SILICON CARBIDE Practical Use of SiC Power Semiconductors

02_PEE_0114_Layout 29/01/2014 14:57 Page 1

CONTENTS

www.power-mag.com Issue 1 2014 Power Electronics Europe

3

Editor Achim ScharfTel: +49 (0)892865 9794Fax: +49 (0)892800 132Email: [email protected]

Production Editor Chris DavisTel: +44 (0)1732 370340

Financial Manager Clare JacksonTel: +44 (0)1732 370340Fax: +44 (0)1732 360034

Reader/Circulation Enquiries Capsule Publishing Services LtdTel: 0333 577 0801 Fax: 0845 604 2327Email: [email protected]

INTERNATIONAL SALES OFFICESMainland Europe: Victoria Hufmann, Norbert HufmannTel: +49 911 9397 643 Fax: +49 911 9397 6459Email: [email protected]

Armin Wezelphone: +49 (0)30 52689192mobile: +49 (0)172 767 8499Email: [email protected]

Eastern US Karen C Smith-Kerncemail: [email protected] US and CanadaAlan A KerncTel: +1 717 397 7100Fax: +1 717 397 7800email: [email protected]

Italy Ferruccio SilveraTel: +39 022 846 716 Email: [email protected]

Japan:Yoshinori Ikeda, Pacific Business IncTel: 81-(0)3-3661-6138Fax: 81-(0)3-3661-6139Email: [email protected]

TaiwanPrisco Ind. Service Corp.Tel: 886 2 2322 5266 Fax: 886 2 2322 2205

Publisher & UK Sales Ian AtkinsonTel: +44 (0)1732 370340Fax: +44 (0)1732 360034Email: [email protected]

Circulation and subscription: Power ElectronicsEurope is available for the following subscriptioncharges. Power Electronics Europe: annual chargeUK/NI £60, overseas $130, EUR 120; single copiesUK/NI £10, overseas US$32, EUR 25. Contact: DFA Media, 192 The High Street, Tonbridge, Kent TN9 1BE Great Britain. Tel: +44 (0)1732 370340. Fax: +44 (0)1732 360034. Refunds on cancelledsubscriptions will only be provided at the Publisher’sdiscretion, unless specifically guaranteed within theterms of subscription offer.

Editorial information should be sent to The Editor,Power Electronics Europe, PO Box 340131, 80098Munich, Germany.

The contents of Power Electronics Europe aresubject to reproduction in information storage andretrieval systems. All rights reserved. No part of thispublication may be reproduced in any form or by anymeans, electronic or mechanical includingphotocopying, recording or any information storageor retrieval system without the express prior writtenconsent of the publisher.

Printed by: Garnett Dickinson.

ISSN 1748-3530

PAGE 8

CIPS 2014

PAGE10

APEC 2014

PAGE12

Industry News

PAGE 26

MOSFETs Make Step Change inPerformance to Meet NewApplication RequirementsLow voltage MOSFET devices (<40 V) are used extensively in the power systems

of portable electronic devices, domestic appliances, data communication servers,

medical equipment and telecom infrastructure deployments. Exacting demands

continue to be placed on the engineers involved in the design of such MOSFETs.

The article will look at the forces at work and how significant, yet potentially

conflicting, technical challenges need to be tackled. Wharton McDaniel,

Product Marketing Manager Power MOSFETs, ON Semiconductor,

Phoenix, USA

PAGE 28

Fine Tuning of Digital Control inPOL Regulators OptimizeDynamic Response and MinimizeComponent CountPower system design encompasses many requirements and criteria. This article is

about perhaps the most basic of all requirements – a stable voltage supply of the

operating circuitry. Patrick Le Fèvre, Ericsson Power Modules, Sweden

PAGE 32

ProductsProduct Update

PAGE 33

Website Product Locator

Practical Use of SiCPower SemiconductorsSilicon Carbide (SiC) power devices are enablingcomponents mainly in the context of higher switchingfrequencies and/or small footprints in powerelectronics. However, this trend imposes newchallenges on the packaging of the chips. Typical strayelements like inductances become crucial elements inthe circuit. In addition, different considerationsregarding the thermal design in power modules arisewhen SiC chips are taken into consideration.Furthermore, the aspects of power density as well asthe general utilization of SiC’s high temperaturecapability are important factors in reliableimplementation of SiC-based power semiconductors inmodern systems. The article will give an insight intohow these boundary conditions can be implementedin innovative solutions using SiC chips. More details on page 21.

Cover supplied by Infineon Technologies

COVER STORY

PAGE 6

Market NewsPEE looks at the latest Market News and company

developments

p03 Contents.QXD_p03 Contents 30/01/2014 09:49

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/pe2 | 508-996-8561

04_PEE_0114_Layout 29/01/2014 14:52 Page 1

OPINION 5


A suitablequote as alead in to

the editorsopinion

According to the crystal ball of market researchers such as IHS, theemerging market for Silicon Carbide (SiC) and Gallium Nitride(GaN) power semiconductors is forecast to grow at a remarkablefactor of 18 up to $2.8 billion during the years 2012 to 2022,energized by demand from power supplies, photovoltaic (PV)inverters and industrial motor drives. Thus market revenue isexpected to rise by double digits annually. SiC power devices are enabling components mainly in the context

of higher switching frequencies and/or small footprints in powerelectronics. However, this trend imposes new challenges on thepackaging of the chips. Typical stray elements like inductancesbecome crucial elements in the circuit. In addition, differentconsiderations regarding the thermal design in power modules arisewhen SiC chips are taken into consideration. Furthermore, theaspects of power density as well as the general utilization of SiC’shigh temperature capability are important factors in reliableimplementation of SiC-based power semiconductors in modernsystems. Our cover story will give an insight into how theseboundary conditions can be implemented in innovative solutionsusing SiC chips. SiC power technology, predominantly in form ofSchottky diodes, is meanwhile established in the market. High-power solutions using power-module technologies have recentlybecome available. Usually, the aim of such components is to enablesystem benefits, for example by increasing the switching frequencyor reducing losses. If this is successful, the high cost of SiC-basedcomponents can be compensated by reduced efforts for passiveelements and/or cooling.While the discrete, unipolar SiC devices as single chips are more

or less ready to achieve higher frequencies (>100 kHz in PFC units),this is still a challenge for power modules. Solutions in a powerrange with traditional modules have high stray inductances and thusthe di/dt is increased if combined with an increase in the frequency.A plug and play between SiC and Silicon at chip level in existingmodules could devaluate the theoretical SiC performance. Thus it isnecessary to improve high-current modules in order to get the fullbenefit of SiC in the frequency range above 20 kHz. Whilst thereduction of parasitic elements in power modules using SiC chips isin line with the approach also used for future Silicon solutions, itmight happen that different optimization criteria will apply for thethermal design. The reason for this is that the cost contribution of

chips in a power module with SiC is different from one with Silicon.Thus, the best solution for a given frame size with respect tosemiconductor area placed in the module could be different.Furthermore, it is expected that the die size of a SiC transistor at1200 V, for example, comes down to one tenth or less of the areaneeded in the current Silicon-based IGBT technology, assuming thesame total losses for both options. This will result in a huge increasein power and current density, and will require more efforts withrespect to an effective heat removal and the connection of chips toterminals.Also Richardson RFPD, an Arrow company focuses on SiC and has

launched its SiC Tech Hub, a micro-website featuring the latest newsand new product releases related to Silicon Carbide technology. Thecompany has an extensive SiC offering, including a selection of ofSchottky diodes, MOSFETs and IGBTs from industry-leadingmanufacturers Cree, Microsemi, Vincotech and Powerex. The SiCTech Hub brings together these products with design resources. Thewide range of design resources offered on the SiC Tech Hubincludes application notes, brochures, datasheets, presentations,selector guides, supplier documentation, technical articles andvideos. Visitors to the SiC Tech Hub may also sign-up for emailedproduct updates, links to technical presentation videos, and otherresources like white papers and technical articles as soon as theyare published.But the most interesting news comes with the GaN merger of US-

based Transphorm and Japan-based Fujitsu Semiconductors. Devicemanufacturers such as International Rectifier, GaN Systems,Panasonic, or Transphorm are competing to commercialize GaNpower devices. Transphorm has already released high voltagecommercial GaN devices, the company continues to makesignificant improvements in performance, quality and productivity.Given this, Fujitsu, Fujitsu Semiconductor and Transphorm will bringtogether their complementary strengths in technology andmanufacturing. Integrating Transphorm’s operations with Fujitsu willenable high-volume and high-performance GaN power deviceproduction at Aizu-Wakamatsu plant in Japan. In addition to being afuture user for GaN devices, Fujitsu has also built customerrelationships with many power management companies in Japan,one of the most important markets for GaN. Recently power supplymanufacturer Delta Electronics has committed its interest in GaNpower.Conferences/exhibitions will focus on these subjects in 2014. First

will be CIPS 2014 in Nuremberg, here Daisuke Ueda, AdvancedTechnology Research Laboratories, Panasonic, Japan, will give akeynote on “Present and Future of GaN Power Devices”. At APEC2014 a dedicated Rap-Session will give an answer to the question“Will GaN and SiC devices become a higher value proposition fordesign engineers and enter mainstream adoption displacing Siliconpower semiconductors in new designs and applications?” And atPCIM Europe 2014 PEE will host a panel discussion on “Si vsSiC/GaN - Competition or Coexistence”. Interesting events to come!

Achim ScharfPEE Editor

The Year of SiC and GaN

05_PEE_0114_p05 Opinion 29/01/2014 15:07

6 MARKET NEWS

Issue 1 2014 Power Electronics Europe www.power-mag.com

Global sales of motion controlproducts recovered in 2013 to reach$12.2 billion, but growth was far tooweak to overcome the 2012 marketdeclines, leaving market revenueswell below the peak level reached in2011. According to marketresearcher IHS strong growth isforecast to return to the market in2014, propelling market revenues toa new record high by the end of theyear. The motion control market

reached a record level of $13.2billion in 2011 following two years ofmarket growth in excess of 20 %. In

Motion Control Revenues Fall Short in 2013

2012, the motion-control marketsuffered from the poor economicsituation in the Eurozone, a weaksemiconductor market in Japan, andthe fallout from overproduction inChina. Market revenues declined inall three of these regions, and theglobal motion-control marketdecreased by 8.3 %. Conversely, theAmerican market provided a brightspot for motion controls, growingover 5 % in 2012. The results for 2013 are expected

to reveal another year of subduedgrowth. Stable industries likepackaging, materials handling, and

food, beverage and tobacco continueto propel growth in machineryproduction and correspondingly, themotion-control market. However,motion-control sales have beenconstrained in 2013 by the market’sdependence on volatile industrieslike semiconductor, machine tool,and robotics, which are known fortheir cyclic growth patterns. Largedeclines in these industries in 2009contributed significantly to the severedecline in the motion-control marketthat year, and high growth in theseindustries in 2010 and 2011boosted the post-recession recoveryof the motion-control market. In 2013, machinery production

declined at the global level for thesemiconductor and roboticsindustries and grew very little in themachine tool industry, according tothe latest research from IHS onmachinery production. Combined,these three industries account forapproximately 50 % of revenues inthe motion-control market, while themore stable industries like packaging,

materials handling, and food andbeverage account for only 16 % ofmarket revenues. Still, the motion-control market is estimated to havegrown between 1.8 % in 2013, assmall declines in the machine toolsand semiconductor machinerysectors were offset by growth in therest of the motion-control market. The outlook for 2014 is more

optimistic. In the Eurozone, GDPgrowth is expected to accelerate andrenewed confidence should spurinvestment in machinery upgrades,leading to an increase in motioncontrol sales. In the Americas,recovery of the semiconductormachinery market will boost growthof motion control revenues in 2014,while in Asia, the machine toolmarket is forecast to recover.Globally, the motion-control marketis forecast to grow by over 8 percentto reach $13.3 billion in 2014,exceeding the 2011 record value by0.6 %.

www.ihs.com

Richardson RFPD offers SiC Tech HubRichardson RFPD launched its SiC Tech Hub, amicro-website featuring the latest news andnew product releases related to SiliconCarbide technology.SiC offers significant advantages in high

power, high voltage applications where powerdensity, higher performance and reliability areof the utmost importance. Industrialapplications like solar inverters, welding,plasma cutters, fast vehicle chargers and oilexploration are a few examples that benefitfrom the higher breakdown field strength andimproved thermal conductivity that SiC offersover Silicon.Richardson RFPD has an extensive SiC

offering, including a selection of of Schottkydiodes, MOSFETs and IGBTs from industry-leading manufacturers Cree, Microsemi,Vincotech and Powerex. The SiC Tech Hubbrings together these products with designresources. The wide range of design resourcesoffered on the SiC Tech Hub includesapplication notes, brochures, datasheets,presentations, selector guides, supplierdocumentation, technical articles and videos.Visitors to the SiC Tech Hub may also sign-up for

emailed product updates, links to technicalpresentation videos, and other resources likewhite papers and technical articles as soon asthey are published. “As we continue to expandour design support resources, a centralizedresource for Silicon Carbide was the next logicalstep,” said Dave Rossdeutcher, Director Global

Marketing. “With the participation of our keysuppliers, the SiC Tech Hub will provide focusedSiC content to support our energy and powermarket customers and their design activitieswith this rapidly expanding technology.”

www.richardsonrfpd.com/sicpower

Market News_Layout 1 29/01/2014 15:05

MARKET NEWS 7


powermodules.danfoss.com

Drive happy while staying cool Reliable power density for tomorrow’s vehiclesIt cannot be stressed enough: e cient cooling is the most important feature in power modules. Danfoss Silicon Power’s cutting-edge ShowerPower® solution is designed to secure an even cooling across base plates. In addition, our modules can be customized to meet your automation requirements in detail, o ering low weight, compact design, extended life and very low lifecycle costs. In short, when you choose Danfoss Silicon Power as your supplier you choose a thoroughly tested solution with unsurpassed power density.

Please go to powermodules.danfoss.com for more information.

MAKING MODERN LIVING POSSIBLE

Fujitsu Limited, Fujitsu Semiconductor Limited, and Transphorm, Inc.,announced that they have reached an agreement whereby FujitsuSemiconductor and Transphorm will integrate their GaN power devices forpower supplies businesses. The three companies have also agreed thatboth Fujitsu and Fujitsu Semiconductor will take a minority equity positionin Transphorm.

Device manufacturers around the world are competing to commercializeGaN power devices. Transphorm has already released high voltagecommercial GaN devices, the company continues to make significantimprovements in performance, quality and productivity. Given this, Fujitsu,Fujitsu Semiconductor and Transphorm will bring together theircomplementary strengths in technology and manufacturing. “IntegratingTransphorm’s operations with Fujitsu will enable high-volume and high-performance GaN power device production at Aizu-Wakamatsu, and wewill also benefit from Fujitsu’s strong technological capabilities underpinnedby Fujitsu Groups’ years of developments in the field of GaN power device.In addition to being a future user for GaN solutions, Fujitsu has also builtcustomer relationships with many power management companies inJapan, one of the most important markets for Transphorm”, commentedTransphorm’s CEO Fumihide Esaka. “From 2009, Fujitsu Semiconductorhas moved forward on developing the GaN mass production technology,and at the end of 2011 began sample shipments of GaN power deviceswith 600 V breakdown voltage, followed by 150 V breakdown voltage GaN.We have positioned GaN power devices as one of our future core productsfor the Aizu-Wakamatsu plant. The business integration will enable us tocollaborate with Transphorm and leverage their technologies to acceleratemass production”, added Fujitsu Semiconductor’s President Haruki Okada.

The employees at Fujitsu and Fujitsu Semiconductor who are directlyinvolved in the GaN power device business will be reassigned to the newcompany, where they will continue development and production work withTransphorm’s employees. After the integration, Transphorm will continueR&D work at both its prototyping line in the US as well as FujitsuSemiconductor’s Aizu-Wakamatsu plant, which will be under exclusivecontract with Transphorm to handle wafer processing.

www.transphormusa.com, http://jp.fujitsu.com/fsl/en/

Transphorm andFujitsu Integrate GaNPower Businesses

Mannheim-based GvA Leistungselektronik and Fuji Electric Europe arepartnering for the sales of power semiconductors in Germany.

GvA as a distributor, developer and manufacturer of custom specific powerelectronics sees in this new product line a complement of its existing programand can thus offer its customers IGBT modules in high diversity. The productrange of Fuji Electric scores with a large product range of IGBT modules in thevoltage classes 600, 1200, 1700 and 3300 V, and rated current of 8 A to3600 A. “With GvA we can not only gain a competent distribution partner forGermany. Their know-how as a developer and producer in almost every fieldof power electronics gives us the assurance that our customers receive thebest possible advice also from a practical point of view”, commented FredEschrich, General Manager of Fuji Electric Europe. “In our new modularinverter system MODIS for example, we use the IGBT modules of Fuji togetherwith the intelligent IGBT drivers of Amantys with very good test results. Webelieve that our broad range of experience also benefits our customers. Withthe wide product range of Fuji we can also reach customers in Germany, wecould not address so far”, GvA Managing Director Werner Bresch added.

www.gva-leistungselektronik.de

GvA partners with Fuji Electric

Market News_Layout 1 29/01/2014 15:05

8 CIPS 2014 www.cips-conference.de


Higher power efficiency, density, reliability, and lower volume and cost: How to reach this goals and what solutions are feasible will be discussed againat the 8th CIPS (Conference on Integrated Power Systems) from February 25-27 in Nuremberg / Germany. 235 engineers and scientists attended theCIPS 2012 at Nuremberg from March 6 to 8. They came from 17 countries in Asia, America and Europe.The conference is organized by VDE ETG and ECPE and technically co-sponsored by the IEEE PELS and ZVEI. The conference papers did undergo a

peer review process which allows their storage in the IEEE Xplore digital data base. The Technical Program Committee has selected 81 papers to bepresented; 29 of them will be poster presentations. The best poster as well as the best young engineers presentation will be awarded.

As usual for CIPS there is a frame of 3 Keynotes and 9 Invited Papers given by worldwide leading experts

Keynote Papers What are the big challenges in power electronics? Johann W. Kolar, ETH Zurich, Switzerland Simulation and test vibration – nonlinear effects in durability of electronic systems Abhijit Das Gupta, University of Maryland, USA Present and future of GaN Power Devices Daisuke Ueda, Advanced Technology Research Laboratories, Panasonic, Japan

Invited papers Power module reliability Transient hygro-thermal-response of power modules in inverters-mission profiling for climate and power loading Power Supply With Integrated PassivEs-POWERSWIPE Practical Aspects and Uses of Thermal Interface Material Testing Methods EMI and Integration Power Electronics - Key Technology for Renewable Energy Systems - Status and Future New applications in power electronics for integrated high-speed magneto-resistive current sensors Packaging very fast switching power semiconductors SiC Power Electronics

Participants especially from universities and research institutions are invited to present results and demonstrators from current research in the TableTop Exhibition.

8th International Conference on Integrated Power Electronics Systems

Toshiba Photocouplersin the SO6 housing delivera big performance in acompact package. Forbig results, think small.

Clearance / Creepage distance of > 5mm

Internal Isolation thickness of > 0,4mm

Low profile package, max. height < 2,3mm

Operating Temperature Range -40 to +125°C

THINK SMALL!

WWW.TOSHIBA-COMPONENTS.COM/PHOTOCOUPLERS

WWW.TOSHIBA-COMPONENTS.COM/VIDEO

CIPS_Layout 1 30/01/2014 08:39

www.apec-conf.orgwww.apec-conf.org

SPONSORED BY

2014March 16–20, 2014

Fort Worth Convention Center

Fort Worth, TX

THE PREMIER

GLOBAL EVENT

IN POWER

ELECTRONICSTM

THE PREMIER

GLOBAL EVENT

IN POWER

ELECTRONICSTM

Visit the APEC 2014

web site for the latest

information!

09_PEE_0114_Layout 30/01/2014 08:35 Page 1

10 APEC 2014 www.apec-conf.org


APEC 2014 will start on March 17 afternoon with the plenary session,traditionally a number of speeches highlighting the actual subjects of powerelectronics within the conference, before the exhibition opens within the lateafternoon. Already on March 15/16 tutorials will be offered.

Plenary session on actual topicsTransforming energy management is the first speech given by Sami Kiriaki, SVPTexas Instruments. The topic of power and energy management continues todraw the interests of semiconductor, business and investment communities.Leading technology companies have always respected power, but it is an areathat has transcended from the “necessary evil” of the past to the “technologyenabler” of today and the future. Power and energy management enable everyportable device to be portable, and without it the run-time alone would havelimited many markets. Advanced power management strategies developed inthe portable world have been transformed into new power technologies usedin many transportation and stationary applications. These include new circuitsolutions that minimize losses, while improving the response to load variations.Developments in leading-edge semiconductors, circuits, packaging andmanufacturing allow designers to now take a systems approach to power andenergy management, which will help fuel growth for power electronics inmarkets, such as portable, transportation, computing and communication. Thispresentation will cover the challenges for these industries, and describe hownew power and energy management technologies are addressing them. Forexample, the expansion of the wireless Internet will be limited by the need forenergy. Efficient power and energy management solutions at each node in thenetwork are required. These solutions will benefit from the new powermanagement technologies to become more power efficient, more powerdense, and more responsive to wide dynamic load variations. New powersolutions will change the type of products and the way we do business in thisdiverse power management industry.Power electronics in emerging applications is the subject to be presented by

Dong F. Tan, Northrop Grumman Corporation. The company’s AdvancedElectrical Power Systems (AEPS) features modularity and scalability to

minimize cost, schedule and technical risk. Its minimum set of five standardelectronic slices allows various units and hence the power system to beconfigured either as regulated or unregulated bus. Its peak power trackingcapability, together with high power conversion efficiency and tightly regulatedbus voltage, reduces significantly the required solar array capacity and payloadpower distribution size and weight for a given mission. It supports either Li-Ionor NiH2 batteries to deliver prime bus power from 1 to 8 kW with scalabilityand modularity. This presentation focuses on simulated and measuredperformances of the peak power tracker (PPT) circuit implemented for a flightprogram. A brief overview of PPT architecture and implementation, a report onsimulation and measurement results on steady-state responses, and keyperformance, such as tracking accuracy and tracking time, will be reported.Transient responses for mode changes under various orbit conditions, such aseclipse entry and eclipse exit, will be presented. Various modes under differentillumination conditions will also be discussed.Mission critical power will be discussed by Miguel Chavez, Director of

Engineering, Eaton. A look at mission critical power and how the evolution ofindustries requiring uninterrupted electric power and advances in powerelectronics have helped shape its path. From early computer rooms to today’shyper-scale data centers, the demands for safe and reliable power have grownsignificantly and are part of a world that expects to be interconnected 24/7.What will the future require? The trends, technological and otherwise, that willimpact how the challenges of supplying the world with mission critical powerare addressed.ARPA-E initiatives in high-efficiency power conversion will be introduced by

Tim Heidel, Program Director, Advanced Research Projects Agency–Energy(ARPA-E). Advances in power electronics promise substantial energy efficiencygains across a wide range of power conversion applications. Innovations inwide bandgap power semiconductor devices and magnetic materialsdevelopments, in particular, hold enormous potential, but, today, are still at asomewhat early stage of technological maturity. ARPA-E is an agency of theU.S. Department of Energy that focuses on funding early stage technologiesthat have the potential to have a transformational and disruptive impact on the

United States’ energy future. Over the past four years,ARPA-E has launched several power electronics focusedprograms including the agency’s ADEPT (2010), SolarADEPT (2011), and SWITCHES (2013) programs. Thistalk will give an overview of some of the successes fromthese programs thus far and some of the challenges thathave yet to be overcome.Significant developments and trends in 3D packaging

will be discussed by Brian Narveson and Ernie Parker, Co-Chairman of the PSMA Packaging Committee. Thepackaged and modular power supply industry isconstantly challenged by its customers to deliver morepower in a smaller volume. This applies whether selling 1W modules or multi kW AC/DC product. The focus onfootprint alone is no longer adequate. The efficient use ofthe z-dimension to minimize the volume of the packageis now in the forefront. Cost effective, volume efficient 3D

APEC 2013 in Long Beach/California closed with a record of nearly 4,000 delegates (+ 27 %) and 187exhibitors, making it one of the greatest power electronics conferences and showing the increasing interestin the role of power electronics for energy efficiency. Wide bandgap technologies such as SiC and GaNwere of major interest, and this will continue in the 2014 event from March 16-20 in Fort Worth/Texas.

More Power in 2014

Fort Worth’s convention center,home of APEC 2014

APEC_Layout 1 30/01/2014 08:41

MARKET NEWS 11


driven by prime

performance

Highly-integrated SCALE-2® ASIC chipsetIGBT driver with plug-and-play functionalityAdvanced Active Clamping (A C)

power. precision. excellence.

www.IGBT-Driver.com

packaging is rapidly being adopted by the power industry. The PackagingCommittee of the Power Supply Manufactures Association (PSMA) hasundertaken a study to obtain an overview of the technology and producttrends that are evolving, what the drivers are and what the marketopportunities will be over the coming years. This study provides valuableinsights into state of the art products and technology roadmaps whileidentifying R&D and manufacturing challenges that need to be addressed. Thestudy encompasses developments from chip level integration and Siliconstacking to high-power modules (up to 10 kW). The study also coversemerging technologies for additive manufacturing such as 3D printing andhow it is being used in the power supply industry. The presentation wouldverbally and graphically present the results of this study, which includes thestate of the industry with identification of key technology, manufacturing andmarket trends in 3D power packaging.

System reliability vs. efficiency, this is the final subject to be discussed byBrian Fortenbery from the Electric Power Research Institute (EPRI). Systemdesigners often strive to increase reliability by building redundancy into theirarchitectures. Since we know that even the most reliable components mustfail eventually, the best approach to overcoming that risk is to add in aredundant component to provide the required functionality in the case of afailure. The problem with this approach is that usually, we like to allow eachcomponent to carry equal parts of the load, and switch over in the event of afailure. As such, the efficiency of the system suffers, and energy use isincreased. This presentation will discuss the pros and cons of this approach,and suggest some alternatives that can increase efficiency and save energy.

Rap sessions Rap (panel) sessions are intended to discuss certain subjects under a broaderscope with a number of penelists. Two sessions will be held on March 18from 5.00 – 6.30 pm.

Session 1 will cover smart grid infrastructures. Once upon a time power gridinfrastructures only concerned large-scale industrial operations, civil engineers,and municipalities. Today we are faced with the convergent pressures of thenear-simultaneous advent of the Internet of Things and the Smart Grid. Thismeans that everyone at every level of today’s users, distributors, andgenerators of power must be aware of the technologies, infrastructures,regulations, and protocols involved. These systems range from consumerhandheld products and remote service subsystems involved in the “Internet ofThings” to the smart-home and micro grid-oriented bridge technologies, to themunicipal grid itself. The real challenge, however, lies in how we integrate notonly the systems, but the management protocols, regulatory issues, servicepriority concerns, and privacy & security. In a world of negotiated power, whodecides what facilities and organizations take priority in a brownout situation?How are the various layers of management, from various sectors of theindustry, negotiate and determine priority in a system? How do we addresssecurity in a system of interleaved and overlapping devices and systems?There are many aspects of the smart grid beyond simply being “smart” thatmust be determined and create plans and strategies for in order to properlyachieve the lofty goals having set for the industry.

Session 2 is a follow-up to a 2013 session now entitled Wide BandgapSemiconductors vs Silicon in Power Electronics. For several years we keephearing about how these semiconductors are going to provide dramaticimprovements in power electronics system performance. Significant resourceshave been and continue to be expended - will this be the year of exponentialgrowth? Will the huge financial investments made thus far pay off soon? Willthe technology performance match the rhetoric and when? Will theprice/performance ratio be where it needs to be? What applications will befirst? Will the innovation, performance, device and circuit techniques andimplementation using Silicon devices continue to dominate new designs. WillGaN and SiC devices become a higher value proposition for design engineersand enter mainstream adoption displacing Silicon power semiconductors innew designs and applications? Will Silicon continue to be good enough at alower price or will wide bandgap devices enter main stream adoption? Perhapsanswers can be given by the panel of experts.

More about APEC 2014 in our next issue.

APEC_Layout 1 30/01/2014 08:41

12 INDUSTRY NEWS


Microstepping Motor Driver Featuring Translator

At power-on the A4992 is configured to drive most small stepper motors withsimple step and direction inputs. The device can be used for high temperatureapplications such as headlamp bending and levelling, throttle control, and gasrecirculation control in automotive applications. It is also suitable for other lowcurrent stepper applications such as air conditioning and venting. It provides aflexible microstepping motor driver controlled with simple step and directioninputs. It can also be switched into an optional serial interface mode, where itcan be configured and driven via an SPI compatible serial interface.The current regulator operates with fixed frequency PWM. It uses adaptive

mixed current decay to reduce audible motor noise and increase stepaccuracy. The current in each phase of the motor is controlled through aDMOS full bridge using synchronous rectification to improve power dissipation.

Internal circuits and timers prevent cross-conduction and shoot-through whenswitching between high-side and low-side drives. The outputs are protectedfrom short circuits. Features for low load current and stalled rotor detection areincluded. Chip level protection includes hot thermal warning, overtemperatureshutdown, and overvoltage and undervoltage lockout. An optional serialinterface mode, using the STEP, DIR, and MS inputs, can be used to configureseveral motor control parameters and diagnostics.

Serial interfacingA three wire synchronous serial interface, compatible with SPI, can be used tocontrol all features of the A4992.The A4992 powers-up in the default step and direction control mode. The

serial mode is only available following a specific sequence on the DIR and MSpins when STEP is high. The sequence starts when STEP is high and MS isheld high and DIR is held low for longer than the Sequence Minimum HoldTime (tSSH). MS must then be held low and DIR high for longer than tSSH,followed by holding MS high and DIR low for longer than tSSH. The final step isto hold DIR high for tSSH, at which time the A4992 will enter the serial controlmode.If STEP is taken low at any time during the sequence then the sequence is

reset and the sequence of MS and DIR must be repeated. When thesequence is accepted by the A4992, the DIAG output will change state for theSerial Mode Acknowledge Pulse duration (tSAP) , to indicate that the switch wassuccessful. For example if a fault is present when switching between modes,DIAG will be low during the sequence and will first go high, to acknowledgethe mode change, then go low after tSAP. If no fault is present, DIAG will be highduring the sequence and will first go low to acknowledge the mode change,then go high after tSAP. When in serial control mode the function of the STEP,DIR, and MS pins change. STEP assumes the function of STRn, the serial datastrobe input, DIR functions as SDI, the serial data input, and MS as SCK, the

Allegro’s new A4992 IC is a microstepping motor driver with integrated phase current control and a built-intranslator for easy operation. It is a single chip solution designed to operate bipolar stepper motors in full,half, quarter, eighth, and sixteenth step modes, at up to 28 V. The A4992 is supplied in a 20-pin TSSOPpower package with an exposed thermal pad (package type LP). This package is lead (Pb) free with 100%matte-tin lead frame plating.

The A4992 is supplied in a 20-pin TSSOP power package with an exposed thermal pad

A4992 functional block diagram Typical A4992 application schematic

Industry News 01_Layout 1 29/01/2014 15:31

INDUSTRTY NEWS 13


serial clock.The A4992 will remain in the serial mode as long as the SER bit remains

set to 1. If a serial transfer occurs when SER is 0, then the A4992 will revert tothe step and direction control mode after the Serial Mode Exit Time, tSSEX,following the rising edge of STRn. The STRn, SDI, and SCK inputs will thenrevert to their default functions, STEP, DIR, and MS, respectively. The DIAGoutput will change state for tSAP to indicate that the switch was successful.The A4992 can be operated without the serial interface, by using the

default settings and the STEP and DIR inputs. Application-specificconfigurations are only possible, however, by setting the appropriate registerbits through the serial interface. In addition to setting the configuration bits, theserial interface can also be used to control the motor directly.

Detecting stall conditionsA PWM monitor feature is included in the A4992 to assist in detecting the stallcondition of the stepper motor. This feature uses the indirect effect of the backEMF on the current rise quadrant to determine the point at which a stalloccurs.When a motor is running normally at speed, the back EMF, generated by the

magnetic poles in the motor passing the phase windings, acts against thesupply voltage and reduces the rise time of the phase current. The PWMcurrent control does not activate until the current reaches the set trip level forthe microstep position. When a motor is stopped, as in a stall condition, the

back EMF is reduced. This allows the current to rise to the limit faster and thePWM current control to activate sooner. Assuming a constant step rate andmotor load this results in an increase in the quantity of PWM cycles for eachstep of the motor. The A4992 uses this difference to detect a motor changing from

continuous stepping to being stalled. Two PWM counters, one for each phase,accumulate the count of PWM cycles when the phase current is stepped fromzero to full current. At the end of each phase current rise, the counter for thatphase is compared to the counter for the previous current rise in the oppositephase. If the difference is greater than the PWM count difference in theConfiguration register, then the ST fault signal will go low.This stall detection scheme assumes two factors: The motor must be stepping fast enough for the back EMF to reduce thephase current slew rate. Stall detection reliability improves as the currentslew rate reduces.

The motor is not being stepped in full step. Although stall detection cannotbe guaranteed using this detection method, good stall detection reliabilitycan be achieved by careful selection of motor speed, count difference, andby conforming to the above factors.

More application information:www.allegromicro.com/en/Products/ Motor-Driver-And-Interface-

ICs/Bipolar-Stepper-Motor-Drivers/A4992.aspx

LEFT: Effectof stallcondition oncurrent rise

RIGHT: Stalldetect by

PWM countcompare

New current transducer from PEMThe new RCTrms 3-ph current transducer fromPower Electronic Measurements (PEM) deliversa convenient, safe and accurate solution formeasuring current in three phases. It features athin, clip-around, flexible sensor coil andprovides accurate true rms measurement with 4-20mA or 0-5V output, enabling simpleinstallation with PLC’s, SCADA systems orautomation equipment.The compact, DIN-rail or panel mountable

RCTrms-3ph extends the design of PEM’s provenRCTrms single-phase transducer, connecting tothree Rogowski coils to capture measurementsfrom three current phases simultaneously. Thissignificantly reduces cost per channel, whileenabling safe and fast installation. With 18 current ratings options from 100A to

50,000A, and a choice of 300mm, 500mm,700mm or custom coil lengths, the RCTrms-3ph

can be used in many applications and connectedto a wide variety of SCADA systems, PLCs, dataloggers, protection equipment or motor controllers.The clip-around coil design allows fast and easypositioning, and provides accurate results withoutneeding to be centralised around the conductor.An isolated BNC-BNC cable-split option is

available, to ease installation such as whenthreading through existing conduit. RCTrms-3ph operates from a 12V-24V supply,

and generates an accurate, true-RMS output as anindustry-standard 4-20mA or 0-5V signal. Thetransducer provides a galvanically isolatedmeasurement. Each unit is supplied with atraceable calibration certificate.The RCTrms-3ph uses non-magnetic materials,

which ensures excellent linearity and preventsdamage from over currents. Typical accuracy isbetter than 1% of reading from 10 to 100% fullscale. The units are safe, with 2kVdc power-supplyisolation and 2kVpeak coil rating, and are CEmarked and compliant with EMC EN 61326-12006 and IEC61010-1:2001.

Further information can be found atwww.pemuk.com.


14 INDUSTRY NEWS


Vicor has introduced a powerful componentpackaging platform – ‘Converter housed inPackage ChiP’ in March 2013. New generationcomponents can provide a staggering 4X increasein power density and 20 percent reduction inpower losses, i.e. power densities up to 3 kW/in3

(183 W/cm3) and up to 98 % efficiency.ChiP technology is made possible by many

stages being analogous to semiconductor waferfabrication. Individual converters are arrayed toutilize 100 % of the PCB material, with no arealost to lead/pin attach. The array is then sawn intoindividual ‘chip-scale’ components with nosuperfluous, non- value- added packaging, thusminimizing space used in customer designs. ChiP utilizes integrated magnetic structures (but

not for the output inductor) penetrating through ahigh density interconnect substrate withsymmetrically two-sided layout to essentiallydouble power density. Similar to howmicroprocessors are sawn out of scalable wafers, amultiplicity of ChiPs are constructed on a panel,molded, and then sawn into as many as 80individual power components, exposing ‘bar codes’from which interconnect terminals are formed. X-and y- dimensions are flexible, allowing low- orhigh- power converters, with additional flexibility inthe z- axis to allow optimized magnetic transformer

or inductor designs.The lineup of components includes five different

package sizes, with more on the way. This isenabled by the underlying scalability of the ChiPmanufacturing process. Process flexibility enables awide range of power component sizes to be builtusing the same manufacturing tooling andprocesses without waste, resulting in cost-effectiveness. This dimensional flexibility isreflected in the package reference, againanalogous to semiconductors. For example, a‘4623’ package measures 46 mm by 23 mm assawn.ChiP technology supports packages as thin as

4.7 mm ranging from ‘1323’ to ‘6123’ with currentcapability up to 180 A, voltage capability up 430 Vand power capability up to 1.5 kW. With anexpanding array of package sizes, voltage capabilityup to 800 V and power capability up to 4 kW arepossible. ChiP packaging offers flexible coolingpaths including single- or dual- sided cooling plusdissipation through the leads, as well as cold- plateoptions and standardized finned heatsinks.

ChiP bus converter The VI Chip Bus Converter is a Sine AmplitudeConverter (SAC), operating from a 260 to 410VDC primary bus to deliver an isolated 32.5 to

51.3 VDC unregulated secondary voltage.The Sine Amplitude Converter (SAC) uses a

high frequency resonant tank to move energy frominput to output. The resonant tank is formed by aparasitic capacitance and leakage inductance in thepower transformer windings. The resonant LC tank,operated at high frequency, is amplitudemodulated as a function of input voltage andoutput current. A small amount of capacitanceembedded in the input and output stages of themodule is sufficient for full functionality and is keyto achieving high power density of 98 %.Low impedance is a key requirement for

powering a high-current, low-voltage loadefficiently. A switching regulation stage should haveminimal impedance while simultaneouslyproviding appropriate filtering for any switchedcurrent. The use of a SAC between the regulationstage and the point of load provides a dual benefitof scaling down series impedance leading back tothe source and scaling up shunt capacitance orenergy storage as a function of its K factorsquared.However, the benefits are not useful if the series

impedance of the SAC is too high. The impedanceof the SAC must be low, i.e. well beyond thecrossover frequency of the system.A solution for keeping the impedance of the

SAC low involves switching at a high frequency,inthis case 1.5 MHz. This enables small magneticcomponents because magnetizing currents remainlow. Small magnetics mean small path lengths forturns. use of low loss core material at highfrequencies also reduces core losses.A major advantage of SAC systems versus

conventional PWM converters is that thetransformer based SAC does not require externalfiltering to function properly. The BCM380y475x1K2A30 offers low noise,

fast transient response, and industry leadingefficiency and power density. In addition, itprovides an AC impedance beyond the bandwidthof most downstream regulators, allowing inputcapacitance normally located at the input of a POLregulator to be located at the input of the BCMmodule. With a K factor of 1/8, that capacitancevalue can be reduced by a factor of 64X.

Current sharingThe performance of the SAC topology is based onefficient transfer of energy through a transformerwithout the need of closed loop control. For thisreason, the transfer characteristic can beapproximated by an ideal transformer with apositive temperature coefficient series resistance.This type of characteristic is close to the

First Products for Vicor’s Converterhoused in Package Technology

LEFT: ChiP power components are formed out of ascalable power platform


INDUSTRTY NEWS 15


impedance characteristic of a DC powerdistribution system both in dynamic (AC) behaviorand for steady state (DC) operation.

When multiple BCM modules of a given partnumber are connected in an array they willinherently share the load current according to theequivalent impedance divider that the systemimplements from the power source to the point ofload.

Thermal managementLeveraging the thermal and density benefits ofChiP’s packaging technology, the BCM module

offers flexible thermal management options withvery low top and bottom side thermal impedances.Thermally-adept ChiP-based power components,enable customers to achieve low cost powersystem solutions with previously unattainablesystem size, weight and efficiency attributes,quickly and predictably.

Reverse operation BCM modules are capable of reverse poweroperation. Once the unit is started, energy will betransferred from secondary back to the primarywhenever the secondary voltage exceeds VIn • K.

The module will continue operation in this fashionfor as long as no faults occur. The BCM380y475x1K2A30 has not been qualified for continuousoperation in a reverse power condition.Furthermore fault protections which help protectthe module in forward operation will not fullyprotect the module in reverse operation.

Transient operation in reverse is expected incases where there is significant energy storage onthe output and transient voltages appear on theinput. Transient reverse power operation of lessthan 10 ms, 10 % duty cycle is permitted and hasbeen qualified to cover these cases.

CHiP supports all known architectures includingisolated bus conversion, DC/DC, buck, boost, andbuck-boost regulation and POL currentmultiplication. Zero-current and zero-voltageswitching topologies are also supported with theaid of the online-tool PowerBench. AC/DC withPFC eliminates one regulation stage compared to aconventional approach and thus increasesefficiency, and the regulator rejects line frequencyfrom input and regulates the point of load.Component-based approach enables differentpartitioning of the system. The Bus Converter canbe part of the Frontend or the Motherboard,

An BCD380x475x1K2A30 1.2 kW EvaluationBoard is available for order from Digikey, FutureElectronics, and Vicor.

http://www2.vicorpower.com/6123BCM

ABOVE: Principle of Sine Amplitude Converter (SAC)topology

BELOW: AC to Point of Load and Post Boost PFCapplication

Cooling options include top, bottom, or both sides plus leads


16 INDUSTRY NEWS


High Efficiency AC/DC Converter IC with Integrated PFCROHM has recently announced the development ofa high-efficiency AC/DC converter that integratesPFC (Power Factor Correction) and QR (Quasi-Resonant) controllers into a single package, makingit suited for 100 W-class equipment such as TVsand industrial power supplies.The BM1C001F is the first to incorporate an ON/OFF

setting function and new PFC output control methodwithin the PFC controller. This improves efficiency duringlight loads and significantly reduces standby power.Power supplies utilizing this IC ensures compliance withEnergy Star 6.0, the latest international energy standard.In addition, the 2-controller-in-one design reduces thenumber of external parts, contributing to power supplyminiaturization.In recent years came a need for greater efficiency

to improve energy savings in electronic devices ofall types. But in order to control harmonics inequipment over 75 W, which can be harmful toother devices, a PFC controller is required, whichmay have a negative effect on efficiency. Also, toreduce design and evaluation loads while enablingeasy repair, power supply adapters are currentlyprogressing in 100 W-class equipment. However,conventional solutions combine an AC/DCconverter with built-in QR controller with a separate

PFC circuit, which can present problems regardingboard space and power supply size.In response, ROHM integrated a special function that

allows for flexible ON/OFF setting of the PFC controller.This improves conversion efficiency during light loadswhile reducing standby power consumption. Inaddition, a new proprietary PFC output control methodis included in the PFC controller for boost converteroperation, significantly increasing efficiency - particularlyin 100 VAC systems. For example, integrating this IC in a100 W class power supply will result in 89 % efficiencyat 100 VAC and 89.5 % efficiency at 230 VAC, clearingthe latest Energy Star standard (6.0) stipulating at least88% efficiency. Also, layout optimization makes itpossible to integrate the PFC and QR controllers in asingle package, reducing parts count by 20 %compared with conventional solutions.

ON/OFF setting function improves conversionefficiencyThe industry’s first PFC controller ON/OFF settingfunction improves conversion efficiency at light loadsand reduces standby power consumption. Itmonitors the load power on the secondary side andswitches the PFC controller ON/OFF, including in theload region where PFC is not required (below 75

W), improving power supply conversion efficiency.Adopting this IC in 100 W class power supplies

results in a standby power of less than 85 mW at100 VAC and 190 mW at 230 VAC, clearing thelatest Energy Star standard (6.0) requiring 210 mWor less.

www.rohm.com/eu

High-efficiency AC/DC converter IC with integrated PFCand QR control function

100 W-class efficiency comparison Source: ROHM

High Performance DT-Triac™ Technology Platform

For more information pleaseemail [email protected] or call Petra Gerson: +49 6206 503249

High Voltage & Current DT-Triac™

FeaturesTriac for line frequency

Three Quadrants Operation: QI - QIII

Planar passivated chip

Long-term stability of blocking currents and voltages

Voltage range: 800-2200 V

Current range: 10-700 A

ApplicationsLine rectifying 50/60 Hz

Softstart AC motor control

DC Motor control

Power converter

AC power control

Lighting and temperature control

New

Products VR/D /V ITRMS /A PackageCLA60MT1200NHR 1200 60 ISO 247

CLA60MT1200NHB 1200 60 TO- 247

CLA60MT1200NTZ 1200 60 D3 Pak

CLA60MU1200NLB 1200 60 SMPD

CMA60MT1600NHR 1600 60 ISO 247Coming

soon


INDUSTRTY NEWS 17


Engineers developing power supplies for single-phase electronics facenew demands for greater energy efficiency and lower line pollution,along with a growing array of government regulations and commercialdemands to reduce energy consumption. As new semiconductortechnologies such as GaN (Gallium Nitride) and SiC (Silicon Carbide)emerge, new test and measurement tools are required to keep pace.“Thus power is one of the most dynamic segments in electronics todaygiven the intense interest on the part of consumers, business andgovernment to reduce overall energy consumption. Improving batterylife is another key driver,” said Curt Willener, general manager, PowerAnalyzer Product Line, Tektronix.Tektronix’s new PA1000 single-phase power analyzer provides

engineers designing and testing power supplies, consumer electronicsand other electrical products with accurate power measurements.Features such as a color graphical display, one-button applicationmodes and intuitive menu system enable optimum instrument set upin seconds, and the PWRVIEW PC software includes comprehensivereporting features such as a full compliance IEC62301 standby powercertificate.A full-color graphics display makes setup and other tasks easy and

intuitive, with one-button access to measurement results, powerwaveforms, harmonic bar graphs, and menus. Application-specific testmodes for standby current, lamp ballast testing and energy integrationhelp to simplify optimization of instrument settings, saving engineerstime and reducing mistakes. PWRVIEW PC software further simplifiestesting with one-click test automation for compliance-test applications.Standard interfaces include LAN, USB and GPIB,

Spiral shuntsThe PA1000 offers 0.05 % basic accuracy and 1 MHz measurementbandwidth. Two internal current shunts are included - one for preciselow-current current measurements up to 1 A, and another for currentmeasurements up to 20 A. The 1 A shunt is particularly useful for maintaining measurements

resolution and accuracy on demanding low-current signals common tostandby power testing. The 20 A shunt design ensures stable, linear

response over a wide range of input current levels, ambienttemperatures, crest factors, and other variables. This new design issuperior to other shunt technologies and contributes to theinstrument’s reliable accuracy and repeatability over a wide range ofsignal conditions common in today’s power conversion technologies.The spiral construction not only minimizes stray inductance (foroptimum high frequency performance) but also provides for highoverload capability and improved thermal stability.

Application-specific test modesSome applications require special instrument settings to ensure

proper measurements. The PA1000 simplifies setup for theseapplications by automatically choosing instrument settings andparameters that are optimized for each type of measurementapplication, resulting in more reliable measurement results with lessopportunity for user setup error. Ballast mode - synchronizes measurements for highly modulated

electronic ballast waveforms. In modern electronic lighting ballasts, it isoften difficult to make accurate measurements because the outputsignals are high frequency waveforms that are heavily modulated bythe power frequency. Ballast mode provides a way of locking themeasurement period to the power frequency. Standby power mode - driven by consumer demand and energy

efficiency regulations (such as ENERGY STAR), there is an ever-increasing need to measure power consumption of products while theyare in standby mode. One of the most widely used standards formeasurement is IEC 62301. Part of this standard requires themeasurement of power over a prolonged period of time withoutmissing any short duration power events. The PA1000 standby powermode provides continuous sampling of voltage and current to producean accurate Watts measurement over the user specified period. Inrush mode - for measuring the peak current during any event.

Typically this is used to measure the peak current when a product isfirst switched on. Integrator mode - used to provide measurements for determining

energy consumption (Watt-hours, Ampere-hours, etc.). The PA1000 20

New Precision Single-PhasePower Analyzer

Tektronix’s new PA1000 offers 0.05 %basic accuracy and 1 MHz measurementbandwidth


Drives &Controls Exhibition & Conference 2014

The UK’s leading exhibition for Drives, Automation, Power Transmission and Motion Control Equipment

8-10 April 2014 NEC Birmingham

BOOK YOURSTAND NOW LAST FEWREMAINING

Drives 2014 Jigsaw ad DPS_Layout 1 19/12/2013 16:17 Page 1

For FREE show registration visit www.drives-expo.com

Just one of sevenco-located shows

To exhibit contact:Doug Devlin +44 (0)1922 644766 [email protected] Nigel Borrell +44 (0)1732 370341 [email protected]

Drives & Controls is the largest UK exhibition forthe Automation and Motion Control industry.Visiting you will see cutting-edge technologyaligned to this dynamic industry. If you are involvedin manufacturing this is a must-see event with over 300 world-class companies exhibiting.

Media Partners

Associates

...The perfect fitCo-located with Fluid Power & Systems, EuropeanOffshore & Energy, Air-Tech, Plant & AssetManagement, Mach and National Electronics Week

PROGRAMME OF FREE TECHNICAL SEMINARS: • Drives Seminar Theatre • EnergyDialogue Conference • Fluid Power & Air Seminar Theatre • Maintenance Seminar Theatre • TechKnowledg-e

Hydraulics&Pneumatics Plant&Works

EngineeringDrives&Controls

Entry is FREE of charge so register now at www.drives-expo.com

REGISTER

FOR FREE

TODAY

www.drives-expo.com

Primary Sponsors

Drives 2014 Jigsaw ad DPS_Layout 1 22/01/2014 17:52 Page 2

20 INDUSTRY NEWS

The PX8000 brings a new dimension

to power analysis with the introduction of

high-accuracy time-based measurement.

Combining Yokogawa’s leadership in power

measurement with its proven expertise in

oscilloscope design, the PX8000 offers

futureproof technology for the next generation

of design engineers.

See more at: tmi.yokogawa.com/px8000or contact +31 (0) 88 4641429

Precision Making

Time-based precision measurementof power inputs

Packed with power-specifi c analysis capabilities

Manufactured to the highest standards of quality and reliability

Highly fl exible confi guration

Familiar and user-friendly interface

Where power meets precisionPX8000 Precision Power Scope

17 features harmonics analysis to the 50th harmonic as a standardfeature. Harmonics, THD and related measurements can all be analyzedsimultaneously with other power parameters.

PWRVIEW PC softwarePWRVIEW is a supporting software application for Windows PCs thatcompliments and extends the functionality of the PA1000. It enablesthe user to do the following: Communicate with the PA1000 over any of the instrument commports

Change instrument settings remotely Transfer, view, and save measurement data in real-time from theinstrument, including waveforms, harmonic bar charts, and plots

Log measurement data over a period of time Communicate with and download data from multiple PA1000instruments

Create formulae for the calculation of power conversion efficiency

and other values Export measurement data to .csv format for import into otherapplications

Automate instrument setup, data collection, and report generationfor key applications with just a few clicks, using wizard-driveninterfaces

Perform automated full compliance testing for Low Power Standbyper IEC 62301, Edition 2

Additional test automation will be added with future releases

www.tek.com/power-analyzer/pa1000

The 1 A shunt isparticularly usefulfor low-currentsignals

The 20 A shuntdesign ensuresstable responseover a wide rangeof input currentlevels


www.infineon.com/sic SILICON CARBIDE 21


Practical Use of SiC PowerSemiconductorsSilicon Carbide (SiC) power devices are enabling components mainly in the context of higher switchingfrequencies and/or small footprints in power electronics. However, this trend imposes new challenges onthe packaging of the chips. Typical stray elements like inductances become crucial elements in the circuit. Inaddition, different considerations regarding the thermal design in power modules arise when SiC chips aretaken into consideration. Furthermore, the aspects of power density as well as the general utilization of SiC’shigh temperature capability are important factors in reliable implementation of SiC-based powersemiconductors in modern systems. The article will give an insight into how these boundary conditions canbe implemented in innovative solutions using SiC chips. Peter Friedrichs, Senior Director SiC, InfineonAG, Erlangen, Germany

SiC power technology, predominantly inform of Schottky diodes, is meanwhileestablished in the market. High-powersolutions using power-moduletechnologies have recently becomeavailable. Usually, the aim of suchcomponents is to enable system benefits,for example by increasing the switchingfrequency or reducing losses. If this issuccessful, the high cost of SiC-basedcomponents can be compensated byreduced efforts for passive elementsand/or cooling.

While the discrete, unipolar SiC devicesas single chips are more or less ready toachieve higher frequencies (>100 kHz inPFC units), this is still a challenge forpower modules. Solutions in a powerrange with traditional modules have highstray inductances and thus the di/dt isincreased if combined with an increase inthe frequency. A plug and play betweenSiC and Silicon at chip level in existingmodules could devaluate the theoreticalSiC performance. It is necessary toimprove high-current modules in order toget the full benefit of SiC in the frequencyrange above 20 kHz.

Current and power densityconsiderations for state-of-the-art SiCdevicesWhilst the reduction of parasitic elementsin power modules using SiC chips is in linewith the approach also used for futureSilicon solutions, it might happen thatdifferent optimization criteria will apply forthe thermal design. The reason for this isthat the cost contribution of chips in apower module with SiC is different fromone with Silicon. Thus, the best solution fora given frame size with respect tosemiconductor area placed in the module

could be different. Furthermore, it isexpected that the die size of a SiCtransistor at 1200 V, for example, comesdown to one tenth or less of the areaneeded in the current Silicon-based IGBTtechnology, assuming the same totallosses for both options. This will result in ahuge increase in power and currentdensity, and will require more efforts withrespect to an effective heat removal andthe connection of chips to terminals.

SiC diodesSeveral new generations of SiC Schottkydiodes have been introduced since thefirst launch in 2001, each leading to afurther increase in power density. Since2006, Infineon has been using a Merged-Pin-Schottky (MPS) structure [1] for600/650 V diodes, mainly in order to offera sufficient surge-current capability. Otherdevices on the market are designed as JBS

(Junction Barrier Schottky) diodes; from adesign point of view, the layout is similar,the difference being that the main purposeof a JBS is to shield the electric field inreverse mode from the Schottky interfacein order to keep the leakage current low,while in an MPS the main purpose is tooffer surge current. In these diodes, only apart of the active area is used for thecurrent flow; the rest is passive, being inoperation in the MPS for only a shortperiod of time (pulse mode). This againrepresents a further local increase in thecurrent density, as shown in Figure 1 usingthe basic principle of the MPS diode as anexample.

The shown values today are close to thehighest current densities for silicon powerparts, which are known from low voltagetransistors (e.g. 25 V at around 1300A/cm²).Taking into account the voltagedrop at operating temperature (which can

Figure 1: Current densities in recent generations of Infineon’s 650 V MPS diodes (3G and 5G devices)

Infineon_Cover Story_Layout 1 30/01/2014 08:48

22 SILICON CARBIDE www.infineon.com/sic


be > 2V), it becomes obvious that heatgenerated by the power density at the chipwith >2kW/cm² can be effectivelyremoved only if effective heat spreading isimplemented between junction and case.

Another aspect of modern device designis the reduction of the die thickness. In thespecial case of vertical SiC devices, theactive layer is just a few µm thick, and thusany additional material below it is merelyan increase in the (differential) on-resistance and the thermal resistance ofthe chip [2]. It is thus reasonable to thinthe wafers down to the smallest possiblevalue, often defined by the handlingcapability in manufacturing. Infineonintroduced thin (110 µm) wafer deviceswith its 5th diode generation (5G), offeringa further step in power density, as shownin Figure 2.

But this benefit has a drawback – thethermal capacitance, which is importantunder pulse current stress, is also reduceddue to the shrunken volume of thesemiconductor. Technologies musttherefore be developed to compensate forthis, otherwise only reduced pulse ratingsare possible. One approach is a goodthermal connection of the chip to the leadframe, e.g. by solderless assemblytechnique, enabling the utilization of theunderlying copper as a support for pulsedoperation [3].

SiC transistorsThe focus today is on the implementationof unipolar high-voltage (600 V … 1700V) transistors, offering threshold-free linearI-V characteristics, integrated body diodes,and negligible dynamic losses comparedto the competing Si-IGBT technology. TheRon x A values expected for 600 Vcomponents that seem to be achievablelong term are around 1mΩcm² [4]. Thebenchmark with the above-mentionedlow-voltage Silicon technology shows that

such values are still a factor of 20 higherthan the best-in in-class devices for 25 Vtoday. So it seems that experience withSilicon is sufficient to handle high currentdensities for SiC transistors.

Nevertheless, one mode of operationwill need closer attention. It seems to bemandatory for the success of SiCtransistors in industrial applications tooperate the internal body diode as afreewheeling diode. There might be someconcerns regarding efficiency in thisoperation mode, because the forwardvoltage drop is too high; however, one canturn on the channel after a short deadtime and then the I-V characteristic inreverse mode is identical to or evenslightly better than in forward mode (seeFigure 3).

Anyway, there might be a critical

situation should the driver circuit fail whendiode mode is required by the system. Tocope with this mode, one may be forcedto define the actual current rating, not fromthe attractive threshold-less forward I-V, butout of the power handling capability inreverse mode. Thus, lowest VF is required,or alternative solutions for designing thediode function in the 3rd quadrant modemust be developed.

High-frequency optimizationReal power circuits with SiC chips insidecontain inductance and capacitance asmajor parasitic elements, and these causeserious deviations from perfect switching.Basic effects of parasitic inductance are: a voltage dip during turn-on of atransistor, caused by rise of current

a voltage spike during turn-off of atransistor, caused by current fall

the parasitic capacitance together withthe inductance form resonant circuits,which show damped oscillations aftereach switching transition

The effect of parasitics is already morepronounced when a combination of anIGBT with an SiC freewheeling diode isconsidered, since the diode performancealso speeds up the transistor and thus,under improper conditions, oscillationscould be observed as seen in Figure 4 (leftside) when no special precautions aretaken. Therefore, an optimum circuitdesign with minimum parasitic inductanceis a prerequisite for the optimization ofpower semiconductors towards lowestlosses [5].

To meet these demands, rules were

Figure 2: Chip thickness shrink effect on the maximum temperature of an SiC diode with 5.1mm2 areaon a copper lead frame and a power stress of 170 W; due to the smaller die thickness, the effective Rth

is reduced

Figure 3: SiCtransistors (exampleSiC JFET fromInfineon) in 3rdquadrant – channelmode possible toreduce losses byturning on thechannel infreewheelingoperation; however,in critical modes theI-V in diode modemight define thecurrent handlingcapability


The inaugural European Offshore & Energy exhibition and conference, beingheld on 8-10 April 2014 at the NEC Birmingham, brings the upstream energymarket together under one roof; providing visitors with the ideal opportunity tolearn about the industry’s latest product innovations, solutions and services.

For FREE show registration visit www.europeanoffshoreenergy-expo.com

Just one of sevenco-located shows

PROGRAMME OF FREE TECHNICAL SEMINARS: • EnergyDialogue Conference • Drives Seminar Theatre • Fluid Power & Air Seminar Theatre • Maintenance Seminar Theatre • TechKnowledg-e

Co-located with Drives & Controls, Air-Tech, FluidPower & Systems, Plant & Asset Management,Mach and National Electronics Week

...The perfect fit

In association withMedia Partners

Shawn Coles +44 (0)1732 370340 +44 (0) 7919 884650 [email protected] Ryan Fuller +44 (0)1732 370344 +44 (0)7720 275097 [email protected]

To exhibit contact:

Drives&ControlsHydraulics&Pneumatics

Plant&Works

Engineering

Primary Media Partner

REGISTER

FOR FREE

TODAY

www.europeanoffshoreenergy-expo.com

And as well as the extensive range of offerings showcased across the exhibition floor, European Offshore & Energy will also feature EnergyDialogue, a three-dayconference covering a host of energy issues together with product/ solution analysis. EnergyDialogue will ensuredelegates receive all the information they need directly fromleading authorities, policy makers and expert practitioners.

Additionally, TechKnowledg-e, a three-day seminarprogramme, will feature a packed programme coveringeverything from detailed analyses of proven existingtechnology to the latest industry innovations. Delegates willalso have the opportunity to attend event organiser DFAMedia’s other highly complementary co-located exhibitions:Drives & Controls, Fluid Power & Systems, Plant & AssetManagement and Air-Tech. So, whether you are involved in oiland gas, power generation, shale gas or renewable energy,European Offshore & Energy, along with its associatedexhibitions, provides a unique one-stop shop for all yourinformation, technology and product requirements. Don’t missout on being a part of the inaugural European Offshore &Energy show and the UK’s largest co-located industry event!

Join us on Twitter@EuOffandEnergy

Join us on LinkedInEuropean Offshore & Energy

OFFSHORE DESIGN & ENGINEERING EUROPE

OfficialRecruiter

Offshore2014_A4_Layout 1 23/01/2014 15:06 Page 1

24 SILICON CARBIDE www.infineon.com/sic


developed at Infineon [6] for the design ofpower modules. It is well known that astrip-line type of circuit design bringsparasitic inductance to a low value,depending on width and distance of thestrip line [7]. The graph on the right inFigure 4 shows the effect on the switchingperformance of power semiconductors ofan improved module layout, taking theseconsiderations into account. Whereas, inthe standard module, a severe over-voltage peak plus oscillations are observed,the new solution can eliminate theseeffects almost completely. It has to benoted that the improvement was achievedwithout changing the chip technology, butsimply by a more careful design of thepackage. These considerations will alsotake into account the fact that, on themidterm scale, paralleling of a highnumber of dies will be the preferredmethod of achieving higher power levels inSiC.

In order to avoid a degradation of thehigh performing SiC chips by an unsuitableenvironment, Infineon will drive theimplementation of SiC semiconductorsonly in module platforms that enable thisapproach.

Thermal considerationsThe final power-handling capability of asemiconductor module is defined by thechip area inside the module combinedwith the thermal performance expressedby the Rth. For Silicon chips, as much aspossible chip area is usually placed on agiven footprint in a power module,because normally the area cost per chip isnot so much different from the cost permodule area, mainly for larger powerratings.

In the case of SiC, differentconsiderations apply. The chips are muchmore expensive than Silicon ones, so it isbeneficial to choose a different ratio of

chip area to module area to get theoptimum solution with respect to theprice/power ratio for a given footprint. Thereason for this is that, by proper design,

the effective Rth per chip can be reducedby a factor of up to 4 due to heatspreading effects, as shown in Figure 5(base-plate-free module).

Figure 4: Left - standard module switched with 400 A (turn-off) and a DC link bias of 700 V; the green trace shows the voltage; blue (left) or yellow (right)the current; and orange the gate signal. Right - new module approach, DC link bias now 900 V

ABOVE Figure 5: Effectivethermal resistance for aconfiguration in a base-plate-free module. Setupshown on the right side. A0.32 mm thick ceramicwith 0.6 mm Cu on bothsides is assumed

LEFT Figure 6: Extractablecurrent from a systemwith a given Rth and TC, thecomponent having adependence of its on-resistance on temperaturewith an R~T2 law


www.infineon.com/sic SILICON CARBIDE 25


In addition, there is a trend towardshigher maximum chip temperatures, inorder to increase the removable heat for agiven die area, and thus its current-handling capability. It must be mentionedthat doing this will increase the absolutelosses, so that efficiency targets could beviolated, depending on the ratio of lossincrease to increase in power-handlingcapability. In contrast to IGBTs, this lossincrease might be huge for unipolar SiCdevices, because now a resistivecomponent with a heavy increase in theRon with temperature is in place, mostlywith a power law in the form of Ron ~ Tx,with x values between 2 and even 2.5, asrecently presented for high voltage SiCMOSFETs.

Taking this into account, one can seethat, for unipolar devices, a continuousincrease of Tj can even lead to a reductionin current-handling capability, becausemore power is dissipated than can beremoved by the increase of Tj. Figure 6shows this result for a typical dataset,assuming x=2 as the exponent in thepower law for the increase of theresistance with temperature.

In addition to these considerations, it isknown that the reliability of power systems,mainly the power cycling capability, is

degraded when the temperature swingbetween off state and maximumtemperature is increased [8]. Thus, for thesake of both efficiency and reliability, itmight be wiser to offer good cooling andkeep the temperature differences in thethermal stack low.

Literature[1] F. Dahlquist, J. Hancock, M. Treu,

R. Rupp, T. Reimann, “2nd Generation600 V SiC Schottky Diodes Use MergedPn/Schottky Structure For SurgeOverload Protection”, presented at APEC2006[2] R. Rupp, R. Gerlach, U. Kirchner, A.

Schlögl and R. Kern, „Performance of a650V SiC diode with reduced chipthickness”, Proceedings of ICSCRM2011, Mater. Sci. Forum, Vols. 717-720,pp. 921-924, 2012.[3] M. Holz, J. Hilsenbeck, R.

Otremba, A. Heinrich, P. Türkes, R.Rupp,„SiC Power Devices: ProductImprovement using DiffusionSoldering”, Mater. Sci. Forum, Vols. 615-617, pp. 613-616, 2009[4] Y.Tanaka, K.Yano, M.Okamoto,

A.Takatsuka, K.Fukuda, M.Kasuga,K.Arai, T.Yatsuo, “Fabrication of 700VSiC-SIT with Ultra-Low On-Resistance of

1.01mΩ•cm², Mater. Sci. Forum, Vols.257-529, pp. 1219-1222, 2006[5] Miller, Gerhard: New

Semiconductor Technologies ChallengePackage and System Setups, CIPS2010, Nuremberg. Germany[6] Bayerer et al: Power Circuit design

for clean switching, CIPS 2010,Nuremberg, Germany[7] Stockmeier, T, et.al: 1200 A, 3300

V IGBT Power Module exhibiting VeryLow Internal Stray Inductance, PCIM,Hong Kong, 1997[8] U.Scheuermann, R.Schmidt,

“Separating Failure Modes in PowerCycling Tests”, published at CIPS 2012,ISBN 978-3-8007-3414-6

To receive your owncopy of

subscribe today at: www.power-mag.com

Power Semiconductor Solutions Power Semiconductors

Quality ProductsImpeccable ServiceConvenient Purchasing

www.pwrx.com001 724-925-7272

www.pwrx.com/support/HowToPurchase.aspx

PurchaseMitsubishi & Powerex

Inventory in StockOnline by Credit Card


26 POWER MOSFETS www.onsemi.com


MOSFETs Make Step Changein Performance to Meet NewApplication RequirementsLow voltage MOSFET devices (<40 V) are used extensively in the power systems of portable electronicdevices, domestic appliances, data communication servers, medical equipment and telecom infrastructuredeployments. Exacting demands continue to be placed on the engineers involved in the design of suchMOSFETs. The article will look at the forces at work and how significant, yet potentially conflicting, technicalchallenges need to be tackled. Wharton McDaniel, Product Marketing Manager Power MOSFETs,ON Semiconductor, Phoenix, USA

Our everyday lives are now almosttotally reliant on the use of various formsof electronics equipment, however thereare major concerns about the rise inworldwide energy consumption that this isleading to - its depletion of fossil fuels, theimpact it has on the environment (in termsof carbon emissions, etc.) and theenlargement of utility bills too. Greaterconsumer awareness, the implementationof hard-hitting legislative measures (suchas EnergyStar), plus OEMs’ testingperformance roadmaps, are allcompounding the acute pressure beingplaced onto power semiconductormanufacturers. Next generation of powersystems will consist of components thatpossess a number of highly favorableattributes.

Important driversThere are two main dynamics that arecurrently responsible for defining MOSFETdevelopment. On one side you have thehigh end processing being required bysever farms. Here the power demand anddensity of the microprocessors found inracks has grown immensely - presentingelectricity consumption, real estate andthermal management obstacles that needto be overcome. Conversely, in computingit is not the processing capacity that is nowthe main concern. As computer platformshave migrated from bulky desktop PCs tolightweight, streamlined portable products,such as tablets and smartphones, this is nolonger the key selling point for customers.The criteria by which the power systemsare outlined have changed here, withbattery life and system compactness nowbeing prioritized. The switching frequenciesof both these types of power systemsmust be raised, so that smaller magnetics

and passive components can be used,with less space thus needing to beallocated.Having established what the drivers are

at the application level, how does thistranslate into the components specs?There are three main parameters whichshould be scrutinized when specifyingMOSFETs for a power system. These are:1. On resistance (RDSON) is vital when itcomes to mitigating a MOSFET’sconduction losses, so this should be aslow as possible.

2. Figure-of-Merit (FOM) - defined by RDSON

x QG (total gate charge), this is anindicator of both the switching andconduction losses of the MOSFET, so itis used as an important selection criteriawhen deciding on which component touse.

3. Switching performance - The betterswitching characteristics of the MOSFETare, the lower the switching losses willbe. With increased switching frequenciesbeing witnessed all the time, this willbecome even more crucial in thecoming years.

MOSFET design and its implicationsMOSFET usage became more common inthe 1980’s, and in the decades that havepassed since then a lot has changed interms of the way these devices arefabricated and the performance specs theycan deliver. Though it is for very differentreasons, the MOSFETs serving both of thevery disparate applications scenarios wediscussed earlier need to dissipate as littlepower as possible (when active and alsowhen inactive), so keeping RDSON low mustbe one of the fundamental goals. Supportvery high conversion efficiencies is also ahuge plus point, as is having a small die

size - as in both portable products andserver implementations board space is at apremium. The question is how to go aboutachieving this performance ‘wish list’ at thecomponent level.There are two elementary constituents

that MOSFET design can be broken downinto. In the past it was possible forcomponent manufacturers to get awaywith just doing one of these really well,then dealing with the other solely as anafterthought. Now though they need to beaddressed simultaneously, with equalemphasis. Process technology - ten years ago a

CPU generally required a current of around10 A to drive it, today it is more likely to be100 A. This has dictated an ordermagnitude reduction in RDSON simply tokeep form factors and heat dissipationlevels in check, otherwise a substantialexpansion in the size and cost of electronicequipment would have been unavoidable.Furthermore, a MOSFET’s overall switchingperformance can be greatly augmentedthrough a reduction in its capacitances.Continued development of newsemiconductor processes have allowed theMOSFET RDSON and capacitance reductionsneeded by OEMs to be realized. Thesereductions should remain scalable -keeping pace with smaller semiconductorgeometries for the foreseeable future. Theadvent of shielded-gate trench topology ishelping to keep the industry ahead of thegame, with features which minimizeswitching overshoot, complementing thelowered RDSON. With switching frequenciesrising, as already stated, overshoot isbecoming more of an issue, so thistopology’s ability to mitigate its effect ishighly advantageous.Package technology - a great deal of

ON-SEMI_Feature_Layout 1 29/01/2014 15:18

www.onsemi.com POWER MOSFETS 27


effort must be put into curbing theinterconnect and thermal resistances ofany power MOSFET package. Theinterconnect resistance has to beminimized with respect to the RDSON valuesof the die (especially as we are nowreaching sub milliohm levels) - otherwiseany improvement in the Silicon will not beworthwhile. The use of clips in place ofbondwires has been the main means ofkeeping interconnect resistance as a smallpercentage of the total RDSON. Theexposed pad on the bottom of surfacemount packages provides the primary lowresistance thermal path. Now there aresurface mount packages which havereduced thermal resistance to the top ofthe package, allowing heatsinking from thetop of the package (see Figure 1). Thistype of package is important where the

PCB cannot be an effective heatsink. It isclearly no good if semiconductor expertisehas managed to provide strongperformance qualities only to then be letdown by poorly implemented packaging.

The continuing reduction in die size overthe years brought on the migration fromleaded power packages, such as TO220, tosurface mount packages for singleMOSFETs and now to dual packages. Highpower phase pairs of MOSFETs now areavailable in 5 mm x 6 mm packages.Performance is enhanced due to thereduction of the interconnect resistancesand inductances. Also, this gives thedesigner a great advantage where space isat a premium as is shown in Figure 2.

The specification of MOSFETs whichoffer engineers an optimized balance ofconduction loss characteristics and

switching efficiency can prove vital to endproduct development. Utilization of thelatest semiconductor processes andpackaging technologies have helped tominimize the losses being experienced aswell as reducing the energy dissipatedfrom system designs so that theaccompanying thermal managementmechanisms take up less space. Thismeans that the end products can beplaced in to sleeker, less cumbersomeform factors. Also greater system reliabilityand longer lifespan can be assured.

ConclusionsIn conclusion, there needs to be continuedinvestigation and then implementation,within the power electronics arena, ofmore effective ways by which to curb thelosses that are related to the MOSFETswitching activity and improve conversionefficiency levels, while simultaneouslyreducing RDSON. Technological progression isenabling processes where both RDSON andthe capacitances associated with theswitching circuit can both be loweredsubstantially. The upshot of this is that thePCB area that needs to be utilized andswitching performance characteristics areboth being improved as more compactpassive components can be employed andhigher switching frequencies can beutilized. Difficulties still lie ahead, however- eventually die sizes will start to get sosmall that the approaches to interconnectand thermal management will have tochange dramatically again.

Figure 1: Thermally enhanced SO8FL packagewith heatsink

Figure 2: Progression of package technology

To receive your own copy of PEE subscribe today at:

www.power-mag.com

ON-SEMI_Feature_Layout 1 29/01/2014 15:18

28 POINT-OF-LOAD OPTIMIZATION www.ericsson.com/ourportfolio/products/power-modules


Fine Tuning of Digital Controlin POL Regulators OptimizeDynamic Response andMinimize Component Count Power system design encompasses many requirements and criteria. This article is about perhaps the mostbasic of all requirements – a stable voltage supply of the operating circuitry. Patrick Le Fèvre, EricssonPower Modules, Sweden

Each electronic circuit on the printedcircuit board (PCB) is designed to operateover a limited range of DC voltage, outsideof which circuit operation or performanceis not guaranteed. A generalized view ofthe static voltage and associated toleranceis shown in Figure 1. In order to achieve high speed and low

power dissipation, recent microprocessor ,DSP and FPGA chips require voltage levels

of 1 to 3.3 VDC with a tolerance of as littleas ±30 mV. These types of requirementsplace severe demands upon the powersystem.

Dynamic response Most types of circuits demand dynamicinput current for short periods of time. Thepower system is designed to handle fastdynamic current by means of decoupling

capacitors located in proximity of theelectronic circuits, while the power supplyhandles low frequency and longer durationdynamic demands (see Figure 2). The power supply control loop, the

distribution impedance and the decouplingcapacitors decides how the dynamiccurrent is shared between the decouplingcapacitance and the power supply.

Digital control of POL regulatorsDigital control is not a magic wand that cantwist the physical laws, but it makes itpossible to improve and optimize thedynamic response performance becausethe digital protocol does not have to takeinto account any tolerance problems whichare the case for the capacitive and resistivenetworks used in analog control loops.Ericsson now offers a useful software

tool that designs, simulates, analyses andconfigures the POL regulator withinminutes. It includes simple tools for robustdesign of control loops, together with moreadvanced design and analysis tools tooptimize the dynamic responseperformance and minimize the number ofdecoupling capacitors needed for a certainload transient requirement. The tools areintegrated in the DC/DC Power Designersoftware, enabling quick and efficientautomated design and analysis.

Modelling of the POL regulator The most common topology for POL

Figure 1: Circuit operating voltage limits

Figure 2: Principals of dynamic responseFigure 3: Model of the power module withexternal decoupling capacitors

Ericcson_Feature_Layout 1 29/01/2014 15:22

www.ericsson.com/ourportfolio/products/power-modules POINT-OF-LOAD OPTIMIZATION 29


regulators is the buck converter. Theoutput filter of the POL regulator consistsof an internal output capacitance and anadditional mix of different types and sizesof decoupling capacitors on the PCB inproximity of the electronic circuits. Figure 3shows the fundamental schematic. As thedecoupling capacitors become a part ofthe system that should be controlled, theymust be included in the model. The modelalso includes a resistance, Rwire, for

modelling of the lead resistance on thePCB. Two different types of capacitors areincluded in the model.

System analyzes and componenttolerances The decoupling capacitors are entered intothe tool with capacitance and ESR,including tolerances. Tolerances areimportant because they define the‘corners’ of the system. This is handled by

calculating all the corner cases for thepower train filter. Figure 4 shows the enterfield for capacitor data.

Using two different capacitors, each withtolerances for the capacitance and the ESR,yields a total of 16 corner systems andone typical system. All 17 systems (16combinations of min/max values, plusone with typical values) are then simulatedand evaluated. Each system’s naturalfrequency and damping ratio are thencalculated and the results are used asinput to the control loop design.

Control loop designThe tool uses a linearized model of thepower train and a system transfer functionbased on a sampled state-space modeland the control loop design is fairly straightforward using standard control theory. Intoday’s digital control circuits, the PIDregulator is often implemented as secondorder digital filter, where the integrator partis fixed and in many cases impossible toremove. Hence, the software tool isdesigned to define the placement of twozeros and the total gain in the loop.

A common tool for open loop design isthe Bode plot, shown in Figure 5, in whichthe compensation zeros are placed toobtaining the proper behavior and adjustthe gain to meet the robustnessrequirements in terms of phase and gainmargins. Common values for theserequirements are 60º, and 6 dB,respectively.

The tool plots the typical system,analyses all 16 corners systems, and plotsthe systems with the minimum andmaximum values for the phase and gainmargin. If all design requirements arefulfilled, a green signal together with theword ‘Stable’ is shown in the upper leftcorner, which gives quick answer of theloop stability.

However, the phase and gain marginsare not enough for a robust control loop.This cannot be observed in the open loopBode plot only, but becomes obvious inthe closed Bode plot, see Figure 6, andcan be handled with additional designrequirements.

If the gain becomes too high the systemattenuation at higher frequencies stops tobe monotonically increasing, and insteadstarts to increase towards the Nyquistfrequency. The Nyquist frequency is fs/2,where fs is the switching frequency.Another indication is that the gain aroundthe system’s resonance frequency forms apeak. This results in a sensitive system andoutput voltage oscillations duringtransients. This can be avoided ifrestrictions are put on the peak gain andthe gain at the Nyquist frequency. Therules (peak gain < 1 dB; gain at Nyquist <

Figure 4: Capacitor description

Figure 5: Open loop Bode plot


30 POINT-OF-LOAD OPTIMIZATION www.ericsson.com/ourportfolio/products/power-modules


-6 dB) have been shown in practice towork well.

An additional key figure that is a littlecoinciding with these former key figures isthe bandwidth, which in the past hasearlier been used as a rule of thumb(bandwidth < fs/10) for avoidingoscillations.

This rule of thumb describes thebenefits of high switching frequency.Besides the advantage of smallercapacitive and inductive components,higher switching frequency will make itpossible to increase the bandwidth of thecontrol loop and thereby improve thedynamic response.

The tool plots the typical system,analyses all 16 corners systems, and plotsthe systems with the minimum andmaximum values for Peak gain, Bandwidth,and Gain at Nyquist frequency.

A common ‘rule of thumb’ forplacement of the control loop’s zeros thatis robust and gives good transient behavioris: Place first compensation real zero at the systems typical resonance frequency

Place the second real zero one octavebelow the first zero

This is implemented in the tool for asimple and robust design, which will inmany cases be good enough.

The tool also has an Auto gain function,which automatically maximize the loopgain for maximum performance with thedesign requirements fulfilled for the typicalsystem. This makes the manual design ofthe zero’s placement much easier andquicker.

Dynamic response optimizationIn the case of a known system the loadtransient can be optimized by placingthe compensation zeros differentlycompared with the rule of thumbdescribed above.

A goal function is defined as theweighted sum, wx, of the voltage deviation, dV, and recovery time, Trec,during load transients and eachnormalized with the corresponding systemrequirements:

The tool picks zero pairs around thesystems resonance frequency and uses aniterative method to find the zero pairwhich gives the smallest Goal value. First acoarse search is made and then a secondfine tuning to find the optimumplacement.

For each zero pair placement the control

loop’s gain is designed based on theclosed loop Bode analyses describedearlier. Then a load transient simulation isperformed with the following user definedload transient key figures: Low current level [A] High current level [A] Slew rate [A/µs] Load transient period [ms]

The zeros placements with the smallestGoal value will result in a design withoptimized dynamic response.

Minimizing the decouplingcapacitanceThe tool can also be used in the design ofthe decoupling capacitance network bydetermining the type and number ofcapacitors and capacitance that is neededin order to meet the load transientrequirements. Decoupling capacitorsoccupy expensive area on the PCB and isalso adding to the system cost. Therefore itis important to optimize the system design

With the default PID settings and a 5-15A load step, with a slew rate of 10 A/µs,gives the voltage deviations andrecovery time shown to the left in Figure7. Using the tool’s rule of thumb forplacement of the controller’s zeros givesthe result shown to the right. This showsthe potential of optimizing PID for aknown load, where in this case, thevoltage deviations are halved and therecovery times are reduced with a factorof three.

Using the load transient optimizationwith equal weights for voltage deviationsand recovery time yields the followingresults shown in Figure 8, which showsthat the recovery time can be improvedfurther without significantly increasing thevoltage deviations.

The second design example shows howit is possible to reduce the decouplingcapacitance and number of capacitors andstill improve the dynamic responseperformance.

Using the tool’s ‘rule of thumb’ PID

Figure 6: Closed loop Bodeplot

Figure 7: Default PID setting (left) vs. “rule of thumb” PID setting

for minimum decoupling capacitance andnumber of capacitors.

Design examplesThe first example shows the differencebetween dynamic response at loadtransients with standard PID settings andoptimized settings. The conditions are: 2 pieces of 470 µF, 10mΩ 8 pieces of 100 µF, 5mΩ Input voltage is 12V, output voltage 1.0 V.

Figure 8: PID settings optimized for bestdynamic response


www.ericsson.com/ourportfolio/products/power-modules POINT-OF-LOAD OPTIMIZATION 31


settings the number of capacitors can bereduced to one 470 µF and six 100 µF. Asthe tool runs the optimization for the worstcase system, the load transient on the

typical system has a margin comparedwith the load requirements, shown to theleft in Figure 9. Further improvementscould be done with the optimized PID

settings as shown to the right in Figure 9.The result shows that the recovery timesare halved and still with maintained voltagedeviations.

Customer application exampleThe tool has been used in actual customerapplications with very good results. Thisexample shows the improvementsachieved using two paralleled EricssonBMR464 POL regulators with optimizedPID settings compared with the defaultsettings.

System requirements: Vin = 12 V Vout = 1.0 V Current transient 26 to 44 A with1.8 A/µs slew rate

Vout tolerance = ±3 % (±30 mVp-p)

With the default PID settings and 5,000µF electrolytic capacitors the measuredVout deviation was 48 mV (see Figure 10).With optimized PID settings and

reduced decoupling capacitance with1,000 µF ceramic capacitors and twoelectrolytic capacitors with 1,000 µFrespective 470 µF capacitance the Voutdeviation is improved to 32 mVp-p (seeFigure 11).ConclusionsThe default wide range PID that EricssonPOL regulators are delivered with isdesigned for robust and stable operation.With the described tool integrated in theEricsson DC/DC Power Designer softwareit is possible for the power systemdesigner to fine tune the POL regulator’scontrol loop in the particular applicationand significantly improve and optimize thedynamic response performance and/orreduce the required decouplingcapacitance.Fewer output decoupling capacitors can

be used to insure a given voltage tolerancewhich means lower component cost andmost importantly it will also have a hugeimpact and pay-off in terms of reducedtime-to-market and increased systempackaging density.

Figure 9: Optimized PID for minimum decoupling capacitance (left), further improvements could bedone with the optimized PID settings as shown to the right

Figure 10: Vout

deviation withdefault PIDsettings and

5,000 µF

Figure 11:Dynamic

response withoptimized PID

settings andreduced

decouplingcapacitance

To receive your own copy of

Power Elecronics Europe subscribe today at:

www.power-mag.com


32 PRODUCT UPDATE


Measuring electrical current in a conductor is critical tomany systems. Classic measurement techniques

use shunt resistors, currenttransformers or magnetic field

sensors. Hall effect sensors arefrequently the choice whenmeasuring DC currents.However the new, improvedMLX91209CA can be appliedin all current sensingsituations. This versatile, high

speed Linear Hall IC, whichfeatures a programmable sensitivity

range from 5 to 150 mV/mT, is suitablefor measuring DC and AC currents up to

200 kHz. Calibration is done using PTC(Programming Through Connector) protocol. This

modulates the supply voltage and does not requireany additional pin for programming, enabling the most

efficient production flows. A linear analog output permits useof the sensor in applications where a very fast response of 3 µs is

required. Custom calibration is best performed in-situ after the sensor is fixed with respect to thecurrent conductor and ferromagnetic core so that a calibrated current sensitivity is achieved. Typicalaccuracy of a current sensing system based on the MLX91209 is better than ±0.5 % at roomtemperature or ±2 % over the full temperature range (from -40°C to 125°C) when applying in-circuit end of line calibration.

www.melexis.com/91209

Linear Hall Current SensorTo receive your

own copy of

Power

Electronics

Europe

subscribe today

at:

www.power-mag.com

Name:

Company Name:

Address:

Post Code:

Tel: Total Number of Copies @ £ p+p Total £

Drives S & S Hyd H/B Pne H/B Ind Mot CompH/B H/B Air

DFA MEDIA LTD,

QUANTITY QUANTITY

Hydraulics&Pneumatics There are now 6 of these handy

reference books from the publishers ofthe Drives & Controls and

Hydraulics & Pneumatics magazines.

Published in an easily readable styleand designed to help answer basicquestions and everyday problems

without the need to refer to weightytextbooks.

We believe you’ll find them invaluableitems to have within arms reach.

From the publishers of

QUANTITY QUANTITY QUANTITY

If you would like to obtain additional copies of the handbooks, please completethe form below and either fax it on 01732 360034 or post your order to:

You may also telephone your order on 01732 370340 Cheques should be made payable to DFA MEDIA LTD and crossed A/C Payee.Copies of the handbooks are available at £4.99 per copy.Discounts are available for multiple copies. 2-5 copies £4.30, 6-20 copies £4.10, 20+ copies £3.75.

QUANTITY

PRACTICAL ENGINEER’S HANDBOOKSPRACTICAL ENGINEER’S HANDBOOKS

HYDRAULICS

INDUSTRIALMOTORS

SERVOSAND STEPPERS

PNEUMATICS

COMPRESSED AIRINDUSTRIAL

ELECTRIC DRIVES

PLEASE ALLOW UPTO 28 DAYS FOR DELIVERY

192 The High Street, Tonbridge, Kent TN9 1BE

Engineers Handbook, DFA MEDIA LTD,192 The High Street, Tonbridge, Kent TN9 1BE

Postage and Packaging:1-3 copies: £2.99 4-6 copies: £4.99 7 or more copies: £6.99

Product Pages.e$S_Layout 1 29/01/2014 15:13

WEBSITE LOCATOR 33


AC/DC Connverters

www.irf.comInternational Rectifier Co. (GB) LtdTel: +44 (0)1737 227200

Diodes

Discrete Semiconductors

Drivers ICS

EMC/EMI

www.microsemi.comMicrosemiTel: 001 541 382 8028

Fuses

Ferrites & Accessories

GTO/Triacs

Hall Current Sensors

IGBTs

DC/DC Connverters

DC/DC Connverters


www.power.ti.comTexas InstrumentsTel: +44 (0)1604 663399


www.neutronltd.co.ukNeutron LtdTel: +44 (0)1460 242200

Harmonic Filters

www.murata-europe.comMurata Electronics (UK) LtdTel: +44 (0)1252 811666

Direct Bonded Copper (DPC Substrates)

www.curamik.co.ukcuramik electronics GmbHTel: +49 9645 9222 0


www.mark5.comMark 5 LtdTel: +44 (0)2392 618616


www.dgseals.comdgseals.comTel: 001 972 931 8463



www.digikey.com/europeDigi-KeyTel: +31 (0)53 484 9584


www.dextermag.comDexter Magnetic Technologies, Inc.Tel: 001 847 956 1140



www.protocol-power.comProtocol Power ProductsTel: +44 (0)1582 477737

Busbars

www.auxel.comAuxel FTGTel: + 33 3 20 62 95 20

Capacitors

Certification

www.powersemiconductors.co.ukPower Semiconductors LtdTel: +44 (0)1727 811110

www.powersemiconductors.co.ukPower Semiconductors LtdTel: +44 (0)1727 811110

Connectors & Terminal Blocks

www.auxel.comAuxel FTGTel: +44 (0)7714 699967

www.productapprovals.co.ukProduct Approvals LtdTel: +44 (0)1588 620192

www.proton-electrotex.com/ Proton-Electrotex JSC/Tel: +7 4862 440642;

www.voltagemultipliers.com Voltage Multipliers, Inc.Tel: 001 559 651 1402



Arbitrary 4-Quadrant PowerSources

www.rohrer-muenchen.deRohrer GmbH Tel.: +49 (0)89 8970120

Website Locator_p41-42 Website Locator 30/01/2014 09:05

34 WEBSITE LOCATOR


Power Modules

Power Protection Products

Power Semiconductors

Power Substrates

Resistors & Potentiometers

RF & Microwave TestEquipment.

Simulation Software

Thyristors

Smartpower Devices

Voltage References

Power ICs

Power Amplifiers

Switches & Relays

Switched Mode PowerSupplies

Switched Mode PowerSupplies

Thermal Management &Heatsinks


www.rubadue.comRubadue Wire Co., Inc.Tel. 001 970-351-6100






www.auxel.comAuxel FTGTel: +44 (0)7714 699967




www.ar-europe.ieAR EuropeTel: 353-61-504300



www.universal-science.comUniversal Science LtdTel: +44 (0)1908 222211

www.power.ti.comTexas InstrummentsTel: +44 (0)1604 663399


www.dau-at.comDau GmbH & Co KGTel: +43 3143 23510

www.power.ti.comTexas InstrummentsTel: +44 (0)1604 663399



www.isabellenhuette.deIsabellenhütte Heusler GmbH KGTel: +49/(27 71) 9 34 2 82

ADVERTISERS INDEX

Mosfets

Magnetic Materials/Products

Line Simulation

Optoelectronic Devices


Packaging & Packaging Materials






www.biaspower.comBias Power, LLCTel: 001 847 215 2427

www.digikey.com/europeDigi-Key Tel: +31 (0)53 484 9584


www.digikey.com/europeDigi-Key Tel: +31 (0)53 484 9584






www.abl-heatsinks.co.ukABL Components LtdTel: +44 (0) 121 789 8686

www.hero-power.com Rohrer GmbH Tel.: +49 (0)89 8970120

www.rohrer-muenchen.deRohrer GmbHTel.: +49 (0)89 8970120


APEC 9Cornell Dubilier 4CT Concepts 11Danfoss 7DFA Media 32

Drives and Controls 2014 18/19European Offshore & Energy 23Fuji IFCInternational Rectifier OBCIxys 16

PCIM 2014 IBC

Powerex 25

Toshiba 8

Yokogawa 20

Website Locator_p41-42 Website Locator 30/01/2014 09:05

Powerful?…then you are right here!

The marketplace for developers and innovators.Future starts here!

International Exhibition and Conferencefor Power Electronics, Intelligent Motion,Renewable Energy and Energy ManagementNuremberg, 20 – 22 May 2014

More information at +49 711 [email protected] or pcim-europe.com

35_PEE_0114_Layout 30/01/2014 08:37 Page 1

Ultra-Low RDS(ON)

Automotive COOLiRFET™

The new International Rectifi er AEC-Q101 qualifi ed COOLiRFET™ technology sets a new benchmark with its ultra-low RDS(on). The advanced silicon trench technology has been developed specifi cally for the needs of automotive heavy load applications offering system level benefi ts as a result of superior RDS(on), robust avalanche performance and a wide range of packaging options.

The COOLiRFET™ Advantage:• Benchmark RDS(on)

• AEC Q101 qualifi ed

• High current capability

• Robust avalanche capability

Key Applications:• Electric power steering

• Battery switch

• Pumps

• Actuators

• Fans

• Heavy load applications

PackageRDS(on) Max@

10Vgs (m )

QG Typ

(nc)

ID Max

(A)

RthjC

MaxPart Number

D2PAK-7P

0.75 305 240 0.40˚C/W AUIRFS8409-7P

1.0 210 240 0.51˚C/W AUIRFS8408-7P

1.3 150 240 0.65˚C/W AUIRFS8407-7P

D2PAK

1.2 300 195 0.40˚C/W AUIRFS8409

1.6 216 195 0.51˚C/W AUIRFS8408

1.8 150 195 0.65˚C/W AUIRFS8407

2.3 107 120 0.92˚C/W AUIRFS8405

3.3 62 120 1.52˚C/W AUIRFS8403

TO-262

1.2 300 195 0.40˚C/W AUIRFSL8409

1.6 216 195 0.51˚C/W AUIRFSL8408

1.8 150 195 0.65˚C/W AUIRFSL8407

2.3 107 120 0.92˚C/W AUIRFSL8405

3.3 62 120 1.52˚C/W AUIRFSL8403

TO-220

1.3 300 195 0.40 ˚C/W AUIRFB8409

2.0 150 195 0.65 ˚C/W AUIRFB8407

2.5 107 120 0.92 ˚C/W AUIRFB8405

DPAK

1.98 103 100 0.92 ˚C/W AUIRFR8405

3.1 66 100 1.52 ˚C/W AUIRFR8403

4.25 42 100 1.90 ˚C/W AUIRFR8401

IPAK

1.98 103 100 0.92 ˚C/W AUIRFU8405

3.1 66 100 1.52 ˚C/W AUIRFU8403

4.25 42 100 1.90 ˚C/W AUIRFU8401

THE POWER MANAGEMENT LEADER

For more information call +49 (0) 6102 884 311or visit us at www.irf.com

36_PEE_0114_Layout 29/01/2014 14:53 Page 1

SILICON CARBIDE Practical Use of SiC Power Semiconductors

Documents

Transcript of SILICON CARBIDE Practical Use of SiC Power Semiconductors