1 or 2 microns of a small heat source - a line about 100 nanome-ters wide - temperature could be one-third to one-fourth what Fourier theory predicts.”
The thermoreflectance thermal imaging approach allows research-ers to create maps of temperature rise at far higher resolution than otherwise possible using light in the visible range.
The work was performed by researchers at the Birck Nanotechnology Center in Purdue’s Discovery Park, Purdue’s School of Electrical and Computer Engineering, Universitat Autónoma de Barcelona, Commis-sariat à l’Énergie Atomique in Grenoble, France, and the Department of Mechanical and Materials Engineering at the University of Cincin-nati. The research was partially funded by the European Commission and the Spanish Ministry of Economy and Competitiveness.
www.purdue.edu
GUEST EDITORIAL
Secrets of ‘Fluid-Like’ Heat Flow in Solid Semiconductor at Nanoscale
Researchers are applying the same “hydrodynamic transport model” used to study flow in fluids to explain heat transport in a solid semiconductor, with potential implications for the design of high-speed transistors and lasers. Thermal imaging of tiny nanoscale
semiconductor heat sources revealed details about vortices of heat-carrying objects called phonons.
By Henning Wriedt, US-Correspondent Bodo‘s Power Systems
Figure 1: Researchers have visualized temperature changes produced by ultra-small heat sources, gold strips formed on top of the semiconductor indium gallium arsenide. The work has potential implications for the design of high-speed transistors and lasers. This image (a) depicts the device structure and experimental setup, an optical image (b) of the fabricated device and (c) an experimental thermal image.
Widest selection of electronic components
ORDER CONFIDENCEWITH
Authorised Distributor of semiconductors and electronic components
mouser.com
C
M
Y
CM
MY
CY
CMY
K
648036-Selection in Stock-A4.pdf 1 16.05.18 18:13
Bodo´s Power Systems® July 2018 www.bodospower.com24
CONTENT
This article focuses on presenting the benefits of a newly developed SiC power module and the utilization of this module in a power train inverter. A comparison of the performance of the SiC based inverter with an Si based traction inverter is also shown in this article.
In motor traction inverter application requirements like space, weight and efficiency play an increasing role. Product development and man-ufacturing expenses should remain low while the design efforts should result in more compact systems and, at the same time, product quality and reliability should be guaranteed. This leads to more demanding design requirements on the system and component level and ulti-mately affects the overall system consisting of power devices, passive components, cooling technologies and PCBs. In order to achieve the required enhanced system properties, semiconductor devices have to enable higher power density, higher efficiency and reliability.
A simplified block diagram of a traction power train is shown in Figure 1. Use of SiC power devices in the power train inverter leads to an increase of the entire power train system efficiency and helps to miniaturize the system thanks to improved switching losses, conduc-tion losses, and thermal conductivity. In consequence it is expected to gain longer driving range with the same battery capacity or a substan-tial reduction of battery size and weight for the same driving range. In this way, the efficiency and weight improvements bring an economic benefit to the users.
Comparison between IGBT Module and SiC ModuleThe new Gtype package module BSM600D12P3G001 from Rohm is used in this application (figure 2). It is a half bridge module which utilizes the device technology of SiC Trench Gate MOSFET [1] [2] and
Schottky Barrier Diode (SBD). 10 chips of SiC MOSFET are used in parallel per switch. The rated DC drain current is 600A at Tc=50°C. In the next diagrams a comparison between the Gtype module and two market available latest generation IGBT modules with the same rated current is shown.
Figure 3 is showing a comparison of output characteristics at a junc-tion temperature of 150°C. In terms of conduction loss the curves for SiC MOSFET and IGBT modules cross at around half of the rated current.
Automotive Traction Inverter Utilizing SiC Technology
In 2010, the first Silicon Carbide (SiC) power transistor was launched to the market. This initiated a revolution in the power electronic field and is leading to an increasing rate of SiC devices adoption in many industrial and automotive application fields which proved that SiC components are strong competitors to classical Si devices in such applications. Electric Vehicle (EV) applications represent a big potential area where the adoption of
high performance SiC power devices can provide benefits.
By Aly Mashaly and Masaharu Nakanishi, ROHM Semiconductor
COVER STORY
Figure 1: Simplified EV Traction inverter system
Figure 3: Output characteristic comparison between Gtype module and two IGBT modules with same rated currents
Figure 2: The new SiC module package Gtype
THE ULTIMATE POWER SOURCE
ANALOG DEVICES’ POWER MANAGEMENT
PORTFOLIO ENABLES YOU TO TACKLE YOUR TOUGHEST
POWER CHALLENGES.
Now you can get to market faster with everything you need to
deliver breakthrough solutions that hold a true competitive edge.
Power by LinearTM gives you:
Industry-leading performance
Proven, trusted reliability
Best-in-class quality
No-obsolescence policy
Premier service and support
EXPLORE THE COMBINED POWER PORTFOLIO OF ADI AND LTC analog.com/power#PowerByLinear
Bodo´s Power Systems® July 2018 www.bodospower.com26
CONTENT
A comparison of switching losses also at a junction temperature of 150°C is presented in figure 4. As can be seen, the total switching losses of the Gtype module with SiC devices are ca. 65% lower than for the IGBT modules at the rated current of 600A. Due to the lower switching losses the SiC module can be operated at much higher switching frequency.
Traction InverterFigure 5 shows two traction inverters. The left hand inverter is rated at 200kW and utilizes power modules with Si IGBTs and Si Fast Re-covery Diodes (FRDs). This inverter is working in the field since 2013. The right hand inverter was newly developed utilizing full SiC modules and is rated for 220kW. In addition to using SiC Trench Gate MOS-FETs and SiC SBDs a better concept of motor control strategy, an efficient cooling system, a low inductive bus bar design and a smaller DC link capacitor are successfully implemented. Both inverters are water cooled systems and both designs can be used with up to 800V battery systems [3].
Table 1 shows a specification comparison of both inverters. The new inverter utilizes variable switching frequency which can be varied depending on the motor operating point from 16kHz to 24kHz. Even in the lower rotor shaft speed range, the inverter utilizes 16kHz switch-ing frequency to avoid the risk of resonance between the DC link ca-pacitor and input cables. While the available maximum output power is 20kW higher the SiC MOSFET based inverter has lower volume and less weight than the Si IGBT based one. The weight is reduced by ca. 6kg and the volume is shrunk by 30% as well. These factors lead to realize a high power density inverter of 22kW/L, which is 57% higher than the conventional IGBT based solution.
InverterSi-IGBT
based
SiC-MOSFET
based
Output Power 200kW 220kW
Inverter Efficiency 1* 98.0% 99.1%
Inverter Efficiency 2** 96.9% 98.2%
Max SW Frequency 16kHz 24kHz
Weight 15.0kg 9.1kg
Volume 14.3L 10.0L
Power Density 14kW/L 22kW/L* Max efficiency ** Max Power Operation
Table 1: Comparison of inverters parameters
Performance of the inverters (Efficiency and Losses)The two Figures (6, 7) describe the inverter efficiency and power transistors losses over the rotor shaft speed and rotor torque in all operating points. For this efficiency and losses comparison, the operating profile based on the same motor design has been used to simply compare the performance.
COVER STORY
Figure 4: Switching Losses comparison between Gtype module and two IGBT modules with same rated currents
Figure 5: Two power train inverters (Si IGBT based 200kW inverter and SiC MOSFET based 220kW inverter)
Figure 6: Comparison of the inverter efficiency of both technologies IGBT and SiC
electrical engineering software
THE SIMULATION SOFTWARE PREFERRED BY POWER ELECTRONICS ENGINEERS
MODELING DOMAINS ElectricalControlThermalMagneticMechanical
KEY FEATURES Fast simulation of complex systemsCode generationFrequency analysisAvailable as standalone program or Simulink blockset
Get a free test license
www.plexim.com/trial
cara
binb
ackh
aus.
com
Bodo´s Power Systems® July 2018 www.bodospower.com28
CONTENTCOVER STORY
Figure 7 is describing power transistors losses per switch for both inverters. At the operating range with high torque, the losses of each SiC MOSFET (per switch losses) are more than 400W lower than for each Si IGBT. Thus, considering that most of the losses occur in the active switches, the SiC MOSFET based inverter has more than 2400W lower losses than the Si IGBT based inverter. These reduced inverter losses equate to more power at the wheels of the car. In ad-dition, lower losses lead to reduced chip temperature with the same cooling system. In these traction inverters set up, the main cooling system outside of the SiC MOSFET based inverter could shrink more than 30% compare to Si IGBT based inverter set up.
Benefit of SiC on System levelA lot of activities to enhance energy density for rechargeable batter-ies are ongoing [4]. At the same time, the high efficiency of the SiC MOSFET based inverter allows the size of an EV system battery to be reduced while keeping the same driving range. The battery capacity reduction can lead to an economic benefit if the overall power train system is considered.
In figure 8, the economic benefit gained from a battery size reduc-tion, enabled by a more efficient inverter, is shown for different battery sizes in 2025. An estimated battery capacity improvement ratio is 3 to 5% based on the inverter efficiencies shown in previous sections of this paper and the standard driving cycle for passenger vehicles [5]. For the SiC MOSFET a chip size of 25mm² at 1200V rated voltage is considered per 100Arms of output current. From the diagram it can
be seen that the 400Arms inverter provided economic benefit at the system level if a battery of at least 32kWh capacity is used. For a 600Arms inverter the benefit exists above a battery size of approxi-mately 48kWh.
With further increasing adoption of SiC devices in the near future and the resulting effects of economy of scale for SiC MOSFETs it is expected that the economic gap to Si IGBTs will get smaller and the
benefits will thus get even larger.
Conclusion The newly developed low stray inductance, high heat dissipation ca-pable package called “Gtype” is suitable for SiC MOSFETs. A traction inverter utilizing this new module reached 220kW output power and a peak efficiency of 99.1%. Furthermore, the SiC MOSFET based in-verter has less weight and volume than a Si IGBT based design. The power density is 22kW/L which is 57% higher than that of the Si IGBT based solution. A high efficiency inverter brings economic benefits for the user as it extends the driving distance for a given battery capac-ity or allows a battery size reduction while keeping driving distance, which can give an economic advantage.
References [1] T. Nakamura, Y. Nakano, M. Aketa, R. Nakamura, S. Mitani, H.
Sakairi, Y. Yokotsuji: “High performance SiC Trench devices with ultra-low Ron”, IEDM 2011
[2] R. Nakamura, Y. Nakano, M. Aketa, N. Kawamoto, K. Ino: “1200V 4H-SiC Trench Devices”, PCIM Europe 2014
[3] M. Nakanishi, K. Hayashi, A. Enomoto, M. Hayashiguchi, M. Ando, K. Ino, ROHM Co., Ltd, Japan, C. Felgemacher, A. Mash-aly, G. Richard, “ Automotive Traction Inverter utilizing SiC Power Module” PCIM 2018
[4] New Energy and Industrial Technology Development Organiza-tion: “Secondary battery technology development roadmap 2013”
[5] World Harmonized Light Duty Test Procedure
www.rohm.com
Figure 7: Comparison of losses per switch of both technologies IGBT and SiC
Figure 8: SiC MOSFET based inverter’s economic benefit vs. battery size in 2025
Find out more at unitedsic.com/sic | [email protected]
How breakthrough 650V SiC FET technology cuts power losses in hard switched systems
• Simple Silicon Substitution™ technology
that works with standard silicon gate drive
• Lowest RDS
(on), down to 27 mOhms in TO-220
• Lowest Qrr < 80 nC (UJC06505K)
If your power system will benefit from
higher frequency operation and higher
efficiency, the simplest solution is to
replace Si Superjunction transistors with
new silicon carbide devices from UnitedSiC.
To learn more about how “SiC cascodes outperform in practical
applications”, download the white paper at
www.unitedsic.com/SiCcascodes1
Electric Vehicles Power Supplies Industrial Drives Renewable Energy
Bodo´s Power Systems® July 2018 www.bodospower.com30
CONTENT
Current trends in power supply unit design are focusing on higher efficiencies and power densities, that go beyond the capabilities of the silicon MOSFET technology. Development engineers need new switching devices that are able to meet these requirements. And so begins the conception of gallium nitride transistors (GaN).
Overview and AdvantagesPanasonic Hybrid Drain-Gate Injection Transistors (HD-GITs) are normally-off GaN-on-silicon transistors (Figure 1). They are based on the HEMT principle, using the highly mobile 2D electron gas forming at an AlGaN-GaN heterojunction as conduction layer. The active part of the transistor is completed on the top side with (ohmic) drain & source contacts, a recessed p-GaN gate (ohmic contact) and a p-GaN “gate like” structure connected to the drain. For cost reasons, the tran-sistor is grown on top of 6 inches silicon wafers by MOCVD process. In order to reduce the tensile stress caused by the mismatched crystal lattices of Si and GaN, to limit the vertical drain-substrate leakage
currents and to prevent deep breakdown paths in the conductive Si substrate, a lattice buffer layer (Figure 2) is inserted between the silicon bulk and the active top side of the transistor.
This buffer plays a central role in the determination of key reliability characteristics of the transistors as will be developed further below. The transistor is turned on resp. off like a field effect transistor, by ap-plying a gate-source voltage above resp. below a threshold voltage. In off state, the p-GaN gate depletes the electron gas underneath by lifting the potential of the AlGaN-GaN junction. In on-state, the gate behaves essentially like a diode. Unlike in MOS transistors, a small (around 10mA) current is injected from the gate into the conducting layers by electrons tunneling through the AlGaN barrier. Due to the low velocity of holes in the GaN material, the current conduction at the AlGaN-GaN interface is only due to the electron gas, and so the transistors are to be understood essentially as unipolar devices in this regard.
Since the HD-GiT gate can be accessed directly, the gate circuit can be designed to control and adjust the transistor’s du/dt and di/dt – a major advantage as compared with the cascode.
x-GaN 2.0: The Destination FiT10 is Demonstrably Exceeded
Optimized Reliability and Robustness Characterize X-GaN Power Devices by Panasonic Industry
For more than 35 years, power MOSFETs have dominated the field of power converter design in the low to medium power range. This has been supported by continuous
innovation in the components structure and related semiconductor technology. Fast switching characteristics and low losses, as well as ease of use in various circuit
topologies also contributed to their success. At the dawn of a new millennium, however, silicon power MOSFETs are reaching their theoretical performance limits, which means that further progress in power supplies and power management systems will no longer be
as easy to achieve with these switching elements.
By Francois Perraud, Teamleader Power & Automotive Solutions, Panasonic Industry Europe GmbH
WIDE BAND GAP
Figure 2: Lattice Buffer
Figure 1: Normally Off
Industry standard power module packages are commodities
today. Supply chain safety for these modules is one of the most
important requirements. Being compatible on the one hand,
SEMIKRON standard modules even exceed these standards on
the other hand.
Our portfolio of sintering and bonding technologies takes
output power capability and reliability to new levels. Module-
integrated current measurement shunts, plug & play drivers
and pre-applied Thermal Interface Materials reduce the
number of system parts, cutting development time as well
as system costs and time-to-market.
www.semikron.com/semitop-e1e2shop.semikron.com
Expect More from Standard Modules.
210x297 mm
SEMITOP® E1/E2 Latest commodity product
Power Modules Portfolio MiniSKiiP, SEMITOP E1/E2, SEMIPACK, SEMiX 3 Press-Fit, SEMiX 5, SEMiX6p, SEMITRANS 2/3/4, SEMITRANS 10
Bodo´s Power Systems® July 2018 www.bodospower.com32
CONTENT
The lateral structure of the GiT is also advantageous for fast switch-ing, since its parasitic capacitances are typically lower than those of vertical structures, such as for example silicon-based super-junction MOSFETs (see Figure 3).
Reverse condition without recovery lossesX-GaN™ GiT transistors allow current to flow in the reverse direc-tion once the source, gate, and drain potential are set in a way that current is fed in at the gate. Unlike with MOSFETs the reverse current does not flow through a parasitic body here, instead is it conducted through the channel. Even though they are reminiscent of a diode, the thresholds in the third quadrant of the static IV curve are not dictated by junction’s behaviour, but are simply the threshold voltage of the transistor plus any negative bias voltage that is applied to the gate. In the same way as a MOSFET, the GiT can be switched on in the reverse direction in order to further reduce the losses by operating under 0V-offset condition. The GiT recovers extremely quickly from re-verse conduction. The recovery energy practically just corresponds to the energy required to charge the output capacitance. The conduction and recovery performances of the GiT in the reverse mode are the same as those of an SiC Schottky diode (see Figure 4).
Electron trapping and current collapse In the absence of specific technical counter-measures, GaN GiT tran-sistors suffer from electron trapping issues. Generally speaking, these effects are related to the build-up of negative-charged regions in the transistors. This is caused by the trapping of electrons in defects of the crystal and/or at the surface of layers, leading to non-optimal repartition of the electric field in the transistor and disturbing the flow of charge carriers in the 2D gas.
Two trapping effects seem to dominate in GiTs, leading potentially to different destruction mechanisms. Specifically it is understood that the traps contributing most to the build-up of the negative regions are i) deep traps in the buffer layer in the drain to substrate region, appearing under high Vds stress and ii) traps located at the AlGaN surface, capturing hot electrons crossing the AlGaN barriers under hard switching conditions in the semi-on state (Figure 6)
First consequence: when a drain-source voltage is applied above a certain threshold – dependent on the device characteristics, a bit above 500V for the transistors we are writing about here – the “cur-rent collapses”, i.e.from the application point of view the Rds(on) of the transistor increases switching cycle after switching cycle, until it reaches a saturation value.This effect, also known as dynamic Rds(on) typically leads to the rapid destruction of the transistor by thermal breakdown (Figure 6).
Second consequence: the hot electrons captured at the surface of the AlGaN layer in the “semi-on” state under hard switching conditions are suspected to trigger a positive destructive feedback loop that will see the electric field at the drain side increase due to the trapped charges, causing more trapping at the AlGaN surface, causing a fur-ther increase of the electric field etc. until the device breaks down.
Panasonic's HD-GiT structure solving the current collapseSo far Panasonic has been the only provider of GaN components to publicly announce the complete elimination of the problem of current collapse. The figures 5 & 6 shows Panasonic’s unique approach for solving the issue. An additional p-doped structure, similar to the gate, is grown near the drain and electrically connected to it. That structure injects holes into the GaN components, that recombine with the trapped electrons and reduces the maximum electric field in the device, leading to less trapping.
The HD-GiT uses a recessed gate so that the thickness of the AIGaN layer is increased in order to avoid depletion of the charge carriers under the p-doped area close to the drain. The HD-GiT was proven to have the same excellent switching characteristics as the conventional GiT structure. Failure mechanisms in GaN transistors in general have
WIDE BAND GAP
Figure 3: Comparison GiT vs. MOSFET
Figure 4: GiT reverse recovery behavior
Figure 5: Panasonic’s New HD-GIT in comparison to GIT – Cross Section
Figure 6: Panasonic’s New HD-GIT in Comparison to GIT – Perfor-mance
www.bodospower.com July 2018 Bodo´s Power Systems® 33
CONTENTWIDE BAND GAP
been closely investigated in the past decade and discussed in many papers. Let’s now review the most important aspects of Panasonic’s X-GaN transistors’ reliability.
ReliabilityIn order to guarantee the reliability of the HD-GiT transistors in mass production, Panasonic not only tests against the usual JEDEC standards for Si components, but also developed its own additional GaN-specific tests to guarantee the long-term stability of the transis-tors, for example with regard to current collapse. Accelerated life tests have showed a worst case FiT rate of better than ~10 FIT can already be achieved. Electron trapping and current collapse.
End of life testThe end of life mechanism in the GiT transistors seems to be well modeled by a so-called percolation degradation model, where the accumulation of defects over time along the path of leakage currents ultimately creates a conductive path that will cause a time dependent breakdown. The life time of HD-GiT transistors is demonstrated to be related to the p-drain leakage current, by comparing physical models with the Weibull plots of transistors failures under High Temperature Reverse Bias (HTRB) conditions. According to this understanding the thickness of the lattice buffer in Panasonic’s X-GaN transistors has been adjusted to limit the p-drain to substrate leakage, aiming at achieve a failure rate over time less than 0.1% for 10 years operation (at Vds=480V and Tj=100°C).
Accordingly the HTRB test is one of the most critical tests to assess the lifetime of the X-GaN transistors. It was therefore per-formed on 10,000 transistors, taken from 20 different lots, far beyond JEDEC require-ments. Data corresponding to 1 billion devices and hours equivalent of operation were accumulated. No failures were observed during the test, proving that the said 10 FiT target set by Panasonic as minimum acceptable level to start the mass production had been achieved.
Robustness in applicationBesides extensive testing of the critical aspects affecting the devices’ intrinsic reliability and life time, Panasonic X-GaN transistors also provide robustness by design, contributing to the safe design of appli-cation circuits. The first, obvious aspect, consists in the X-GaNs being normally off transistors, and appeals no further comments here. There is more however …
Diode gate, doesn’t break The gate of the HD-GiT transistors in on state behaves like a diode. This means that there is no breakdown destruction to fear in case of voltage spike on the gate. Gate noise will be clamped by the diode and “absorbed” as transitory peak current.
The safety of the gate circuit is therefore insured by design. This is a significant advantage as compared with other devices implement-ing different gate contact concepts, that impose the gate signal to be kept within a very narrow voltage range in on-state, and that are very sensitive to overvoltage breakdown. The benefit comes at a very small cost in the HD-GiT: the small gate injection current drawn by the gate under normal turn-on condition will typically causes gate losses around 10mW, which is neglectable from the point of view of the ef-ficiency of the system.
Figure 7: Extra margin on breakdown voltage
ZH Wielain Electronic (Hangzhou Co.,LTD)Hangzhou, China
E-mail: [email protected]://www.zhwielain.com
+86 571 28898137
Power Module Product and Packaging & Interconnect Power Module Products Press Fit Solderless Connection Solution
Connect teminals
DBC connectors
IGBT body
[email protected] · www.hivolt.de
» High Voltage Cable
Bodo´s Power Systems® July 2018 www.bodospower.com34
CONTENT
Extra margin on the break-down voltageUnlike MOSFETs, GiT’s lateral structures have no junctions that will avalanche and clamp voltage spikes above the rat-ed breakdown. The GaN material’s wide bandgap property however allows the design of small dies with high break-down voltage. Therefore, to warrantee a fail-safe operation of the transis-tors for example in the event of line surges, Panasonic X-GaN devices have been designed with a with a big margin on the drain-source breakdown limit. Indeed the static field-dependent break-down voltage of the currently available transistors qualified for 600V operation is in the range of 900V to 1kV (Figure 7). As a side effect, it allows the X-GaN transistors to be qualified with a Vds spike voltage rating of 750V for one microsecond. The vertical field dependent breakdown is also in the same 1kV range; the lattice buffer design (again) is playing here a major role in set-ting this value.
Panasoniy Gate Driver ICPanasonic brought out its own X-GaN™ gate driver at the end
of 2016 for developers who want to quickly deploy a solution using GiTs. The X-GaN driver IC is optimized for high switching frequencies up to 2 MHz and provides an easy way to unlock the full performance of the transistors. Besides optimized gate control terminals, additional integrated functions are provided – such as a charge pump for (op-tionally) generating negative gate voltages, or safety features against under-voltage and gate oscillations (Figure 8).
Advantages in applicationsPanasonic’s GiT transistors are aimed at power converters in the ~100W to ~5-6kW range, where MOSFETs with 600V to 650V are typically used today. Depending on the requirements of the applica-tion, developers can target maximum efficiency, maximum power density, or a compromise between the two. Thanks to their “0 reverse recovery” behaviour, GaN transistors make some topologies practi-cally usable, such as for example totem-pole PFCs, which require fewer parts as conventional designs and exhibit state-of-the-art ef-ficiency performance. Increased switching frequency enables passive components to be miniaturized – in particular magnetic components – whereas the power density of circuits such as resonant DC-DC con-verters can be increased. Last but not least, GaN bring significant ef-ficiency improvements under partial load operation in resonant circuits of this type Panasonic has used and demonstrated these capabilities in a highly compact and efficient AC-DC demo unit (Figure 9).
Applications like power supplies for IT, telecoms servers and AC adapters should benefit most from these in the short term. The au-tomotive industry has also demonstrated significant interest in being able to use such components in on-board chargers or DCDCs in the medium term.
www.industry.panasonic.eu
Figure 8: x-GaN gate driver SC
Figure 9: Ac-DC demo unit
WIDE BAND GAP
Rethinkingconverters!converters!
• Design according to customer speci cations
• Extremely low inductance• 10 percent higher capacity volume• No corrosion of contacts• Easy to assemble• Very long service life
FischerLink
DC-Link capacitors
in a robust and
low-inductive
module
CapacitorsMade in Germany
www.ftcap.de
01_DU__FischerLink_BodosPowerSystems_43x130_EN.indd 123.01.18 16:09
Automotive Home Appliances Industrial Power Transmission Renewables Railway
Home appliances are becoming more and more demanding regarding function- ality, reliability and efficiency. In the field of Power Semiconductors Mitsubishi Electric had created the necessary basis already 20 years ago as the pioneer of the DIPIPM™ Transfer molded package intelligent power modules, followed by the continuous development and expansion of this series. Consequently, with the new MISOP™ a surface-mount package Intelligent Power Module has been added to the line-up to realise compactness and easy assembling in small power inverters for pumps and fans. Also low power servos in industrial appli- cations can be covered. The versatile integrated features are designed to give the benefit of reduced development time for the complete inverter system.
Scan and learn more about our Power Semiconductors on YouTube.
Power Devices from Mitsubishi Electric.
More Information: [email protected]
One of our key products:Trust.
New surface-mount package type IPM with latest RC-IGBT Chip technology
The MISOPTM Surface-mount Package IPM – Optimized terminal layout enabling simple and compact PCB design – Integrated driver ICs (HVIC and LVIC) – Integrated bootstrap diodes and capacitors – Short circuit protection through external shunt resistor – Power supply under-voltage protection: Fo output on N-side – Over temperature protection – Analog temperature voltage signal output – Interlock function
ME_SC_PowerDevices_HomeAppliances_AZ_A4_QR_E_RZ.indd 1 14.06.18 15:48
Bodo´s Power Systems® July 2018 www.bodospower.com36
CONTENT
While USB-C is relatively new, indications are promising that there will be significant uptake and sales for the new standard. Market research firm, IHS, are forecasting that 500 million devices with USB-C will ship this year, increasing rapidly to 2 billion devices in 2019.
The consumer benefits of USB-C include high speed data transfer, more power available for charging and powering devices and flippable connectors, making plugging simpler than ever before. In this techni-cal article, Toshiba Electronics Europe will discuss how delivering consumer convenience brings challenges for designers and look at some of the solutions reaching the market.
USB Type-C provides a small, versatile connection scheme based around a bidirectional cable and an unkeyed flippable connector that connects to hosts and peripherals, replacing old-style Type-A and Type-B connectors and cables. Data rates of 10Gbps match the Su-perSpeed USB 3.1 and the USB Power Delivery (USB PD) specifica-tion is incorporated, increasing power capability to 100W. However, these advances require careful selection of key components for power and data switching as well as interface protection.
Increased power to charge devices simplyUSB PD allows devices to act as either a source or a consumer of power via the USB connection. It specifies profiles at 10W, 18W, 36W, 60W and 100W that operate at voltages between 5V and 20V with currents up to 5A. Devices that require power negotiate for power after the connection has started up in the default 5V/2A mode, thereby ensuring the safety of older USB2.0 devices.
USB PD allows for faster charging of peripherals and allowing devices such as smartphones, tablets and laptops to be powered from their USB connection, reducing the need for mains power adapters. A mains monitor can power a laptop via USB-C, act as a hub for pe-ripherals such as external HDDs and also receive and display video information from the laptop. In this scenario, only the monitor requires an AC/DC power adapter, removing cost and bulk from the set-up.
High speed data and flexibilityIn USB-C, a single style of plug and socket containing 24 pins on a 0.5mm pitch in an 8.4mm x 2.6mm form factor is defined meaning that non directional cables can be used, allowing power and data to flow in either direction.
The USB-C plug is un-polarised, and can be inserted either way. In addition, the physical layer interface for USB-C contains two data pairs (D+/D-), maintaining backwards compatibility with USB 2.0, while supporting the 10Gbps rates for USB 3.0 and USB 3.1 and providing a 20Gbps capability for the future.
A USB-C cable can accommodate legacy USB type-A/B, Mini-USB and Micro-USB connectors via an adapter as the new standard is electrically compatible with USB 2.0, while catering for USB 3.1 and PD specifications. Additional Configuration Channel (CC) pins deter-mine cable orientation and allow for PD negotiation.
Via alternate modes, standards such as DisplayPort, HDMI, MHL, Audio and Thunderbolt can all be delivered through USB-C allowing a single cable type to be used for multiple purposes.
POWER SUPPLY
USB Type-C: Consumer Convenience Comes with Design Challenges
It has been 20 years since Universal Serial Bus (USB) made its debut on an Apple iMac G3 computer, beginning a fundamental change in the way that computers and peripherals connect. During that time the standard has evolved through various iterations to the latest version - USB Type-C - that, again, promises to make fundamental changes to the way we
connect computers and peripheral devices.
By Robert Heinzelmann, Toshiba Electronics Europe GmbH
Figure 1: USB PD spec contains power profiles up to 100W
Figure 2: USB Type-C connector pin designations
www.bodospower.com July 2018 Bodo´s Power Systems® 37
CONTENT
Design challenges within USB-C There are three principal areas of improvement within USB-C, each of which bring design challenges:• The inclusion of USB PD requires power switches capable of with-
standing at least 20V while having low on-resistance to minimise power loss and thermal load.
• High speed transceivers are susceptible to electrostatic discharge (ESD), requiring external protection on data lines often accom-plished with Transient Voltage Suppression (TVS) diodes. However, the extreme 10Gbps data speed means a trade-off between TVS terminal capacitance that avoids data distortion and the available ESD protection level is needed.
• The new reversible connector interface requires switching devices to ensure that the data flow is mapped according to the connector insertion, while maintaining high-speed data transmission
Solving the power delivery challengeA simple dual-MOSFET power switch can be implemented with P-channel MOSFETs such as Toshiba’s SSM6J50xNU series with +20/-25V VGSS eliminating the need for a driver IC. The devices can be used in any of the USB PD profiles and offer RDS(on) as low as 19mΩ, ensuring efficient operation with minimal heat generation and voltage drop.
Alternatively, an integrated load switch such as Toshiba’s TCK30xG 28V device can be used. This is a very compact solution with built-in protection features including integrated thermal shutdown, under-voltage lockout, adjustable overcurrent protection and overvoltage protection.
In addition to the low on-resistance of the switch IC valuable features in USB-C applications are integrated. This includes adjustable over-current protection up to 3.0A (set by a single external resistor), and adjustable slew rate (determined by an external capacitor). This com-bination of capabilities makes the TCK30xG family the most compact solution for USB charging up to Profile 3.
The most efficient power solution is to use discrete N-channel MOS-FETs with an external driver IC. Toshiba’s TCK40xG family of driver ICs can withstand 40V and has a built-in charge pump to provide the gate drive voltage for an external N-channel MOSFET. This external MOSFET can offer the lowest RDS(on) in a small package, like the SSM6K513NU, which is rated at only 8mΩ in a 2x2mm package. In tests, this setup ran 16°C cooler than that of a competitor’s approach when simulating Profile 5 (100W @ 5A).
In terms of space the dual N-channel MOSFET plus TCK40xG solu-tion requires 15mm2 while the dual P-channel MOSFET driven from the open drain output of the charger IC requires 10mm2. The inte-grated TCK35xG solution is significantly smaller at 2.3mm2, but does offer reduced performance.
Protection and switching solutionsIn addition to devices for efficient power switching solutions, power lines also require protection, which can be realised with Toshiba’s new DF2SxxP TVS diode series, The diodes in this line-up are capable of protecting the VBUS lines from surge as they provide contact discharge protection up to ±30kV in a compact packages and allow for VBUS voltages between 5-20V, making them an ideal choice for all profiles.
Furthermore, when protecting data lines it is important to meet IEC-61000-4-2 while not unduly distorting data signals. This can be achieved with TVS diodes such as the single DF2B5M4SL (0.2pF) or the DF5G5M4N array that can provide protection for two signal pairs (RX1+/RX1-, TX1+/TX1-).
A bus switch such as the TC7PC13212MT is ideal for switching the data path to match connector orientation and will maintain full data integrity up to 10Gbps.
SummaryUSB-C is clearly a significant step forward in convenience for us-ers with additional power available and increased cable simplicity. However, this brings challenges for designers who have to cope with increased currents, higher voltages and switching data paths.
Fortunately, Toshiba offer a variety of products to ease this task including several options for power management, including fully inte-grated designs that maximise simplicity and space saving as well as discrete designs for the ultimate efficiency.
Alongside this, bus switches allow easy management of data paths while advanced TVS diodes protect data lines from ESD without introducing signal distortion.
https://bit.ly/2lfERTL
Figure 3: Simple dual-MOSFET power switch solution
POWER SUPPLY
Figure 4: Highly efficient USB Type-C power solution using a control-ler and discrete MOSFETs
Bodo´s Power Systems® July 2018 www.bodospower.com38
CONTENT
Broadcom Inc. (formerly Avago Technologies) gate drive optocouplers are used extensively in driving Silicon-based semiconductors like IGBT and Power MOSFETs. Optocouplers are used to provide rein-forced galvanic insulation between the control circuits and the high voltages. The ability to reject high common mode noise will prevent erroneous driving of the power semiconductors during high frequency switching. This paper will discuss the benefits of GaN, its gate drive requirements, the gate drive designs, tests and performance.
Benefits of GaNGallium Nitride is a wide bandgap (3.4 eV) compound made up of Gallium and Nitrogen. Bandgap is a region formed at the junction of materials where no electron exists. Wide bandgap GaN has high breakdown voltage and low conduction resistance. It has higher electron velocity and lower parasitic capacitance which improves its switching speed.
The benefits of GaN over Silicon can be summarized into 3 main points:- Smaller system designs- Lower system costs- Higher system efficiency
The smaller and lower costs are results of fewer and smaller periph-eral components. GaN can operate in reverse conduction mode which can eliminate external freewheeling diode. It can operate in high frequency which results in smaller filters and magnetics like inductors
and transformers. GaN operates 60°C cooler than Silicon which will help to reduce the size of the heat sink.
Higher efficiency is the result of lower switching and conduction loss. GaN has higher electron velocity and lower parasitic capacitance for low switching loss. It is also smaller in size than Silicon at the same breakdown voltage and hence lower conduction resistance.
Figure 3 shows the different types of GaN and their gate drive require-ments. Brand E for example manufactures 200V GaN, used mainly for low voltage applications like 12V DC-DC converter. Brand T
TECHNOLOGY
Gate Drive Optocouplers for GaN Power Devices
Gallium Nitride (GaN) power devices are gaining popularity over Silicon power devices as its faster switching capability can improve overall system efficiency and reduce the size and costs. The technical benefits coupled with lower costs due to increase in GaN
production have increased the adoption in applications like industrial power supplies and renewable energy inverters.
By Robinson Law, Applications Engineer and Chun Keong Tee, Product Manager, Broadcom Inc.
Figure 1: Silicon vs. GaN, smaller and lower system costs
Figure 2: Silicon vs. GaN, higher system efficiency
Figure 3: Types of GaN and Gate Drive Requirements
www.bodospower.com July 2018 Bodo´s Power Systems® 39
CONTENTTECHNOLOGY
manufactures 600V GaN but is a normally-on switch. It requires a low voltage Silicon MOS in cascode connection to turn it into a normally-off switch which will be safer to use. Due to the cascode structure, the switching speed cannot be controlled by adjusting the gate resistance. This will lead to complication in fine tuning the EMI (Electromagnetic interference) and switching loss.
Panasonic and GaN Systems manufactured normally-off switches by using P-type barrier structure under gate to deplete the high mobility electrons during 0V gate bias. Due to the high electron mobility, the threshold of GaN, VTH is relatively lower than that of Silicon MOS or IGBT. The input capacitance is also very small, less than 1nF and only requires 5nC to switch on.
GaN switches very fast and care should be taken when designing with a high switching dv/dt. It is important to control the high dv/dt noise coupling from the GaN to the gate driver.
Otherwise, it is required that the gate drivers must have noise immu-nity of more than 100kV/µs to prevent false switching of the GaN.
Since GaN devices from Panasonic and GaN Systems are normally-off and easy to use, the gate drive requirements are very much similar to Silicon MOS. Panasonic GaN has a robust gate, which allows a high gate voltage of 12V for fast turning on of the gate. GaN Systems recommends 6V to charge the gate. Due to the small gate capaci-tance and gate charge needed, the gate current needed is relatively low at less than 1.5A.
For Panasonic GaN, one thing to take note is that the gate requires DC holding current of approximately 10mA to maintain it in “ON” sta-tus. For GaN Systems, special care is needed to ensure the absolute maximum gate voltage of 7V is not exceeded.
Gate Drive Design for Panasonic GaN
Figure 4 shows a half bridge evaluation board featuring Panasonic 600V 70mΩ X-GaN transistor, PGA26E07BA. The gate drive is designed using two gate drive optocouplers, ACPL-P346 to drive the GaN transistor directly.
The ACPL-P346 is a basic gate driver optocoupler used to isolate and drive the GaN operating at high DC bus voltage. It has a rail-to-rail output with 2.5A maximum output current to provide fast switching high voltage and driving current to turn-on and off the GaN efficiently and reliably. The ACPL-P346 has a maximum propagation delay times of less than 110ns and typical rise and fall times of around 8ns. The very high CMR, common mode rejection of 100kV/µs(min.) is required to isolate high transient noise during the high frequency operation.
Figure 5 shows the schematic of the half bridge evaluation board and the ACPL-P346 gate drive design. The GaN transistors, QB and QA would require about 12.5mA on-state current to continuously bias the transistor in on-state. This is done by the gate driver through 680ohm resistors RB1 and RA1.
The initial in-rush charging current to switch on the GaN quickly is provided by ACPL-P346 and the peak current limited by resistors RB2 and RA2. Capacitors CB3 and CA3 are used to turn on the GaN faster by increasing the charging current momentarily. The board has the flexibility to be powered by 2 isolated DC-DC supplies for the top and bottom bridges or 1 DC-DC with bootstrapping.
Gate Drive Design for GaN Systems GaN
Figure 4: Half bridge evaluation board with Panasonic GaN and ACPL-P346
Figure 5: Schematic of half bridge evaluation board with Panasonic GaN and ACPL-P346
Figure 6: Half bridge evaluation board with GaN Systems GaN and ACPL-P346
Bodo´s Power Systems® July 2018 www.bodospower.com40
CONTENT
Figure 6 shows another half bridge evaluation board featuring GaN Systems’ 650V E-HEMT GS66508T (30A/50m Ω) GaN transistor. The half bridge evaluation board uses two gate drive optocouplers ACPL-P346 to drive the GaN transistors directly. The schematic shows the bottom bridge gate bias and driver circuit. The top bridge uses the same circuitry. The isolated DC-DC, 5V-10V converters are used to provide +6V and -4V bipolar gate drive bias for more robust gate drive and better noise immunity. The 10V is then split into +6.2V and -3.8V bias by using 6.2V Zener diode. The ACPL-P346 gate drive output is a combination of 10Ω gate current limiting resistor (for charging) and 10Ω paralleled with 2Ω plus a diode for discharging.
GaN Half Bridge Evaluation Board Test and PerformanceUsing the half bridge evaluation board from Panasonic and GaN Systems, the slew rate, switching power loss and efficiency tests were performed on the GaN and ACPL-P346.
An inductor of about 120 to 160µH was connected between VDC+ and VSW to form the boost configuration also known as low side test. The low side GaN transistor Q2 was active in boost mode. 400V Bus voltage was applied to VDC+/VDC-. Double pulse test was used for easy evaluation of device switching performance at high voltage and current without the need of actually running at high power.
The period of first pulse TON1 applied to Q2 defined the switching current ISW. t1 (turn-off) and t2 (turn-on) were the measurement points as they were the hard switching transients for the half bridge circuit when Q2 is under high switching stress.
The slew rate tests were conducted at 400V DC and around 30A hard switching.
The Q2 turn-on and off slew rates (dv/dt) were measured at t1 (turn-off) and t2 (turn-on) respectively. The highest slew rates of more than 110kV/µs were measured when the GaN hard turned off at 400V, 30A.
ACPL-P346 has a minimum CMR, common mode rejection of 100kV/µs. In other words, ACPL-P346 can isolate high transient dv/dt noise from the GaN switching. The scope pictures show the GaN fast slew rates were not affecting the gate drive outputs and gate voltages.
The switching loss test used the same boost configuration with a current sensor installed for ID measurement. The same double pulse signals and timing of the slew rate test were used for the power loss measurement. The measurement was done at the monitoring points when the GaN turned on or off at the target current level. The math function on oscilloscope was used to find the multiplication of VDS and IDS. The measure function on the oscilloscope was then used to find power loss which was the area under the curve.
TECHNOLOGY
Figure 7: Schematic of ACPL-P346 gate drive circuit for GaN Systems half bridge evaluation board
Figure 8: Slew rate and switching power loss test setup and wave-forms
Figure 9: Panasonic GaN and ACPL-P346 slew rate test
Figure 11: Switching power loss test setup and waveforms
Figure 12: Switching power loss measurements
www.bodospower.com July 2018 Bodo´s Power Systems® 41
CONTENT
The turn-off power losses are represented by the red line and both GaN power losses were kept below 15µJ regardless of inductor load current. The turn on power losses are represented by the blue line, both GaN show a low loss of around 40µJ at 15A.
The half bridge evaluation boards were connected as DC-DC convert-ers to test the efficiency of GaN in hard switching operation. The Panasonic GaN DC-DC was connected in boost 200V to 380V con-figuration while GaN Systems DC-DC was connected in buck 400V to 200V configuration (Q1 turned on to charge the inductor, and turned
off to allow inductor current to continue discharging through output ca-pacitor and through Q2 as freewheeling diode). Both converters were operated at 100 kHz frequency, room temperature and the efficiency at different power tested.
Both converters showed high conversion efficiency of approximately 99%.
AcknowledgementBroadcom would like to thank the Application Teams of Panasonic Semiconductor Solutions Singapore and GaN Systems for their tech-nical support.
References• “ACPL-P346/ACPL-W346 2.5-Amp Output Current Power and
SiC MOSFET Gate Drive Optocoupler with Rail-to-Rail Output”, Broadcom Inc., AV02-4078EN.
• “ACPL-P346 Panasonic X-GaN Transistor PGA26E07BA Half Bridge Evaluation Board”, Broadcom Inc., ACPL-P346-X-GaN-RM100.
• “ACPL-P346 GaN Systems GaN E-HEMT GS66508T Half Bridge Evaluation Board”, Broadcom Inc., ACPL-P346-RefDesign-RM101.
• “PGA26E07BA Datasheet”, Panasonic Semiconductor.
• “GS66508T Top-side cooled 650 V E-mode GaN transistor Prelimi-nary Datasheet”, GaN Systems.
• “GN001 Application Guide Design with GaN Enhancement mode HEMT”, GaN Systems.
www.broadcom.com
TECHNOLOGY
Figure 13: Efficiency test setup
Figure 14: Efficiency test measurements
Bodo´s Power Systems® July 2018 www.bodospower.com42
CONTENT
There is a symbiotic relationship between power WBG devices and Near Chip-Scale SMD packages, or more specifically these pack-ages enable power WBG and power WBG enable Near Chip-Scale packages. Inevitability of Near Chip-Scale power WBG packaging can be better understood by looking at the example of today’s digital microprocessor’s integration and packaging. In principle, hundreds of thousands of transistors could be packaged as discretes, and soldered together, but such a maze of leads and interconnects would have unacceptable parasitics, size, reliability and cost. Discrete micro-processors would just not be possible. Going one step further, the proprietary µMaxPak package architecture is an optimum example of Ideal Near Chip-Scale packaging. The following Frequently-Asked-Questions provide insight into the benefits of WBG Near Chip-Scale package capability and its Ultimate form, the μMaxPak package technology.
Q1: Are Near Chip-Scale SMD Packages Really Inevitable for Power WBG devices?A1: Yes they are Inevitable. Near-Chip-Scale SMD packages are small, Ieadless & wire bondless, approaching the size of the chips enclosed. They are essential for Full performance of power WBG devices, because they reduce parasitics and thermal resistance (Rjc). Thereby improving WBG speed, efficiency, reliability & costs, and they allow similar improvements to the overall system. The µMaxPak is an example of Ultimate Near Chip-Scale WBG packaging.
Q2: What are the Primary Performance Advantages of the µMaxPak packages?A2: The µMaxPak is the ultimate implementation of the Near Chip-Scale packaging concept, because it minimizes parasitics and maxi-mizes heat transfer. This is accomplished by minimizing current loop path length with leadless and wire bondless packages. Thus enabling power WBG devices to operate at maximum efficiency by reducing both conduction & switching losses. Furthermore, the exceptionally low package thermal resistance Rjc enhances reliability and increases maximum current rating. Table 1 quantifies potential performance ad-vantages for µMaxPak half-bridges with WBG die size between 5mm x 5mm and 7mm x 8mm.
Q3: If more Efficient WBG devices reduce Power Dissipation, why is lower Package Thermal Resistance required?A3: Although higher WBG efficiencies are reducing power dissipation significantly, the WBG die power densities are increasing about twice as fast as efficiencies. Therefore, thermal resistance per unit area must be reduced. The basic µMaxPak packages utilizes robust Cu conductors and reduced solder interfaces to more effectively remove heat from both top & bottom of the die. For very high power dissipa-tion products, more heat can be removed with heatsinks or cold-plates on both the top & bottom of the package. Looking to the future, potentially GaN & SiC operating temperatures will increase, further enhancing heat dissipation.
Q4: Why have Chip-Scale SMD packages not been used for Power Si IGBT modules?A4: Si IGBT modules require very large IGBT & diode die. Such large die cannot be soldered directly to Cu leadframes, because of large coefficient of thermal expansion (CTE) mismatch between Cu and die. This is acceptable for smaller die where we see smaller discrete IGBT & diode die soldered directly to Cu leadframes or to the bases of TO220 & TO247 packages, but the much larger IGBT die in modules require a DBC CTE buffer between die and base plate.
Q5: Why can Higher Current SiC and GaN die be used Directly on Cu? A5: Power SiC and GaN die can have 10 times the current density of Si IGBT die, which enables die to be 1/10th the size. Therefore, WBG devices in Near Chip-Scale SMD packages can be soldered to Cu leadframes and operate at hundreds of amperes.
WIDE BAND GAP
Inevitability of Near Chip-Scale SMD Packaging for Power GaN & SiC
Frequently-Asked-Questions About Ultimate WBG Near Chip-Scale SMD Packaging
Today’s performance of power Wide Band-Gap (WBG) devices is severely limited by conventional power packages (i.e. TO220, TO247 & IGBT DBC-modules). We are seeing
some early Power QFN and LGA WBG packages with modest improvements, but they are still performance limiting and are costly. Future higher performance power WBG
evolution will simply not be possible without Near Chip-Scale SMD packages. Such Near Chip-Scale SMD packages are absolutely required for power WBG device’s potential
power density, performance, speed, efficiency and costs.
By Courtney R. Furnival, Semiconductor Packaging Solutions, Inc.
Table 1: Primary QFN-µMaxPak Performance Advantages.
Bodo´s Power Systems® July 2018 www.bodospower.com44
CONTENT
Q6: What are the basic Types or Structures of µMaxPak Archi-tecture?A6: The µMaxPaks are molded, leadless and wirebondless SMD packages, like the QFN and LGA packages. They uniquely can accommodate double-sided assembly of the internal leadframe or substrate. The double-sided assembly is made possible by the Propri-etary Bottom-Side Cavity(s) Architecture.
Standard Type 1 QFN-µMaxPak packages have the power die soldered into the bottom cavity(s) of the QFN leadframe (or LGA substrate) with the die gate (G) & source (S) pads soldered to the leadframe, leaving the die drain (D) pad(s) exposed at the bottom of the package. The Type 1 configuration is Ideal for Vertical Integration with additional components on the top of the leadframe (or substrate). These components can include bump-chip gate-drivers (GD), isola-tors and passive components. The cross-section in Figure 1 is a Type 1 example with a vertical WBG die in the bottom cavity and GD IC on top of the leadframe.
Inverted Type 2 QFN-µMaxPak packages have the power die sol-dered into the cavity(s) in the bottom of the leadframe (or substrate), but with the D pad(s) soldered to the leadframe and the G & S pads exposed at the bottom of the package. The Inverted Type 2 µMaxPak allows horizontally integrated gate-drivers, which can be external or internal (Horizontal gate-driver options not shown). See cross-section in Figure 2.
Thin Type 3 QFN-µMaxPak packages are Inverter Type 2 µMaxPak with the leadframe also exposed at the top of the package. They are robust, and accommodate both top & bottom heatsinks. They are well suited for EV Inverters, which are usually clamped between top & bot-tom cold plates. See cross-section in Figure 3.
Q7: Why use modified QFN µMaxPak packages for power WBG packaging?A7: The QFN structure is simple, robust, reliable and low cost. The mature commercial QFN/DFN package technology is flexible, and available at many contract assemblers. Its flexibility allows easy size and layout customization without hard tooling, allowing packages to be optimized for each product configuration, optimizing performance and reducing tooling NRE, design/process risk, and new product time-to-market. These advantages are built on inherent features like being leadless and wire bondless, with a heavy single piece Cu leadframe that provides connections from the top of the die to the bottom-side pads. These features contribute to the exceptionally low inductance, electrical resistance and thermal resistance.
Q8: Why are Chip-Scale µMaxPak QFN packages so Cost Effec-tive?A8: The QFN packages are small, and they have simple internal structures with one-piece leadframe. They do not require internal DBC for CTE buffer, and they do not require complex external leads and terminals. Existing Power QFN packages often replace wire bonds with solder clips, increasing assembly complexity, NRE, interfaces, cost, size and reliability risks. The proprietary µMaxPak architecture eliminates solder clips and easily accommodates multi-chip configura-tions like half-bridge (HB), full-bridge (FB) and 3-phase bridge (3P).
Q9: What are the differences between Leadframe-Based and Substrate-Based µMaxPak?A9: Leadframe and Substrate µMaxPak use the same proprietary architecture having die cavities(s) on the bottom of the package and components or supplementary heatsink/cold-plate on the top. Al-though QFN are LGA packages, and LGA can be QFN packages, it is common to call Leadframe µMaxPak packages “QFN” and Substrate µMaxPak packages “LGA.”
Q10: What are pros & Cons of QFN versus LGA µMaxPaks?A10: Leadframe-Based QFN µMaxPaks utilize thicker and wider Cu interconnects, creating an optimum 3-D geometry. This provides the lowest current-loop-inductance, lowest current-loop electrical resistance, and is mechanically more robust. It also provides the lowest thermal resistance (Rjc), and highest heat capacity to better accommodate repetitive peak power pulses. The simple one-piece leadframe assembly provides lower costs than the LGA-µMaxPak.Substrate-Based LGA µMaxPaks also have low current loop-inductance and low electrical resistance, but not as low as the QFN µMaxPak. Their substrates are PCBs or laminates, and use high tem-perature laminate materials with Cu foil and Cu via for electrical and thermal connections. Since Cu foil and via are not as heavy as solid Cu leadframes, electrical and thermal internal connection of the LGA µMaxPaks can never perform quite as well as QFN µMaxPaks. That said, since both µMaxPak packages have exposed die pads on the bottom of the package as primary heat dissipation path, the thermal resistance Rjc of the LGA µMaxPak is still nearly as good. LGA sub-strates can provide thinner traces and more complex interconnects, which allows more complex circuitry and vertical integration. Circuit density can be further increased in LGA µMaxPak by embedding pas-sive component in the laminate substrate.
WIDE BAND GAP
Figure 1: Type 1 example with a vertical WBG die in the bottom cavity and GD IC on top of the leadframe.
Figure 2: Inverted Type 2 µMaxPak allows horizontally integrated gate-drivers.
Figure 3: QFN-µMaxPak packages are Inverter Type 2 mMaxPak Table 2: Comparison, QFN-µMaxPak versus LGA-mMaxPak.
www.bodospower.com July 2018 Bodo´s Power Systems® 45
CONTENT
Q11: Can µMaxPak Architecture accommodate both Lateral and Vertical WBG die?A11: Yes it can. The Vertical die structure is most common for power SiC, and ideal for Near Chip-Scale packaging, because all power pads provide both electrical and thermal external connections, and they provide the best high-voltage separations with D pads on one side and S/G pads on the other.Power GaN die are typically Lateral structures with all electrical con-nections on the top-side of the die, while the other side contain only thermal pads. With G, S & D on the same side, the pads are smaller and the high voltage spacing (S to D) must be large, which can limit maximum current and voltage. An exception is GaN-on-SiC die, which can accommodate thru-SiC via to bring the D to the other side. GaN-on-SiC is more expensive today, but better thermally.
Q12: Can µMaxPak accommodate both Normally-Off and Normally-On WBG die?A12: Yes it can, but Normally-On die does increase package complexity. Normally-On die are typically managed with a Cascoded circuit, which requires a low-voltage MOSFET die in series with the high-voltage WBG die, and extra internal connection(s). Normally-Off die are easiest to manage. Today most power WBG die are Normally-Off. However, Cascoded switches can have advantages in some applications?
Q13: How are Signal Pads and Connections managed in µMaxPak?A13: WBG power and signal die pads vary by manufacturer, ratings and product. Therefore, it is recommended that the die pads and µMaxPak pads be co-designed for performance, manufacturability and reliability. Ideally all connections will be soldered to the QFN lead-frame pads or LGA foil pads. Signal pads often include gate, sense and current sense pads. External signal pads will be different with integrated gate driver(s).
Q14: Can µMaxPak packages accommodate Multiple WBG power die?A14: Yes it can. The µMaxPak architecture can accommodate single or multiple WBG switches, and each switch can accommodate paralleled WBG die and anti-parallel diodes. Common configurations for multiple switch µMaxPak can be HB, FB and 3P, and they can be easily customized for special configurations. FB and 3P are best suited for lower power levels, and at higher power levels it can be more practical to create them with multiple HB packages. In general, the Half-bridge provides higher power density and lower current-loop inductance than two single-switches. An example of a SiC µMaxPak
HB is shown in Figure 4A (Cross-Section) and Figure 4B (Bottom-side pads). This HB structure has a non-inverted high-side power die and an inverted low-side power die enabling it to be directly soldered on the positive bus & negative bus to minimize loop-inductance.
Q15: Are µMaxPak packages suitable for Higher WBG Operating Temperatures?A15: GaN can operate at higher temperatures than Si, and SiC at higher temperatures than GaN. Today‘s power GaN and SiC devices are usually not operated above 150°C, being primarily limited by increased Rds(on) at higher junction temperatures, but both WBG materials can accommodate much higher operating temperatures as WBG devices evolve.The µMaxPak package operating temperature can be limited by the choice of solder and molding compound, and by CTE mismatch. LGA µMaxPak packages can additionally be limited by choice of substrate laminate material.Today’s typical QFN & LGA materials can accommodate temperatures Tj(max) of 185°C, with typical Sn/Ag/Cu solders, Hi-Tg molding com-pounds, and laminates like BT resins. There are higher temperature materials available today, which can allow µMaxPak Tj(max) up to about 225°C in the µMaxPak packages
Q16: Is it true that µMaxPak technology Improves System Archi-tecture and Performance?A16: Yes it does. The small & thin packages significantly reduce system size and weight. Smaller power SMD packages can be placed close together eliminating heavy leads, bulky terminals and large spacing/creepage distances. This reduces Integrated System size, weight and cost. The total current loop-inductance and resistance of integrated system interconnects are typically larger than that of the packages. The µMaxPak packages virtually eliminate most high cur-rent/voltage internal connectors and interfaces. Furthermore, tighter control, feedback and protection circuitry are more easily implement-ed.
Q17: Are UL/EN isolation & Safety Regulations met by Near Chip-Scale packages?A17: Yes, the smaller UL minimum spacing & creepage distances are already used inside traditional power Si IGBT modules, where die, traces and passive chips are small & thin, allowing easier coating or potting. Likewise the small and thin µMaxPak packages make coating, potting and under-fill much easier, especially in enclosed integrated systems. Under-fill, coating and potting materials must be suitable for Pollution Degree 1 minimum spaces & creepage distances, like those used inside traditional high power Si IGBT modules.
In Summary:Power WBG (GaN & SiC) devices need Near-Chip-Scale SMD pack-ages to reach today’s WBG devices full performance, efficiency and power density potential. Inevitably such packages will be absolutely required as power WBG devices evolve toward the inherent potential of GaN & SiC materials. Increased WBG power density and efficiency has enabled new packages like the µMaxPak, and the µMaxPak packages enable the power WGB full performance, efficiency and power density. Furthermore, Near Chip-Scale µMaxPak packages reduce unit cost, tooling NRE and new product time-to-market. These packages are compatible with standard QFN & LGA assembly technologies, which are reliable, robust and available today. Small SMD packages can also simplify System structures increasing system performance, efficiency and power density, and reducing system cost.
POWER MANAGEMENTWIDE BAND GAP
Figure 4a: HB Cross-Section View
Figure 4b: HB Bottom Pad View
Bodo´s Power Systems® July 2018 www.bodospower.com46
CONTENTMEASUREMENT
IntroductionMany switching converter designs are driven by stringent output noise requirements. The demand for low output noise has pushed designers to implement heavy output filtering, such as using several capaci-tors at the output. With increased capacitance across the output rail, excessive inrush current may become an issue during startup that can potentially lead to inductor saturation or damage of the power switch.
The power switch of a monolithic switching regulator is internal to the chip, as opposed to a switching controller. This is an ideal approach on point-of-load switching converter applications, because of advan-tages like smaller PCB footprint and better design of the gate-drive circuit. This means that protection against overcurrent becomes a necessity to avoid damaging not only the switch, but the regulator chip itself. The ADP5070 dual, high performance dc-to-dc monolithic switching regulator is an example, as illustrated in Figure 1.
To prevent damage during output overload condition or startup when high current flows through the internal switch, switching regula-tor manufacturers employ different current-limiting techniques on monolithic switching regulators. Despite the existence of current-limit protection, the switching regulator may not properly operate as intend-ed, especially during startup. For instance, with hiccup mode as the current-limit protection, at initial power-up when the output capacitor is still fully discharged, the switching regulator may enter hiccup mode, causing a longer start-up time or potentially not starting up at all. The
output capacitor may pull excessive inrush current that, in addition to the load, causes the inductor current to go high and hit the hiccup mode current-limit threshold.
Overcurrent Protection SchemesIntegrating the power switch inside switching converters makes the current-limiting protection a basic function. Three commonly used current-limiting schemes are: constant current-limiting, foldback current-limiting, and hiccup mode current-limiting.
Constant Current-LimitingFor a constant current-limiting scheme, the output current is held con-stant to a specific value (ILIMIT) when an overload condition occurs. As a result, the output voltage drops. This scheme is implemented by using cycle-by-cycle current-limiting that utilizes the peak inductor current information through the power switch to detect the overload condition.
Figure 2 shows a typical inductor current of a buck converter during normal and overload conditions for the peak current-limiting scheme. During overload condition, as illustrated by ILIMIT, the switching cycle is terminated when the peak current detected is greater than the predetermined threshold.
In the constant current-limiting scheme, the output current is main-tained at ILIMIT, resulting in high power dissipated in the regulator. This power dissipation causes the junction temperature to increase, which may exceed thermal limits.
POWER SUPPLY
Preventing Start-Up Issues Due to Output Inrush in Switching
ConvertersSwitching converters on applications demanding reduced output noise may encounter
delayed startup, or may not startup at all, due to excessive output inrush. Output inrush current, attributed to inappropriate design of output filters and its impact, can be minimized by increasing the soft start time, increasing the switching frequency, or
decreasing the output capacitance. In this article, practical design considerations toward preventing start-up issues due to excessive output inrush will be presented.
By Fil Paulo Balat, Jefferson Eco, and James Macasaet, Analog Devices
Figure 1: Switching converter using ADP5070 regulator.
Figure 2: Cycle-by-cycle constant current-limiting.
www.fujielectric-europe.com
Lower system costs
Less harmonics
Higherefficiency
Reduction of noise
FUJI RB-IGBT implements RB capability without additional diode
Lower losses
Higher power density per module
Authentic RB-IGBT technology is only available by FUJI Electric
Available as 600 V, 900 V, 1200 V
MAIN FEATURES REVERSE BLOCKING IGBT
Fuji 3-Level IGBT ModulesHighEfficientSolutionsfor3-LevelApplications
RB-IGBT = Reverse Blocking IGBT
RB-IGBT
AC-Switch
Bodo´s Power Systems® July 2018 www.bodospower.com48
CONTENT
Foldback Current-LimitingThe foldback current-limiting scheme partially solves the issue with constant current-limiting, helping to keep the transistor in its safe operating area under fault or overload conditions. Figure 3 shows the comparison of the VOUT vs. IOUT response curves between the constant and foldback current-limiting schemes. The reduction in out-put current (IOUT), as opposed to constant current-limiting, reduces power dissipation, thus reducing the thermal stress on the switching converter.
The disadvantage of this scheme is that it is not fully self-recoverable. Due to its foldback nature and depending on the nature of the load, the operating point could fall into the foldback region toward the short circuit operating point once the current-limit threshold was reached or exceeded. This would require power cycling the part or re-enabling the part to get back to the normal operating condition.
Hiccup Mode Current-LimitingIn a hiccup mode current-limiting scheme, the converter switching goes into a series of short burst of pulses followed by sleep time—hence the name hiccup. Once an overload condition occurs, the switching converter enters hiccup mode, where sleep time refers to the switch being turned off for a predefined period of time. At the end of the sleep time, the switching converter attempts to start again from soft start. If the current limit fault is cleared, the device resumes normal operation—otherwise, it re-enters hiccup mode.
The hiccup mode current-limiting scheme overcomes the drawbacks of the two overcurrent protections discussed. Firstly, it solves the thermal dissipation problem, as the sleep time reduces the average
load current that allows the converter to cool down. Secondly, it allows smooth autorecovery once the overload condition is removed.
However, some issues may arise if the hiccup mode detector is ac-tive during startup. Excessive inrush current, in addition to the load current, may cause the inductor current to go beyond the current limit threshold, which triggers hiccup mode and prevents the converter from starting up. For example, the negative output of the inverting regulator of ADP5071, configured to have an output voltage of –15 V and 100 mA output current with around 63 μF of total output capaci-tance, is not starting up after powering from a 3.3 V power supply. The negative rail is under hiccup mode, as shown in Figure 4, which is triggered by the large output inrush current. Inductor current peak goes to around 1.5 A, exceeding the typical current-limit threshold of around 1.32 A.
POWER SUPPLY
Figure 3: VOUT vs. IOUT curve of constant and foldback schemes.
Figure 4: ADP5071 inverting regulator in hiccup mode.
Figure 5: ADP5070 inverting regulator delayed startup.
Figure 6: Basic switching topologies>
www.bodospower.com July 2018 Bodo´s Power Systems® 49
CONTENTPOWER SUPPLY
Also, if there’s excessive inrush due to large output capacitance, the converter may get unexpected longer start-up time, as shown in Figure 5.
Inductor Current in Switching ConvertersInductor Current AverageIn nonisolated switching converters, the location of the inductor defines the converter topology. With a common ground reference be-tween input and output, there are just three distinct rails possible for the position of the inductor: the input, the output, and the ground rails.
Refer to the three basic switching topologies shown in Figure 6. When the inductor is at the output rail, the topology is a buck. When it is at the input rail, the topology is a boost. And when the inductor is at the ground rail, the topology is an inverting buck-boost.
During steady-state condition, the average current (IOUTRAIL) on the output rail must be equal to the output current since the average current on the capacitor is zero. For a buck topology, IL-AVE = IOUT. However, for the boost and the inverting buck-boost topologies, ID-AVE = IOUT.
For boost and inverting buck-boost topologies, it is only during switch-off time that current flows through the diode. Therefore, ID-AVE = IL-AVE during switch turn off. Refer to Figure 7 in deriving the average inductor current with respect to the output current. The rectangular area in green during switch-off time is the average diode current ID-AVE, with height equal to IL-AVE, and width equal to TOFF. This current all goes to the output and, therefore, can be translated into a rectangular area averaged to a width of T and with height IOUT.
Table 1 shows a summary of the average inductor current IL-AVE and switching duty cycle D. Based on the equations, the inductor cur-rent will be at its maximum when the input voltage is at its minimum providing maximum duty cycle and when the output current is at its maximum.
Inductor Current PeakFigure 8 shows inductor voltage and current waveforms of a buck-boost inverter in a steady-state condition in continuous conduction mode of operation. As to any switching topology, the amount of induc-tor current ripple (ΔIL) can be derived according to the ideal inductor Equation 2.
In switching converter applications where the inductor current is trian-gular and exhibits a constant rate of change, and therefore constant induced voltage, (ΔIL/Δt) can be used in the inductor equation, as found in the rearranged Equation 3. Inductor current ripple is deter-mined by the applied voltseconds to the inductor and the inductance.
Switch turn-on time can be easily related to duty cycle and switching frequency as in Equation 4. It is therefore more convenient to use voltsecond products during switch turn-on than switch turn-off in the succeeding formula.
Table 2 shows a summary of the inductor current ripple in the three different topologies. The voltseconds product term tON, based on Equation 3, is replaced by Equation 4, and the term VL-ON is replaced by the induced voltage across the inductor according to topology.
Topology Inductor Current Ripple
Buck
Boost
Buck-boost inverter
Table 2: Inductor Current Ripple
Looking back at the steady-state inductor current in Figure 8, it will be observed that the inductor current average simply lies at the geometri-cal center of the ramp or the swinging of waveform at the point ΔIL/2.
Figure 7: Diode current of boost or buck-boost inverter.
Table 1. Average Inductor Current and Duty Cycle Inductor Current Peak
Topology Inductor Current Duty Cycle
Buck
Boost
Buck-boost inverter
Figure 8: “Swing” of the inductor current.
Bodo´s Power Systems® July 2018 www.bodospower.com50
CONTENT
Therefore, the inductor current peak is the sum of inductor current av-erage and half of the inductor current ripple, as shown in Equation 5.
(5)
Capacitor Inrush CurrentThe charging current or displacement current equation of the capaci-tor is defined in Equation 6. It states that current flows through a capacitor in correspondence to a rate of change of voltage across it.
The capacitor charging current should be considered when choosing output capacitor values for switching converters. At startup, assuming that the capacitor voltage is equal to zero or no capacitor charge, the output capacitor will begin to charge and draw as much current depending on the total capacitance and rate of change of the capacitor voltage, until the capacitor voltage reaches steady-state.
The rising of the output voltage in switching converters is a controlled ramp with constant slope so the rate of change equation can be simplified, as shown in Equation 7. Change in output voltage (ΔV) corresponds to the output voltage at steady-state and Δt corresponds to the time it takes for the output to reach its final value during startup, or what is commonly called soft start time.
If there is too much output capacitance (COUT) or if the soft start time is small, the current demanded from the regulator ICAP may be too high, which may cause problems with the converter operation. This large amount of current impulse is referred to as the inrush current. Figure 9 shows capacitor inrush current and output voltage during the startup of an inverting buck-boost converter with an output of 15 V, 10 µF output capacitor, and 4 ms soft start time.
Inductor Current Peak at StartupA simple boost converter circuit is shown on Figure 10. When the transistor switch is on close switch, current flows through the inductor while no current flows through the output rail. It is the discharging phase of COUT where the discharging current (ICAP) goes to the output while none goes through the reverse-biased diode. When the transistor’s open switch is off, current ID flows through the diode.
By Kirchhoff’s current law, the current through the output rail (ID) must be equal to the sum of the current flowing through the output capaci-tor (ICAP) and output load (IOUT).
This is described by Equation 8.
This equation applies during every charging phase or when voltage is rising across the capacitor. Therefore, it is also applicable during the startup of a switching converter when the initial state of the output capacitor is discharged or when the output voltage is not yet in the steady-state value.
The inductor current peak during startup can be defined using Equa-tion 5 and includes the impact of inrush current due to output capaci-tor. Equation 8 will be applied into the IL-AVE equations in Table 1, replacing IOUT with IOUT + ICAP. Inductor current peak equations during startup are summarized in Table 3.
Topology Inductor Current Peak
Buck
Boost
Buck-boost inverter
Table 3: Inductor Current Peak at Startup
For any of the three topologies, the inductor current peak is propor-tional to IOUT. In terms of output current, the output capacitor must be designed at full load conditions.
Most applications require operation within a range of input voltage. So against input voltage, there is a difference between the buck and the other two topologies in terms of the magnitude of the dc and ac components voltage of the inductor current. This can be understood better through Figure 11. For the buck, as input voltage goes up, ac component voltage goes up. Average current is equal to output cur-
POWER SUPPLY
Figure 9: Output capacitor inrush current.
Figure 10: Boost dc-to-dc converter circuit.
www.bodospower.com July 2018 Bodo´s Power Systems® 51
CONTENTPOWER SUPPLY
rent, so the dc component voltage remains constant. Inductor current peak is therefore maximum at maximum input voltage.
For the boost and buck-boost inverter, as input voltage goes up, ac component voltage goes up—but dc component voltage goes down because of the impact to the average current by the duty cycle, as shown in Table 1. The dc component voltage dominates, so the induc-tor peak current is at its maximum rating at minimum input voltage. In terms of input voltage, design of the output capacitor must be done at the maximum input voltage for buck and at minimum input voltage for boost and buck-boost inverter.
Mitigating Impact of InrushOutput Capacitor FilterAs shown in previous sections, too much capacitance at the output causes high inrush current that may cause the inductor current peak to reach the current-limit threshold during startup. Therefore, the right amount of capacitance is necessary to achieve smallest output volt-age ripple, while maintaining good converter start-up performance.
For buck converters, the relationship between COUT and the peak-to-peak voltage ripple is defined by Equation 9.
For boost and inverting buck-boost converters, the relationship between COUT and the peak-to-peak ripple is defined by Equation 10.
Note that these equations neglect the effect of parasitic elements on the capacitors and inductors. These, in line with the rated specifica-tions of the converter, can help the designer in limiting the capacitors added to the output. A good balance of filtering level and output inrush current are key considerations.
Second-Stage LC FilterIn certain cases, switching transients occur on the output voltage, as shown in Figure 12. If the magnitude is significant, it becomes an issue to the output load. The switching spikes are primarily caused by the switching transitions of the current on the output rail, which is the diode current for boost and buck-boost inverters. They can be magni-fied due to the stray inductance on the PCB copper traces. Because the spikes are in much higher frequency than the switching frequency
of the converter, the peak-to-peak ripple cannot be reduced by the output filter capacitor alone—additional filtering is needed.
Figure 12 shows the periodic switching action of the inductor in a boost converter represented by the blue trace, and the output voltage ripple represented by the yellow trace. High frequency transients are observed within the ripple voltage upon the switching transitions of the inductor current.
A great article on analog.com that provides more insight on how to reduce the high frequency transients by second-stage LC filtering is “Designing Second Stage Output Filters for Switching Power Sup-plies” by Kevin Tompsett.
Ripple MeasurementThe right measurement method is also important when getting the output voltage ripple. Incorrect measurement setup can result in inaccurate and high voltage ripple readings, potentially leading to over-design of the output capacitor. It is easy to make the mistake of putting too much capacitance at the output in the hopes of reducing voltage ripple without realizing the tradeoffs.
An application note done by Aldrick Limjoco entitled “Measuring Out-put Ripple and Switching Transients in Switching Regulators” should be of help. See the references for details.
Soft Start FeatureFor boost and inverting buck-boost, a bigger impact is dictated by the increase of the dc component voltage of the inductor current. At lower input voltage, the increase in the duty cycle causes a big increase in the inductor current average as shown in the (1-D) factor in equations in Table 3—this is also illustrated in Figure 11. This means that the inrush current of the output capacitor has to be significantly reduced. It is achieved by increasing the soft start time (tSS) in Equation 7.
Most switching regulators (tSS) have a soft start feature that refers to its capability, in order to give designers option to adjust the rise time of output voltage during startup. Changing the value of a single resistor is often the convenient method of adjusting the soft start time. Figure 13 shows the start-up waveforms of a buck-boost inverter. A significant 25% decrease in inductor current peak can be seen by a change in soft start time from 4 ms to 16 ms.
Increasing the Switching FrequencyFigure 14 illustrates the impact to inductor current by change switch-
Figure 11: Inductor current against input voltage.
Figure 12: Output voltage ripple with switching transients.
Bodo´s Power Systems® July 2018 www.bodospower.com52
CONTENT
ing frequency (fSW). Assuming that duty cycle D and output current are constant, the ac component voltage of the inductor current or Δ IL/2 is affected by change in fSW, while the dc component voltage is
not. Inductor current peak being inversely proportional is therefore lower at higher switching frequencies.
ADP5070: An Example
How Large Can the Output Capacitance Be?ADP5070 is a monolithic, dual-boost and inverting buck-boost regula-tor with hiccup mode current-limiting scheme as the overcurrent protection. Some customers forgot to consider the trade-off of putting too much capacitance at the output, especially at high duty cycle operating condition or at the minimum input voltage. This usually has led to start-up issues at the inverting output, because the inverting buck-boost regulator is designed with lower current-limit threshold than the boost regulator.
Figure 15 can be used as aid for application engineers as to how much capacitance is allowed at the output of ADP5070 to avoid start-up issues. Max COUT is shown vs. max IOUT on different input and output voltage combinations, using the direct relationship of inductor peak current to output current, including inrush in Table 3’s equation. It will help in the design limits of the output capacitor values after having considered the optimum VOUT ripple performance using either Equation 9 or Equation 10.
Both graphs were computed based on the shortest tSS and the current-limit threshold of the regulator. External components were
Figure 13: Inductor current vs. soft start time.
Figure 14: Factors affecting inductor current peak.
Figure 15: Maximum COUT vs. maximum load current.
POWER SUPPLY
www.bodospower.com July 2018 Bodo´s Power Systems® 53
CONTENT
chosen to be of much higher current handling capability than the regulator. In other words, the numbers in these graphs will definitely increase in magnitude if the tSS were increased.
For applications requiring higher output load current, ADP5071 should be considered. ADP5071 is designed with higher current-limit thresh-old than ADP5070 for both boost and inverting buck-boost regulators.
Computed vs. Measured DataFigure 16 shows the start-up waveforms of the inductor induced volt-age and current of the inverting regulator, while the data that follows in Figure 17 show the inductor current data both by computation using equation in Table 3 and measured bench data.
The data proves that inrush current is greatly reduced if tSS is increased, thereby lowering inductor peak current. At 4 ms tSS, the inverting regulator is already hitting the current-limit threshold of 0.6 A and has a tendency of having start-up issues. The remedy is to increase tSS to 16 ms to give enough inductor peak current margin.
ConclusionThis article has shown that careful design of the output filter capacitor is important in designing switching converters. Good knowledge of the factors influencing the inductor peak current during startup helps avoid start-up issues. Boost and inverting buck-boost converters are more prone to these issues, especially those using the hiccup mode current-limiting scheme.A direct relationship between the inductor peak current and output inrush current has been provided. It will prove to be useful when de-signing the output capacitors while keeping track of the inductor peak current against current-limit threshold. For the same output condi-tions, output inrush current can be minimized by increasing the soft start time or the converter switching frequency.This article comes as a reference material when designing a dc-to-dc switching converter using the ADP5070/ADP5071/ADP5073/ADP5074/ ADP5075 series of monolithic switching regulators of Analog Devices.
References• Erickson, R.B. and D. Maksimovic. Fundamentals of Power Elec-
tronics, Second Edition. Springer, 2001.
• Kirchhoff, Gustav. “Kirchoff’s Current Law.” Electronic Tutorials.
• Limjoco, Aldrick S. Application Note AN-1144 Measuring Output Ripple and Switching Transients in Switching Regulators. Analog Devices, Inc., January, 2013.
• Tompsett, Kevin. “Designing Second Stage Output Filters for Switching Power Supplies.” Analog Devices, Inc., February, 2016.
Fil Paulo S. Balat [[email protected]] works as an applications engineer in ADI for power management, dc-to-dc products and has 17 years of experience in switching power supply design. His vast experience also involves flyback, quasi-resonant ac-to-dc convert-ers for mobile phone chargers. Fil received his bachelor’s degree in electronics engineering from the Xavier University Ateneo de Cagayan, Philippines.
Jefferson Eco [ [email protected]] joined Analog Devices Philippines in May 2011 and currently works as an application development engineer. He graduated from Camarines Sur Poly-technic College Naga City, Philippines, with a bachelor’s degree in electronics engineering.
James Jasper Macasaet [[email protected]] received his B.S. degree in electronics engineering from the University of Santo Tomas, in Manila, Philippines, in April 2013 and is currently continuing his master’s in engineering at the University of Limerick. He is currently a systems applications development engineer in Analog Devices Philippines, in Metro Manila, Philip-pines, focusing on characterization of ADI high performance power management products over a range of mixed-signal products, such as DACs and ADCs. He has coauthored an application note on the topic of powering dual-supply, precision DACs in single-supply systems.
www.analog.com
POWER SUPPLY
Figure 16: Inductor current and induced voltage at startup.
Figure 17: Inductor current: Computed vs. measured.
Bodo´s Power Systems® July 2018 www.bodospower.com54
CONTENT
ABSTRACT Achieving 99% efficiency for PFC front-end stages has been a domain of wide band gap (WBG) devices due to reverse recovery challenges associated with high-voltage (> 500V) silicon superjunction MOSFETs. Such limitation, however, doesn’t hold true for low-voltage (< 150V) silicon variants deployed in multi-level bridgeless totem pole PFC which has been researched and developed by ICERGi. This article demonstrates low-cost, extremely-high-efficiency, and Si-based implementation of PFC stages enabled by innovative ICERGi gate drive and digital control technologies.
Totem Pole topology for PFCRecent commentary has pointed out the rel-evance of the totem-pole topology for power factor correction. In a topology context this is a very simple circuit, and is as shown in figure 1. It also has the advantage of using a single PFC inductor as compared with the dual semi-bridge approach, which requires a dual inductor.
The active switches are stacked as shown – which immediately rules out silicon super-junction technology due to the Qrr reverse recovery challenge – and the extreme shoot-through current and associated losses that
are a consequence of the high Qrr values. This is the main challenge to its usage and may explain the historical preference for con-ventional “bridged” boost solutions in spite of the power loss due to the extra diode drop in these “bridged” types.
The logical approach is to use devices with low Qrr as the active switches – and hence usage of GaN or SiC (collectively “wide bandgap”, or “WBG”) types. Getting soft switching in a PFC context is difficult – given the wide range of operating conditions – and hence hard switching is the norm. The low capacitance of WBG devices assists here, in limiting losses under hard switching, albeit with challenges in dVdt and dIdt manage-ment. Getting 99% efficiency with such device types is no longer seen as a “rare” ac-complishment! Design concerns here revolve around the newness of WBG devices, the lack of dual sources and the very high cost of devices.
99% Efficiency with Silicon!Getting 99%+ efficiency performance with low-cost, mature, multisourced silicon switch technology is the desire, rather than having
sole reliance on WBG devices. And it IS possible, and well proven in multiple designs.
A topology set that can use silicon VERY well (performance AND cost) in a totem-pole Power Factor Corrector using multilevel approaches. Multilevel approaches have been used for some time at higher volt-age and power levels, and are well proven with a considerable body of accumulated industry experience and analysis captured in academic publications. Usage of multilevel technologies in a totem-pole power-factor correction stage is as in figure 2.
Multilevel usage has many additional attrac-tive characteristics including:- Inherent frequency multiplication – where
each cell switches at a low frequency with the effective frequency being a multiple of the cell frequency
- Low-voltage switching – where each cell is switched at low voltage. Alongside the low switching frequency property as cited above, this can result in very efficient operation with low switching losses
- Inherently lower dVdt and dIdt – of value in limiting EMI/EMC effects
- The topology operates with phase cancel-lation, implying reduced voltage applied to the PFC inductor for reduced time. As inductor size tends to be a function of ap-plied volt-seconds, there is the opportunity for making this key component small and neat as compared with the traditional “large and bulky” part.
Multilevel conversion also uses mid-voltage FETs – so a universal-line power factor cor-rection stage with a nominal output voltage of ~400V, can use for example 2x300V, 3x200V, 4x150V or 6x100V types. Such FET types are readily available from multiple suppliers.
POWER CONVERSION
99% PFC Efficiency – with Silicon, at Low Cost!
Power factor correction (“PFC”) functionality is a fundamental building block in all midrange power supplies for electronic equipment. Our work primarily relates to power levels from about 300W – below which efficiency rarely gets commercially rewarded –
through to 3kW, which corresponds approximately to the limit for practical single-phase operation.
By Garry Tomlins and Dr. Trong Tue Vu PHd, ICERGi Ltd.
Figure 1: “Basic” Totem Pole Circuit Imple-mentation Figure 2: Usage of Multilevel Technology in
Totem-Pole Power Factor Correction
100% of our 947D DC Link capacitors are conditioned
at accelerated temperature and voltage before they
leave the factory. We’ve demonstrated that this process
eliminates infant mortal failures
and produces the most thermally
and electrically stable DC Link
Capacitors available today.
If that’s the kind of reliability
you’re looking for, trust the
one company where it’s
burned in.
Call for samples with thermocouples.
www.cde.com | 508-996-8561
CDEENERGIZING IDEAS
CORNELLDUBILIER
100% of our 947D DC Link capacitors are conditioned
at accelerated temperature and voltage before they
leave the factory. We’ve demonstrated that this process
eliminates infant mortal failures
and produces the most thermally
and electrically stable DC Link
Capacitors available today.
If that’s the kind of reliability
you’re looking for, trust the
one company where it’s
burned in.
Call for samples with thermocouples.
www.cde.com | 508-996-8561
CDEENERGIZING IDEAS
CORNELLDUBILIER
Bodo´s Power Systems® July 2018 www.bodospower.com56
CONTENTPOWER CONVERSION
Most characteristics of lower-voltage FETs are also considerably “nicer” than those of their 600V high- voltage cousins. Specifically, reverse recovery charge is much lower (as well as which it can be recovered more easily), turn-off losses are low, and one can take advantage of standard manufacturing-friendly 5mm x 6mm solder-tab packages. And the “raw” power train silicon cost usually “benchmarks” at a lot less than values for SuperJunction silicon – even if this latter type were feasible for usage in a totem-pole configuration in the first instance! 4x15mR devices with 150V rating will typically be lower in cost than a 60mR 600V device. Naturally a more realistic comparison for high efficiency is with good-quality GaN or SiC devices, where the cost differential in FAVOUR of the multilevel silicon can be VERY large. So – the cost equation can work out very much in FAVOUR of multilevel technology usage. On the favourable side one has lower device costs and reduced magnetic cost. On the other side one has the cost of the flying capacitor and the cost of control and drive – and more later on this topic!
PerformanceThis is as summarised in figure 3 for two deployments. These are initially a 700W industrial-medical supply with thyristor-based inrush management, and a 3kW server-optimised deployment. This unit uses line-frequency synchronous rectification – basically using silicon active device for the “return path” as was shown in figure 2. In both cases 3-level solutions are used with similar control hardware and scaled power hardware. 300V switches have low market demand and are difficult to obtain with attractive characteristics, and hence a composite device using commodity 2x 150V devices, series driven, is used. Optional switching energy recovery magnetics can be used to limit dVdt and dIdt in the context of system EMI.
These implementations also reflect very compact designs, with the small magnetics and compact control allowing high implementation density without efficiency tradeoffs.
Figure 4(a) shows an overall end-end 700W design for industrial and medical usage (the PFC stage of which is characterised above) and figure 4(b) shows the 3kW implementation.
Drive and ControlRecent developments allow driving of the multilevel switches with low component count and at low cost, integrating drive and control into a single “daughter card”. These ICERGi® developments have allowed the inherent advantages of multilevel technologies to be realised in a practical and cost-effective fashion at relevant power levels, typically the 300Wà3kW range as needed for most of the single-phase power factor correction marketplace. A low-cost commodity ARM-based digital controller works well in control of multilevel converters, and is designed for optimal usage of the driver characteristics. The overall control and drive functionality is captured at low cost in this com-ponent-type “daughter board” as shown in figure 5. This connects directly to the power MOSFETs and effectively integrates control and drive elements in a 35mm x 26mm card.
SolutionsA key ICERGi “mission” is to make multilevel solutions practical and very cost effective for usage in high-volume ACDC power conversion.
ICERGi makes available:- Overall control modules – essentially as shown in figure 5.- A “Chip Set” for user implementation, along with printed circuit
board samples and Gerber layout detail. The chip set comprises the controller chip along with 8 driver chips, one for each FET in a “4+4” device stack.
Figure 2: Showing Reported Performance
Figure 4(a): 700W Universal Line to 12V Converter in 76mm x 127mm (3”x5”) footprint
Figure 4(b): 3kW PFC Stage in 1U/80mm format
www.bodospower.com July 2018 Bodo´s Power Systems® 57
CONTENTPOWER CONVERSION
The ICERGi devices are priced such that high efficiency using multilevel silicon is considerably more cost-effective than using wide-bandgap approaches.
A particular point of comparison relates to a multilevel implementation vs the established interleaving approaches, an example of which is in Figure 6.
If one takes for example a two-phase interleaving design, there are two magnetics, each of which has the full volt-seconds. In the multilevel approach, the volt-seconds effectively cancel before the magnetic element – and thus the magnetic element can be made a lot smaller. And switching in the conventional interleaved approach is at full-frequency and full-voltage. This involves materially greater loss than in the multilevel case, with reduced cell switching frequency and voltage.
The interleaved approach also requires a low-Qrr Diode for each phase – again part of the cost equation. Finally – the conventional interleaved approach has an extra diode in the conduction path – add-ing further to losses. It is thus to be expected that a multilevel solution can be at cost parity or below the cost of an interleaved design, with losses reduced by 50% - typically going from 95% stage efficiency to 97.5% at low line. So – winning on the three criteria of cost, size and efficiency is a property of the ICERGi multilevel approach when benchmarked against the most relevant alternative topology!
SummaryYes, established, low-cost silicon CAN achieve better than 99% ef-ficiency in power factor correction usage! Multilevel solutions also have sustainable competitive advantage in magnetics size and EMI/EMC performance, as well as an expected long-term cost advantage over WBG (SiC and GaN) devices.
Contact Detail ICERGi is implementing its technology – modules and/or chip sets – through alliances with leading semiconductor and power conversion businesses.Contact details are:Garry Tomlins – [email protected] McCarthy – [email protected]
www.icergi.com
Figure 5: Control Card and Interfaces
Figure 6: Showing the established Interleaved-Boost Implementation
ABB is a world’s leading suppliers of high-power semiconductors, setting standards for quality and performance. Our unique knowledge in the field now expands to industry standard medium-power IGBT. The 62Pak phase leg IGBT modules, featuring 1700 V are designed for low-loss, high performance and the best in class operating temperature capability of up to 175 °C for the most demanding medium-power applications.abb.com/semiconductors
— 62Pak High-quality products for industrial applications.
Bodo´s Power Systems® July 2018 www.bodospower.com58
CONTENT
As today’s power electronic systems move to more complexity and higher dynamics it is essential for the control platform to keep up with this trend. Modern power electronics is useless without algorithms and vice versa. Multi-level inverters, for example, provide lower THD and enable a reduction of passive components but in contrast need increasing efforts for gate drivers and signal transmission. Changing from a two-level to a five-level inverter equals at least a quadrupling of the involved active components. However, the complexity does not just apply to the hardware, but also to the algorithm for generating the switching pulses. Another example like model predictive control (MPC) requires extensive algorithms as well. It is used to overcome the disadvantages of common linear control algorithms. Here, the effect of each possible switching operation is precalculated on-line for several prediction steps, which in turn requires a high computing power. With novel, fast-switching semiconductors (e.g. SiC, GaN), the computing platform used for generating the pulse width modulation (PWM) pattern is also confronted with higher requirements as carrier frequencies rise. Selected microcontrollers have PWM or even space
vector modulation (SVM) cores implemented in hardware already, however with the disadvantage that no significant changes can be made to them. This is not satisfying for R&D purposes, especially if the modulation scheme is to be modified.
Because of the aforementioned challenges it is worth considering a FPGA as the central control unit for the power electronic system. FPGAs usually provide a high IO-pin count that allows connections of extensive topologies with all their different control and protection signals. Since FPGAs imply programmable logic they are best suited for paralleling algorithms in order to boost computation speed for dedicated partitions. Ordinary procedural programming can be used as well by implementing a microcontroller as softcore into the FPGA. Recent progress in model-based code generation (e.g. by means of MathWorks® HDL Coder™) greatly facilitates the rapid prototyping implementation of desired algorithms. This not only enables easy and complete system-simulation and -validation but also simplifies getting started for FPGA newcomers.
Today there are several real time control platforms available on the market. For control and power electronic use-cases most of them lack in: • Re-usability: Re-using one control-platform type for various projects
succeeds in few cases only because of different hardware require-ments
• Performance: The performance is crucial for R&D projects in order to be able crossing the border of state of the art setups and look ahead to what’s probably off-the-shelf in the coming years
• Extensibility and flexibility: These features are crucial and they are aimed at using one platform over a long period of time by con-stantly improving it
• Costs: Costs should be as low as necessary
For these reason TTZ-EMO developed a modular FPGA control platform within the past two years to eliminate these shortcomings. The system itself is divided into several plug-in cards, which are connected to each other via a backplane, all together mounted into a standard 19-inch rack. The core system consists of a FPGA card distributing more than 250 IOs to the backplane via one single high-density connector. They form an interface for extension cards, i.e. a modular expansion of the platform to implement various functions like data acquisition, communication or interfacing with an inverter. Figure
DESIGN AND SIMULATION
Modular Real-Time Development Platform for Power Electronic Systems
FPGAs provide an impressive computing power, which is increasingly required in power electronic systems. A modular rapid prototyping system based on an FPGA with individual expansion cards is the basis for the implementation of a wide range of high performance
power electronic R&D projects.
By Julian Endres, Fabian Bayer, Andreas Linke and Ansgar Ackva, TTZ-EMO - University of Applied Sciences Würzburg-Schweinfurt
Figure 1: FPGA card with backplane and several extension cards, 19″ case is left for better presentation
www.bodospower.com July 2018 Bodo´s Power Systems® 59
CONTENTDESIGN AND SIMULATION
1 depicts a sample configuration without the 19-inch case for better presentation.
The link between the FPGA-card and any extension card is an individual point-to-point connection without bus protocol for minimum latency. The FPGA card itself comes up with several communica-tion and storage capabilities such as dual Gigabit-Ethernet, 1 GB DDR3-SDRAM, SATA, MicroSD-Card and a dual port RAM interface. It incorporates an up-to-date Xilinx® Artix®-7 FPGA with approx. 215.000 logic cells and 740 DSP blocks, enough computation power even for extensive algorithms in the field of power electronics. The backplane currently offers slots for up to ten extension cards, entirely sufficient and flexible for state of the art R&D applications. Neverthe-less for even more computational power the back plane offers a slot for one additional FPGA (or other) card with then both FPGA-cards being able to communicate via four 6 Gb/s links. As a starting point we developed seven extension cards to satisfy our needs for current research projects. They contain the functions for getting connected to the gate drives of the power electronics (power stage), for reading analog measurement probes and converting internal FPGA signals to analog values to be displayed e.g. on an oscilloscope. In more detail, selected specifications for major extension cards are as follows:• There are inverter interface cards for 3-phase 2-level and 3-level
inverters and a H-bridge, the latter also being able to be used for 5-level inverters. Each card provides a true galvanic isolation to withstand 2.5 kV in case of an inverter fault. Gate drive signals are distributed via 15 V logic levels for high immunity against noise, while switching frequencies beyond 100 kHz are easily possible. Most cards are equipped with a low resolution ADC for measure-
ments of inverter phase currents, DC-link voltages and tempera-tures.
• The ADC extension card is equipped with four independent chan-nels, each 14 Bit resolution at a sampling rate of 40 MSPS. The voltage input range is ±1 V at an impedance of 50 Ohms. With an individual active or passive on board matching network the input channel can be adapted to the respective measurement sensor. Digital data are transferred to the FPGA over the backplane via high speed serial LVDS data streams. Inside the FPGA the serial stream is converted back to parallel words for further processing.
• For monitoring and debugging internal controller signals a DAC ex-tension card is available that represents the counterpart to the ADC extension card. It provides four channels, each 14 Bit at 40 MSPS. The full scale output range is mapped to ±1 V by default.
Each extension card carries an EEPROM that is interfaced via a one wire connection to provide storage for manufacturing data like card type, serial number etc. This also enables individual card detection within the FPGA logic to avoid damage to the hardware.
Before implementing a code into the FPGA the time discrete model of the control algorithm is simulated together with continuous hardware models to get a whole system simulation result. Since the simula-tion uses a model identical to what is coded and implemented into the FPGA, we are able to analyze and improve the control algorithm solely by means of simulation. Also, fast model-based code genera-tion, as available today, supports developing, testing and improving of the final control algorithm. Such an approach of rapid prototyping may include state machines or particular user codes as well. The
Dr.-Ing. Artur Seibt - Consultant - Electronics Design Lab
Former R & D manager and managing director in D, USA, NL, A, author of 156 publications and patent applications, offers consultant's services and a fully equipped design lab. 30 European and US customers (firms) to date. Assistance in all stages of development, design of complete (or parts) of products, tests of designs, failure analysis, evaluation of products for cost reduction or/and performance improvement.Specializing in power electronics (SMPS, lamp ballasts, motor drives, D amplifiers including EMI, 5 years experience with SiC and GaN), measuring instruments, critical analog circuitry.
Articles, books, lectures, in-house seminars for engineers e.g. about active and passive components, SMPS design, measuring instruments. Critical translations German-English and vice versa.
Lagergasse 2/6 Tel. +43-1-505.8186A 1030 Wien (Vienna) email: [email protected] HP: http://members.aon.at/aseibt
Bodo´s Power Systems® July 2018 www.bodospower.com60
CONTENTDESIGN AND SIMULATION
modular design of the hardware as well as the code allows engineers to choose those software modules for needed IO-Cards which best fit their demands. For example, the user only needs to know which type of IO-Card is suitable for his needs. The respective handwritten high-performance code for IO-Cards is then packed into so called IP-Cores. These custom building blocks fit in a graphical user interface (GUI) to build up the whole FPGA software simply by visual program-ming. This simple approach makes it easier for beginners to create complex logic circuits for FPGAs.
Last but not least we completed the design workflow by integrating a highly customizable LabVIEW based real-time user interface for tuning, monitoring and logging the internal model parameters. For this purpose, an UDP-Core is installed into the FPGA to establish a connection based on a widely available interface: Gigabit-Ethernet. Making use of our appropriate design blocks for model-based pro-gramming it is getting a breeze to connect up to 62 signals in each direction with an accuracy of 32 bit. These real-time signals are trans-ferred bidirectionally at a rate of 50 kHz between the FPGA and a PC. Figure 2 shows the whole toolchain involved in a test bench setup. Furthermore, a keep-alive mechanism ensures that user defined safety-critical processes are shut down and put in a safe state on connection break. The modular LabVIEW-Dashboard offers several visualization and interaction elements including up to ten 4-channel software oscilloscopes with on-line FFT calculation and various trigger modes suitable for almost every need.
As a result, we are able to serve different R&D projects, each having its distinct requirements on a control platform. Model based code generation is an essential ingredient for rapid prototyping of complex controllers in the field of power electronics (see Figure 3 and Figure 4). A versatile real-time Ethernet interface allows connection to vari-ous GUIs on a standard PC for operation of the control platform. ttz.fhws.de
Figure 3: FPGA System inside 19″ rack (top) used to control a test bench for wireless charging with SiC.
Figure 2: Toolchain for rapid prototyping with the proposed FPGA system
Figure 4: FPGA System (back) integrated in a five-level three-phase inverter (front) test bench
www.bodospower.com July 2018 Bodo´s Power Systems® 61
CONTENT
Vacuumschmelze had showcased first cores and common mode chokes made of the nanocrystalline material VITROPERM® 550 HF at PCIM. The newly developed alloy enables next-generation compo-nents with optimized size and weight.
The constant demand for smaller and lighter systems is also a chal-lenge in power electronics. With the material innovation VITROPERM 550 HF low-cost cores, chokes or power transformers in compact design for demanding requirements can be realized. The components show a significantly improved high-frequency behaviour compared to solutions made of VITROPERM 500 F or typical EMC ferrites. Applications for the high-performance materials are e.g. switched mode power supplies, wind or solar inverters and frequency convert-ers. In the automotive sector, common mode chokes in the drive train offer enormous advantages in size - components that are up to 60% smaller can be produced cost-effectively. Furthermore, power transformers with significantly reduced core losses can be realized for the use in wall boxes or in on-board charging systems for charging electric vehicles.
www.vacuumschmelze.com
ITROPERM 550 HF, a New Family Member
Powerbox announces the launch of four new series of extra-wide input voltage range, 8W to 20W board-mounted DC/DC converters for railway and transportation industry. With a 13:1 input voltage range of 12V to 160V, the MAD33 (8W), MAD32 (10W), MAE35 (15W) and MAF35 (20W) have been developed to provide systems designers with a single part number that is able to power a large range of trans-
portation industry applications (e.g. railway, industrial automation and automotive, remote radio-control), reducing inventory, time to market and documentation.The MAD-MAE-MAF series are designed to meet railway specifica-tions EN50155, EN50121-3-2, EN61373 and EN45545, but as well ISO7637-2 for 24V vehicles, and EN12895 for industrial trucks and other industrial applications such as robotics. The 8W MAD33 is pack-aged in an industry standard DIP24 case, and the MAD32, MAE35 and MAF35 are housed in a 2x1 inch case. All four series have a typical efficiency rating of 86%, which considering the extra-wide up to 13:1 input range is an outstanding figure.With the development of connected devices and Internet of Things (IoT), the number of points-of-interaction and sensors are rapidly increasing. Systems designers are facing challenging situations by having to develop products able to operate in many different applica-tions powered by a large range of voltages – versatility is the name of the game.
www.prbx.com
Four Series of Extra-Wide Input DC/DC Converters for Railway and Transportation
Infineon Technologies AG offers the Eice-DRIVER™ 1EDC Compact 300 mil family of single channel gate driver ICs. With an in-sulation test voltage of V ISO = 2500 V(rms) for 60 sec these galvanically isolated drivers are recognized under UL 1577. Due to their high switching frequency of up to 1,000 kHz they do not only drive IGBTs, but can also be integrated in demanding SiC MOSFET to-pologies. A wide variety of applications profits from these characteristics, e.g. photovoltaic string inverters, uninterruptable power sup-ply , EV charging stations, industrial drives, welding equipment, and CAVs.The 1EDC Compact family provides a 300 mil wide body package for an increased creepage distance of 8 mm leading to the
1EDC Compact Gate Driver Family Qualifies for UL 1577 Certificate
www.infineon.com/1edcompact
UL1577 acknowledged insulation. Further-more, the driver ICs feature improved ther-mal behavior and can drive power devices with up to 40 V output voltage. They are marked by a well-matched propagation delay of 120 ns (typical, max 150 ns). Based on Infineon’s 1200 V coreless transformer tech-nology, the gate drivers enable a world class common mode transient immunity great than 100 kV/μs. Additionally, the1EDC Compact can provide current drive strengths of up to 10 A on separate output pins for sourcing and sinking.
NEW PRODUCTS
Bodo´s Power Systems® July 2018 www.bodospower.com62
CONTENTNEW PRODUCTS
Microsemi Corporation announced it will be expanding its Silicon Carbide (SiC) MOSFET and SiC diode product portfolios early next quarter, including samples of its next-generation 1200-volt (V), 25 mOhm and 80 mOhm SiC MOSFET devices; next-generation 700 V, 50 A Schottky barrier diode (SBD) and corresponding die. These SiC solutions, along with other recently announced devices in the SiC SBD/MOSFET product families, have been shown at PCIM Europe 2018.As Microsemi continues to expand development efforts for its SiC product family, it has become one of the few suppliers providing a range of Si/SiC power discrete and module solutions to the market. These next-generation SiC MOSFETs are ideally suited for a number of applications within the industrial and automotive markets, includ-ing hybrid electric vehicle (HEV)/EV charging, conductive/inductive
onboard chargers (OBCs), DC-DC converters and EV powertrain/trac-tion control. They can also be used for switch mode power supplies, photovoltaic (PV) inverters and motor control in medical, aerospace, defense and data center applications. Microsemi’s next-generation 1200 V, 25/40/80 mOhm SiC MOSFET devices and die as well as its next-generation 1200 V and 700 V SiC SBD devices offer customers attractive benefits in comparison to competing Si/SiC diode/MOSFET and IGBT solutions, including more efficient switching at higher switching frequencies as well as higher avalanche/UIS rating and higher short-circuit withstand rating for rug-ged and reliable operation.
www.microsemi.com
Expanding SiC Portfolios with 1200 V SiC MOSFET and 700 V Schottky Barrier Diodes
Designers of Power Factor Correction stages (PFCs), Active Frontend Rectifiers, LLC converters and Phase Shift Full Bridge converters can now upgrade existing system performance by using the new UJ3C1200 series of SiC JFET cascodes from UnitedSiC. With a voltage rating of 1200 V and ON-resistances of 80 and 40 milliohms, these devices offer a ‘drop-in’ replacement solution for many exist-ing IGBT, Si-MOSFET and SiC-MOSFET parts, with no change to gate drive circuitry. This simplifies design upgrades and provides an alternative-purchasing source for existing parts. Applications include Power Factor Correction stages, Active frontend rectifiers, LLC converters and Phase Shift Full Bridge Converters where improvements to efficiency and/or power density are required. End-uses include on-board EV chargers, battery charging for forklifts, PV inverters, welding and more.Based on UnitedSiC’s Gen 3 SiC transistor technology, the UJ3C1200 series integrates a SiC JFET with a custom-designed Si-MOSFET to produce the ideal combination of normally-OFF operation, high performance body diode and easy gate drive of the MOSFET with the efficiency, speed and high temperature rating of the SiC JFET. As a result, existing systems can expect a performance increase with lower conduction and switching losses, enhanced thermal properties and integrated gate ESD protection. In new designs, the UnitedSiC
UJ3C1200 series delivers increased switching frequencies to gain substantial system benefits in both efficiency and reduction in size and cost of passive components, such as magnetics and capacitors.
www.unitedsic.com
1200 V SiC FETs Deliver Highest-Performance for IGBT, Si and SiC-MOSFET Users
Toshiba Electronics Europe announced the launch of the TC78B025FTG, a three-phase brushless motor driver IC with a
rotation speed control (closed loop control) function. The new device is intended for small fan applications in servers, home appli-ances and industrial equipment. The cooling fans used in servers and other must be small and rotate at high speeds with uniform accuracy.
Uniformity is best achieved by using a speed-feedback method to suppress fluctuations in rotation speed caused by changes in the power supply voltage and load. Until now, this has required the use of microcomputers, but Toshiba’s new solution achieves flexible rotation speed control without external microcomputers by incorporating a non-volatile memory (NVM). This allows easy system configuration and supports motor speeds from hundreds to tens of thousands of rotations per minute (RPM) especially as Toshiba’s intelligent phase control (InPAC) motor drive technology realizes efficient driving across a wide rotation range without requiring adjustment. The device operates from a power supply in the range 4.5V to 16V and provides a sine-wave drive with 150-degree commutation.
www.toshiba.semicon-storage.com
Three-Phase Brushless Fan Motor Driver IC
www.bodospower.com
TDK Corporation has extended its range of EPCOS film capacitors for DC link applica-tions. The B3277*H series now provides
types for particularly harsh environmental conditions. The robustness has been veri-fied in a temperature, humidity, bias (THB) test at 60 °C, 95 percent relative humidity, and the rated voltage for 1000 hours. Spe-
cial types are also available that are THB tested at 85 °C, 85 percent relative humidity, and the rated voltage for 1000 hours. The capacitors are designed for rated volt-ages of between 450 V DC and 1100 V DC and offer capacitance values of between 1.5 µF and 120 µF. The maximum operating temperature of the RoHS-compatible com-ponents is 105 °C. Depending on voltage and capacitance, the lead spacing is 27.5 mm, 37.5 mm, or 52.5 mm. The types with the latter grid size are offered exclusively in a 4-pin version. At the rated voltage and an operating temperature of 85 °C the service life of the self-healing capacitors is 50,000 hours. Applications include the DC link circuits of converters for photovoltaic systems, frequency converters and power supplies for industrial applications. AEC-Q200 qualified capacitors are also available on request.
www.epcos.com/film_mkp
Range of Rugged DC Link Capacitors Extended
Rogers Corporation’s Power Electronics So-lutions (PES) group has introduced at PCIM Europe ROLINX® CapLink Solutions. They are custom assemblies consisting of a num-ber of capacitors integrated to a laminated busbar and are ideal for any power-man-agement application requiring high-power handling capabilities, low equivalent series inductance (ESL), low equivalent series
resistance (ESR) in small, light-weight as-semblies such as DC link systems in EV/HEV as well as in industrial, solar-power and wind-power systems. ROLINX CapLink Solutions are an essential solution to sup-port IGBT/SiC technology devices at very high switching frequencies with increased voltages and higher temperatures.ROLINX PowerCircuit Solutions are the ideal alternative when the current and power
requirements of the application exceed the capabilities of a standard PCB solution. Their compact 3D design and superior ther-mal management result in space and weight savings. ROLINX PowerCircuit Solutions are highly integrated structures that are solder-process compatible. These products help engineers design power circuitry with the smallest possible footprint.
ROLINX Hybrid busbars are especially designed to reduce installation time of a battery module by combining power and signal lines for voltage and temperature measurement, in a one piece solution. Integrated surface mount components, such as connec-tors, enhance the functionality of the assembly. ROLINX Housing Solutions are designed to simplify the design and assembly of electronic equipment
housings. These injection-molded struc-tures combine connectors with busbars for reduced weight and low-inductance power connections in electronic equipment hous-ings. Ideal for high-volume applications in automotive and aerospace products, they are rated up to 100 kW at continuous oper-ating temperatures from -40 to +125ºC.
www.rogerscorp.com
ROLINX® CAPLINK Solutions Intro-duced at PCIM 2018
Bodo´s Power Systems® July 2018 www.bodospower.com64
CONTENT
MEMORY HiCORDERMR6000
ExceedAll LimitsNew FlagshipWaveform Recorder
・ 200MS/s sampling・ Fit up to 8 user exchangeable input modules for a maximum of 32 analog channels・ 1GW built-in memory + Superior real-time recording・ Intuitive operability with 12.1-inch touch screen
[email protected] / www.hioki.com
HIOKI EUROPE GmbH Stand 917-21 Sept. 2018
Kipsala International Exhibition Centre, Latvia
ABB Semi C3ABB Semi 57Analog Devices 25Broadcom 15Cornell Dubilier 55Dr.-Ing. Seibt 59Electronic Concepts 1EPE ECCE 43FTCAP 34Fuji Electric Europe 47
GvA C2Hioki 64Hitachi 9hivolt 33Infineon C4Kikusui 63LEM 5Microchip 17Mitsubishi Electric 35Mouser 23
Plexim 27Rohm 7Semikron 31SMT-Wertheim 41UnitedSiC 29Vincotech 11Würth Elektronik eiSos 3ZH Wielain 33
Advertising Index
Mouser Electronics, Inc. is stocking the LMZM33602 and LMZM33603 power modules from Texas Instruments (TI). These completely integrated 4 V–36 V power module solutions combine a step-down DC-to-DC converter with power MOSFETs, a shielded inductor, and passive compo-
nents in a compact, low-profile package. To simplify design and offer even greater ease-of-use, the LMZM33602/LMZM33603 power modules require as few as four external compo-nents and enable engineers to eliminate the loop compensation and magnetics part selection from the design process. The TI LMZM3360x power modules, available from Mouser Electronics, are 2 A (LMZM33602) or 3 A (LMZM33603) synchronous step-down DC-DC converters. The LMZM33602 offers an output voltage range from 1 V to 18 V at 2 A, whereas the output volt-age range of the LMZM33603 is 1 V to 13.5 V at 3 A. Both converters allow synchronization to an external clock and feature an adjustable switching frequency from 200 kHz to 1.2 MHz. The switching frequency can also be adjusted by either an external resistor or a sync signal, which allows the solution to accommodate a variety of input and output voltage conditions as well as optimize efficiency up to 95 percent.
www.mouser.com/ti-lmzm33602-lmzm33603-power-module
LMZM3360x Integrate 36V Buck Converter at Mouser,
— Rectifier diodes Exceeding nominal and surge current capabilities.
ABB Semiconductors’ high-power rectifier diodes are the first choice in many de-manding applications in industry and traction. We offer two families of high-power rectifier diodes: Standard and avalanche diodes, both featuring reverse repetitive voltage up to 6000 V and junction temperature from up to 190 °C. abb.com/semiconductors
www.infineon.com/econo
EconoDUAL™ 3 with shuntsIntegrated shunt resistors for direct current measurement – reduced system costs by higher integration.
EconoDUAL™ 3 modules are now available with integrated shunt resistors for current monitoring in the AC path. The integration of additional functionality into the IGBT module is a very e ective way to optimize the overall system costs of an inverter. External current sensors are no longer needed. This saves space in the system, reduces material costs and lowers manufacturing e orts.
› Cost reduction› Eliminate needed space of Hall Eect sensors
Power module with integrated shunts
› Digital bitstream data transfer for high robustness against disturbances
› Galvanic isolation
∆∑ modulator
› No external ∆∑ demodulator IC needed › Cost advantage › High degree of integration
XMC4400 … XMC4800 with integrated∆∑ demodulator
shunt
XMC™
Digital mapping of the current by bitstream
Isolationbarrier
∆∑demodulator
∆∑ modulator
Imeas
Key features › Integrated shunt resistors for direct current measurement
› High current measurement accuracy across a wide temperature range
› Optimized heat spread guarantees e ective thermal management
› PressFIT control pins and screw power terminals
› TIM – pre-applied thermal interface material to achieve longest lifetime