featured product: by RTC Media

featured product:

BitWave BW1102 Softransceiver RFIC

HARDWARE/SOFTWARE CO-DESIGN

wireless communications: Multi-mode Transceivers technology focus: Next-Generation Portable Displays portable power: Powering LED Backlights

May 2008

www.portabledesign.com

An RTCâ&#x20AC;&#x2C6;Group Publication

CEO Interview: John East, Actel

Redefine the portable world. We’ve got the broad line of performance driven analog products to make it happen. Today’s breakthrough portable products demand analog solutions that are compact, low profile and highly efficient. Our high performance analog ICs are key in many of  Power management ICs today’s market-defining portable products. We provide the highest quality power  Battery chargers management and signal processing components for handheld consumer, industrial and  Display & LED drivers communications applications. All our products are backed by the hands-on expertise  DC/DC converters  RF PA & power controllers of our highly trained applications team. They’ll work with you to design the most efficient portable solutions, ensuring that your next product works right—the first time. Get your FREE high-performance portable analog solutions brochure at www.linear.com/designtools/brochures/index.jsp Performance Driven Products

LT, LTC and

are registered trademarks of Linear Technology Corp.

contents

departments

editorial letter daveâ&#x20AC;&#x2122;s two cents industry news analystâ&#x20AC;&#x2122; pages product feature products for designers second opinion

cover feature

5 6 8 14 40 42 46

11 industry news Variable Time (analog, RF)

Hardware

Hardware Software Co-Design Accelerates 18 Development of Smart Wireless Devices Matti Kattilakoski, Navicron

Model-Based Design Accelerates 22 Functional Verification

David Lidrbauch, The MathWorks

Event-Based

25 cover feature open state

collapsed stat

Finite State Machines

Mechan

wireless communications Glass substrate

Developing Highly Integrated 26 Thin film stack Solutions to Meet the Changing Air gap Demand of Multi-Mode Devices

Duncan Pilgrim, Sequoia

Reflective membrane

portable power Backlight LED Driver System in 30 Small Format Liquid Crystal Displays

Dario Nurnad, National Semiconductor

technology focus

Discre

38 reflective pixel

Red Subpixel

Defining the Display Technology 36 Needs of the Future

Jim Cathey, Qualcomm MEMS Technologies, Inc.

CEO Interview

John East 48

Actel

44 flash memory

MAY 2008

Gre

team editorial team

Editorial Director Editor-in-Chief Managing Editor Copy Editor

Creative Director Art Director Graphic Designer Director of Web Development

Web Developer

Associate Publisher Product Marketing Manager Western Advertising Manager Western Advertising Manager Eastern Advertising Manager Eastern Advertising Manager

Warren Andrews, warrena@rtcgroup.com John Donovan, johnd@rtcgroup.com Marina Tringali, marinat@rtcgroup.com Rochelle Cohn

art and media team Jason Van Dorn, jasonv@rtcgroup.com Kirsten T. Wyatt, kirstenw@rtcgroup.com Christopher Saucier, chriss@rtcgroup.com Marke Hallowell, markeh@rtcgroup.com James Wagner, jamesw@rtcgroup.com

management team

Untitled-1 1

4/13/08 3:57:12 PM

Circulation

Chief Executive Officer Vice President Vice President of Finance Director of Corporate Marketing Director of Art and Media

Marina Tringali, marinat@rtcgroup.com Aaron Foellmi, aaronf@rtcgroup.com Stacy Gandre, stacyg@rtcgroup.com Lauren Trudeau, laurent@rtcgroup.com Nancy Vanderslice, nancyv@rtcgroup.com Todd Milroy, toddm@rtcgroup.com Shannon McNichols, shannonm@rtcgroup.com

executive management

HOW WELL DO YOU KNOW THE INDUSTRY?

John Reardon, johnr@rtcgroup.com Cindy Hickson, cindyh@rtcgroup.com Cindy Muir, cindym@rtcgroup.com Aaron Foellmi, aaronf@rtcgroup.com Jason Van Dorn, jasonv@rtcgroup.com

portable design advisory council Ravi Ambatipudi, National Semiconductor Doug Grant, Analog Devices, Inc. Dave Heacock, Texas Instruments Kazuyoshi Yamada, NEC America

corporate office WWW.EMBEDDEDCOMMUNITY.COM

The RTC Group 905 Calle Amanecer, Suite 250 San Clemente, CA 92673 Phone 949.226.2000 Fax 949.226.2050 www.rtcgroup.com

For reprints contact: Marina Tringali, marinat@rtcgroup.com. Published by the RTC Group. Copyright 2008, the RTC Group. Printed in the United States. All rights reserved. All related graphics are trademarks of the RTC Group. All other brand and product names are the property of their holders. Periodicals postage at San Clemente, CA 92673 and at additional mailing offices. Postmaster: send changes of address to: Portable Design, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673. Portable Design(ISSN 1086-1300) is published monthly by RTC Group 905 Calle Amanecer, Suite 250, San Clemente, CA 92673. Telephone 949-226-2000; 949-226-2050; Web Address www.rtcgroup.com.

embeddedcommad_14v.indd 1 PORTABLE DESIGN

11/13/06 5:55:59 PM

C Block

On February 17, 2009, analog TV in the U.S. will go off the air. The digital TV channels will be compacted into channels 2-51 (54-698 MHz), freeing up the 700 MHz portion of the spectrum (Channels 52-69, 698-806 MHz). The FCC auctioned off that real estate in March. Google weighed in early on with a letter to the FCC requesting that access to this spectrum be open to all devices. When the FCC ruled in Google’s favor, Verizon filed suit challenging the ruling; faced with a firestorm of bad PR, they later withdrew their suit. Google, meanwhile, anted up $4.7 billion to bid for Block C (746-757 and 776-787 MHz). Verizon eventually outbid Google and paid $9.36 billion for six C Block licenses, promising that they would abide by the open access provisions (details to follow). This spectrum will now be used primarily for wireless telephony and data services, as well as TV broadcasts optimized for mobile and handheld devices. AT&T, the other big auction winner, then announced that they too would open up their network (details to follow). Skeptics and conspiracy theorists abound, but implementation issues aside, Google’s having “put in the fix” for open wireless access is bound to spur wireless innovation and even faster adoption of portable wireless devices. Google got what they wanted for next to nothing. “Gaming the system!” as their critics charge, or just good business? Both, it seems to me, and I couldn’t care less, since it’s smart, ethical and works out well for both designers and consumers.

White Spaces

With the 700 MHz spectrum now off the table, the scramble is on for the remaining “white spaces” or guard bands between the new digital TV channels. In 2004 the FCC issued a notice of proposed rulemaking (NPRM) that would open up these chunks of spectrum for unlicensed low-power, portable devices as well as higher-power fixed devices that could provide broadband wireless access to por-

editorial letter

In real estate, value depends on “location, location, location.” The same is true for RF real estate, with $19.6 billion changing hands at the FCC’s recent auction of the 700 MHz portion of the spectrum. The results of that auction will have a major impact on mobile wireless access, as will the FCC’s pending ruling on the use of the remaining white spaces in the compacted broadcast TV spectrum (channels 2-59). Where there’s money involved, there’s politics—and in this case there’s plenty of both.

tables—as long as they include a “spectrum sensing” capability that would enable them to avoid interfering with TV broadcasts and wireless microphones. The IEEE quickly started to work on IEEE 802.22, a standard for wireless regional area networks (WRANs) that would work in these white spaces. The SDR Forum got very focused on the spectrum sensing problem, since that’s one area where cognitive radio holds considerable promise. Google wrote a letter to the FCC stating, “Coupled with the ‘Android’ open-source platform for mobile consumer devices, TV white spaces can provide uniquely low-cost mobile

Spectrum Politics john donovan, editor-in-chief

broadband coverage for all Americans.” Joined by Microsoft, Dell, HP, Intel, Philips and others—now calling themselves the White Spaces Coalition—Google called for unlicensed, open access to this spectrum. It didn’t take long for the Empire to strike back. The CTIA, representing carriers and broadcasters, suddenly thought it would be a great idea to license and auction off these spaces to the highest bidder. The CTIA was joined by the NAB, the NFL, NBC, Disney and Shure—leading “a broad coalition of highprofile wireless microphone users, organized as the Microphone Interests Coalition (MIC).” Microsoft didn’t help out here. In August 2007 they proudly sent the FCC a cognitive radio they developed that they claimed “detected DTV signals at a threshold of -114 dBm in laboratory bench testing with 100 percent accuracy.” The test device failed for FCC engineers not once but twice. This brought demands from broadcasters to give up on spectrum sensing and just license the frequencies (presumably to them). Philips is hoping to ride to the rescue with its own cognitive radio system before the FCC ends field testing in June. Look for a ruling in October. So while you toil away over a hot oscilloscope or compiler, your wireless future is being worked out in C Block and the White Spaces. As Dave Barry would say, hey, what a great name for a rock band!

MAY 2008

dave’s two cents

Maybe it’s just me, but it seems that straight answers are getting harder to find. While teenagers and spouses have developed “not giving straight answers” into an art form, it seems to have spread into society in general. Recently, a flight attendant provided an example of a less than straight answer. Two airline passengers who seemed to be very infrequent fliers were the last to board the plane and moved to First Class as compensation for some airline sin. One asked the flight attendant for an adapter to plug his earphones into the seat plug. The attendant responded with “There is no audio on this aircraft.” When the other asked for an

dave’s two cents on...

Straight Talk

adapter, she got the same answer. When I advised the passengers that the plug was only a power outlet, they replied, “Why didn’t she say that instead of saying there was no audio on the plane?” I was amazed that the passengers even had found the under-seat power outlet. This year’s 25th International Battery Seminar and Exhibit was started with papers discussing the safety aspects of Lithium and LithiumIon (Li-Ion) batteries. The first speaker, Dan Halberstein, presented: Lithium Battery Public Awareness Initiative. The Web site “safetravel. dot.gov” contains much of the material Dan presented. The idea is that an informed traveler can help prevent on-aircraft battery incidents, which are reportedly on the rise. Since January 2008, new rules have been implemented on board an aircraft regarding lithium batteries. Spare or uninstalled lithium batteries and battery packs can no longer be packed into checked luggage. Given the confusion around carrying liquid or gels on board an aircraft, I am not sure the general public clearly understands what the battery ruling means to them. Admittedly, I have not checked any luggage this year, so it’s possible that airport screeners are asking about PORTABLE DESIGN

batteries. I recommend new labels for all new lithium or Li-Ion battery packs warning owners about this new requirement. The public needs some help if we expect them to cooperate. Another interesting presentation was from Brian Barnett of TIAX: Battery Safety Overview and the Safety Implications of Various LiIon Cell Components (You may be surprised!). Brian described the differences between field failure and abuse. All batteries are subject to failure— given sufficient abuse. Actually, field failures are rare events (less than one part per million) resulting in smoke or flames. Failures are usually initiated by an internal short. No known methods exist today to determine if a properly designed and manufactured cell will end its life as a field failure. There-in lays the problem. So, predicting a field failure must be done near the time the event will occur. Since typically these events are internal short circuits, methods of dealing with the situation are limited. During the conference there were many discussions about early detection of cell self-heating, or spontaneous changes in cell voltage as methods to detect an internal short. Once detected, about the only thing that can be done is to try to reduce any power to the short. Ideas included creating an external short. Of course, the energy would have to be dissipated externally without causing harm. Whether or not an answer is straight is subjective. It depends on the knowledge or experience of the individual asking the question. I am not sure whether or not the public is asking if batteries containing lithium are safe. It is likely that there is an assumption that these products are safe. For the vast majority of the public’s experience, Li-Ion batteries are very safe. We need to help inform the public about battery safety without undue alarm. For my two cents, straight answers are often the simplest answer. As you might predict, it is unlikely that passengers wanting to listen to the power outlet also want a lot of technical direction. We need to provide simple means for passengers to be part of the battery safety solution. If this can be done, we could move on to getting teenagers to give straight answers—yeah, right!

by Dave Freeman, Texas Instruments

Tired of making trade-offs in your portable video design?

F RSRO G YEGAIN EE IN EN

ESS

SUCC

Maxim’s patented video filter technologies minimize power and space consumption without sacrificing performance

Video DirectDrive™ provides 2VP-P from a +1.8V supply

SmartSleep enables smallest video filter

Consumes 70% less power

Consumes 30% less space

LPF

OUT

DirectDrive CHARGE PUMP

LINEAR REGULATOR

SHDN

Full-scale output from +1.8V

Minimizes device power consumption

SmartSleep VIDEO SENSE

LOAD DETECT

AUTOMATIC SHUTDOWN

LPF

OUT

-UCSP T

Internally generated negative supply

2VP-P VIDEO

MAX9515

BUFFER

0.25VP-P TO IN 1VP-P VIDEO

Automatic shutdown if no video input or load is detected

MAX9516

m x 1m

UCSP is a trademark of Maxim Integrated Products, Inc.

Technical Support www.maxim-ic.com/support

DIRECT

™

www.maxim-ic.com/shop

www.avnet.com

www.maxim-ic.com/Portable-Video

For a free sample, visit www.maxim-ic.com/samples or call 1-800-998-8800 Distributed by Maxim Direct and Avnet Electronics Marketing. The Maxim logo is a registered trademark of Maxim Integrated Products, Inc. © 2008 Maxim Integrated Products, Inc. All rights reserved.

news BLOG: Apple Buys PA Semi— What Was That About?

Apple sent shock waves through the industry with its acquisition of PA Semi for a reported $278M. Does this mean that Intel—who had great hopes of seeing next-generation iPhones and iPods based on Atom—and Samsung, who makes the iPhone’s ARM-based applications processor, are hosed? Basically, yes.

PA Semi’s PA6T-1682M is a 64-bit low power dual-core version of the PowerPC architecture that Apple used in its computers before switching to Intel. Capable of delivering 8,800 Dhrystone Mips while running at 2 GHz, nd this chip is clearly much better suited to server blades and high-end embedded applications er exploration ether your goal than it is for portable designs. Apple coyly let speak directly on that it was acquiring the firm for its IP and ical page, the ght resource. design expertise and had no interest in productechnology, ing chips, at least for the merchant market. That es and products caused a chorus of cries from mil/aero contraced tors, who have been scooping up PA’s chips as fast as they could acquire them for a wide range of (presumably rack-mounted) applications. Some have reportedly been whinging to the Department of Defense (DoD), asking them to block the acquisition so that the processors companies providing solutions now will to be available. exploration into products, technologies and companies. Whether your goal is to research the continue latest datasheet from a company, mp to a company's technical page, the goal of Get Connected is to put you in touch with the rightlawyers resource. Whichever level of deal with any Apple’s can probably gy, Get Connected will help you connect with the companies and products you areDoD searching for. objections by licensing generic versions of onnected PA’s design to silicon manufacturers. What Apple apparently has in mind here is getting tighter control of the hardware, not to mention importing the profits that Samsung and ARM have been enjoying to date. Steve Jobs has long stated that Apple’s winning edge is tight integration between hardware and software, and this acquisition will enable new levels of hardware/software co-design—not to mention a scaled-down version of PA’s current chip that will fit perfectly into Apple’s next-generation of products. I’m Get Connected with companies mentioned in this article. sure TMSC will welcome them with open arms. www.portabledesign.com/getconnected What does this mean for the rest of the sup-

End of Article

PORTABLE DESIGN Get Connected with companies mentioned in this article.

ply chain? Mostly angst. This hits Intel at a bad time; it needs some big design wins for Atom, and the iPhone/iPod socket would have really launched the architecture. Intel is counting on Mobile Internet Devices (MIDs) as the Atom launch platform, but MIDs are barely beyond the concept stage, and it’s unclear that they’ll get any real traction. OTOH, sub-notebooks seem to be coming back, thanks to the Asus Eee PC 701 (currently based on a Celeron processor), and this is a market where Atom could take off. So don’t sell your Intel stock just yet. Same for ARM, who has the luxury of being ubiquitous in the embedded and portable space. Shed a tear for Samsung, however. PortalPlayer can feel your pain. by John Donovan – Editor-in-Chief

STMicroelectronics and NXP Merge Wireless Businesses

NXP and STMicroelectronics today announced their agreement to combine key wireless operations to form a joint-venture company with strong relationships with all major handset manufacturers.

According to NXP and ST, the new company will have the scale to better meet customer needs in 2G, 2.5G, 3G, multimedia, connectivity and all future wireless technologies. The combined venture will be created from successful businesses that together generated $3B in revenue in 2007 and will own thousands of important communication and multimedia patents. The new company will be a solid top-three industry player and among the few companies with the scale and expertise to pursue the R&D investments necessary to establish itself as a leading player in the wireless and mobile-multimedia market. The new organization will combine key design, sales and marketing, and back-end manufacturing assets from both companies into a streamlined worldwide joint venture that will rely on its parent companies and foundries for wafer fabrication services. This new leading

news

player will be well positioned with all of the vital technologies for UMTS (Universal Mobile Telecommunication System); for the emerging 3G Chinese standard; as well as other cellular, multimedia and connectivity capabilities, including Wi-Fi, Bluetooth, GPS, FM Radio, USB and UWB (Ultra-wideband), to effectively serve its global customers with complete wireless and mobile solutions across the spectrum of applications. The JV will also integrate the Silicon Laboratories’ wireless and GloNav’s GPS operations recently acquired by NXP. Both parent companies contribute strong businesses generating comparable revenue—each with 2007 operating profit of approximately $100 million. In order to create a clear ownership structure, STMicroelectronics will take an 80% stake in the joint venture. NXP will receive $1.55 billion from ST, including a control premium, to be funded from outstanding cash (cash and cash equivalents balance for ST at year end 2007 were $3.5 billion). The new organization is nd designed to be in a very healthy financial position, without debt, and able to grow its business er exploration ether your goal with all of the leading cellular handset manufacspeak directly turers. The parents have also agreed on a future ical page, the ght resource. exit mechanism for NXP’s ongoing 20% stake, technology, which involves put and call options, exercisable es and products beginning three years from the formation of the ed JV, at a strike price based on actual future financial results, with a 15% spread. The new company will be incorporated in the Netherlands and headquartered in Switzerland with approximately 9,000 employees worldwide. These people, almost equally contributed companies providing solutions now bythe STlatest anddatasheet NXP, will in position to serve the exploration into products, technologies and companies. Whether your goal is to research from abe company, mp to a company's technical page, the goal of Get Connected is to put you in touchJV’s with large the rightand resource. Whicheverglobal level ofcustomer base. demanding gy, Get Connected will help you connect with the companies and products you areNot searching for. owning any wafer fabs, the joint venture is onnected designed with low capital intensity, while having access to secure leading-edge manufacturing capacity from both parent companies and foundries; and will operate its own assembly and test facilities in Calamba, Philippines and Muar, Malaysia. NXP’s Calamba site as a whole will be transferred to the JV. In addition, part of ST’s back-end operations in Muar will be separated from the parent company’s existing facility in the area and transferred to the JV. The new company will also benefit from a dedicated Get Connected with companies mentioned in this article. worldwide sales and customer support team. www.portabledesign.com/getconnected The Joint Venture will be governed by a

End of Article

10 PORTABLE DESIGN Get Connected with companies mentioned in this article.

board of Directors on which both Carlo Bozotti and Frans van Houten will participate, looking after the best interest of its customers and the success of the JV. Aiming for a closing in Q3 of this year, the deal is subject to regulatory approvals and labor council consultations. The parent companies expect over $250 million in annual cost synergies from the JV by 2011. In financial impact, ST expects the transaction to be accretive to its non-GAAP cash EPS in 2009. Commenting on the impact to NXP, Frans van Houten said, “This deal transforms the portfolio of NXP and strengthens our cash position. We will continue to pursue building leadership positions through innovation and investment in our remaining focus areas: Multimarket Semiconductors, Automotive, Identification and Home electronics.” “This transaction strengthens our wireless business and enhances our leadership position in an important market segment we have targeted for expansion and external growth,” added Carlo Bozotti. “Coupled with our recent deconsolidation of Flash memory, it further proves our execution in reshaping ST’s product portfolio towards value and leadership. This, together with our recently announced decisions on distribution to shareholders, demonstrates our commitment to improving shareholder value.” According to iSuppli, the global handset market was 1.15 billion units in 2007 and is forecasted to grow at about an 8% compound annual growth rate through 2011. The handset semiconductor market represented 14% of the global semiconductor TAM in 2007, making up the second largest segment of the industry. STMicroelectronics, Geneva, Switzerland. +41 22 929 29 29. [www.st.com]. NXP Semiconductors, San Jose, CA. (408) 474-8142. [www.nxp.com].

Online PCB Prototyping Design Environment

Sunstone Circuits has launched the Sunstone ECOsystem Design Environment, a one-stop resource for the necessary knowledge, capabilities and expertise needed to produce a PCB. The Sunstone ECOsystem relieves pain points and breaks down traditional barriers in the PCB prototype design, manufacture and assembly process by providing a collaborative, open en-

vironment in which any PCB designer, vendor or supplier can easily participate.

Four discrete segments of PCB prototyping comprise the Sunstone ECOsystem: Knowledge—PCB designer expertise; Tools—Design packages and software; Libraries and Parts— Sourcing and tracking of PCB components; and Manufacturing—PCB manufacturing capability and expertise. These four segments of the Sunstone ECOsystem enable prototype and quick-turn PCB designers to find the knowledge, vendors, tools and libraries necessary to undertake design projects and scale up to production with Sunstone’s manufacturing capabilities and partner fabrication and assembly services. Through the Sunstone ECOsystem, designers can make use of low- or no-cost design tools, thereby freeing up R&D capital equipment budget dollars for other purposes. Quality working prototypes in fewer turns and shorter time are made possible through this open environment. Sunstone offers PCB123, a complete, complimentary circuit board design solution for the electronics design industry; DFM Rule Decks, which work with CadSoft’s EAGLE PCB Layout tool and Altium Designer; and agreements with ECOsystem strategic partners to streamline the entire PCB prototype production process from board design to final assembly. Sunstone Circuits, Mulono, OR. (503) 829-9108. [www.sunstone.com].

Renesas, IMEC Collaborate on Reconfigurable RF Transceivers

Renesas has entered into a strategic research collaboration with IMEC, Europe’s leading independent research center in the field of nanoelectronics, to perform research on 45nm RF transceivers targeting Gbit/s cognitive radios. To this end, Renesas has joined IMEC’s software-defined radio (SDR) front-end program. This research program includes reconfigurable RF solutions, high-speed/low-power analog-todigital converters (ADCs) and new approaches

to digitize future RF architectures. Researchers from Renesas will reside at IMEC to closely collaborate with IMEC’s research team. In this way, they will build a fundamental understanding and develop robust solutions for Renesas future mobile electronics products. On the near term, IMEC’s SDR-front-end program targets the development of a new generation cost-, performance- and power-competitive reconfigurable radio in 45nm digital CMOS technology. This radio will have a programmable center frequency from 100 MHz to 6 GHz and programmable bandwidth from 100 kHz to 40 MHz covering all key communication standards, with a merit comparable to state-of-the-art single mode transceivers.

The research program builds on IMEC’s previous 130nm RF transceiver results, the world’s first prototype of a true SDR transceiver IC (SCALDIO). Also, further evolutions of IMEC’s record breaking ADCs (merit record by IMEC at ISSCC 2008 of 40 Msamples/s, 9 bit, 54fJ/conversion step) will be developed within this collaboration. “The ability to develop an innovative RF architecture with scaled-down CMOS technology and circuit technologies in transceiver products supporting next-generation cellular standards such as 3GPP-LTE and 4Gs is one of the key differentiators for our products that are superior in cost advantages, performance and power,” Masao Nakaya, board director and executive general manager of LSI product technology unit at Renesas Technology Corp. said in a statement. “We are pleased to be a part of IMEC’s SDR-front-end program, collaborating on the research to explore new technologies for multi-standard RF transceivers. We aim to contribute to the early realization of next-generation mobile phones by combining our advanced

semiconductor solutions with IMEC’s R&D expertise in RF technology.” Renesas Technology America, Inc. San Jose, CA. (408) 382-7500. [www.renesas.com]. IMEC, Leuven, Belgium. +32 16 28 12 11. [www.imec.be].

DRAM Makers Developing NextGeneration Memory Interface

ARM, Hynix Semiconductor, Inc., Silicon Image, Inc., Sony Ericsson Mobile Communications AB and ST Microelectronics have announced the formation of a working group committed to creating an open standard for nextgeneration memory interface technology. This first-of-its-kind memory standard for dynamic random access memory (DRAM), named Serial Port Memory Technology (SPMT), enables bandwidth flexibility, significantly reduced pin count, lower power demand and multiple ports by using a serial interface instead of the parallel interface commonly used in today’s memory devices. This technology is ideal for mobile handset manufacturers who are incorporating more media-rich features requiring higher performance at lower system cost. “As DRAM content continues to grow for the new generation of mobile devices, it is becoming increasingly difficult for current technologies to keep up with the demand for greater bandwidth, more power reduction, and design flexibility without significantly increasing cost,” Jim Venable head of Silicon Image’s Advanced Memory Technology Products told Portable Design. “Serial Port Memory Technology will be a gamechanger in the way mobile device developers design new products that deliver significant performance improvements and a higher quality user experience. The SPMT Working Group’s goal is to define a technology that reduces pin count by a minimum of forty percent, provides a bandwidth range from 3.2 Gbytes/s to 12.6 Gbytes/s and higher, reduces input/output power by fifty percent or more over other currently available DRAM offerings, and provides the ability to use either a single port or multiple ports in a single SPMT-enabled memory chip. Venable sees SPDRAM as the natural successor to LPDDR2, extending the DRAM technology roadmap out to 2016 and beyond. While initially targeted at the mobile handset mar-

MAY 2008

news lates,” claims Nam Hyung Kim, memory analyst at iSuppli. “It makes sense to develop an interface standard for DRAM integrating serial technology that offers a way to achieve higher bandwidth, pin count reduction and scalability not achievable with current interface technologies.”

ket, the technology will also be in demand by other markets such as portable media players, digital still cameras and handheld gaming devices. The Working Group came together to develop a new technology to meet the growing demand by manufacturers to increase the performance and functionality of handsets while maintaining competitive or even reduced system cost for the devices. This is in response to mobile service providers’ demand for solutions enabling them to give consumers more data-intensive, media-rich capabilities such as video (including high-definition video), GPS, gaming, Internet access, e-mail, multimedia applications and music at a competitive price. nd Data paths in portable designs are going more serial in order to reconcile higher data rates, the er exploration ether your goal demand for greater functionality and the need to speak directly lower, not raise, chip pin count. The latter will ical page, the ght resource. enable DRAM manufacturers, who watch their technology, margins erode daily, to reduce the overall chip es and products cost by going to smaller packages. ed The SPMT Working Group has been meeting since the third quarter of 2007 and is expected to organize a formal consortium later this year consisting of handset, memory and system-on-chip manufacturers and semiconductor IP providers with the intention of bringing the SPMT specificompanies providing solutions now cation to the industry the end of 2008. exploration into products, technologies and companies. Whether your goal is to research the latest datasheet fromby a company, mp to a company's technical page, the goal of Get Connected is to put you in touch with the righttold resource. Whichever level of Venable Portable Design that the group is gy, Get Connected will help you connect with the companies and products you aredivided searchingover for. whether to submit their specification onnected to the IEEE or other standards body for formal verification. If not, through an “open standard” each vendor would be free to implement it in their own fashion. This could lead to the numerous “forks” that are common to Linux, so don’t expect pin compatibility between SPMT DRAMs from Hynix and ST devices any time soon. Still, a recognized serial DRAM standard would make life easier for portable designers and device makers alike. “The need for faster, denser DRAM chips for Get Connected with companies mentioned in this article. handsets will continue to grow, particularly as the www.portabledesign.com/getconnected requirement for media-rich functionality esca-

End of Article

12 PORTABLE DESIGN Get Connected with companies mentioned in this article.

Silicon Image, Sunnyvale, CA. (408) 616-4000. [www.siliconimage.com].

Mentor Graphics Aligns Product Groups to Address IC Implementation Challenges at 45nm and Beyond

Mentor Graphics Corporation has announced it has aligned its integrated circuit (IC) implementation product lines under the Design-toSilicon division to better address the design and manufacturing challenges of 45nm and smaller process nodes. The division, which will be headed by vice president and general manager, Joseph Sawicki, will now include Mentor’s industry-leading IC products: the Olympus-SoC place-and-route system, the Calibre physical verification and DFM platform, and the designfor-test (DFT) product line. “This change positions Mentor to more effectively help our customers tackle the com-

plex design and manufacturing challenges they are experiencing as ICs move to smaller processes,” said Gregory K. Hinckley, president of Mentor Graphics. “Joe has been instrumental in building the Calibre franchise into a full back-end flow. His leadership and experience will serve the newly aligned organization well as Mentor continues its IC implementation business strategy.” Sawicki joined Mentor Graphics in 1990 and has held previous positions in applications engineering, sales, marketing and management. He holds a BSEE from the University of Rochester and an MBA from Northeastern University’s High Technology Program. Mentor Graphics Corporation, Wilsonville, OR. (503) 685-7000. [www.mentor.com].

analysts’ pages Mobile Device Market Exceeds Expectations to Deliver 289 Million in 1Q 2008

“Operators and distributors continued to top up their inventories in 1Q 2008 after a particularly strong 4Q 2007,” says ABI Research vice president Jake Saunders. “Year on year, 1Q 2008 was up 13.7%, but 2Q 2008 is likely to be softer than in previous years.” Shipment volumes in the developed markets have softened slightly due to the credit crisis, but emerging markets such as Asia-Pacific, South America and the Middle East/Africa are delivering growth rate percentages in the mid-20s. Mobile devices have proved to be a “lifestyle necessity,” rather than a mere luxury accessory.

er exploration ether your goal speak directly ical page, the ght resource. technology, es and products

Despite all the global concern about price inflation in food, rent, clothing, oil and utility bills, device manufacturers are not benefiting. The ASP (Average Selling Price) has shown comprehensive price erosion for all manufaced turers. In the market-share stakes, Nokia increased its share to 39.9% while Samsung and LG were also net winners with 16% and 8.4%, respectively. Motorola continued to lose market share (falling 2.6%) to 9.5%, while it was more of a companies providing solutions now surprise Sony exploration into products, technologies and companies. Whether your goal is to research the latestthat datasheet fromEricsson a company,lost market share mp to a company's technical page, the goal of Get Connected is to put you in touch(down with theto right resource. level case of 7.7%). It Whichever may be the that Sony’s gy, Get Connected will help you connect with the companies and products you areWalkman searching for.and CyberShot lineups are in need of onnected a substantial refresh. Additionally, WCDMA continues to build momentum with shipment volumes that are up 44% YoY—but no one should count out GSM, as GSM-enabled handsets grew 17% in the quarter. The market is still very much dominated by the “Big Five” manufacturers, but an innovative tier of manufacturers (RIM, HTC, Apple) has been stirring up interest in smartphones and will soon be joined by a new class Get Connected with companies mentioned in this article. of mobile device: MIDs (Mobile Internet Dewww.portabledesign.com/getconnected vices). MIDs made their debut last year, but

End of Article

14 PORTABLE DESIGN Get Connected with companies mentioned in this article.

vendors such as Lenovo, Aigo and Asus are expected to drive growth rapidly. “Mobile device shipment volumes show no sign of abating in growth, despite the uncertain economy,” says research director Kevin Burden. “ABI Research expects 2008 to top out at 1.28 billion devices shipped—a 12% increase YoY—but these volumes could be subject to the overall global economic climate.” ABI Research, Oyster Bay, NY. (516) 624-2500. [www.abiresearch.com].

Steady Growth Expected for Portable Consumer Electronics

As consumers continue to demand content personalization and portability in their electronics products, In-Stat predicts steady shipment growth of portable consumer devices through 2011. The markets for digital radio receivers, edutainment toys, portable media players and portable navigation devices will all thrive over the next five years, the high-tech market research firm says.

“The most noticeable trends in these categories are the continued expansion of features, such as portable media players that can access the Internet, as well as the potential threat mobile phones pose to stand-alone personal navigation devices and personal media players,” according to In-Stat analyst Stephanie Ethier. Recent research by In-Stat found the following: • Worldwide Digital Radio Receiver shipments will reach 32.2 million units by 2011. • Worldwide Edutainment Toy shipments will reach 114.4 million units by 2011. • Worldwide Portable Media Player shipments will reach 275.0 million units by 2011. In-Stat, Scottsdale, AZ. (480) 483-4440. [www.in-stat.com].

iSuppli Upgrades DRAM Market Rating to “Neutral”

The DRAM market finally is showing some signs of improvement, prompting iSuppli Corp. to upgrade its rating of near-term conditions for suppliers. iSuppli is raising its DRAM rating to “neutral,” up from its “negative” assessment issued in November 2007. iSuppli’s neutral rating comes with a positive bias regarding future conditions for the market, given indications that the worst is over for DRAM suppliers.

appear, iSuppli is likely to further upgrade its rating of market conditions to “positive.” “The average per-megabyte price for commodity DRAM has dropped by more than 80 percent during the last 12 months,” Kim observed. “Thus, in order to return to pricing levels from a year ago, suppliers would need to increase prices by 500 percent, boosting them to $6, up from $1—an impossible occurrence. The troubled NAND market received some slightly positive news with word of a price rally in the spot memory market. However, the fundamental underlying market conditions have not changed significantly. iSuppli predicts the current spot market price rally will be short lived. iSuppli Corporation, El Segundo, CA. (310) 524-4000. [www.isuppli.com].

DRAM market conditions since the beginning of 2007 have been characterized by oversupply, bloated inventories and weak pricing. The average megabyte DRAM selling price is estimated to have dropped by 17 percent in the first quarter, following a 31 percent decline in the fourth quarter of 2007. While still an aboveaverage decline in the first quarter, the slower rate of decrease indicates the market now is bottoming out. Amid reductions in channel inventories, and lower levels of expected capital spending among memory manufacturers, the supply/demand balance is coming into better alignment, a development that should bolster pricing. Permegabyte DRAM prices will rise by 2 percent during the second quarter, iSuppli projects. “Although the DRAM suppliers themselves are still carrying more inventory than normal, stockpiles in the channel have been reduced significantly,” said Nam Hyung Kim, chief analyst at iSuppli. “Furthermore, OEMs including the PC makers now are at optimal DRAM inventory levels, meaning their orders will increase during the critical third-quarter holiday build season.” This is paving the way for an expected market recovery. When indications of a noteworthy decline in DRAM supplier inventories

HDD Shipments Cross the 500Million-Unit Mark in 2007

How good was the Hard Disk Drive (HDD) market in 2007? So good that: • Worldwide HDD unit shipments grew by a prodigious 18.9 percent for the year, according to iSuppli Corp. • Annual HDD shipments crossed the halfbillion-unit shipment threshold for the first time ever. • Revenue rose by 4.6 percent. Indeed, 2007 was a year to remember for HDDs, with shipments reaching 516 million units, up from 434 million in 2006. Factory revenue rose to $32.8 billion, up from $31.2 billion in 2006. Beyond strong demand growth from notebook PCs and consumer-electronics products, HDD makers benefitted from a second-half cease-fire in the price war that had dictated market conditions over the past three years. Although the price war continued to rage in the first half of the year, the second half brought a rebound and an appreciation of the value of HDDs. Pricing was supported by the arrival of high-capacity 1 Tbyte HDDs, a key milestone in the face of technology challenges from NAND flash memory.

Seagate Still No. 1 Seagate Technology LLC maintained its lead in the global HDD market in 2007, with its share of shipments rising to 34 percent, up from 33.1 percent in 2006. The company shipped 175 million HDDs in 2007, up 22 percent from 144 million in 2006. Second-ranked Western Digital Corp. in 2007 shipped 113 million HDDs and increased its share to 22 percent, up from 19.6 percent in 2006. No.-3 Hitachi Global Storage Technologies increased its share to 17.3 percent, up from 16.1 percent in 2006. Hitachi reversed its operating loss of $93 million in the fourth quarter of 2006 to achieve an operating profit of $95 million in the fourth quarter of 2007. Despite the competitive threat posed by NAND-type flash memory in products ranging from MP3 players to notebook computers, the industry’s strong performance in 2007 indicates the HDD isn’t ready to quit any time soon. Still, the threat of NAND is going to haunt the industry in the coming years. Only cost reductions and the continued advancement in HDD technology, driven by profit-funded research and development activities, will keep the industry afloat. iSuppli Corporation, El Segundo, CA. (310) 524-4000. [www.isuppli.com].

MAY 2008

analysts’ pages Sprint-Clearwire Joint Venture “Very Positive News” for WiMAX, Says ABI Research

Today Sprint and Clearwire announced a new joint venture to build an advanced mobile network. Based on WiMAX technology, the new network will bear the Clearwire name, but the $12 billion value of the new company includes $3.2 billion from partners Google, Intel and three cable companies, Comcast, Time Warner and Bright House. “This is very positive news for WiMAX globally,” comments ABI Research principal analyst Philip Solis. “WiMAX has been moving forward in developing countries. But in the industrialized nations, the industry was waiting for Sprint to get the ball rolling, lifting volumes and driving down ASPs. Sprint chose WiMAX in 2006: it had the vision, and it pulled together an ecosystem of chipset and equipment companies. But there was one missing piece: money. They needed financing and the confidence of the nd investor community. The formation of this joint venture completes the picture.” er exploration ether your goal “What’s particularly interesting is the particispeak directly pation of the cable companies,” Solis continues. ical page, the ght resource. “Competing against telcos that have moved technology, into their video markets, they have long wanted es and products wireless in their portfolios. Now they are parted owners of a mobile wireless network.” However, the cable companies still need a mobile voice solution, which will not be immediately resolved by deploying a WiMAX network, since VoIP is planned for future upgrades. Senior analyst Nadine Manjaro adds that, companies providing solutions now “This announcement exploration into products, technologies and companies. Whether your goal is to research the latest datasheet from provides a company, the kick-start that mp to a company's technical page, the goal of Get Connected is to put you in touchWiMAX with the right resource. Whichever levelwas of the company needed. Since Sprint gy, Get Connected will help you connect with the companies and products you aredriving searchingWiMAX for. so hard, a waiting world was onnected getting disappointed that it hadn’t really done anything with it yet. Verizon recently announced its plans to deploy LTE in 2010, and since WiMAX was said to enjoy a two year time-tomarket lead, everybody was waiting for Sprint to solidify that advantage. With this announcement, Sprint can claim that it is doing so.” “Clearwire will be taking a lead operational role and the new joint venture will be under its name,” she says. “Sprint has to a degree suffered from a perceived deficit in customer service. Get Connected with companies mentioned in this article. This way, Sprint has some ownership without www.portabledesign.com/getconnected that baggage.”

End of Article

16 PORTABLE DESIGN Get Connected with companies mentioned in this article.

Other winners from this announcement are the WiMAX chipset makers, and Samsung, Nokia and Motorola, the three vendors selected to provide infrastructure for the new WiMAX network. However, the challenge will be how each party in the joint venture deals with issues such as the cannibalization of some of their existing products. ABI Research also believes that there is a potential downside for Sprint: it does dilute the company’s control and ownership of the network that it considers to be its future. Telecom joint ventures (especially those with multiple parties) have a very poor track record, and this strategy is not without significant risks. ABI Research, Oyster Bay, NY. (516) 624-2500. [www.abiresearch.com].

Fragmentation of Mobile TV Broadcasting Here to Stay

There is a fragmentation of standards for digital mobile TV broadcasting—just as we have NTSC, PAL and SECAM for analog TV—that appears to be the way of the future, reports InStat. In digital mobile TV broadcasting, several standards are being used, several mainly in one country, and that situation is likely to continue in the future, the high-tech market research firm says. Recent research by In-Stat found the following: • The most popular mobile TV broadcast services are those that are offered without a subscription, as in Japan and South Korea. However, mobile TV broadcast services are viewed by many as a way to generate revenue, so many of the mobile TV broadcast services will be subscription-based. • DVB-H is the most widely used digital mobile TV broadcast standard, in terms of number of operators having launched services. • Analog mobile TV broadcast viewers will outpace digital mobile TV broadcast viewers and subscribers in 2009. • Worldwide mobile TV broadcast subscription revenue will reach US$12 billion in 2012. In-Stat, Scottsdale, AZ. (480) 483-4440. [www.in-stat.com].

cover feature hardware/software co-design

Hardware Software Co-Design Accelerates Development of Smart Wireless Devices Handset designers are increasingly using open source code platforms to develop next-generation wireless devices.

by Matti Kattilakoski, Chairman of the Board, Navicron Trends in handheld wireless devices can change quickly. Driven by fickle consumer tastes, technologies and applications can drift in or out of fashion as easily as a pair of sunglasses. To capitalize on market demand for “what’s hot,” embedded device designers must apply new technologies rapidly and time-to-market becomes the Holy Grail. Recent technology trends have been driven by the success of the Apple iPhone and competitive products in the smartphone category. Increasingly popular features include: touch screen displays with better quality and higher resolution; embedded megapixel cameras; exceedingly thin form factors and various Internet-based services such as location-based mapping. These complex features have a significant impact on both hardware and software choices for embedded designers of wireless devices and add a great deal of complexity to an already tedious and time-consuming process. During that process, selected features are translated into detailed technical requirements, and plans and key components must be selected, integrated and tested. The functionality of a handset is implemented with a set of interconnected system components such as RF, ASICs, memories and various peripherals. Designers must select a set of system components and partition the system’s

18 PORTABLE DESIGN

functionality among these components. Final implementation has to satisfy a set of constraints such as cost, performance and power consumption. In the design of cell phones, there are only a couple of major handset manufacturers who actually do system design from “scratch” that requires systems partitioning and handling of algorithms. Outside of those few powerhouses, designers are now using open source code platforms, new rapid development program languages and pre-tested software and hardware platforms. These standardized platforms with off-the-shelf processors from companies such as Texas Instruments and Freescale have changed the embedded design landscape by lowering development costs and providing a faster time-to-market.

Linux for Smart Wireless Devices

For mobile wireless device designers there are three principal operating system (OS) choices available: Microsoft Mobile, Symbian and Linux. For an open source code approach, savvy designers are choosing Linux for its rapid development capabilities and cost savings. With Linux’s open architecture, achieving connectivity becomes straightforward and intellectual propriety rights issues can be efficiently handled with the proper use of licenses.

cover feature RAIS UI FRAME GTK+ More

GSM

Bluez

Gstreamer

Wireless device designers making use of Linux can select a number of high-level programming languages and development environments for development projects. The Graphical Tool Kit (GTK) and Python programming languages are two popular options. GTK+ is a feature-rich toolkit for implementing graphical user interfaces and features an easy-to-use application programming interface (API). GTK+ was originally written in

As another option, Python is a powerful and fast object-oriented programming language and a great choice to figure 1 speed application development for wireless devices. Python offers strong support for integration with other languages and tools, and performance imNavicron provements can be Application Suite Native Apps wrapped in C or C++, making it flexible. Implementation of Python is also under the open source license, which makes it freely usable and distributLinux Services able and supported by Linux Kernel a set of extensive standard libraries. Python HW is popular with developers due to the fact Linux software stack diagram that is can be learned in just a few days. In fact, many programmers report substantial productivity gains when using Python. Python runs in many software environments such as Windows, Linux, .NET, Java, and most interestingly, the Nokia Series 60. A major trade-off to consider when designing with Python is that applications may require more memory and processing power in order to run seamlessly.

X-windows

Linux Development Tools and Languages

C language, but has ties to many other popular programming languages such as C++ and Python. GTK+ is licensed under the general public license GNU LGPL 2.1, allowing development of both free and proprietary software without any license fees or royalties. GTK+ has a number of features that today’s wireless developers need including an easy look and feel for easy programming, an object-oriented approach, theme support, internationalization, localization and accessibility. GTK consists of four libraries: • GLIB, which is a low-level core library that is the basis of GTK • Pango, a library for layout and rendering of text • Cairo, which is a library for 2D graphics • ATK, a library for a set of interfaces providing accessibility

More VOIP MSG Media Browser More Network PIM MSG Call

Linux was originally written in 1991 by Linus Torvalds as an improvement over earlier public domain versions of Unix. Torvalds merged his efforts with a group of visionaries who organized the “free software movement” under what was known as a General Public License. His kernel eventually was adopted as the cornerstone of this movement, which became known as “Open Source.” The Open Source Initiative (OSI) has been gathering steam ever since it was first officially organized in 1998. Bolstered by the fact that many governments have adopted policies favoring and often requiring “open source” solutions for their computer solutions, the notion of Open Source was taken seriously mostly because of Linux. The use of Linux as an OS in the server market has proliferated because of its modular extensibility, open source code community, vast libraries of existing applications, robustness and, not surprisingly, favorable licensing. The obvious benefits of these same features suggest that Linux will also become the dominant OS for wireless devices. A typical Linux software stack appears in Figure 1. The recent introduction of the Linux-based Android software stack by Google is a milestone event in the wireless device development community. Android provides a designer with everything they need to build and run Android applications including a device emulator and advanced debugging tools. Android has been criticized by some designers who feel it is not really free or open-source software despite Google’s claims to the contrary. Some parts of the software development kit are still proprietary and closed source, and some believe it is a conscious decision to control the platform by Google. Despite the criticism and uncertainty of its future, the Android SDK is an option a designer must also consider.

MAY 2008

figure 2

nd Navicronâ&#x20AC;&#x2122;s fusionplatform block diagram

er exploration ether your goal speak directly ical page, the ght resource. technology, es and products

Hardware Selection Is the Key for Success in the Long Run

Software development is critical, but the selection of the correct hardware for wireless device design is paramount. Wireless devices are some of the most complex equipment that people use on a daily basis. Todayâ&#x20AC;&#x2122;s mobile companies providing solutions now phones tensa company, of millions of calculaexploration into products, technologies and companies. Whether your goal is to research the latestprocess datasheet from mp to a company's technical page, the goal of Get Connected is to put you in touchtions with the right resource. Whichever level of and receive every second to transmit gy, Get Connected will help you connect with the companies and products you aresignals. searching for. The basic construction of a mobile onnected phone includes hardware components such as: an antenna; battery; keys or touch pad integrated to a LCD; microprocessor and memory; wireless receivers and transmitters; and electromechanical components such as speakers, microphones and buzzers. The hardware design for a wireless device may look like a simple task but it can become a complicated exercise due to the increasing amount of wireless connections required. Today, it is not uncommon to equip smartphones Get Connected with companies mentioned in this article. with 3G, WLAN, GPS and Bluetooth capabilwww.portabledesign.com/getconnected ity. This means that several receivers and trans-

End of Article

20 PORTABLE DESIGN Get Connected with companies mentioned in this article.

mitters have to be integrated into a compact form factor. The 3G specification alone covers five frequency bands. Most mobile phone antennas today are internal and highly dependent on the handset size, thickness and the materials used for maximum effectiveness. Today an ARM11-based application processor runs at over a 600 MHz clock frequency, and with several megabits of available 16/32bit SDRAM, DDR, NAND Flash and PSRAM, it has the performance comparable to PCs on the market just a few years ago. Many application processors have vector floating point coprocessors, hardware-accelerated multimedia coders with graphics acceleration, advanced power management, a hardware-accelerated image processing unit for H.264 coding supporting VGA 30 fps and security features enabling a hardware- and software-based fuse box solution. Image processing and graphics processing units can also be used to lower power consumption and perform picture resizing, blending and rotation. Most application processors also support high-speed USB, MMC or memory stick external cards, ATA-6 hard disk

Co-Designing Hardware and Software

Traditionally, the approach to wireless device design has been to separate it into hardware and software projects that work independently of each other, while trying to deliver a common result. In order to ensure that critical system design requirements are being met, companies today are relying on silicon solutions and ready-made tested platforms to support unified product development. In this approach, the importance of having the right development tools and rapid prototyping tools, including both a software and hardware platform that works in tandem, is critical to simplify potential issues early in the development stage. These tools enable hardware and software designers to work closely together. In the design process today, the questions have become “What platform do I use as a baseline for my development?” and “What applications do I need to have?” What software development tools are selected has a direct impact on how much processing power and memory is needed. Hardware and software plans should be developed according to the features needed, the required form factor, the type of applications required and, of course, time-to-market. With high volume designs, component costs need to be minimized, but development cycles are longer and designs cannot necessarily be easily adapted to changing market requirements. A pre-tested hardware development platform with a number of built-in functionalities gives designers an early prototyping opportunity to identify issues and avoid problems in later stages. Independent wireless device development companies such as Navicron are building such platforms from the ground up to optimize wireless device design (Figure 2). Selecting the right technology platform to speed up wireless device development is not an easy task. First and foremost, designers should be able to integrate software and hardware features quickly. When considering hardware requirements, designers need to look at platform features such as overall maturity, component selections and component costs. A platform must have the ability to serve as the basis of mass manufacturing with the ideal key component selection and optimal bill of materials cost. From a software perspective, available operating systems as well as middleware and applications are also primary considerations in the platform selection process. The return on

cover feature

interface, various ways to MUX audio signals and several SPI/SSI/UART ports.

the investment in the right platform is often realized in the second and third development projects when the hardware is already built and ready for software application development. A technology platform needs to be generic enough to be application-independent, and at the same time, specific enough to be relevant for the application supported. Software application development can slow down tremendously without having the right environment available in the early stages of the design process. Having a platform that provides support for early software development becomes critical. Software applications can be acquired in different ways. First, the necessary applications can be developed in house for the specific project and purpose. Second, open source communities have a wide selection of applications available. Before using open source code applications, designers must verify that the maturity of the code is what is expected and the license agreement is acceptable for the purposes of the project. And finally, applications can be acquired from various commercial sources.

Summary

To enable wireless device designers to respond quickly to changing trends in the industry, new development tools and methods are emerging. Use of modern high-level programming languages will shorten the development time but at the same time, increase processing power and memory size requirements. Selection of the right development tools has to be made based on the specific design requirements and business objectives of the project. Designers must understand how they can legally use open source software in commercial products and the implications their choices have on hardware. Designers can use high-level graphical toolkit software frameworks such as GTK+ to make the application development process faster and easier. The trade-off of using high-level languages is that they will consume a great deal more memory and processing power. Designers can speed development time, but will have the dilemma of adding development costs for processing power and memory. Although a hardware-centric approach to design of wireless devices is advisable, implications on software must be considered early on in the development cycle. The use of a pre-tested hardware platform optimized for open source software can be a powerful tool for achieving faster timeto-market and a cost-effective consumer device that will satisfy market demands with the current “must have” technology features.

MAY 2008

cover feature hardware/software co-design

Model-Based Design Accelerates Functional Verification Using model-based design lets the development team move through the process in a systematic way. It also minimizes verification headaches later. by David Lidrbauch, Product Marketing Manager, The MathWorks

Functional verification consumes from 50% to 65% of the development time and budget for today’s system-on-chip (SoC) projects. With fragmented tools and development methodologies that rely on manual scripting, collections of disparate verification tools, inter-tool incompatibilities with almost-standard transaction languages, and mismatched database sources, functional verification can itself become a project that deflects attention from the main development goals. Instead of optimizing and debugging product designs, engineers spend time debugging their workflow and compensating for gaps in their verification tool chain. Chip-level functional verification must be examined in a larger perspective. Gary Smith identified this issue succinctly prior to the 2006 DAC: “It’s the software, stupid.” He and Daya Nadamuni pointed out the critical, integral role of software in today’s SoC projects. “You can have elegant algorithms, first-pass

22 PORTABLE DESIGN

silicon and fancy intellectual property (IP),” said Nadamuni. “But without software, the product goes nowhere.”[1] Most handheld devices today illustrate this larger perspective, as do the development challenges in Software Defined Radio (SDR) and its siblings: the Joint Tactical Radio System (JTRS) and Cognitive Radio. Given the onsite reconfiguration intelligence demanded by these systems [2], one could call them the perfect storm of hardware/software co-design challenges. The question you have to ask is, “Is each task an expense or is it an investment?”

Model-Based Design

In Model-Based Design, the system model is an executable specification that becomes the gold standard for design and verification activities. During the design phase, this model is shared among the various design teams, in-

cover feature

cluding systems, software and hardware engineers. Each team uses the model to elaborate its components, all the while ensuring that the design intent of the original system is still maintained. Figure 1 shows a typical modelbased design workflow.

figure 1 Requirements • Excel, Word, HTML, etc. • Specialized req. mgmt tools • (e.g. DOORS)

Simulation • Application-specific block libraries • Digital and A/M-S EDA simulators • MCAD tools • Other domain-modeling tools

Requirements, System Specs

Data Analysis

Design Validation Lifecycle Mgmt • Config Mgmt • PLM Design Verification

Operating Environment

Algorithm Development

Analog/RF Control Components Logic Timed Algorithms ne Embedded Software

Digital Electronics

C, C++

VHDL, Verilog

MCU

Code Verification

r ne Ge

ra te

Data Modeling

DSP

FPGA

ASIC

} Integration

Test Environments

Implement

Implementation • Software IDEs: Green Hills, Eclipse • HDL Hardware Synthesis • Xilinx, Altera, TI, ADI, ARM, etc

Continuous V&V

Verification and Validation • Hardware-in-the-loop Systems • T&M Instrumentation • Test Case Generation

Model-Based Design workflow

Model-Based Design proves essential to merging the efforts of verification and design. The model of the system and its operating environment, minus the component being designed, becomes the test bench and analysis environment for each component. This approach ed compensates for the difficulty in verifying market-leading systems that lack standardsbased, well-defined interfaces between internal components and subsystems. It also eliminates nonproductive effort often expended to create tests for individual blocks and interfaces. companies providing solutions now Anlatest engineer HDL or embedded exploration into products, technologies and companies. Whether your goal is to research the datasheetverifying from a company, mp to a company's technical page, the goal of Get Connected is to put you in touchsoftware with the right resource.asks: Whichever typically “Is level my ofimplementation gy, Get Connected will help you connect with the companies and products you aregood searching for. enough?” The critical follow-up question onnected should be: “What is good enough?” The answer to this latter question is already contained in the algorithm specification created by the algorithm or systems engineer. Using co-simulation, that executable specification is employed as a test harness by the verification team to ensure correct system functionality. The co-simulation allows algorithm engineers as well as the hardware and software engineers to work in their own specialized environments, yet makes the impact of their deGet Connected with companies mentioned in this article. sign and implementation decisions clear to the www.portabledesign.com/getconnected entire team. The original effort associated with

er exploration ether your goal speak directly ical page, the ght resource. technology, es and products

End of Article

24 PORTABLE DESIGN Get Connected with companies mentioned in this article.

designing the executable algorithm specification is transformed from an expense into an investment with on-going payback.

Adopting Model-Based Design Incrementally: Component-Block Verification

Component-block -level development can be done independently and in parallel for high initial productivity. Algorithms can be individually explored and optimized but component interdependencies will be left untested. Divide-and-conquer approaches worked well for functional verification in the past because the system building blocks were naturally separate and independent. Conversely, because of functional convergence within today’s products, the analog, digital, interface and processor blocks in SoCs are extremely interdependent. Fully verifying any block requires full exploration of its independent functions and its interactions with other parts of the design. Verifying the interactions across domain boundaries is where the divide-and-conquer approach critically fails. Figure 2 illustrates some of the difficulties involved.

Verifying the Design in the System Context

No amount of component-block-level verification can explore the interfaces and dependencies between the various domains and uncover issues that could show up once the product is in actual use. Merging the algorithm blocks with the rest of the system is essential to prove out the design. Broadcom used Model-Based Design to accelerate the development of the SPINNER family of WCDMA semiconductor products for 3G mobile devices, easing the challenge of translating the sheer complexity of the WCDMA system into a functioning piece of hardware. Broadcom used MATLAB and Simulink to model various subsystems, including the base station, a wireless channel model and the handset. Each subsystem served as a unit test harness for the others. They were able to quickly partition the system model into architectural subsystems to enable the hardware designers to implement each piece more easily. Evaluating the optimal implementation for portions of a system is often best done after the algorithms are defined and validated. The system architecture may need to be explored to identify the most appropriate target implementation. For example, at Rice University in Hous-

Mixing Levels of Abstraction for Accuracy and Completeness

Development rarely proceeds linearly from high-level abstractions to ever more detailed representations of the entire design. ModelBased Design facilitates mixing levels of detail at will, which streamlines the ability to improve the real-world detail of the functional simulation, because the appropriate abstraction level can be chosen as needed to maintain very fast system-level simulation speeds while exercising the block under test with enough detail to verify functionality and prove out the implementation choices in the context of the entire system. In their paper entitled “DVB-T System Analysis using System Level Co-simulation,” [4] the authors describe how the co-simulation interface between MATLAB, Simulink and Cadence Virtuoso AMS Designer offers the system designer a wider choice of mixing abstraction levels by integrating the system design with the Virtuoso AMS Designer mixed-signal simulation, which also provides fast turnaround cycles and design insight. They also describe the difficulty inherent in non-Model-Based Design modern design flows: “Each tool is optimized for a specific level of abstraction and application area. Even though there is a certain range of overlap between the tools, difficulties arise when effects need to be analyzed that span different levels.”

Model-Based Design ROI Multiplies When Applied Across Teams and Projects

Using Model-Based Design lets the development team move through the process in a systematic way. ROI with Model-Based Design increases significantly when the specialized skills of each team member are leveraged by other

cover feature

ton, Brogioli, Radosavljevic and Cavallaro explored the partitioning trade-offs in a mixed DSP/FPGA implementation for 3.5G HSPDA mobile handsets and found performance gains of 90% with the appropriate partitioning. [3] This process of partitioning and implementing from the system model can be dramatically accelerated. Automatic code generation can be used to generate the embedded code for the processor and the HDL and test bench for the FPGA. Co-simulation links close the verification loop to compare each implementation against the golden reference system model. That original system model with the developed analysis code, system metrics and other tests can be reused as the verification platform.

teams during component and system verification. This leveraging occurs both downstream in the flow toward implementation and final verification and upstream to enhance the earlier design exploration; it ultimately works to improve understanding figure 2 at all stages of development. The collaborative use and reuse of a common system Discrete Time Variable Time (analog, RF) model prevents gaps in the functional verification process, improves Hardware the quality of the final product, reduces duplicate and wasted efEmbedded Software Event-Based fort, and improves the entire development process. Model-Based Finite State Machines Design can be adopted Mechanical Systems (MEMS) in an incremental way and the maximum benefits are realized as all Systems are multidiscipline; your verification environment should be too. of its elements are adopted across the entire project. In this way Model-Based Design effortDiscrete Time lessly emerges as an investment that continues to pay benefits in all aspects of complex system development. References: [1] “Dataquest to EDA: ‘It’s the software, stupid’,” Rich Goering, EE Times, July 24, 2006. [2] “Next-Generation Design Issues in Communications,” Bruce Fette, Mieczyslaw M. Kokar, Mark Cummings. Portable Design, March, 2008, p. 20. [3] “Hardware/Software Co-design Methodology and DSP/FPGA Partitioning: A Case Study for Meeting Real-Time Processing Deadlines in 3.5G Mobile Receivers.” Michael Brogioli, Predrag Radosavljevic and Joseph R. Cavallaro. Rice University. http://www.ece.rice. edu/~rpredrag/doc/MWCAS.pdf. [4] “DVB-T System Analysis using System Level Co-simulation.” CDNLive! Conference paper 2008. Munich, Germany, April 28-30. Hans-Werner Groh – Atmel Germany GmbH, Walter Hartong – Cadence Design Systems GmbH, Uwe Eichler – Fraunhofer Institute for Integrated Circuits. The MathWorks, Inc., Natick, MA. (508) 647-7000. [www.mathworks.com].

MAY 2008

wireless communications multi-mode transceivers

Developing Highly Integrated Solutions to Meet the Changing Demand of Multi-Mode Devices As the number of 3G/4G wireless permutations increases, a multi-mode 2-point polar modulation scheme may point the way forward. by Duncan Pilgrim, Director of Product Marketing, Sequoia Communications

With an increasing number of travelers, both nationally and internationally, the world continues to become a smaller place and the desire for a single universal communication device grows. As the demand grows, so do the requirementsâ&#x20AC;&#x201D;voice is still king, but the use of packetbased services, such as Web browsing, e-mail, file downloads and multimedia applications, is on the rise. These packet-based services require increased data rates that go beyond the capabilities of 2G technologies (GSM and CDMA) and require 3G and 4G standards. However, in order to maintain acceptable voice coverage, the devices must be backward compatible with established 2G networks. These multi-mode devices introduce a new set of challenges to the industry that will continue to increase as consumers demand more functionality. A few years ago, it appeared that network carriers would continue to follow roadmaps based on their installed network base. For the

26 PORTABLE DESIGN

carriers that had GSM networks, the progression to EDGE, WCDMA and LTE was relatively stable. For CDMA networks, the progression to EVDO Rev. A through to Rev. C also seemed to be well-defined. Then, a number of decisions occurred that really changed the dynamics. The WiMAX consortium developed traction in the cellular space when Sprint decided to use WiMAX as its 4G network, and a WiBro network was rolled out in Korea. Having traditionally followed the CDMA path, the Korean carriers have one of the largest established CDMA user bases; however, some of these carriers have migrated to WCDMA for 3G. Verizon Wireless, a carrier that is focused on CDMA, announced that the company would jointly develop LTE with Vodafone, which suggests that they will not migrate beyond EVDO Rev B. The last major resolution was the initial roll-out of TD-SCDMA by China Mobile just in time for the Olympics.

The Newest

Wireless Solutions New Products from:

ZigBee

RFID

RCM4000 RabbitCore® Modules www.mouser.com/ rabbitsemi/a

Bluetooth™ GPS WLAN

RN-41 Bluetooth® Module www.mouser.com/ rovingnetworks/a

Cellular

��

ISM Short Range RF Modules www.mouser.com/ rfmonolithics/a

The ONLY New Catalog Every 90 Days Experience Mouser’s time-to-market advantage with no minimums and same-day shipping of the newest products from more than 335 leading suppliers.

Connect® ME & Connect® Wi-EM www.mouser.com/digi/a

The Newest Products For Your Newest Designs

www.mouser.com Over 965,000 Products Online

(800) 346-6873

wireless communications

These initiatives led to the introduction of two new cellular standards, TD-SCDMA and WiMAX (WiBro). With these two additional standards, the number of potential 3G/4G technologies that still require 2G backward compatibility for voice coverage has grown. The possible combinations include GSM/EDGE with WCDMA, TD-SCDMA, WiMAX and/or LTE. For the CDMA networks, the options are CDMA2000 with EVDO, LTE and potentially WiMAX. This complexity is further compounded by the different options associated within each standard—WCDMA includes HSPA with various data rates, and LTE and WiMAX support various channel bandwidths. The increasing number of modes and combinations challenge the capability of the industry to develop highly integrated solutions for each different combination. For products to meet the strict power, size and cost requirements that are expected by the consumer, alternative solutions are required that can address the changing demand of multi-mode devices.

The RF Problem

previous work. This can be compared to discrete products, where in each case all modes need to be tested and all of the integration problems must be resolved.

A Flexible RF Solution

For a single RF device to function for different modulation schemes, it needs to support narrow band modulation (GSM and EDGE) and wide band modulation schemes (WCDMA, LTE, WiMAX, etc). They each have their own challenges. Until recently, the use of a narrow band polar modulator and a wide band linear modulator has proven to be the only solution that provides acceptable performance, despite being a suboptimal solution for size and cost. If a common architecture cannot be realized, then the bill of materials (BOM), die size and solution cost will never approach that of GSM/ EDGE solutions. Polar modulation has been proven to provide the most efficient transmit path for narrow band modulation schemes such as GSM and EDGE. It is the ideal architecture for GSM/EDGE as it provides a common low-noise, high-efficiency architecture for both modes. With the improved noise performance of the architecture, it is possible to remove the transmit SAW filters, which are required in order to support multi-band power amplifiers. The basic architecture is shown in the block diagram in Figure 1.

It is clear that the individual development of integrated circuits (IC) for all of these different modes is not practical. Beyond the IC er exploration ether your goal development costs, evaluating and integrating speak directly an RF solution requires significant resources ical page, the ght resource. from a handset team in both the hardware and technology, software engineering groups. The initial tasks es and products are ensuring that the solution will theoreticalfigure 1 ed ly meet the performance requirements based on the data sheet and other documentation. A PA VCO physical evaluation of the hardware is then LPF PFD TX performed and can take many weeks of test- REF ing. The next step is to integrate the RF with the companies providing solutions now baseband platform, involves developexploration into products, technologies and companies. Whether your goal is to research the latest datasheet fromwhich a company, Counter ∆∑M FM mp to a company's technical page, the goal of Get Connected is to put you in touching withthe thehardware right resource. Whicheverand level the of software to platform AM gy, Get Connected will help you connect with the companies and products you areallow searching PLL allfor. of the devices to “talk” to each other. onnected This often requires weeks of performance optimization beyond the integration phase. The GSM/EDGE Polar Modulator entire process can take anywhere from three to nine months depending on the complexity and the teams’ familiarity with the standard. If this The concept of polar modulation is simple. process has to be repeated for many devices, The transmit signal is split into AM and FM vast engineering resources would be required components—the FM component is digitally and the time-to-market for many of these prod- modulated onto the PLL, and the AM path (for ucts would extend considerably. EDGE) is time aligned and recombined with Developing an RF platform that can support the FM at the PA driver or the PA stage. The multiple modes and combinations would mean linearity requirements of the FM path have Get Connected with companies mentioned in this article. that the efforts necessary to add an extra mode, been effectively negated, which means that the www.portabledesign.com/getconnected such as TD-SCDMA, would be incremental to architecture’s linearity is only really defined

End of Article

28 PORTABLE DESIGN Get Connected with companies mentioned in this article.

wireless communications

to support GSM, EDGE, WCDMA, HSUPA and TD-SCDMA. Further VGA VCO testing is underway to TX REF CP LPF DIV PFD show support for CDMA, LTE and WiMAX. A true single multiLPF D/A mode device reduces the overall footprint that the AM RF section consumes, Counter providing either imPLL proved form factors or FM ∆∑M allowing additional features to be added into the 2-Point Polar Modulation handset. For the development teams, it means by the VGA or PA stage of the design. Hence, that more time can be the support of higher peak to average ratio sig- spent evaluating a single device rather than nals, such as those used in HSUPA, LTE and multiple devices, and that they only need to WiMAX, presents less of a challenge compared develop a single RF platform in order to supto a standard linear modulator. port all of the different requirements. There are, however, significant challenges in extending polar modulation to support wide Summary band modulation schemes, such as WCDMA The cellular market is rapidly changing and and LTE. It is not possible to simply widen will continue to do so as network operators dethe bandwidth of the digital modulator as the termine the best solutions to meet subscriber increased loop bandwidth would degrade the performance and capacity needs. Regardless overall noise performance and cause stability of their decision, many options and variants concerns. The timing alignment for the AM and will need to be supported, and it is not realistic FM path has significant effect on the spectral for semiconductor suppliers to develop custom purity of the output signal. For EDGE, the sig- devices for each possible variant. Significant nals need to be aligned to less than 10 nS, but benefits can be realized by handset engineerfor WCDMA, this reduces to less than 2 nS. ing teams with a single RF platform capable of There is little doubt in the industry that as meeting their varied needs. with current GSM/EDGE solutions, polar For this to occur a different approach to RF modulation also provides the best architecture is required, and Sequoia Communications has for WCDMA and beyond; however, the diffi- proven that multi-mode 2-point polar moduculties have proven to be a deterrent for most lation provides a solution to many of these companies. Sequoia Communications is one problems. A single device has been developed company that has conceived a solution to the that supports both narrow band and wide band issues surrounding polar modulation and devel- modulation schemes creating a flexible multioped a 2-point polar modulator that provides mode product. This product (and architecture) an architecture to support increased modulation will allow development teams to cost-effecbandwidths, as shown in Figure 2. In 2-point tively sustain the myriad of devices that are exmodulation there are two modulation paths: pected to be released into the market over the the digital path, which is similar to a standard coming five years. GSM/EDGE path, and the analog path, which extends the bandwidth that the PLL can sup- Sequoia Communications, port. Maintaining the gain match between the San Diego, CA. (858) 946-7400. analog and digital paths is critical in order to [www.sequoiacommunications.com]. prevent higher EVM and degraded ORFS. This architecture has been proven in Sequoia Communications’ SEQ7400 multi-mode transceiver. The device has been tested and has proven its flexibility—it has been verified

figure 2

MAY 2008

portable power powering backlights

Backlight LED Driver System in Small Format Liquid Crystal Displays It pays to take a system-level approach to powering LED backlights in small-format LCDs.

by D ario Nurzad, Product Marketing Engineer, Mobile Devices Power, National Semiconductor

Light-emitting-diode (LED) technology is widely used to provide illumination for the pixels in small-format liquid crystal displays (LCDs) in battery-powered applications. White light, emitted by the LEDs, is transmitted through a polarizer to the LCD where the light can be blocked or attenuated and sent on to RGB color filters to create colored light. Figure 1 shows a system-level view of a backlight LED driver, which consists of a DC/DC converter and one or more regulated current sources. Moreover, RGB-LED-based backlights require a temperature-based feedback control, which translates to a much higher cost than white-LED-based backlights. How much PCB area can be used? What features are needed? How much power will the system consume? Answering these questions can guide the designer in the selection of the appropriate backlight LED driver.

PORTABLE DESIGN

DC/DC Converters for LED Backlight

In portable applications with single-cell LiIon source, the sum of the voltage drop across the white, green or blue LED and the current source can be lower or higher than the battery voltage. This means that, whereas a red LED can be powered directly from a single-cell LiIon battery, a white, blue or green LED requires the battery voltage to be sometimes boosted. The first aspect to consider when selecting an LED driver for a battery-powered application is the total area that the IC driver together with the external components will occupy (Figure 2). Two boost technologies are widely used: step-up DC/DC converter, also called inductive boost, and switched capacitor converter, also called charge pump. The charge pump implementation requires only four ceramic capacitors and one low-power resistor, which results

portable power

in a typically much smaller solution size. The capacitor value recommended for these applications is 0.47 μF to 1 μF with a voltage rating of 10V (to help with DC bias losses). These capacitors can be found in 0402 or 0603 case sizes from a number of capacitor manufacturers. Total solution sizes less than 21 mm2 are fairly common with also having the advantage of being very thin, less than 1 mm. Depending on the LED driver package, the capacitors can be the tallest component in the solution. Inductive-boost-based LED drivers tend to have larger solution sizes compared to the switched-capacitor drivers. A typical solution size for an inductive-boost-based LED driver is closer to 30 mm2 of board area. Inductive drivers typically require two capacitors, one on the input and one on the output, with capacitor values of 1 μF to 2.2 μF and are available in 0603 and 0805 case sizes. Inductive boosts require a rectification element that can handle the peak inductor current and output voltage. In a synchronous boost, a pass PFET can be integrated into the IC. However, this integration often causes the size of the IC package to increase over an asynchronous solution. The power conversion is also less efficient by about 10% in presence of an integrated high-voltage PFET or Schottky diode. In an asynchronous topology, the pass element consists of a Schottky diode. The main area increase of the inductive boost

figure 1

RGB Backlight Only

Temperature Sensor

SDA

Current Source

µP/µC SCL

DC/DC Converter

Backlight LED driver system

compared to the switch capacitor boost is the inductor itself. Applications with 6-8 LEDs with currents of 15 mA to 20 mA typically require one inductor of 10 μF to 22 μH with a saturation current between 0.4A and 0.5A. These inductors can be found in footprints smaller than 3.0 mm x 3.0 mm. The inductor is also the tallest component in the solution with heights in the range of 0.8 mm to 1.2 mm. The easiest way to boost the battery voltage is to use a step-up DC/DC converter (Figure 3). The advantage of this method consists of very high efficiency over all load and input voltage conditions as the input voltage can be boosted to the sum of the LED forward voltage and current source headroom voltage. This significantly optimizes efficiency at the expense of cost and PCB area as previously explained. Figure 3 shows the principle of operation of a magnetic boost regulator. When the NFET switch is closed (solid arrow), the inductor current iL(t) ramps upward from a minimum value Ia at t = t0 toward a maximum value of Ib at t = t1. During this time the Schottky diode is reverse biased and the load is supported by the energy stored in the output capacitor. At t = t1 the NFET switch is turned off and the energy stored in the inductor L is now delivered to the output capacitor and to the load (dashed arrow) through the Schottky diode. Consequently the inductor current ramps down

figure 2 VOUT

3.1mm R

COUT

4.4mm

4.1mm

CIN

Main LED String SUB LED String SDA

SCL

6.5mm

Typical PCB layout examples: charge pump (left), inductive boost (right)

PORTABLE DESIGN

RGPIO

GND

RSET

Dx Pins

GPIO

GND

SET

COUT

Enable/PGEN/GPIO

VC1 VC 1 VC 2 VC 2

In this way voltage boost operation is accomplished. The duty cycle of the switching signal is usually 50% as this value generally yields the optimal charge transfer efficiency. A voltage conversion with a gain of 1x is achieved by closing switch S8 and leaving switches S1-S7

figure 3 10µH (RL=0.62Ω)

3.6V

Schottky (VF=0.4V) 1µf

iL(t) Ib Ia

1µF

RDS_ON=0.55Ω

PWM

t0 SDA SCL

I2C

LM3509, Inductive boost LED driver

figure 4 S8

V IN 2

The output load current is provided by the output capacitor CHOLD. As this capacitor discharges and the output voltage falls below the desired output voltage, the second phase is activated in order to boost the output voltage above this value. During the second phase C1and C2 are in parallel, tied between VIN and VOUT. Switches S4-S7 are closed, whereas switches S1-S3 and S8 are open. Because the voltage drop across the capacitor cannot change instantaneously, the output voltage jumps to one and a half times the value of the input voltage:

portable power

during time t2 to the previous value of Ia. The output voltage must be greater than the input: if this voltage relationship were not true, the inductor would not discharge into the output network. In other words, when the NFET is turned off the voltage across the inductor is reversed because the current discharge can not occur instantaneously. The input voltage added by the reversed magnetic voltage leads to the output voltage being higher than the input voltage. When driving 10 LEDs in series the supply voltage required can be as high as 35V. Another advantage of the boost topology is the simplified PCB routing: only two connections between the driver and LED strings are needed. The second method of boosting the battery voltage is to use a charge pump (of which a simple implementation is shown in Figure 4), which utilizes the following property of capacitors: the capacitor charge build-up cannot occur instantaneously, which means that the initial voltage variation across the capacitor is equal to zero. The voltage conversion is achieved in two phases. During the first phase switches S1, S2 and S3 are closed, whereas switches S4-S8 are open. Therefore C1 and C2 are stacked and, assuming C1 is equal to C2, charged to half the input voltage:

S7 S5

VIN

C1 +

C2 S6

VOUT CHOLD

VC1 VC1 VC 2 VC 2 VOUT VIN

VIN 3 VOUT VIN 2 2

Charge pump circuit with 1x and 1.5x gains

MAY 2008

portable power

open. The benefit of the charge pump approach is the absence of the inductor. The inductor is a source of EMI noise, which can affect performance of the display or of the radio in a mobile phone.

figure 5

Input Power and LED Efficiency in Charge Pumps

Efficiency (%)

100

In a charge pump LED driver, the output power relation that is to be used in the efficiency calculation, assuming all LEDs are identical, is given by:

POUT 4 VLED I DX 80

The LED drive efficiency is given by:

3.0

3.5

4.0

4.5

5.0

5.5

where

VIN - Input Voltage (V)

I IN Gain 4 I DX IQ and I

Charge pump LED efficiency

er exploration ether your goal speak directly ical page, the ght resource. technology, es and products

4 VLED I DX VIN I IN

ILED = 15mA VLED = 3.3V

figure 6 VOUT

ed + ISET

IDX

VLED -

is the supQ ply current of the LED driver. Figure 5 shows a typical efficiency plot with the step indicating a gain transition. However, for a given a LED current, the forward voltage can vary with process and temperature. This means that the efficiency of the LEDs can vary still maintaining brightness constant, since the latter is dependant only upon the current. In order to make things clear let us consider an adaptive charge-pump-based LED driver circuit with the following specifications:

VLED 3.0V

Error Amplifier companies providing solutions now

15mA

3.7V

exploration into the latest datasheet from a company, VREF =Iproducts, XR1 technologies +and companies. Whether your goal is to researchDX SET mp to a company's technical page, the goal of Get Connected is to put you in touch with the right resource. Whichever level of gy, Get Connected will help you connect with the companies and products you are searching for. IN

onnected

Neglecting the quiescent current IQ, the input current, when the charge pump operates with a gain of 1x, is given by:

I IN 1 4 I DX 60mA

End of Article

Regulated current source

Get Connected

with companies mentioned in this article. www.portabledesign.com/getconnected

Thus efficiency and input power are equal to:

4 IDX VLED VIN IIN

81.1%

PIN VIN IIN 222mW

PORTABLE DESIGN

Get Connected with companies mentioned in this article.

Let us now calculate efficiency and input power for the same circuit but with LEDs having 3.3V forward voltage:

4 I DX VLED VIN I IN

89.2%

PIN 222mW It is clear that efficiency has increased, but input power has remained the same; this means that efficiency of the LEDs, given by 4 VLED I DX , does not affect the power drained from the battery but the amount of power dissipated in the driver circuit. Efficiency is, therefore, not an adequate figure of merit to evaluate power consumption: what must be considered is input power versus LED brightness, i.e. LED current. Input power is the true measure of how many electrons are drained from the battery for a given LED brightness. With a gain of 1.5x, under the previous conditions, the input power is equal to 333 mW regardless of VLED. Since a charge pump converter has a finite number of voltage gains there will always be, based on the application specifications, a certain amount of wasted power in the driver circuit. It is, therefore, important, in order to minimize input power, to operate the charge pump with the smallest possible gain.

Constant-Current LED Driver

The LED characteristics determine the forward voltage needed to achieve the needed level of current, which determines the amount of emitted light. Due to the variation in LED voltage versus current characteristics, controlling only the voltage across the LED leads to variability in light output. Therefore, most LED drivers use current regulation. A circuit that accomplishes current regulation is a low-dropout regulator and is illustrated in Figure 6. The error amplifier takes the voltage across R2, V2, compares it against the reference voltage VREF, and adjusts the LED current, IDX, via the series-pass element, an NFET, to the value required to drive the error signal (VERR=VREF-V2) as close as possible to zero. VREF is equal to:

VREF I SET R1

I SET

1.25V RSET

Setting VREF =V2, the current through the LED is given by:

I DX

in small-format LCDs. A switched capacitor voltage regulator is desirable for applications requiring low cost and small solution size. The constant current regulator approach adds to these features excellent current matching and brightness control. Last but not least the inductive boost solution yields the lowest

R1 I SET R2

This holds true only if VOUT-VLED is sufficiently high to keep the pass element from saturating. As a matter of fact, the current sources require a minimum voltage, called headroom voltage VHR, across them in order to provide the desired regulated current through the LED. The headroom voltage is typically modeled with a resistor:

figure 7 Temperature Feedback Loop

Summary

This paper described the different building blocks needed to power LED backlights

2.9-5.5V CVDDD

CVDDA

+ _

5-20 V

4.7 µH

COUT

10 µF

10 µF 1µF

CVLDO

VDDA

VLDO S1_IN

ADC

S2_IN

MCU WITH I2C

NRST SS/SDA SCK/SCL SI SO IFSEL

CVDDIO

VDDD CALIBRATION MEMORY

SW FB

BOOST

VLDO

LP5520

LM 20

SPI/I2C INTERFACE

VDDIO

100nF

CIN

100nF

VHR R HR I DX Brightness can be controlled directly by changing the LED current (analog control) or indirectly by shutting on and off the LED fast enough to create the perception of dimming to the human eye (PWM control). In most portable applications analog brightness control is preferred, as the backlight controller is normally distant from the LED driver. Therefore, having to route a PCB trace that carries a PWM signal close to noise-sensitive systems such as a radio transmitter, speaker or display, can cause problems. Finally, in applications requiring superior color gamut, red, green and blue LEDs are utilized. The red LED is made of InGaAlP, whereas both the blue and green are made of InGaN. When the ambient temperature is varied, the dominant wavelength changes dramatically for red compared to blue and green thus requiring some kind of temperature-compensated feedback loop. The LP5520 (Figure 7) adjusts the RGB LED currents for perfect white balance (color accuracy ΔX and ΔY < 0.003) by using an internal calibration memory, which stores the intensity vs. temperature data of the LEDs, and an external temperature sensor.

portable power

where

PWMR PWMG PWMB BRC

COLOR AND BRIGHTNESS PWM LOGIC

LED DRIVERS

ROUT GOUT BOUT

GND

LM20 Temperature Sensor

LP5520, Backlight RGB LED driver

power consumption over the entire supply voltage range. We emphasized the criteria to follow when evaluating power consumption of the different topologies: input power consumption, for a given brightness level, is the key factor to consider when evaluating a charge-pump-based LED driver. In switched capacitor circuits, input power is directly related to the gain of the device, whereas in magnetic topologies input power is dependant on forward voltage of the LED (diode efficiency). National Semiconductor Corporation Santa Clara, CA. (408) 721-5000. [www.nsc.com].

MAY 2008

technology focus display technology

Defining the Display Technology Needs of the Future Reflective micro-MEMS displays lose the backlight, cut power dramatically in portable applications.

by Jim Cathey, VP of Business Development, Qualcomm MEMS Technologies, Inc.

The role of the cellular phone handset within the consumer electronics space is evolving rapidly. A simple voice device now functions more like a Swiss Army knife, as it now must behave as a PDA, digital camera, video camera, gaming system, Web browser, television set, GPS navigator, mp3 player and more. As these applications converge onto the mobile handset platform, a new generation of component technologies must accompany them to meet their escalating demands. As designers of portable devices know, the most power-hungry component of a cell phone is the display. The backlight alone can account for up to 50 percent of the total battery budget. The next generation of display technologies must account for increased display on time as well as accommodate the increased demands of applications running on the device while also providing users ever slimmer and sleeker designs so the device stays relevant as dictated by style.

PORTABLE DESIGN

The implication of this increased convergent use of multimedia applications is that as handset power is increasingly applied to multimedia applications, a form of â&#x20AC;&#x153;energy crisisâ&#x20AC;? occurs. Increasing data usage, Web browsing and GPS applications all are contributing to this widening energy gap. This energy dilemma would be mitigated if battery capacity increased by a commensurate amount; however, it seems that will likely not be the case. To the contrary, there is a growing energy gap (Figure 1) between greatly increasing energy demand and moderately increasing battery capacity. Sound industrial design, solid feature integration and a bigger battery alone will likely no longer suffice in providing energy consumption balance.

Reflective Displays

One compelling solution to this issue, and one that is gaining significant traction, is a reflective display technology pioneered by

¥ !CTEL #ORPORATION !LL RIGHTS RESERVED 3PARTAN AND #OOL2UNNER ARE REGISTERED TRADEMARKS OF 8ILINX )NC #YCLONE AND -!8 ARE REGISTERED TRADEMARKS OF !LTERA #ORPORATION

7(!4 $/ 7% (!6% 4/ $/ $2!7 9/5 ! 0)#452% /NLY !CTEL GETS YOU THIS CLOSE TO ZERO !NY OTHER CLAIMS OF LOW POWER SUPERIORITY ARE JUST THAT !CCORDING TO THEIR OWN DATA !LTERA AND 8ILINX USE BETWEEN AND TIMES THE POWER OF !CTEL )',// &0'!S DEPENDING ON DEVICE AND MODE 7ANT SPECIFICS 6ISIT US TO GET THE WHOLE PICTURE INCLUDING A VIDEO OF ACTUAL MEASUREMENTS

MORE PROOF AND PICTURES AT ACTEL COM POWER

technology focus

figure 1 Convergence = Growing Power Gap Energy demand of handset applications is growing • Device performance rising • Complexity rising • Functionality rising

Power

Handset Demand

Power Gap

Battery Supply

Battery technology falling behind er exploration ether your goal 2000 2003 2006 2009 2011 speak directly ical page, the ght resource. technology, The “energy gap” not closed by improved battery capacity must be closed with components boasting es and products improved efficiency.

Qualcomm called mirasol displays (Figure 2). This technology provides consistent viewing quality in varied environments and dramatically reduces the energy consumption from the display. Qualcomm’s mirasol displays deliver substantial performance benefits over emissive display technologies. These reflective displays, based on interferometric modulation (IMOD) technology, provide a significant reduction in power consumption, no backlighting, and viewing in bright sunlight. Inspired by the simplicity of natural iridescent colors, the mirasol display physically manipulates light using micron and sub-micronsized mechanical elements. Each pixel within a mirasol display is composed of micro-electromechanical-system (MEMS) elements. Color displays are made by composing a single pixel from MEMS elements of different thickness. Varying the cavity depth results in the resonance of wavelengths to yield varying colors. This pixel construction utilizes no color filters, polarizers, or organic compounds. Such simplicity leads to high energy efficiency, brightness and environmental stability, which are the hallmarks of the mirasol design.

figure 2 open state

collapsed state

Glass solutions substrate now companies providing

1 Pixel

Thin film stack

exploration into products, technologies and companies. Whether your goal is to research the latest datasheet from a company, gap the goal of Get Connected is to put you in touch with the right resource. Whichever level of mp to a company's technicalAir page, Reflectivewill membrane gy, Get Connected help you connect with the companies and products you are searching for.

onnected Red Subpixel

Green Subpixel

Blue Subpixel

A representation of a reflective pixel of a mirasol display.

End of Article Get Connected

with companies mentioned in this article. www.portabledesign.com/getconnected

I Don’t Have the Energy

The above design impacts how long a consumer can actually use their device. As low-power displays make their way into the marketplace, it will be important for cell phone marketers to acknowledge that converged devices act differ-

PORTABLE DESIGN

Get Connected with companies mentioned in this article.

ently and so should be measured differently. Today’s cell phone packaging focuses on two commonly published metrics for cell phones: Talk Time and Standby Time. These metrics are relics of the non-convergent handset era. A handset with a power-hungry display will yield the same Talk Time and Standby Time as will an otherwise identical handset with low-power display technology, thereby implying usage times that are not true to the actual functionality of the device. Metrics should focus on real live usage models where multimedia is central to the activity on the device. The metrics in Figure 3 are based on the concept of residual energy. A phone’s residual energy is defined to be the amount of energy remaining in its battery after a typical day’s use (after using the phone in a typical manner, how many minutes of use are left for the multimedia applications). Not surprisingly, the phone featuring the reflective mirasol display has a residual video advantage over the other technologies. Reflective displays seem the most likely candidate to close the “energy gap” discussed above, while also enabling new opportunities across the entire wireless handset value chain. Industrial design options and cost reductions accrue to handset manufacturers. The application and feature space expands for product managers—including the availability of previously unachievable applications. The increase in available revenue time (ART) means higher ARPU for carriers and operators. Users gain the use of valuable new features while being greatly freed from worries of dead batteries and the nuisance of over-frequent phone recharging.

Looking Good

Energy is one of the big metrics to clearly define, but what about the ultimate measure of any display technology: viewability? Emissive displays, for instance, generate their own illumination, yet the available light output consistently obscures in the face of modest room lighting and is significantly degraded in bright office lighting. Worse, emissive displays be-

technology focus figure 3 Industry Needs Benchmark Reflecting New Handset Usage

“The display technologies

mirasol

of tomorrow... need to be energy efficient and provide greatly extended battery life in diverse handset usage models.”

come washed out in diffuse sunlight and are completely overpowered in direct sunlight. Consequently, the viewing quality of emissive displays deteriorates to the same extent that the display light emission is washed out. The nature of reflective displays allows consistent contrast ratios, and hence consistent viewing quality, across the full range of illumination environments. JND, or “just noticeable difference,” is a recognized method of expressing the number of separately discernable image levels available to the viewer. One can associate the ability to discern small levels of difference in an image with perceiving the image as having high quality. Conversely, a decrease in JND count represents a decrease in the quality level of the viewed image.

A Trick Done With Mirrors

The display technologies of tomorrow have very high standards to live up to. They need to be energy efficient and provide greatly extended battery life in diverse handset usage models. They must support the onslaught of new features available to users, the design space available to developers, and the revenue streams available to operators. Qualcomm’s MEMS technologies mirasol display brings bright, vibrant color without the use of liquids, polarizers, organics, or semiconductor materials.

Generic Phone with mirasol display Video Time

Average Use

MMedia Use

4.5 hours

206 mins

160 mins

Total Video Watching Time Your Time May Vary

End-of-Day Video Watching Time Your Time May Vary

LCD

Generic Phone with TFT LCD Video Time

Average Use

MMedia Use

3.3 hours

70 mins

24 mins

Total Video Watching Time Your Time May Vary

End-of-Day Video Watching Time Your Time May Vary

Comparative residual energy metrics for LCD and mirasol-equipped phones.

The lack of these optical enhancers combines with the reflective nature of the display to offer consistent viewing quality across all environments, including bright sunlight. Additionally, the consistent viewing quality facilitated by reflective displays provides a significant advantage to consumers who desire to use their cellular phones anytime and in any viewing conditions. Qualcomm Incorporated San Diego, CA. (858) 587-1121. [www.qualcomm.com].

MAY 2008

product feature First Programmable CMOS RF Transceiver for Handsets SDR RFIC supports all bands and modes between 700 MHz and 3.8 GHz; integrates RF chain in a single CMOS chip. by John Donovan, Editor-in-Chief RF designers have a serious problem with handsets. Consumers expect their cell phones to deliver seamless voice, data and multimedia at high data rates while communicating over cellular, Bluetooth and Wi-Fi links; they want Ferrari performance but they don’t want to pay for it; and they want unlimited battery life. Handset manufacturers need to build multi-band/multiprotocol designs that are scalable, work in all geographies, won’t explode their BOM cost and won’t blow their market window. The latter is either “the Holy Grail” or a “pipe dream,” depending on how much residual optimism you’ve managed to retain. There are two approaches to solving this problem, neither of which is attractive. You can add multiple RF chains; this approach adds considerable cost, complexity and power consumption. The alternative is software-defined radio (SDR), but traditional SDR architectures have been too large, costly and powerhungry for portable designs. Still, the logic of SDR is compelling: storing multiple protocols in memory, then loading and running them when needed on optimized hardware would save a lot of complexity, and therefore cost and power. BitWave claims to have solved the RF designer’s conundrum with the announcement of “the world’s first programmable CMOS RF transceiver.” BitWave is currently sampling its BW1102 Softransceiver RFIC, a single transceiver component chain with dynamic component reconfiguration: digitally controlled components facilitating tunable performance. This means one LNA, one mixer, one amplifier, one A/D converter, and so on—not multiple functional blocks switched in and out based on mode. This makes it possible to replace the multiple fixed-application transceivers found in most cell phones today with one programmable transceiver. All of this is implemented in UMC’s 130nm digital CMOS with no extra “RF-friendly” steps. The BW1102 is fully software configurable across all channels in the frequency bands between 700 MHz and 3.8 GHz with a protocol bandwidth of 25 kHz to 20 MHz. This enables

PORTABLE DESIGN

the BW1102 to be configured for multiple wireless protocols such as GSM, GPRS, EDGE, WCDMA, HSDPA, CDMA2K, EVDO, 802.1 1b/g and 802.16d/e. The BW1102 is a single lowcost, low-power chip targeted at the cellular handset and femtocell markets. The BW1102’s architecture differs substantially from the classic SDR software control architecture (SCA). The chip itself is preceded by antennas, power amplifiers (PAs), SAW filters and duplexers that are selected by the 1102’s internal RF switches, which in turn feed an LNA that precedes a configurable receiver. The BW1102 avoids placing a power-hungry, highbandwidth ADC close to the antenna, instead opting for a more traditional—though fully programmable—superheterodyne receiver; the receiver features one analog down-conversion that can be programmed either to a low IF or down to baseband. The transceiver partitions the CMOS circuitry into the RF, analog, mixed signal and digital domains. The RF/analog/mixed signal circuitry includes an LNA, several VGAs, analog filters, mixers, and an ADC and DAC. Digital circuitry adds digital filtering, offset correction, decimation, phase compensation, sync compensation, up sampling, encoding, pulse shaping, GSM and EDGE modulation; as well as control, calibration and a programmable digital interface. These domains are linked through SPI chains, clocking circuitry comprised of a DCXO, a fractional-N or integer PLL, and separate transmit and receive fractional-N synthesizers. The BW1102 stores protocols in “software mode files” that can be updated over the air—a major advantage when RF air interfaces are developing as rapidly as they are. The chip is capable of dynamic reconfiguration between cellular network, wireless data network and personal area network protocols. According to Erik Org, BitWave’s business development & marketing manager, “With our single RF chain, we can’t do simultaneous full duplex protocols. However, we can switch fast enough—from GSM to Wi-Fi, for example—so that two TDD protocols can appear simultaneous to the user. Switching times vary depending on the protocols but are typically limited by the synthesizer settling time. In the case of a GSM/compressed mode to WCDMA handoff, we can switch in 125 microseconds.” The BW1102 is sampling now, with volume production planned for 3Q08. The chip comes in a 7x7 mm PBGA package and requires about 40 passives. BitWave Semiconductor, Inc., Lowell, MA. (978) 888-0200. [www.bitwavesemiconductor.com].

Where electronic design meets... Mark your calendars for the 45th Design Automation Conference (DAC) at the Anaheim Convention Center, June 8-13, 2008, in Anaheim, California, USA. • A robust technical program covering the latest research developments and trends for the design of SoCs, FPGAs, ASICs, digital ICs and more. • Worldwide attendance from leading electronics companies and universities. • Over 250 companies displaying electronic design technologies and services.

Highlights Include:

• Wireless Theme sessions • Management Day on Tuesday, June 10 • Paper sessions featuring leading technical research • New “iDesign” and WACI sessions • Dynamic Panels Sessions • 19 collocated events and workshops • Full-day and Hands-on Tutorials

2008 Keynotes: • Justin R. Rattner, Intel Senior Fellow, Vice President, Director, Corporate EDA for Digital, Programmable, Multi-Radios Tuesday, June 10 • Sanjay K. Jha, COO and president of Qualcomm CDMA Technologies Challenges on Design Complexities for Advanced Wireless Silicon Systems Wednesday, June 11 • Jack Little, President, CEO, and aCo-founderofTheMathWorks,Inc. Thursday, June 12

products for designers High-Performance, Low-Power Platforms for Wireless, Multimedia Applications

Oscilloscopes for Embedded Systems Designers

CEVA has announced its next-generation DSP subsystem platforms for developers using the CEVA-X family of DSP cores. The robust solutions build on CEVA’s extensive track record of powering complex, multi-function communications products and offer a comprehensive and verified approach for efficiently integrating its cores into complex system on chips (SoCs). The platforms come in two versions, the CEVA XS1100A optimized for wireless baseband applications, and the CEVA XS-1200A aimed at multimedia and other applications requiring high-performance signal processing. These configurable, highly efficient hardware platforms reduce development effort, the risk of costly silicon re-spins and, ultimately, time-to-market for embedded processor applications. It uses industry standard system buses, offering designers the ability to add their own hardware blocks or connect the DSP to other systems present on chip, making integrating CEVA cores a very straightforward, efficient proposition. Both platforms support critical low-power design requirements through CEVA’s smart Power Management Unit (PMU) technology, which includes automatic sleep/wake of each resource and matrix separately according to transaction type, source, destination, initiator and duration. The two platforms—which have evolved through numerous customer engagements in targeted application areas—feature architecture enhancements that can significantly lower die size and power consumption, without compromising on performance. They are geared for the most complex and highly integrated SoCs, and feature a complete AHB matrix, DMA, TDM ports, power management, external master and slave ports, complete lineup of DSP-oriented peripherals and interface to L2 memories. Compared to previous-generation platforms from CEVA, the new platforms allow designers to develop chips that offer 10% higher speed; 20% smaller die size; 20% lower leakage power and 10% lower dynamic power; 50% fewer clock-tree cells, ensuring higher production yield; and 5% decrease in MHz requirements for video codecs such as H.264.

Tektronix has announced the availability of the DPO3000 Digital Phosphor Oscilloscopes (DPOs), a portable series of models that extends the advantages first introduced with the award-winning DPO/MSO 4000 series. The new DPO3000 provides triggering and decode support for the most popular serial buses used in embedded design applications. Additionally, the new oscilloscopes offer 5M standard record length and utilize the popular Wave Inspector navigation and search controls to simplify finding events of interest. The new portable DPO3000 simplifies debug and sets new standards for performance, value and ease-of-use. Virtually every electronic product designed and produced today is an embedded system and may contain microprocessors, microcontrollers, DSPs, RAM, flash, EPROMs, FPGAs, A/Ds, D/As and other I/O. Increasingly, embedded system designs are replacing internal parallel buses with serial buses such as I2C, SPI and CAN. On a serial bus, a single signal may include address, control, data and clock information; the complexity presents significant debug challenges for design engineers. The DPO3000 Series addresses these problems with the most complete set of tools for working with industry standard serial buses such as I2C, SPI, RS-232/422/485/UART, CAN and LIN. The DPO3000 oscilloscope family consists of six models, ranging from 100 MHz to 500 MHz. All models provide 2.5 Gsample/s sampling with >5x oversampling on all channels. 5M record length is standard on all channels enabling the capture of long time windows of signal activity with very high resolution. The new DPO3000 complements the recently introduced DPO4000 and MSO4000 Series, ranging from 100 MHz to 1 GHz bandwidths and providing the most complete and capable portable oscilloscope offering in the market for embedded systems engineers. U.S. list prices range from $4,450 for a two-channel DPO3012 to $10,900 for a four-channel DPO3054. All DPO3000 models are available for purchase and delivery. NI LabVIEW SignalExpress Tektronix Edition LE base version is included free of charge with the DPO3000 Series.

CEVA, Inc., San Jose, CA. (408) 514 2900. [www.ceva-dsp.com].

Tektronix, Inc., Beaverton, OR. (503) 627-4027. [www.tektronix.com].

Lowest Power Large-Memory 16-Bit USB Microcontroller Family Microchip Technology has announced the 12-member PIC24FJ256GB1 microcontroller (MCU) family, which it claims is the lowest power (100 nA standby current) large-memory (up to 256 Kbyte flash and 16 Kbyte RAM) 16-bit USB microcontroller family in the world. As the only 16-bit microcontroller family with integrated USB 2.0 device, embedded-host, dual-role and On-theGo (OTG) functionality, the PIC24FJ256GB1 makes it cost-effective and easy to add advanced USB features to embedded designs. Applications that previously required a high-end chip can utilize the cost-effective, low-power 16-bit PIC24FJ256GB1 family to easily incorporate both advanced USB OTG and capacitive-touch functionality. Additionally, Microchip provides complete software support, via free USB class drivers and USB applications. And, the PIC24FJ256GB1 has ample code space for these advanced applications, while providing up to four UARTs, three SPI ports and three I2C ports to expand control capabilities and eliminate the space and cost of support chips. Key features include USB Implementers Forum certified for Peripheral and Embedded-Host functionality; USB 2.0 OTG—supports device, embedded host, dual role and OTG; Low power—standby current of 100 nA; Peripheral Pin Select flexible pin mapping; CTMU peripheral for capacitive touch; Expanded peripherals—4 UARTs, 3 SPI, 3 I2C ports; and 23 independent timers. The 12-member PIC24FJ256GB1 family is offered in 64-, 80-, or 100-pin TQFP package options, and all are available now for general sampling with volume production expected in May 2008. Pricing starts at $3.47 each in 10,000 unit quantities. Microchip Technology Inc., Chandler, AZ. (480) 792-7200. [www.microchip.com].

PORTABLE DESIGN

Virtutech has announced Simics 4.0, the new release of the company’s flagship VSD platform. Simics 4.0 embraces a completely new approach to delivering performance and scalability improvements to simulation capabilities as well as additional APIs to support more use cases and a repository that leverages the thousands of models accrued since Simics’ first release. With Simics 4.0, Virtutech is leading the charge for widespread simulation adoption by delivering the speed and scalability necessary to enable software development not only well ahead of hardware availability but throughout the complete development cycle to realize quality, productivity and time-to-market advantages. The most significant new capability of Simics 4.0 is the addition of Simics Accelerator, a set of new features that boost performance and scalability, such as page sharing and multi-threading. This allows one Simics process to simulate several machines in parallel. The ability to spread a simulation across multiple host CPU cores is unique to commercially available, fast, functional simulators and improves performance simulation by up to the number of host CPU cores across which the simulation can be distributed. Simics 4.0 also provides an expanded set of APIs, which enables developers to integrate third-party models into Simics, and an enhanced version of its Device Modeling Language (DML), which is a simple and elegant declarative language for Transaction Level Modeling (TLM). Leveraging the pre-existing capabilities of three previous versions of Simics, Simics 4.0 features an up-to-date repository that provides users with easy access to information on models for more than one thousand processors and devices. Simics 4.0 includes Simics Accelerator, with page sharing and multi-threaded simulation for performance and scalability; IPv6 and Hot Plug-in support for networking simulation; Visualization Console with performance monitor and memory viewer provides in-depth perspective on how a customer’s system is running; Device Modeling Language enhancement for C/C++ and debug interface; Eclipse integration with Wind River Workbench 3.0; integration with Freescale CodeWarrior development tools; Documented API and coding conventions for integration of third-party models; repository of thousands plus of models of processors and devices. Virtutech, Inc., San Jose, CA. (408) 392-9150. [www.virtutech.com].

Secure Device Management Enables Remote Diagnostics, Upgrade Green Hills Software has announced the world’s first secure device management solution. Green Hills Software’s Device Management Solution enables customers to securely perform in-field and remote diagnostics, debugging, upgrade and management of deployed software. Traditionally, embedded systems have been relatively immune to remote tampering: most devices simply lack a built-in capability for remote patching, upgrading, or modifying software. However, device management solutions preload the exact mechanism that hackers desire: a channel by which to infect, disable and commandeer the system. Existing device management solutions are integrated with operating systems such as VxWorks and Linux that meet EAL4+ (Common Criteria) security levels or lower. This level of security is understood by experts to be inappropriate when protection is required against determined attempts by hostile and well-funded attackers. Device management software, intended to improve serviceability, actually creates serious security vulnerabilities. Devices are increasingly networked, perform critical functions requiring in-field maintenance, and discharge an evolving role requiring in-service software upgrades. Designers require a device management solution that cannot be hacked. Green Hills Software’s Secure Device Management Solution is built upon the INTEGRITY separation kernel, the first and only operating system to be accepted by a U.S. NIAP lab into a high-assurance (EAL6+) security evaluation. This is the same level of security required by high-value U.S. government computing and networking infrastructure. Communications are secured by a certifiable suite of cryptographic protocols—not the open source implementations frequently found to contain security flaws. Furthermore, attestation (using hardware-based mechanisms such as a Trusted Platform Module, if available) of device management software and any remotely loaded patches is enforced. Green Hills Software’s solution ensures that only authorized users and applications can access, control and modify the device. Green Hills Software’s Secure Device Management can be deployed in products that make use of any general-purpose and/or real-time operating system. Green Hills Software, Inc., Santa Barbara, CA. (805) 965-6044. [www.ghs.com].

Proof Accelerators for Rapid and Exhaustive Verification of Intractable Datapath Designs Jasper Design Automation has unveiled a set of modeling extensions—Formal Scoreboard Proof Accelerator, Clock Domain Crossing (CDC) Proof Accelerator, Cache Proof Accelerator and FIFO Proof Accelerator—for the rapid and exhaustive verification of intractable datapath designs. Delivering orders of magnitude greater coverage than traditional simulation alone, the JasperGold Verification System Proof Accelerators reduce complexity, improve performance and dramatically increase formal capacity. While used successfully on production computing, consumer electronics, networking and microprocessor chips, JasperGold’s Proof Accelerators can also be used to accelerate the functional verification of any complex chip where datapaths, multiple clock domains, caches and FIFOs pose a verification challenge. Jasper is the first and only company to deliver a Formal Scoreboard, a formal-optimized equivalent of a simulation scoreboard. This collection of checks and techniques exhaustively ensures datapath design functionality. For blocks containing multiple asynchronous clock domains, Jasper delivers a Clock Domain Crossing (CDC) Proof Accelerator. The Proof Accelerator enables exhaustive formal verification of design blocks historically known to be particularly challenging to verify. Despite the rapid increase in the number and complexity of clock domains in today’s system chips, with the CDC Proof Accelerator, JasperGold users can now exhaustively verify the correctness of a design across all clock edge combinations, including clock variation and jitter. While both cache and FIFO verification typically result in state-space explosion, Jasper’s Cache and FIFO Proof Accelerators provide a “formal safe” way of rapidly and successfully modeling complex caches and FIFOs. These JasperGold modeling extensions are uniquely able to manage complexity, thereby ensuring formal functional equivalents of the cache and FIFO blocks in the design. The set of four JasperGold Verification System Proof Accelerator extensions—Formal Scoreboard, Clock Domain Crossing, Cache and FIFO—is currently available with the JasperGold Verification System. Jasper Design Automation, Mountain View, CA. (650) 966-0200. [www.jasper-da.com]. MAY 2008

products for designers

First Deterministic Multithreading VSD Platform

products for designers

Serial Quad I/O Flash Memory

Cortex-M3 Optimized for Low-Power Designs

Silicon Storage Technology (SST) has announced the new 26 Series Serial Quad I/O (SQI) family of 4-bit multiplexed I/O serial interface flash memory devices. The 26 Series’ low pin count and enhanced serial interface architecture make the family an ideal code storage solution for applications such as ultra-low-cost handsets, Bluetooth headsets, optical disk drives and GPS devices, where small form factor, low power consumption and high data rate are a must. With a fast 80 MHz operating frequency and a specialized instruction set, SST’s 26 Series SQI flash devices target execute-in-place (XIP) applications, allowing programs to be stored and executed directly from the flash memory without the need for code shadowing on an SRAM, thus further increasing overall system performance. The 26 Series SQI devices feature a 4-bit multiplexed serial interface architecture and up to a 350 Mbit/s sustained burst data rate, which provides performance exceeding typical parallel flash memory. To provide additional system speed and performance, the 26 Series supports 8-, 16-, 32- and 64-byte burst mode operation with wrap around, which reduces the memory’s read access time, as well as memory index jumping, which reduces the number of input clock cycles. Products in the 26 Series are based on SST’s proprietary, high-performance SuperFlash technology. The 26 Series SQI devices offer superior reliability of 100,000 cycles endurance (typical) and greater than 100 years data retention. The active read current of the devices is only 15 mA (typical at 80 MHz) and standby current is a mere 10 tA (typical). All 26 Series SQI devices are available in both commercial and industrial temperature ranges, and are offered in an 8-contact WSON (6 mm x 5 mm) or 8-lead SOIC (200 mil) package. The first product in the 26 Series SQI family is the 16 Mbit, 3.0V SST26VF016. Samples of the product will be available in May, with volume production scheduled for Q4 2008. Pricing for the SST26VF016 device is $1.16 each in 10K unit quantities.

ARM has announced a significant upgrade to the widely licensed Cortex-M3 processor, which is utilized in a wide range of high-performance, low-cost applications. The latest release includes a new Wake-Up Interrupt (WIC) controller, which allows almost instantaneous return to fully active mode from an Ultra-Low Leakage (ULL) retention state and introduces enhanced power management features that address the ongoing need in the embedded market for increased performance and longer battery life in next-generation designs. The Cortex-M3 processor is now also supported by the ARM Power Management Kit (PMK) and is tightly integrated with, and optimized for, low-power physical IP standard cell libraries and memories from the ARM Artisan physical IP family. Additional enhancements include the ability to integrate solutions for safety-critical and fault-robust applications in industrial, medical and automotive applications. To enable ultra-low-power standby operation, the processor has a new integrated Wake-Up Interrupt Controller that, when utilized with low-power cells such as the ARM Power Management Kit (PMK) and standard cell libraries and memories, enables the core and the interrupt controller to be rapidly placed in an ultra-low leakage retention mode, returning to fully active mode almost instantaneously on critical events. When combined with Partners’ unique insight into power-efficient design techniques, the WIC will lead to a new wave of extremely power-efficient microcontroller solutions. By combining the ARM PMK, memories and standard cell libraries with the Cortex-M3 processor, designers can also gain significant time-to-market savings when targeting ultra-low leakage manufacturing processes for major power-saving benefits with very low area impact. The Physical IP memories feature byte-write capability and allow multiple power-down (with retention) modes when coupled to the PMK. Other new benefits of the latest release of the Cortex-M3 processor include significantly extended configurability of debug and trace capabilities to enable the creation of extremely cost-sensitive, high-volume 32-bit devices, such as electronic toys and consumer electronics systems; improved operating frequency for highest performance embedded applications; and enhanced signaling to power management controllers. Enhanced support features for the Cortex-M3 processor will be available in Q2 2008.

Silicon Storage Technology, Inc. (SST), Sunnyvale, CA. (408) 735-9110. [www.sst.com].

ARM Inc., Sunnyvale, CA. (408) 734-5600. [www.arm.com].

Continuous Verification Workflow The MathWorks has announced that it now offers a continuous verification workflow that connects systemlevel models and algorithms developed in MATLAB and Simulink with digital hardware simulators from the three major EDA companies. With the availability of EDA Simulator Link DS, which supports co-simulation between MATLAB and Simulink and the Synopsys VCS MX functional verification solution, The MathWorks completes its EDA Simulator Link portfolio, which also includes EDA Simulator Link MQ (for Mentor Graphics’ ModelSim and Questa) and EDA Simulator Link IN (for Cadence Incisive Simulator). EDA Simulator Link products from The MathWorks offer support for VHDL, Verilog and mixed-language simulators, enabling engineers to connect MATLAB and Simulink to their choice of hardware description language (HDL) and register transfer level (RTL) simulator for their hardware design and verification tasks. The products also work seamlessly with Simulink HDL Coder from The MathWorks to automate integration of legacy RTL IP with designs developed in MATLAB and Simulink. The EDA Simulator Link products support design teams across FPGA and ASIC markets that are striving to reduce development time, design flaws and verification costs. The success of EDA Simulator Link products in improving product quality and cutting verification time has fueled demand for additional interfaces to hardware workflows. As a result, the EDA Simulator Link portfolio has expanded and has prompted EDA vendors to deliver similar tools for analog and mixed-signal simulators such as Synopsys Discovery AMS and Saber, Cadence Virtuoso Multi-Mode Simulation, Cadence PSpice and Cadence Allegro AMS Simulator, and Mentor Graphics ADVance MS (ADMS). The MathWorks, Inc., Natick, MA. (508) 647-7000. [www.mathworks.com].

PORTABLE DESIGN

Express Logic has announced the introduction of BenchX, a comprehensive Integrated Development Environment (IDE) for embedded systems. Expanding beyond the RTOS and middleware markets it has served for eleven years, Express Logic designed BenchX IDE to meet customer need for low-cost, industrial-grade software development tools. BenchX is an ideal solution for developers using the ARM, ColdFire, Power or MIPS architectures. BenchX is based on the Eclipse Europa CDT Release (4.0) and contains a complete windows-hosted graphical IDE; project builder with new project wizard; language-sensitive editor and code browser; GNU C/C++ compilers; GDB debugger with graphical interface; target instruction set simulator; user tutorials for quick learning; target connectivity hardware probe; and no license keys. Standard Eclipse downloads do not include the GNU tools, nor do they support any hardware debug probes, making them short of what is required to be useful to an embedded developer. BenchX addresses these shortcomings, yet it is priced at a very affordable level, making it accessible to all developers on a project with minimal cost impact. Each BenchX component has been configured for embedded use, enhanced by Express Logic with added-value, and integrated together, creating a polished commercial product that can be used out-of-the-box. Express Logic’s BenchX IDE is not only available for use with Express Logic’s ThreadX RTOS, but can also be used to develop embedded systems without any RTOS at all, with an in-house RTOS, or with another commercial RTOS. BenchX is available for order now, with initial production deliveries planned for May 15. BenchX licenses are priced from $1,000 per seat, including debug probe, documentation and three months of technical support. Express Logic, Inc., San Diego, CA. (858) 613-6640. [www.expresslogic.com].

32-Bit MCUs Target Embedded Applications NXP Semiconductors has announced the expansion of the largest portfolio in the ARM7 and ARM9 microcontroller space with the introduction of the LPC3200 family. The NXP LPC3200 family is based on the popular ARM926EJ processor and targets consumer, industrial, medical and automotive applications providing designers with a high-performance power-efficient microcontroller. Designed in 90nm process, the new NXP LPC3200 family combines the performance of an ARM926EJ core, a Vector Floating Point (VFP), an LCD Controller, an Ethernet MAC, On-The-Go USB, an efficient bus matrix and a large set of standard peripherals. These features offer embedded designers the ability to reduce on-chip components and maximize power savings without sacrificing performance. Designed for flexibility in applications that require fast, simultaneous communications, the LPC3000 combines high-performance, low power consumption and a myriad of peripherals. The features include I2C, I2S, SPI, SSP, UARTs, USB, OTG, SD, PWMs, A/D with touch-screen interfaces, 10/100 Ethernet MAC and a 24-bit LCD controller that supports STN and TFT panels. The family supports DDR, SDR, SRAM and flash memory devices and provides the option of booting-up from NAND Flash, SPI memory, UART or SRAM. The NXP LPC3200 family of microcontrollers will be sampling in April with volume shipping in Q3 2008. Pricing for the LPC3200 ranges from $6.95 to $8.25 in 10 Ku volumes. NXP Semiconductors, San Jose, CA. (408) 474-8142. [www.nxp.com].

Crossware Adds Memory Management Unit Support to Arm Debugger Crossware has added support for the ARM9 memory management unit (MMU) and fast context switch extension (FCSE) to its ARM Development Suite. This allows developers to rapidly exploit the advanced memory management features of ARM9 chips such as the Atmel AT91SAM9263 and Atmel AT91RM9200 microcontrollers. Full debugging is possible when the memory management unit is enabled even when complex virtual to physical memory translation is being used. The Crossware debugger understands the memory address translation process and can determine which physical memory is being used for different virtual addresses. Full debugging is also possible when the fast context switch extension is being used. The FCSE allows different processes to easily share the same memory space. The Crossware debugger knows which executable code belongs to each process and can select the appropriate debugging information even though the processes share the same execution address. Crossware’s ARM simulator will also simulate the operation of the MMU and the FCSE. This enables code that uses these features to be tested on the developer’s PC before being tested on the target hardware. It also allows developers to explore the operation of the MMU and FCSE without having any hardware at all. The complex memory organization required when virtual memory translation is being used or when the fast context switch extension is in operation, necessitates the use of a linker script to specify the detailed code and data layout of the target system. In order to simplify the process of creating the linker script, Crossware has added a drag-and-drop GUI linker script editing facility to its Embedded Development Studio. This allows the developer to rapidly create and modify linker scripts and therefore avoid the need for a detailed understanding of the linker script command language. Crossware has created a simple example program that uses the MMU and FCSE. This runs in the Crossware simulator and on an Atmel AT91RM9200 microcontroller target board and demonstrates how easy it is to use these advanced memory management features from within the Crossware environment. A short video shows the drag-and-drop linker script editor being used, illustrating the simplicity of this approach to linker script creation. The Crossware ARM Development Suite provides a complete and extremely user friendly development environment for the ARM family of microprocessors and microcontrollers with its advanced C/C++ compiler, libraries, wizards, simulator, source level debugger and the Jaguar USB JTAG debugger interface. Crossware Products, Morgan Hill, CA. (360) 812-2397. [www.crossware.com].

MAY 2008

products for designers

Eclipse-Based IDE for Embedded Systems

second opinion CTO Interview: Lisa Su, Freescale Semiconductor

As semiconductor companies increasingly absorb the R&D costs of box designers, their R&D roadmaps need to reflect where their end markets are going. They are moving up the food chain, acquiring software and systems-level expertise in a wide range of fields. This has forced semiconductor firms to narrow and focus their product lines; some have reorganized or even reinvented themselves to better align with new market realities, which are particularly demanding in the portable consumer space. Few companies have given the problem more thought than Freescale, whose reliance on the handset market is reflected in the products that populate its R&D roadmap. Lisa Su is the senior vice president and chief technology officer for Freescale Semiconductor, responsible for Freescale’s technology roadmap and their global R&D operations. Lisa joined Freescale in June 2007 from IBM, where she was vice president of the Semiconductor Research and Development Center and was responsible for the strategic direction of IBM’s silicon technologies, joint development alliances and semiconductor R&D operations. Portable Design sat down with Lisa recently in her office in Austin to discuss Freescale’s pending products as well as where she sees the portable market going. Portable Design: I’m curious about MRAM. You’ve mentioned making MRAM into a commercial product, and Freescale is the first to be turning it out in commercial quantity. Where do you see it making the most impact, and where does it fit into your long-term plans? Do you see it as a universal memory, or does it have a certain niche? 46

PORTABLE DESIGN

by John Donovan, Editor-in-Chief

Su: Our philosophy on innovation has been to choose a few things and see if they can really become game changers. MRAM was one of those things that many people in the industry have worked on but no one else has been able to commercialize. The challenge with any of these new memory technologies is that they’re always competing against something that is already incumbent. MRAM from a technology standpoint has done very well. We are shipping commercial volumes today at 4 Mbits and we have announced a 1 Mbit version. MRAM is still in a relatively small niche application like battery-backup SRAM, but I think that’s a good place for it to enter. When I look longer term I think MRAM-specific areas for us are in the embedded space and particularly integrated with some of our microcontrollers; but that market has yet to develop. I don’t necessarily see it as replacing all flash, it would be difficult to do that. But there are some places where it does things that flash can’t do and where MRAM can provide product differentiation. Portable Design: An early promise of MRAM was the instant-on notebook computer; this is an area where flash has made some inroads. Do you see that as a viable possibility? Su: It’s a possibility. You just need to get to the right cost point. We look at it much more from the standpoint of the applications that MRAM enables, applications that require its true nonvolatile nature and the fact that it doesn’t degrade over time. I think the biggest challenge to overcome is really that flash has progressed so far. I don’t see MRAM replacing all of it, but there are some areas where it can provide value.

Portable Design: Two years ago Freescale announced that it was close to having a CMOS-compatible silicon nanocrystal solution. Where do you stand with that? Su: We still have a very active research program on silicon nanocrystals, and we’re working on it aggressively at the 90nm node. With any new technology you need to go through quite a bit of reliability testing and qualification and we’re going through that phase right now. I think it’s one of those very promising technologies. Portable Design: But then there’s always the chasm between the lab and the fab... Su: There is, and particularly in our marketplace. Our microcontroller business is very much automotive centric, and the very high reliability requirements of automotive requires that we do substantial reliability and quality evaluations. MRAM will also have applications in the consumer and industrial space as well, but our view is that it has to be good enough for auto, and that’s how we view any of these technologies. Portable Design: What other interesting R&D projects does Freescale have underway? Su: One of the things I love about Freescale is that since we have such a breadth of portfolio, there’s a lot that’s interesting. I would put things in a few categories. For our base technologies we are pretty much in partnership with some of the big players; we’re in the IBM alliance going from 45 to 32nm. We’re putting a lot of investment in two areas. The first is what we call differentiating

Su: And we are in that category. The question is, How do we make silicon systems easier to use? It’s more an ease of use discussion than anything else. One of the things we’ve been looking at is how do you get pin-compatible solutions. Our view is that you need software application compatible solutions. In our microcontroller business we’ve been pretty much focused on that.

Portable Design: It does seem fairly absurd to have a 45nm chip with tens of millions of transistors and wire bonds coming out at the end. That harks back to the old days of individual transistors. Su: You got it. It’s the idea that you can’t be working at small dimensions and have wire bonds that are tens and hundreds of microns across. It just doesn’t work. So now you get to the micron level using a lot of the technology that we’ve developed for the silicon side. So we believe this technology is very important; we like to think of it as the next evolution of packaging, and it’s important in our low-end microcontrollers and analog components because it will give us a better footprint, 30% lower cost and much more dense capability in these packages. So that’s another sample. Another area where we’re starting to invest a lot more is software. That’s really a view of where we see the market going. As our components get a lot more complicated, what’s limiting their adoption rate is really the number of lines of code that people can write. So more and more we’re attempting to provide a package of solutions, which includes software.

Portable Design: Let’s talk about RF interfaces. The number of RF interfaces is exploding in portable products. What are the major challenges you see in integrating these interfaces into silicon, and what is Freescale doing along those lines? Su: I’d generalize the problem to all types of mixed-signal interfaces. You have communications technologies, and you have the base digital technologies. It is becoming more and more important to integrate them. The issue is always, do you do it on a single chip or do you try to integrate to a system-in-package? I think the challenge of integrating everything on chip is that the standards actually are not so stable. And so what you find is that you may want to upgrade your processing element on a different timeline than your comms element. So I think there’s a place for both, the fully integrated on-chip as well as silicon-inpackage type of solutions. It really depends on the cost point in the marketplace that you’re trying to address. We’re attacking both very aggressively. The first one as I mentioned is with our chip-scale packaging technology. It’s really nice for what you’re talking about because you can take a very advanced 45nm digital chip and integrate it with a 130nm analog communications chip and be just fine. In addition to that there is innovation going on in terms of just getting to full singlechip solutions as well. I don’t think that’s a technical challenge frankly, I think it’s more from a business standpoint whether it makes sense to do that given where the markets are going.

Portable Design: I see more and more semiconductor companies essentially becoming systems houses. They’re hiring more software engineers than silicon engineers.

Portable Design: What emerging technologies do you see having the most impact on portable designs over the next few years?

Su: In portable designs the challenge right now is time-to-market. The ‘lab to fab chasm’ that you referred to has been true on the silicon side, and I think it’s now starting to appear more on the solutions side. That’s what we are forced to solve. What you’re going to see is that where you used to be limited by processing, you’re not anymore. The key is how quickly we can get products out there. So I think the key thing is going to be a much more sophisticated solutions infrastructure to really utilize the silicon technology. You have a software infrastructure that is able to quickly upgrade your devices from generation to generation, with the idea of really trying to put that into place in emerging markets. And so you look at China as a place where there can be a lot of experimentation; you don’t need experts to design these things anymore, you can do it with a relatively simple infrastructure. That’s what I see in portable designs. I’d like to see how we as semiconductor providers can provide a much more complete solutions environment. Portable Design: The more you have to do that, the more you’re putting yourself on the line because you have to guess what customers are going to want and hope you’re right. Say you’re developing a chipset for a cell phone, you’re going to have to guess 12 months out. Under the best of circumstances it may take you six months from concept to completion on the chip alone. If you’re wrong you can really be out of luck. Su: That puts the challenge on us. One, you do have to be ahead. But two, I think you need to be a lot more flexible in how you do designs. Portable Design: It almost sounds like the argument the programmable logic folks have been making for years. Su: It’s almost like that. What you’ll find is the importance of being much more flexible. The key is you don’t have to do it all in hardware. Freescale Semiconductor, Austin, TX. (800) 521-6274. [www.freescale.com].

MAY 2008

second opinion

technologies. You’ve mentioned two of them, MRAM and silicon nanoocrystals. We also have a very active effort in chip-scale packaging called ‘redistributed chip packaging’. It’s some pretty neat stuff because we look at where the roadmaps are going in the future. There’s been so much investment on the silicon side, but when you really look at cost and size in consumer markets you find that you’re very much limited by packaging; in some cases the packaging is more than half of the overall product cost. You also find that as you’re trying to make these devices smaller and smaller, you’re limited by how much space you have on a board. So we’ve been working on this technology doing wafer-level packaging using a lot of silicon techniques. We have a 300 mm panel where we’re actually building the package on the panel instead of doing things on strips.

ceo interview John East Actel

With Xilinx and Altera dominating the programmable logic space, you’d think that smaller players like Actel would be an endangered species. You’d be wrong. Actel shipped its first product twenty years ago, an antifusebased FPGA. They’ve long been strong in rad-hard mil/aero applications where SRAMbased products don’t hold up (though margins do). While not abandoning antifuse, Actel has moved aggressively into the consumer market with the introduction of its flash-based IGLOO family of FPGAs, which it claims are “the industry’s lowest power programmable solution.” While still hardly a Goliath, right now their sales and stock are doing nicely in a down market, thank you. Actel’s CEO John East is a true believer about low power, and not just in chips. Along with Mark Thompson at Fairchild and T.J. Rogers at Cypress, East believes that energy conservation on a larger scale—going “green”—is also good business. Portable Design’s Editor-in-Chief John Donovan talked with East at ESC in April and asked him about the logic behind programmable logic as well as green technology. Portable Design: We see a lot of CPLDs in low-power portable designs—for example, Xilinx’s CoolRunner—but I’m yet to see much FPGA penetration except for limited production runs. Do you see your flash-based FPGAs going where no 48

PORTABLE DESIGN

SRAM-based FPGA has gone before—and if so, where? East: We do see that, without any doubt. In fact our most recent .13µ flash family is already our highest volume [line] by a considerable margin. A lot of that’s going into Asia, so you might not have had a chance to see the products that are using them yet. Where is it going to go? In the long run it will probably go everywhere, because there’s a worldwide need to use less energy. We’ll start off seeing [flash-based FPGAs] in circuits that are battery-powered, in applications that are portable and small. But I believe it will spread pretty well everywhere, because low power should not be constrained to portable applications. We all need to be worried about power. Portable Design: At last October’s Portable Design conference, Martin Mason made the case for using your Fusion devices for smart battery management; you even offer a Smart Battery Evaluation Board. Do you plan to offer customized versions of your products to different vertical markets? East: We don’t as yet have plans to do custom hardware, custom silicon. We talk about it often, and I wouldn’t be surprised if sooner or later one day we do that. But right now the customization is with respect to the applications. We have done a lot of work with power management. We’ve also done quite a bit of work with battery charging. We’ve done a lot of work with motor control; there’s a lot of power wasted by motors or motor controllers that are not particularly efficient. So we have a lot of custom applications work, but it’s not done on custom silicon, at least not yet.

Portable Design: As more and more functions in handsets migrate into SoCs, aren’t you concerned that your chips will be squeezed off the boards? East: I see it going the other way. There was a time when it was hard to get arrested if you were selling a portable device for a handheld application, because of the cost, the power, those sorts of things. Now I see more and more interest instead of less. There is an ever widening array of processes, control panels, storage, displays and keyboards that are being used. The issue today is that no processor that I’m aware of has a native interface to all the different choices with respect to peripherals. So there is a need there for interface circuitry that can be implemented easily, quickly, inexpensively and with low power by using an Actel IGLOO circuit. So we’re seeing a lot of interest. I don’t think it’s going to decrease over time, it will increase. Portable Design: Like your friend Mark Thompson of Fairchild, you’ve long been an advocate of ‘green technology.’ What is your definition of green technology and what are you doing about it at Actel? East: Let’s take a step back here. A year or two ago people had to preach that there was global warming; in fact there were a lot of skeptics, but I think the skeptics are now gone; people are aware that there is global warming and that it’s caused by humans. You can argue about degrees and about the long-term effects—these are rational arguments—but humans are causing global warming, so we just need to create less temperature rise; we need to generate fewer greenhouse gases.

There are two ways to do that. One is by generating the energy more efficiently, and the other is by using less energy. A lot of the effort over the first half decade of taking this seriously has gone into more clean sources of energy. But if you question each of those sources ,there is a gaping hole in each one of them with respect to being able to turn the situation around in a decade or so. I think nuclear is a clean source, but it takes forever to get these things built and they’re not even legal in many parts of the world. You can talk at length about generating power by corn or sawgrass or other things, but if you look at the math of that it turns out not to be very helpful. You can talk about electric cars; that’s not a source of generation but a source of saving power. What you don’t think about is, ‘How do you charge a battery in an electric car?’ You either turn on the gasoline engine, which charges up the battery and pollutes, or if it’s purely electric, you drive it home, you plug it in, and then your local utility will turn on generators that they run with fossil fuel. So there are problems in generating all the energy we need more cleanly, and unfortunately, we’re needing more and more energy as time goes on. So to me the short-term solution is, Let’s just use less! You then say, ‘Who are the culprits, who is using energy?’ Roughly half of the greenhouse gases come from cars, so you will want to swing to electric cars. The other half of the energy is generated for electric motors. They tend not to be very efficient. The fastest-growing consumer of electricity, which soon will be huge, is electronics. So what do I mean by clean? Well, we all need to help with legislation that dictates that power be generated more cleanly. But between now and then we have to figure out ways to use less of it. In my field that means low-power circuits. You can get the same job done today with a low-power circuit as you can with the old standard or high-power

circuit. Use the low power; it will make a big difference. There are other things we can do. I now have a solar installation on my roof, and we’re in the early stages of installing a 220 kW system on Actel. My own solution to the car problem is that if you get an electric car, and you have solar on your own roof, then you charge your own car with your own solar power and you do not pollute. I’m waiting for the Tesla, which is a very cool looking electric car. I’m going to buy one and I’m going to charge it from a solar system that I’ve already installed. And that, my friend, is what I mean by green. Portable Design: Does green technology, while admirable, make good business sense? East: Sure—though all things in moderation. If you are trying to build businesses on ‘green at any cost,’ it’s probably not wise. But if you’re trying to build a business as I am based on ‘I can make a low-power circuit with the same features and at the same cost when somebody else is making a high-powered version,’ why isn’t it going to be good business for me? I think the world is ready for that, and I think it will be a great business. Properly chosen, I think there are terrific opportunities in green. Portable Design: How do you see the programmable logic industry changing over the next 3-5 years? East: My view is that in the early days, and the medium days of program logic, it was based pretty much on telecom, and more specifically on the data paths in big telecom systems—big switches, big add/drop muxes, big cross-connects, base stations, communications systems. That’s a huge market today for programmables, nobody is going to turn those design wins back in. But it appears to me as though we’ve reached sort of a static equilibrium; for every new win that comes along

an old one goes away, so it’s been hard to get growth in that area. I also think the conventional CPLD area is fraught with peril these days; similar logic, but the conventional 2000-gate or 2000-macrocell CPLD—high-power, good speed—this is pretty much a saturated market. So in my view the growth in the programmable logic market is going to come in the consumer and other handheld and batterypowered systems that I see proliferating around the world. And I think the geocenter of that is going to be in Asia, probably China. I think you’ll see all the programmable logic people spending a lot of time in China. A lot of the action is going to be there in these consumer and low-power applications. Portable Design: How do you see Actel changing over the next few years? East: We’ve bet everything on a couple of market shifts, a couple of sea changes. The first market shift I see is toward low power, and we want to say that we were there first. We’re putting all our eggs in that basket. I do think eventually everybody will get on board that wagon, and the competition will steepen. At that time we will be looking for yet another sea change, and I predict that second sea change will be analog/mixed-signal devices, particularly programmable ones, because as we’ve seen in the past, people prefer programmable circuits. If two circuits are available and one is programmable and the other is not and they do roughly the same job, the programmable one wins out. So I see in the market over the next few years big progress in low-power, consumer-oriented FPGAs. Following that I think you’ll see a major shift toward mixed-signal FPGAs. Actel Corporation, Mountain View, CA. (650) 318-4200. [www.actel.com]. Watch the video of this interview at www.portabledesign.com/news/251

MAY 2008

The RTC Group is a media services company specializing in bringing companies and their products to a focused group of electronic and computer manufacturers. RTC is proud of its track record of blazing new trails in search of marketing value for our clients. Portable Design magazine is the newest addition to RTC Groupâ&#x20AC;&#x2122;s collection of publications.

advertiser index Actel Corporation

www.actel.com

event calendar

Altera Corporation

www.altera.com

05/28-30/08 MicroTCA Summit East

Design Automation Conference - DAC

www.dac.com

embedded Community

Express Logic

Chantilly, VA www.microtcasummit.com

06/16-19/08 NXTcomm08 Las Vegas, NV www.nxtcommshow.com

06/17/08 (NEW date) Real-Time & Embedded Computing Conference Denver, CO www.rtecc.com/denver2008

06/24/08 EDA Tech Forum Moscow, Russian Federation www.edatechforum.com

09/18/2008 (NEW event) Portable Design Conference & Exhibition San Jose, CA www.portabledesignconference.com

If your company produces any type of industry event, you can get your event listed by contacting

sallyb@rtcgroup.com This is a FREE industry-wide listing.

PORTABLE DESIGN

Linear Technology

2,52

www.embeddedcomunity.com

www.expresslogic.com

www.linear.com

Linx Technologies, Inc

www.linxtechnologies.com

Maxim Integrated Products

www.maxim-ic.com

www.mentor.com

www.micrel.com

Mouser Electronic

www.mouser.com

NXTcomm

www.NTXcommShow.com

Qualcomm MEMS Technologies

www.qualcomm.com

Mentor Graphics

Micrel Semiconductor

MAKING ONE OF THESE?

...THEN YOU NEED THREADX ThreadX is Express Logic's small, fast, royalty-free RTOS that powers over 500 million electronic devices, with millions more produced each month. That's because ThreadX is easy-to-use, making your development job easier and more likely to ﬁnish on time or even ahead of schedule. Developers of consumer, medical, networking, industrial, aerospace, and automotive electronics products rely on ThreadX for their RTOS. Call today to ﬁnd out how ThreadX can help you bring your next electronic product to market faster. Make it better - make it with ThreadX.

T H R E A D

• Small footprint • Low overhead • Full source code • Easy-to-use • Field-proven • Low cost • Royalty-free

For a free evaluation copy, visit www.rtos.com , 1-888-THREADX Order today on Amazon Real-Time Embedded Multithreading Using ThreadX and ARM by Edward L. Lamie Copyright © 2008, Express Logic, Inc. ThreadX is a registered trademark of Express Logic, Inc. All other trademarks are the property of their respective owners.

Pick the Rails You Need

Optimized Topologies: Buck-Boost, Buck, Boost & LDO Power designs for multiple output rails in battery-powered portable devices are challenging – especially when these rails can be above, below or equal to the input supply. Linear’s new family of multiple output and multitopology converters enable straightforward and compact solutions. Our family of DC/DC regulators features up to four outputs with conversion topologies that include buck, boost, buck-boost and LDOs. Furthermore, most provide 95% efficiency, fast transient response and 2.25MHz switching in tiny QFN packages.

Info & Free Samples

Selected Multifunction Output ICs Part No.

Buck (IOUT)

Boost (ISW)

300mA + 200mA + LTC 3544 200mA + 100mA 600mA + 600mA + LTC3562 400mA + 400mA 600mA LTC3445 400mA LTC3670/72 250mA LTC3100 800mA 1A LTC3446 500mA LTC3541 800mA + 800mA + LTC3545 800mA 200mA LTC3522 600mA LTC3520 LTC3537 600mA 400mA LTC3523 600mA LTC3527 600mA + 400mA 300mA + 300mA LTC3547 600mA + 600mA LTC3419 800mA + 400mA LTC3548 LTC3407A-2 800mA + 800mA 1.5A + 1A LTC3417A-2

Buck-Boost (IOUT)

LDO (IOUT)

Package (mm) 3 x 3 QFN-16

www.linear.com/multioutputregs 1-800-4-LINEAR

3 x 3 QFN-20 50mA + 50mA 4 x 4 QFN-24 150mA + 150mA 2 x 3 DFN-12 / 2 x 2 DFN-8 100mA 3 x 3 QFN-16 300mA + 300mA 3 x 4 DFN-14 300mA 3 x 3 DFN-10

Free Portable Solutions Brochure

3 x 3 QFN-16 400mA 1A

LDO Controller 100mA

3 x 3 QFN-10 4 x 4 QFN-24 3 x 3 QFN-16 3 x 3 QFN-16 3 x 3 QFN-16 2 x 3 DFN-8 3 x 3 DFN-10, MS10 3 x 3 DFN-10, MS10E 3 x 3 DFN-10, MS10E 3 x 5 DFN-16, TSSOP-20E

www.linear.com/portsolutions

, LTC and LT are registered trademarks of Linear Technology Corporation. All other trademarks are the property of their respective owners.