FEATURED The magazine of record for the embedded

FEATURED

in this issue

The magazine of record for the embedded computing industry

June 2007

www.rtcmagazine.com

MEDICAL SYSTEMS Pack power and portability

National Instruments’

Fred Ruegg:

"We strongly believe in VXS." An RTC Group Publication

PCI Express • Fast Lane to the Future Smart Sensors = Smart Networks For Testable Software • Automate!

GE Fanuc Embedded Systems

Buy some time.

Find the AdvancedMC™ products you’re looking for at GE Fanuc Embedded Systems.

We took a leadership position in AdvancedMC™ products right from the beginning, and we’re continually adding new products. Whether you’re building a MicroTCA™ system or adding features to an AdvancedTCA™ blade, there’s a good chance we have the AdvancedMC you need.

In I/O we have the traditional WAN products, including a card that offers iTDM capabilities. We have Fibre Channel cards, including a 4Gb optical card. We also have a SATA hard disk storage card. And a single board computer based on the Intel® Pentium® M processor. We have a

Cavium-based packet processor. We even have a GPS clock, and a VGA card. Not to mention our carrier cards for AdvancedTCA and IBM® BladeCenter™.

When time is short and deadlines are looming, start you search with our terrific line of AdvancedMCs and quickly assemble your system components. That way, you won’t just be buying boards, you’ll be buying time.

Telum™ NPA-38x4 High-performance AdvancedMC packet processor

www.gefanucembedded.com/amc

© 2007 GE Fanuc Embedded Systems, Inc. All rights reserved.

Departments 5

Editorial: Say, Didn’t the Automobile Used to be All About Freedom?

7

Industry Insider

www.rtcmagazine.com

62 Products & Technology

T RX-EYE 0.25 0.20 0.15 0.10 0.05 0.00 -0.05 -0.10 -0.15 -0.20 -0.25

VRX-EYE

0

20

40

60

80 100 120 140 160 180 200

Time (ps)

Features Technology in Context

Receiver Eye Margin. • Pg. 10

PCI Express

10 P CI Express 2.0: The Next Frontier in Interconnect Technology Ali Jahangiri, PLX Technology

Solutions Engineering

I/O and Sensor Technology

14 D eploying Standards-Based Wireless Sensor Networks Brian Bohlig, Arch Rock

20 F rom FPDP to VPX: Back-End Management and Processing of Sensor-Derived Data Robert Nokes, GE Fanuc Embedded Systems

Industry Insight Embedded Choices for Medical Systems 26 M ulticore Blade Servers Help Advance 3D Ultrasound Technology Frank Setinsek, TechniScan Medical Systems

The ultimate goal of TechniScan’s UltraSound CT Imaging System is to differentially characterize normal, benign and malignant tissue to help physicians decide on appropriate clinical management. • Pg. 26

30 C hoosing the Processor Brain to Control Heart, Lungs P.J. Tanzillo, National Instruments

Executive Interview 36 RTC Interviews Fred Ruegg, President, Elma Americas Special Supplement: Automation and Control in a Networked World 42 T he Guts That Make It Go: Processor Control Modules Get Smaller, More Connected Ann R. Thryft

48 C onnecting the Brains: Connecting Embedded Devices Moves Toward an All-IP World Tom Williams

Software & Development Tools Embedded Data Management 56 E mbedded Software Development Needs a More Automated, Test-Driven Approach Mark Underseth, S2 Technologies

WinSystems’ LBC-GX500, a highly integrated SBC designed for machine-to-machine connectivity, features a wide variety of wired and wireless options. These include 802.11 wireless Ethernet, GSM/GPRS cellular modem, CDMA cellular modem, ZigBee wireless RF module, 10/100 wired Ethernet port, global-compliant dial-up modem, six USB ports and six COM channels.

• Pg. 42 June 2007

June 2007 Publisher PRESIDENT John Reardon, johnr@r tcgroup.com EDITORIAL DIRECTOR/ASSOCIATE PUBLISHER Warren Andrews, warrena@r tcgroup.com

Editorial

EDITOR-IN - CHIEF Tom Williams, tomw@r tcgroup.com SENIOR EDITOR Ann Thr y f t, annt@r tcgroup.com MANAGING EDITOR Rebecca Bauer, rebeccab@r tcgroup.com COPY EDITOR Rochelle Cohn

Art/Production

CREATIVE DIRECTOR Jason Van Dorn, jasonv@r tcgroup.com SENIOR GRAPHIC DESIGNER Kirsten Wyatt, kirstenw@r tcgroup.com GRAPHIC DESIGNER Christopher Saucier, chriss@r tcgroup.com DIRECTOR OF WEB DEVELOPMENT Marke Hallowell, markeh@r tcgroup.com WEB DEVELOPER Brian Hubbell, brianh@r tcgroup.com

Advertising/Web Advertising WESTERN REGIONAL ADVERTISING MANAGER Lauren Hintze, laurenh@r tcgroup.com (949) 226 -2014 EASTERN REGIONAL ADVERTISING MANAGER Nancy Vanderslice, nancy v@r tcgroup.com (978) 443 -2402 EMEA SALES MANAGER Marina Tringali, marinat@r tcgroup.com (949) 226 -2020 BUSINESS DEVELOPMENT MANAGER Jessica Grindle, jessicag@r tcgroup.com (949) 226 -2012

Billing

Mini-ITX systems & solutions for embedded applications, industrial & mobile computing.

Mag gie McAuley, mag giem@r tcgroup.com (949) 226 -2024

Mini-ITX Mainboards with VIA, Intel, or AMD processors. x86 platform small form factor design Linux and Windows XP compatible fully customizable systems

To Contact RTC magazine: HOME OFFICE The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Phone: (949) 226-2000 Fax: (949) 226-2050, www.rtcgroup.com EASTERN SALES OFFICE The RTC Group, 96 Dudley Road, Sudbury, MA 01776 Phone: (978) 443-2402 Fax: (978) 443-4844

Fanless Mini-ITX Systems Utilizing heat pipe technology, these completely fanless Mini-ITX systems offer durability, reliability, and power-efficiency for a wide range of embedded applications in harsh, remote environments.

LOGICSUPPLY.COM info@logicsupply.com 802 244 8302

June 2007

LogicSupply_3.5x4.75.indd 1

4/16/2007 3:10:17 PM

Editorial Office Warren Andrews, Editorial Director/Associate Publisher 39 Southport Cove, Bonita, FL 34134 Phone: (239) 992-4537 Fax: (239) 992-2396 Tom Williams, Editor-in-Chief 245-M Mt. Hermon Rd., PMB#F, Scotts Valley, CA 95066 Phone: (831) 335-1509 Fax: (408) 904-7214 Ann Thryft, Senior Editor 15520 Big Basin Way, Boulder Creek, CA 95006 Phone: (831) 338-8228

Published by The RTC Group Copyright 2007, 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.

Editorial June 2007

Say, Didn’t the Automobile Used to be All About Freedom?

by Tom Williams, Editor-in-Chief

O

K, now my car knows me. I walk up to it and reach for the door handle and it unlocks for me. I get in, push the “on” button, it starts and away I go. Pretty soon they’ll all be like this, but wait! There’s more! If you think we are knitted into an inextricable network with our cell phones, BlackBerries, laptops and mobile Internet devices of all kinds, you ain’t seen nothin’ yet. Wait until the Vehicle Infrastructure Integration initiative achieves full deployment.

Vehicle Infrastructure Integration (VII) is aimed at providing communication links between vehicles on the road and between vehicles and a roadside infrastructure over a 5.9 MHz dedicated short-range communications (DSRC) link. Connection to the VII network will take place via roadside units placed along major highways at first, but doubtless eventually spreading to the whole nation. Each vehicle will be equipped with a suite of onboard equipment (OBE) consisting of a GPS unit, a radio module and a set of standard and optional devices and applications.

The VII network will—as presently conceived—be its own “cloud” based around the roadside units, but it will also have public and private gateways to the Internet. The stated goal of VII is to increase transportation safety and efficiency. Indeed, the potential benefits are many. Communication between vehicles (be careful what you shout at the other guy with an open mike) can alert drivers to potential rear-end collisions, speed alerts for upcoming curves, approaching speed of other vehicles and the like. Communication with roadside units can alert drivers to road conditions and provide real-time traffic information to the entire network. That could result in things like intelligent intersections where the timing of the stop lights is automatically adjusted based on the volume and flow of traffic, commuter lanes are turned on and off, signals are sent to intelligent highway signs, etc.

There is a huge potential for optional services such as point of interest information and the location of services. Obviously, this includes the opportunities for advertising and many services that have not been dreamed up yet. Then there is the emerging market for equipment OEMs, both for the roadside units and the OBE. The radio and networking technology will be unique on

some layers and standard TCP/IP on others. The installation of GPS units will be two-way in order to enable the location of individual vehicles by emergency services. And here we get into the sticky area…

So far, the people I have spoken to who are presenting this technology seem to bristle at the suggestion that there could be a Big Brother element to all of this. “Well, that is possible,” they say, “but the goal of the system is road safety and efficiency.” Yeah, right. We all know that the data will not be anonymous, no matter how much we are assured it will be. It is too attractive for use by law enforcement to do things like locate stolen vehicles—something everybody can support. But that means that anyone with access to the network can access it to track neighbors or spouses or to enable the creep of government monitoring of peoples’ lives.

There is also potential for automatic ticketing or even the ability to override driver control to slow down or stop a vehicle. Arguments for this can be made from a safety perspective, but technology does not make policy decisions. What is available to civil authorities is also potentially available to hackers or to terrorists.

Already, Eastern Bloc nations with less concern for privacy and civil rights are testing similar systems.

Test sites have already been established and pilot projects are underway in California, Florida, Michigan and Minnesota. Field operational testing is due to be completed and a decision to deploy made by late 2008 with first installations of access points beginning in 2009. OEM equipment in vehicles could be expected to appear by 2011, ramping to all vehicles by 2016. All this amounts to a huge market opportunity, a technology challenge and the potential for great benefits as well as for potential abuse. It is not yet clear that all the issues surrounding installation (retrofit older vehicles? How to build out the roadways) and security have been adequately addressed to date. In the latter regard, security means, in addition to network security, the physical security of roadside access points. They can be run into, vandalized or otherwise tampered with if not adequately protected. In any event it looks like that old cliché about the Chinese word that means both crisis and opportunity. In other words, it’s up to us what we make of it.

June 2007

con:card+

70

the most reliable AdvancedMC connectors

A new standard in reliability con:card+ is a new quality seal that identifies connectors possible offset between connector contacts and module pads providing the highest level of mating reliability for by 60%, thus significantly increasing the mating reliability of con:card+MicroTCA EvenTMunder high and shock and vibration, con:card ® backplanes AdvancedTCA carrier blades. + connectors AdvancedMCTM modules.

will keep the AdvancedMC™ modules in the correct position ™ ™ con:card+ connectors feature the unique GuideSpring By preventing the connector, con:card+ connecmodules. helping tomodule assurefloat the in reliability of your MicroTCA backplane technology. The GuideSpring pushes and holds the tors also enhance the reliability of AdvancedMCTM systems under ® and carriade.

AdvancedMCTM module against the connector housing. shock andAdvancedTCA vibrations. ™ module precisely within the slot prior to By systematically positioning the module precisely in the connector,™con:card+ connectors reduce the maximum

system tremendously.

innovation by +

HARTING + ept – at

www.concardplus.com

Learn more about the 5 con:card+ features at www.concardplus.com

Industry Insider

June 2007

ATCA Deployment Finally Gaining Momentum According to a study by Venture Development Corporation, deployment of merchant AdvancedTCA (ATCA) integrated platforms, configured as full systems, is beginning to increase its growth and momentum. The deployment of ATCA platforms and associated growth of the market were constrained by major obstacles during the period of 2004-2006. Each has been either overcome or is far along.

The first obstacle was the constraint associated with different technical interpretations relating to the ATCA specification. There were key questions surrounding hardware and software interoperability between products from different suppliers. Significant progress has been made to resolve these interoperability concerns. The interaction and development of a fuller ecosystem between critical special interest organizations such as PICMG, the SA Forum, CP-TA, SCOPE and the Mountain View Alliance have created an important dialog between users and suppliers to solve these issues.

Total Market, Merchant ATCA Integrated Platforms, 2006-2008 (US$ in Millions) 2006 - 2008 CAGR: 81% 400

$ 339.8

300

200

$ 198.4 100

$ 104.8 0

2006

2007

2008

The second obstacle surrounded the implications of ATCA as a standardized communications platform associated with value realized. Within its target market of Tier 1 Network Equipment Providers (NEPs) and Telecommunication Equipment Manufacturers (TEMs), ATCA has engendered a great deal of central engineering and management discussion. Many of these organizations needed to confront how best to deal with standardized and proprietary platform architectures and the implications for engineering manpower investments. It has been more difficult for the Tier 1s versus the Tier 2s and Tier 3s since the former have invested more heavily in hardware platform engineering.

Finally, ATCA requires supporting subsystem and component technologies. These include such diverse areas as processors, switching fabrics, chassis and operating systems. The hurdles of price and technical maturation for enabling technologies such as Gigabit and 10Gigabit Ethernet and multicore processors added to delays within ATCA evaluation and adoption. As the capabilities of these subsystems became better understood their adoption for deployment has increased.

According to VDC’s study titled, “AdvancedTCA and MicroTCA Components and Solutions: Global Market Demand Analysis, 2nd Edition,” 2006 shipments of Merchant ATCA Integrated Platforms were at $104.8 million and forecast to reach $339.8 million by 2008. This indicates that NEPs and TEMs are now moving from the evaluation and pilot stages to selected production and development stages for ATCA.

Automotive Consortium to Develop Equipment and Infrastructure for Vehicle Safety

The Vehicle Infrastructure Integration Consortium (VIIC) was established in 2004 to support the National VII Coalition effort to determine the feasibility of nationwide deployment of a Vehicle Infrastructure Integration (VII) program, and to establish a strategy for implementation, communications standards and capabilities. VIIC’s Onboard Equipment initiative will enable a standards-based communications infrastructure that supports vehicle-to-infrastructure and vehicleto-vehicle communications in an effort to improve vehicle safety, vehicle mobility and enable consumer and commercial services.

The National VII Coalition consists of the U.S. Department of Transportation, ten State Departments of Transportation and light vehicle manufacturers. The VIIC’s members are: BMW of North America, LLC, DaimlerChrysler Corporation, Ford Motor Company, General Motors Corporation, Honda R&D Americas, Inc., Nissan North America, Inc., Toyota Motor Engineering & Manufacturing North America, Inc. and Volkswagen of America, Inc. The VIIC will use Wind River Systems’ expertise in open source and automotive solutions to deliver a board support package (BSP) and configure an emulation platform to accelerate development of OBE. In addition, Wind River’s global support organization will deliver local support, enabling the decentralized development and field testing for the initiative’s disparate members. VIIC has selected Wind River’s General Purpose Platform, Linux Edition as the platform for developing the Vehicle Onboard Equipment (OBE) for their proof-of-concept activities.

Vehicle OBE delivers an OSGi/Java-based application host June 2007

Industry Insider platform; vehicle interface, human/ machine interface (HMI) and global positioning services; and embedded Dedicated Short Range Communications (DSRC) radio, WAVE stack and Java communications API. If deployed, VII will require the build-out of networks, digital radios, pods and communications systems on major US roadways. Such a deployment of roadside infrastructure may incorporate various Intelligent Transportation Systems (ITS) technologies into the transportation network and integrate ITS communications and sensors in vehicles.

create a highly flexible core computing and connectivity solution for communications, military/government, medical and industrial control markets. The securePMC series security board is a security offload solution designed for integration into Linux-based systems.

Advanced encryption processors accelerate SSL and IPsec cryptographic operations, significantly improving security, performance and availability of applications.

One Stop Systems Completes Purchase of SBE, Inc.’s Embedded Products Division

Fluffy Spider Technologies, a provider of software development platforms for Graphical User Interfaces in embedded devices, and 3ivx Technologies (3ivx), specialists in MPEG-4 video and audio software, have announced a joint research, development and marketing alliance.

FancyPants is a platform, created

by FST, for embedded application development. It enables OEMs to quickly and easily develop innovative devices with advanced Graphical User Interface capabilities and multi-media, while also reducing hardware resource requirements and development costs.

FST and 3ivx are integrating the 3ivx MPEG-4 software in the FancyPants platform, which will allow developers of products such as smart phones, TV set top boxes, point-of-service terminals, in-car systems and building automation devices to leverage the power of stateof-the-art MPEG-4 compression as part of the FancyPants platform.

The alliance also offers OEMs the option of outsourcing the complete development lifecycle of these next generation products. FancyPants is easily integrated into embedded operating systems and micro-kernels and meets the demands required in embedded environments.

One Stop Systems has completed the acquisition of the Embedded Products division of SBE. As a provider of Compact PCI systems and PCI Express-based cable expansion products, One Stop Systems designs and manufactures many board-level and packaging products for the industrial computing marketplace. This acquisition expands One Stop Systems’ family of high-speed communication board-level products. Newly acquired communication products include Wide Area Networking (WAN), Local Area Networking (LAN), Carrier and Security board products. The WAN solutions are designed to support a variety of I/O interfaces and hardware architectures. They focus on satisfying the need for WAN interfaces in data communications products residing between the middle and the edge of the Internet used in routers, firewalls, virtual private networks (VPN) servers and Voice over Internet Protocol (VoIP) gateways.

The carrier boards include HighWire series of communications controller products providing high-bandwidth intelligent connectivity to servers designed to act as gateways and signaling points within communication networks and network devices. For embedded Linux applications, HighWire products can be coupled with our WAN and LAN PCI mezzanine cards or other third-party PMC products to

June 2007

Alliance Combines MPEG-4 with GUI and Multi-Media Platform

Wind River and CurtissWright Expand Strategic Partnership with Linux Offering

Wind River Systems and Curtiss-Wright Controls Embedded Computing have an-

nounced that Curtiss-Wright’s Linux Center of Excellence, located in San Diego, will support Wind River General Purpose Platform, Linux Edition and the Wind River Real-Time Core for Linux solution as their standard Linux operating environments for use on the company’s x86based platforms. Curtiss-Wright has long offered VxWorks as its standard out-of-the-box RTOS solution for its rugged COTS board product line. As a result of this expanded strategic partnership, embedded system integrators now will be able to more easily and rapidly deploy Wind River Linux and VxWorks solutions for rugged deployed military systems, significantly mitigating their risk and speeding their time-to-market.

Curtiss-Wright’s COTS Continuum initiative supports interoperability across its product lines and enables customers and solution providers to leverage new computing technologies without the cost and risk associated with adopting new software and hardware interfaces. Curtiss-Wright expands by offering Wind River Linux and Real-Time Core on its Intel x86based board products. With the addition of Wind River Linux and Real-Time Core as standardized solutions, Curtiss-Wright is better able to offer its end-customers an integrated solution.

Offered as an add-on to Wind River Linux, Real-Time Core, acquired from FSM Labs, has been regarded in the software industry as a mature realtime Linux solution. Real-Time Core provides a guaranteed realtime executive that coexists with Wind River Linux. Together, they combine the guaranteed real-time responsiveness with the broad capabilities of the Linux operating system to enable mission-critical applications.

Mellanox Surpasses 2 Million InfiniBand Ports Milestone

An indication of the strength of the InfiniBand market was the announcement by Mellanox

Technologies that it has shipped over 2 million InfiniBand (10 and 20 Gbit/s) ports to leading server, storage, communications infrastructure and embedded system OEMs. These ports span four generations of InfiniBand adapter and switch families and are an integral part of the IT infrastructure that services a wide range of end-user markets, including automotive, bio and geosciences, communications, database, defense, economic and financial services, educational research, electronic design automation, entertainment, government, health services, retail, transportation services, weather analysis, Web services, and more. Leading OEM suppliers that deliver InfiniBand-based server and storage hardware and software solutions to these end-user markets include Cisco, DataDirect Networks, Dell, HP, IBM, LSI, Network Appliance, QLogic, SGI, Sun and Voltaire, among others.

“Shipping products that include over 2 million ports to leading OEMs is an important milestone for Mellanox and for the entire InfiniBand industry as it demonstrates the establishment of a significant end-user customer base,” said Eyal Waldman, chairman, president and CEO of Mellanox Technologies.

CD Recycling Center Calls for Support

The CD Recycling Center of America has announced a call for companies that use or distribute CDs and DVDs to join the Center’s disc recycling awareness campaign, which was launched this past Earth Day. The campaign promotes the importance of recycling compact discs and of increasing community awareness of related environmental issues caused by unwanted discs being placed in the trash. The Center invites consumers and companies to send in their unwanted CDs and DVDs for recycling.

When compact discs are placed in the trash, it is harmful

Industry Insider to the environment. When discs are recycled properly, it will help stop unnecessary pollution, conserve natural resources, and slow global warming. The plastic used in compact discs can be recycled into other everyday items, including household products, building materials and auto parts.

The Center asks that companies visit their recycling Web site, www.cdrecyclingcenter. org, and sign in as a supporter.

The Center encourages new members to place The CD Recycling Center logo on their own Web site and promotional materials, thus promoting to end users how and where to recycle unwanted discs when they are finished using them. Discs should be collected, boxed and sent to The Recycling Center.

The Center provides a logo, a plan, promotional materials and a place to send the discs.

Event Calendar 08/21/07

09/25/07

10/11/07

Real-Time & Embedded Computing Conference Longmont, CO www.rtecc.com/longmont

Real-Time & Embedded Computing Conference San Diego, CA www.rtecc.com/sandiego

Real-Time & Embedded Computing Conference Tyson’s Corner, VA www.rtecc.com/tysons

08/23/07

09/27/07

Real-Time & Embedded Computing Conference Colorado Springs, CO www.rtecc.com/coloradosprings

Real-Time & Embedded Computing Conference Long Beach, CA www.rtecc.com/longbeach

09/11/07

10/02-04/07

Real-Time & Embedded Computing Conference Ottawa, ON www.rtecc.com/ottawa

ARM Developers’ Conference Santa Clara, CA www.rtcgroup.com/arm/2007

09/13/07 Real-Time & Embedded Computing Conference Pointe-Claire, QC www.rtecc.com/montreal

10/03-04/07 NEW Portable Design Conference & Exhibition Santa Clara, 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.

HARTING sets new standards in AdvancedMC™ connector reliability HARTING offers the widest product portfolio of AdvancedMCTM and Power connectors for AdvancedTCA® and MicroTCATM applications.

HARTING now brings the reliability of signal connectors to new heights.

Introducing con:card+, a quality seal that identifies press-fit connectors that provides the highest level of mating reliability for AdvancedMCTM modules.

Untitled-4 1

The con:card+ GuideSpring systematically positions the AdvancedMCTM module precisely in the connector, reducing the maximum possible offset between connector contacts and module pads by 60%.

This significantly increases the mating reliability of MicroTCATM backplanes and AdvancedTCA® carrier blades.

HARTING provides complete design in support, including signal integrity services and 3D modeling.

5/30/07 3:26:38 PM June 2007

TechnologyInContext PCI Express

PCI Express 2.0: The Next Frontier in Interconnect Technology The widespread success of PCI Express continues with the rollout of the 2.0 specification. The spec brings with it wider bandwidth for users and tighter design conditions for developers.

by A li Jahangiri PLX Technology

d

exploration er your goal eak directly al page, the resource. chnology, and products

P

CI Express (PCIe) is now the de Application Layer facto standard for I/O in the server and PC interconnect arena. The PCI Special Interest Group (PCI-SIG) has recently released the updated PCIe 2.0 Base Transaction Layer Specification, which offers significant enhancements over its predecessor, PCIe Data Link Layer 1.1, at the physical, access control, software notification and system levels. This PIPE Interface was accomplished while maintaining full backward compatibility with PCIe 1.1 PHY panies providing solutions now hardware and software. PCIe 2.0 doubles ration into products, technologies and companies. Whether your goal is to research the latest Logical Layer PCIe 1.1’s rated speed, to page, 5 gigatransfers lication Engineer, or jump to a company's technical the goal of Get Connected is to put you perwhatever second effectively increasing ice you require for type(GT/s), of technology, Electrical Layer ies and productsthe youaggregate are searchingbandwidth for. of a 16-lane link to approximately 16 Gbyte/s. Gigatransfers are used to measure PCIe 2.0 bandwidth because PCIe uses 8b/10b encoding, whereby every eight bits are encoded into a 10-bit symbol.

Figure 1 PCIe layers and PIPE. PCIe is a layered protocol composed of physical (PHY), data link (DLL) and transaction (TL) layers (Figure 1). The though the PHY bit rate was doubled in layered architecture has provided PCIe PCIe 2.0, it did not affect the upper laywith a modular structure such that even ers. The Physical Interface for the PCI Express Architecture (PIPE) defines the interface between the PHY and the DLL. Get Connected with companies mentioned in this article.

PIPE inherently preserved its 10-bit/20-bit www.rtcmagazine.com/getconnected wide data path, while the frequency of the

End of Article

10

June 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

PIPE clock was doubled to accommodate 5 GT/s bit rates.

To maintain backward compatibility, the PHY in PCIe 2.0 has to operate at both 5 GT/s and 2.5 GT/s. (The software can also dynamically switch the link speed.)

These high frequencies pose a new set of signal-integrity challenges for the PCIe PHY: There are substantial differences between 2.5 and 5.0 GT/s transmitter (Tx) and receiver (Rx) timing, based on the need to account for additional jitter effects and differences in jitter budgeting methodology (Table 1).

Jitter is placed in two categories: random sources (Rj) and deterministic sources (Dj). Total jitter (Tj) is the convolution of the probability density functions for all the independent jitter sources, Rj and Dj. While the allocation to Rj and Dj was not specified in the PCIe 1.1, PCIe 2.0 now explicitly defines jitter tolerance for receivers.

The PCIe 2.0 transmitter phase lock loop (PLL) bandwidth and jitter peaking have to be tightly controlled, whereas PCIe 1.1 had sufficient margin to allow a wide (1.5-2.2 Mhz) PLL range. The key design factor on the Tx side is to control

TechnologyInContext

the jitter and noise sources. The PLL and REFCLK have to be designed with very low jitter. Power supply design also has to be optimized for noise reduction as the noise tolerance on the supply is around 30 percent less than it is for PCIe 1.1.

On the Rx side, the minimum receive eye voltage aperture, or VRX-EYE (Figure 2), defines the range over which a receiver must operate at all times. This range has been reduced to 120 mV (PCIe 2.0), from 175 mV (PCIe 1.1) (Table 1). Additionally, Rx jitter compliance, which requires a low-latency clock recovery loop, has been reduced by around 50 percent from that of PCIe 1.1. These changes have tightened the Rx budget significantly. To accommodate these to a certain extent, the PCI Express Card Electromechanical (CEM) 2.0 specification has reduced the allowable impedance to 85 ohms, from 100 ohms (1-2 connector channels). All in all, board layout, material variances and ISI effects will play an important role in determining the system jitter for PCIe 2.0 designs.

In PCIe 1.1, there was no notification mechanism if the intended link width changed because of hardware-autonomous link retraining. Such a downnegotiated link in a system can cause bandwidth throttling without software knowledge and adversely impact system performance. PCIe 2.0 addresses this issue by providing native support for generating an interrupt on link width or speed change. The interrupt notifies the PCIeaware software if link bandwidth (speed or width) changes due to link width re-negotiation. This robust mechanism enables the software to take corrective action in case certain lanes in a link fail. The software’s ability to dynamically change the link speed can be used to maintain control over bandwidth allocation.

Also added to PCIe 2.0 were access control services (ACS), which determine how a transaction level packet (TLP) should be routed. ACS also includes features such as source validation and peerto-peer (P2P) control. In source validation, the downstream ports validate the

Symbol

Parameter

2.5 GT/s

5.0 GT/s

Units

UI

Unit Interval

399.88 (min) 400.12 (max)

199.94 (min) 200.06 (max)

Ps

T TX-EYE

Transmitter Eye including all jitter sources

0.75 (min)

0.75 (min)

UI

T TX-HF-DJ-DD

Tx deterministic jitter > 1.5 MHz

Not specified

0.15 (max)

UI

T TX-UF-RMS

Tx RMS jitter < 1.5 MHz

Not specified

3.0

Ps RMS

T TX-RISE-FALL

Transmitter rise and fall time

0.125 (min)

0.15 (min)

UI

V RX-EYE

Receive eye voltage opening for Common Refelk Rx Architecture

175

120

mV

VRX-DIFF-PP-CC

Differential Rx peak-peak voltage for common Refelk Rx architecture

0.175 (min) 1.2 (max)

0.120 (min) 1.2 (max)

V

VRX-DIFF-PP-DC

Differential Rx peak-peak voltage for data clocked Rx architecture

0.175 (min) 1.2 (max)

0.100 (min) 1.2 (max)

V

Table 1

2.5 GT/s and 5.0 GT/s Transmitter and Receiver Specifications.

T RX-EYE 0.25 0.20 0.15 0.10 0.05 0.00 -0.05 -0.10 -0.15 -0.20 -0.25

VRX-EYE

0

20

40

60

80 100 120 140 160 180 200

Time (ps) Figure 2

Receiver Eye Margin.

Source: PCI Express Base Specification. REV. 2.0

June 2007

11

TechnologyInContext

requester ID in the TLP, after which the ID is directed to the root complex (RC).

Peer-to-peer controls determine whether to forward directly, block, or redirect peerto-peer request TLPs to the RC for access validation. ACS functionality is reported and managed via ACS extended capability structures. PCIe components are permitted to implement ACS structures in some, none, or all of their applicable functions.

Finally, a function level reset (FLR) mechanism is introduced. FLR enables the software to stop and reset individual functions within an end point. Such function-level granularity guarantees that external I/O operations performed by the card are stopped and the functionâ&#x20AC;&#x2122;s hardware returns to its default state on a FLR.

Graphics, Storage Applications to Gain from PCIe 2.0

Among the applications most likely to benefit from the performance gain in PCIe 2.0, graphics cards are going to take advantage of its power-limit value, which has been redefined to accommodate devices that consume higher power. Other protocols such as Serial ATA and Serial Attached SCSI standards are poised to move to 6 Gbits/s, from 3 Gbits/s, and in the future up to 12 Gbits/s. The primary advantage of PCIe 2.0 in these applications is that, to achieve similar performance, it would require half the number of lanes with half the latency than that of PCIe 1.1. This translates into less real estate, fewer routing resources and smaller form-factors. By extension, Ethernet controllers, InfiniBand and Fibre Channel could be serviced via faster system links.

The industry is gearing up and introducing PCIe 2.0-based systems in 2007.

Intel has released its Stoakley platform featuring the Seaburg chipset supporting PCIe 2.0 for the workstation market. AMD is set to release a trio of chipsets supporting PCIe 2.0: the high-end RD790+, the mid-range RX740+ and the budget RS740+. And nVidia is addressing the PCIe 2.0 market with its MCP72 single-processor chipset for the AMD HT-3 / AM2+ architecture.

Finally, I/O interconnect chip companies such as PLX Technology are expanding their PCIe 1.0 products, and developing PCIe 2.0 switching devices to meet ever-increasing demand for more ports and lanes.

In summary, one challenge designers could face in the deployment of PCIe 2.0 is signal integrity of the links, but this can be rectified with careful design and layout methodology. Additionally, software likely will play a more significant role in bandwidth and functionality management, with tools such as dynamic-link-speedcontrol, ACS and FLR. Despite these challenges, the one thing that is certain is that PCIe 2.0 will usher in a new generation of applications and systems with dramatically increased performance and improved response times.

PLX Technology Sunnyvale, CA. (408) 774-9064. [www.plxtech.com].

12

June 2007

Do Over?

FAST-FORWARD YOUR PROJECT WITH WINDOWS EMBEDDED. New devices mean new challenges. Speed your design to market with end-toend development tools backed by the long-term commitment of Microsoft ® and the support of the global Windows® Embedded partner community.

See how CoroWare reduced development hours by more than 60% vs. Linux at microsoft.com/embedded © 2006 Microsoft Corporation. All rights reserved. Microsoft, Windows, the Windows logo, and “Your potential. Our passion.” are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.

Done.

SolutionsEngineering I/O and Sensor Technology

Deploying Standards-Based Wireless Sensor Networks IP-based networking all the way down to the individual sensors makes wireless sensor networks more easily deployable, configurable and usable.

by B rian Bohlig Arch Rock

W

irelessly connected battery-operated servers transmitting data over a reliable mesh network to any Internet Protocol (IP) device in the world…reality or fantasy? As with any emerging technology, there’s been a fine line. But the last several years have seen significant developments in scalable standards, which are now pacing adoption and presenting wireless sensor networks in formats familiar to both IT and plant workers. Wireless sensors can be deployed quickly in an ad hoc fashion and used to report environmental changes in South American rainforests, ensure the efficiency of industrial processes in an oil refinery, determine how much power the blade servers in a data center are using, or tell you if your refrigerator is still as energy efficient as when it was purchased.

In the nearly 10 years that wireless sensor networks (WSNs) have been around, improvements in the “ingredients” have continued to push applications to the mainstream. Semiconductor technology continues to follow Moore’s law, providing smaller, more powerful and cheaper wireless devices. There is now an established and reliable low-power link (IEEE 802.15.4) supported by a growing number of quality CMOS radios. Furthermore, with the recent emergence of the IETF 6LoWPAN draft standard, WSNs—and even individual sensor nodes—for the first time look just like other Internet devices.

This is good news for users and manufac14

June 2007

Figure 1

Sensor Network Commissioning and Deployment in a manufacturing facility displaying the configuration information for the node titled “Entrance.”

turers alike. No longer is adoption gated by users having to learn another proprietary protocol that may or may not be around in a few years. IP is already the foundation of the world’s largest network, the Internet, so extending it into the world of low-power embedded devices makes perfect sense.

Users have complained that WSNs have been unreliable and difficult to use.

Furthermore, to put a sensor network to-

gether, a potential user or developer had to be adept in multiple disciplines—hardware, embedded software programming, RF and enterprise integration—further widening the gap between application concept and deployed network. What was needed was a way to abstract the complexity of setting up and commissioning a WSN from the ongoing management and data mining of the sensor data itself.

If WSNs could be made easily accessible

Intel. Igniting Innovation.

The Engines of Innovation Building on its storied legacy as the inventor of the microprocessor and microcontroller, today Intel offers one of the broadest lines of Intel速 processors, chipsets, standards-based processor boards and software components for the embedded market segment. With Intel速 multi-core processors serving as the building block foundation for industrial automation, interactive clients, automotive, communications and the newest in-vehicle infotainment applications, Intel remains the catalyst for innovation. The universal motor for electronics has now become the engine for change. www.intel.com/go/embedded

SolutionsEngineering over conventional IP-based networks, their potential user base would automatically become far vaster and more diverse.

A Logical View

Figure 2

Sensor Network Connectivity Display. This listing does not correspond exactly to Figure 1 because new nodes have been added. These can be seen as those listed by their MAC addresses because they have not yet been assigned names.

A key attribute of wireless sensor networks—and the reason they represent the future of intelligent embedded devices—is their ability to be deployed in diverse and varied physical-world environments. What has been lacking is a logical representation of that physical space on users’ computers.

With no computer-based map of sensor locations, users were often left to remember (or guess) where their sensors had been deployed. Only recently have sensor network applications emerged that bind the physical to the logical, allowing users to upload an existing floor plan, map or image into the WSN user interface and then see an individual sensor node positioned on that map. Once the nodes begin to monitor and collect data on a particular space, thing or interaction, the map provides context, meaning and the ability to easily manage the WSN deployment (Figure 1).

Deploying the Mesh Network

Figure 3

16

Web Services API and Response. The list of returns in the lower section is the response to the selection of option 4 in the top section.

June 2007

Today’s wireless sensor networks use mesh networking techniques to automatically determine the most efficient routing path from the sensors to a gateway server; that path may be direct or through one or more other nodes. The self-configuring nature of such networks makes WSNs flexible and surprisingly easy to set up. But all mesh networks are not created equal.

Many, for example, require powered, stationary nodes acting as packet forwarders for other nodes that simply sense. This segregation of functionality not only imposes constraints and limitations on the network, but it actually runs counter to the original design objectives of mesh networking and wireless sensor networks.

A better and more flexible mesh networking approach is to treat all nodes as equals: every single node is battery-operated and has both sensing and routing capabilities.

Unlike PCs and servers, which exist in protected environments, sensors out in physically varied conditions must be able to “take care of themselves.” Just as humans have networks of people they can contact in case of emergency, good mesh networks and sensor nodes have a failover plan in the event of a change in the RF environment, such as a truck pulling up next to a building or a node failure.

SolutionsEngineering When there is no hierarchy of different types of nodes, it is much easier to deploy nodes wherever they are needed to create a sensor network. Once the “what” and “where” of the thing or things to be monitored are defined, it becomes a matter of simply activating the sensor nodes and letting them discover other nodes within range. As that discovery takes place, the network begins to build a routing tree that will add redundancy (a choice of two or three paths) and network robustness.

Figure 2 shows the sample deployment’s mesh networking in detail. Each sensor node has one parent node and two alternate nodes. The radio signal strength as well as the link-layer data transmission success rate to the parent and the alternates are also shown. Connectivity to any node from the gateway server can be easily verified using the “Ping” facility.

management and security tools. More important, the WSN now can be viewed and managed as just another IP device, making it accessible and familiar to many more people and applications.

Using the Web Services Model

Once the network is formed and the sensor nodes start collecting data, that data needs to be accessible, either in a database or directly to an application that displays it or performs analysis. This is where the latest sensor networks have taken a lesson from the enterprise IT world. Modern enterprise applications communicate and d1

1

design.

2

develop. d

2

3

d3

deploy.

Are You Board With Me?

Connecting a WSN

Since the end goal of a sensor network is to gather useful data and analyze it, the ability to take the collected sensor data and import it into an enterprise application—or, at the very least, a spreadsheet— is a critical requirement of any deployment architecture. Until recently, all embedded WSN-to-Internet integration happened via some kind of gateway device that sat between the IEEE 802.15.4 network and the IP network. The gateway server’s role was to “translate” the sensor network traffic and provide it in a consumable form for another network—either IP or an industrial network such as BacNet, Modbus or LONworks.

Recently the 6LoWPAN working group of the Internet Engineering Task Force (IETF) submitted the first implementation of IP for low-power, low-bandwidth networks. 6LoWPAN defines IP communication over low-power wireless IEEE 802.15.4 personal-area networks.

The proposed standard, approved by the IETF in March 2007, incorporates IPv6, the latest and most scalable version of the IP protocol. Because of IP’s pervasiveness as a global communication standard across industries, vendors can now create sensor nodes that can communicate directly with other IP devices, whether those devices are wired or wireless, local or across the Internet, on Ethernet, WiFi, 6LoWPAN or other networks. Network managers gain direct, real-time access to sensor nodes and the ability to apply a broad range of Internet

share information using the Web services model, which provides a convenient and scalable way for sensor networks to pass collected data to an end-user application or remote database. Figure 3 shows how sensor data can be accessed from a corporate IT network using Web services to build Web pages. The example API calls to collect data from the WSN are listed, along with the well-formed XML returned as a result of calling this API. The ability to access the data in a number of different ways through Web services APIs or by running SQL queries against a database allows data to be used

design.

The next generation of connected devices.

Start with an AMCC PowerPC 405 Processor.

Specify your I/O requirements.

develop.

Your products based on our platform.

deploy.

Your solution faster.

EP885 Stack Advert V4.indd 1

Choose your operating system.

Relax while Embedded Planet delivers your rugged and reliable system.

4760 Richmond Rd / Cleveland, OH 44128 Tel: 216.245.4180 / Fax: 216.292.0561 www.embeddedplanet.com

June 2007

17

1/11/2006 1:24:11 PM

SolutionsEngineering for trending analysis or fed directly into an existing ERP or analytics package. Information from the physical world can now be used to drive decisions and actions by increasing visibility into previously closed worlds. However, because sensor networks are distributed and largely unattended, the network that supports them must be robust and the data integration schema well thought out and sufficiently generalized to accommodate the diverse sources of information being generated, such as temperature, humidity, light, motion, pressure, etc.

There is no shortage of current and potential applications for sensor networks, and as a result a wide array of sensor and actuator devices have come on the market.

Accommodating that variety of devices, however, is not a trivial challenge. Users should look for an embedded operating system that supports a wide range of leading hardware platforms, while preserving the full capabilities of each. The leading open-source embedded operating system designed for wireless sensor networks, TinyOS, is such an operating system. The OS should include a simple driver framework to support the incorporation of new sensors across multiple platforms. External

expansion ports and drivers should also be available to add new kinds of sensors after the fact. Accommodation for analog sensors of different types (resistive or inductive) as well as digital sensors (contact switches) is key. This makes sensor nodes and WSNs ideal for proof-of-concept and pilot networks where functionality and ROI must be proved before finalizing industrial design and the appropriate enclosure for the deployment environment.

Deploy a WSN in Under an Hour

Deploying such a wireless sensor network can be done by anyone who has installed a TCP/IP- addressable device, such as a home gateway or a WiFi access point.

With today’s standards-based products and tools, the whole process can take less than an hour. The typical steps are: 1. Install and power up the gateway appliance.

2. If the gateway acquires an IP address via DHCP, locate the gateway’s IP using a service such as Bonjour. Otherwise, assign the gateway a static IP address.

3. Select a subnet address and mask for the IP subnet to be formed by the

WSN nodes. If installing in a corporate environment, obtain these from your IT department. Add a static route to this subnet (via the gateway) in your PC or an upstream IP router.

4. Install batteries in the wireless sensor nodes and turn the nodes on.

5. Register the sensor nodes that are allowed to participate in the WSN (MAC authentication).

6. Set the link-layer encryption password.

7. The sensor nodes will now form a mesh network, with the gateway as its router, and begin collecting sensor data.

The WSN should now be ready to use.

To access it, go to the gateway’s Web portal user interface, or connect directly using the TCP and UDP services implemented by the sensor nodes, such as ping, traceroute, systat and netstat. Alternatively, access the data using a Web services-enabled tool such as the .NET tool suite.

Customizing the WSN with additional sensors can be very straightforward depending on the type of sensors required.

For example, adding a two- or three-wire resistive thermal device (RTD) can be as simple as connecting the passive sensor to the expansion ports of the sensor node, with the possible addition of a simple resistor. Adding a binary open-close sensor such as a magnetic Reed switch for doors and windows is even simpler. Sensor nodes can also be built for outdoor deployment with suitable NEMA-rated enclosures and Omni-directional wireless antennas.

Once the network is deployed, it is critical to have a way to commission, configure, manage and monitor individual sensor nodes—and the entire network. Once again, the industry-standard Web services model provides the most flexible and standard way to build applications on top of sensor networks. With the nodes presented as web services, not only can new sensor applications be built in most popular programming languages, but the Web-enabled sensor networks and application can also be viewed anywhere, anytime—from the browser at your desk, or from a remote computer on the other side of the world.

Arch Rock San Francisco, CA. (415) 692-0828. [www.archrock.com].

18

June 2007

Your ONE STOP for Professional Development Tools and Embedded Solutions Visit www.ultsol.com and use our ToolFinder to find the right tools for your next project!

Please choose a processor:

Contact Ultimate Solutions today at: Toll Free: 866.455.3383

info@ultsol.com

www.ultsol.com

SolutionsEngineering I/O and Sensor Technology

From FPDP to VPX: Back-End Management and Processing of Sensor-Derived Data From custom-built to COTS-based solutions, rapidly evolving technology is transforming the way sensor-captured data is managed and processed— notably due to the advent of serial switched fabrics and FPGA technology.

by R obert Nokes GE Fanuc Embedded Systems

d

exploration her your goal peak directly al page, the t resource. chnology, and products

S

ystem design is all about the art of in its initial version, offered a sustained the possible. After the governments throughput of 160 Mbytes/s. It is busable of the USA and other countries with multi-drop capability so that blocks mandated the use of COTS in the early of data can be either added to, or extracted 1990s, one of the first tasks was to devel- from, a frame of FPDP data on a timeslot op VME-based COTS components suit- basis. Connection to the port is made via able for upgrading large sonar systems on an 80-way ribbon cable connector located existing naval vessels. The requirements on the front panel of each board. of such systems are complex because they Dedicated DSP devices were the opgenerally involve a large number of ana- timum choice for sonar signal processing log channels, high sample rates, complex since they easily outperformed the genmpanies providing solutions now interfacing and data formatting, large eral-purpose processors of that era. By oration into products, technologies and companies. Whether your goal is to research the latest computing power for beam forming, fil- optimizing the DSP board architecture plication Engineer, or jump to a company's technical page, the goal of Get Connected is to put you andtype demodulation for sonar operations (such as FIR filtering/ vice you requiretering for whatever of technology, and a high data nies and products you are searching transfer rate. for. decimation, complex demodulation, beam Data flow management was a signifi- forming and replica correlation) over 600 cant challenge in early COTS systems due MOPS (million operations per second) to the extremely limited bandwidth of the sustained computing power could be shoeVME backplane. To overcome this barrier, horned into a single 6U VME slot. The proa number of VME board vendors—led by vision of two FPDP ports per DSP board, ICS Sensor Processing, now part of GE for separate input and output data paths, Fanuc Embedded Systems—pioneered meant that an effective data throughput of a unidirectional data flow architecture 2.5 Gbits/s could be supported. called the Front Panel Data Port, or FPDP. Figure 1 shows an active-passive hull This is a 32-bit synchronous port that, sonar receive processing system developed using a set of COTS products based on these techniques. The system digitized Get Connected 216 analog hydrophone inputs at 18 kHz with companies mentioned in this article. www.rtcmagazine.com/getconnected sample rate by using seven 32-channel,

End of Article

20

June 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

24-bit ADC boards. Beams could be formed using any 72 elements and three DSP cards could produce over 36 simultaneous beam outputs. Two further DSP cards provided low pass filtering of passive beams and demodulation of active beams.

An optional single DAC card provided up to 32 analog beam outputs. Buffer cards were used to route beam data to the sonar post-processor either via VME64 or FPDP interface.

Impact of Technology Advances

The spectacular growth of the personal computer market over the last decade has had a profound influence on sensor processing system design since it has been a major driving force in semiconductor technology improvement. This has driven down component costs and also led to significant improvements in processor architecture and speed, signal interface data rates and data acquisition silicon performance.

Consumer interest in high-speed serial interfaces started with the introduction of FireWire and USB ports to provide a low-cost, high-speed and flexible way to attach computer peripherals. At around the same time, low voltage differential

Looking for the right Linux development environment?

You've just found it.

NexusWare® is a suite of Linux-based software products that enables users of our application-ready platforms and blade products to rapidly develop and deploy lawful intercept, monitoring, wireless, IMS and VoIP solutions.

Why waste valuable resources trying to piece together hardware and software from different vendors!

NexusWare ®, Per f ormance Technologies' CGL 3.2 registered Linux-based operating system and development environment, is a comprehensive Linux distribution that is fully pre-integrated with our industry-leading application-ready platforms and robust blades. Customized to fit your application development needs and extendable through installable Protocol software packages (MTP2, SIP, HDLC, Frame Relay, X.25…), NexusWare is the development package that meets your needs today, tomorrow and beyond.

NexusWare is here. Stop searching start getting your products to market on time!

Learn more at:

www.pt.com/nexusware

NexusWare® An integrated software/hardware solution

SolutionsEngineering

ICS110A

216 DIFF. ANALOG INPUTS

ICS110A

ICS110A

{

ICSICSICS2200S 2200S 2200S

ICS115

ICSICS2200S 2200S

BEAM OUTPUT

••• FPDP

PHONE INPUT FPDP ICS

ICS ANALOG & ACQUISITION (7 x ICS-110A)

FPDP ICS

ICS

ICS

ICS

ICS

ICS

ICS

ICS

ICS

BEAM OUTPUTS VIA FPDP OR

BUFFER BEAMFORMING (3 x ICS-2200S) ANY 32 ANALOG BEAM OUTPUTS (OPTIONAL)

BEAM OUTPUTS VIA VME64 OR VSB

COMPLEX DEMOD OF ACTIVE BEAMS LOWPASS FILTERING OF PASSIVE BEAMS

Active-passive hull sonar receive processing system.

signaling (LVDS) technology was originally proposed as a “future-proof” means of providing inter-device communication because its small amplitude and low DC offset could cope with anticipated future reductions in CMOS supply rails.

After more than a decade of development, high-speed serial point-to-point interfaces based on LVDS techniques have finally emerged as a viable alternative to traditional parallel bus standards.

High-speed serial standards such as PCI Express and Gigabit Ethernet have already been implemented in PC products thanks to highly integrated custom chips.

However, the much lower volume military embedded market has only been able to exploit the benefits of high-speed pointto-point communication since serial fabric bridge silicon and sophisticated FPGA core and I/O functions have only quite recently become available.

Acoustic System Design

COTS-based sonar systems are normally operated in a benign environment and so the current generation designs adopted PC-based technology and a more distributed architecture in order to minimize cost while maintaining or improving performance.

Hydrophone data acquisition today uses a synchronized array of multi-chan22

ICS7220

FPDP

SYNC. BUS FOR SYNCHRONIZATION

Figure 1

ICS7220

June 2007

nel ADC boards, controlled by a local CPU via a PCI-based backplane and housed in a rack-mounted 19” chassis. The digitized datastream is then transported via highspeed links to a separate data processing engine, typically a networked array of high-end PC servers.

Applying state-of-the-art technology to the multi-channel data acquisition function has led to the creation of a new concept: the acoustic acquisition server, a high-performance networked system component that radically reduces sonar system size and cost by packing up to 192 analog I/O channels into a single 19” rack-mounted, 1U enclosure. The server uses Gigabit Ethernet as an LVDS high-speed serial point-to-point interface to transfer twoway data and control information to or from a remote system controller, as illustrated in Figure 2.

In this example, the system controller, which incorporates the digital data processing engine, communicates with each acoustic server via Ethernet. Data flow uses UDP since it is a low-latency, reduced overhead transport method and can handle the data throughput of an acoustic server fitted with its maximum number of I/O channels in a low-frequency sonar application. The use of a managed Gigabit

Ethernet switch offers greater flexibility in how UDP packets are routed. For example, during a beam forming operation, packets from one or more acoustic servers could be multicast to a subset of system controller processors. Remote configuration and status monitoring operations for each acoustic server are all performed using SNMP transactions.

Figure 3 shows a simplified view of the 1U acoustic server architecture. Internally, the design of the acoustic server has been optimized to provide high-performance, high-speed signal conversion capability for up to 192 channels of analog I/O or 240 channels of digital I/O using any combination of four I/O modules of three types. Each of the I/O modules slots is assigned a unique UDP port number, allowing for distributed processing such as forming partial beams using multiple distributed data processing engines.

Each I/O module communicates via two bidirectional high-speed serial links to the embedded controller motherboard.

This provides sufficient bandwidth to operate an analog input module with 48 state-of-the-art performance, 24-bit ADC chips running at a sample rate of up to 2.5 MHz. At the heart of the embedded controller is the FPGA-based data manage-

SolutionsEngineering

The Harsher the Environment, the More You Need MEN Micro!

DATA / CONTROL SYNC SYNC

ETHERNET GB SWITCH

SYNC

T/R SWITCH

NETWORK I/F SYSTEM DISPLAY

SYNC

SYSTEM CONTROLLER

POWER AMP

SYNC

DIGITAL OUTPUTS (CONTROL)

DATA / SYNC CONTROL

Figure 2

SONAR ARRAY

Intel® Pentium®M up to Core™(2) Duo CompactPCI®/Express

Sonar system built on 1U 19” rack units.

ment and processing engine, which uses the Aurora protocol to manage I/O module data transfer, IP cores to implement the dual redundant Gigabit Ethernet interface, an embedded Linux-based software subsystem to control all functionality of the acoustic server and core logic to create a variety of hardware features used in normal operation and for test purposes. The 1U acoustic server development has been enabled by today’s technology. This becomes clear when you consider that a single high-speed serial link can handle the maximum throughput of the FPDP interface used in the first COTS sonar solutions.

Max. 60 Ch.

I/O Module

I/O Module

Military radio and radar systems require sampling and real-time digital processing of a small number of high-bandwidth analog channels and generally operate in a harsh environment. VME-based COTS components are widely used in rugged applications and now come equipped with high-speed serial links via the P0 connector. However, the bandwidth available for backplane communication is constrained by the limited availability of P0 signal pins.

The beam forming application example in Figure 4 illustrates this problem.

In this architecture, a beam is formed from six channels of data, sampled at 200

Max. 60 Ch.

Max. 60 Ch.

Max. 60 Ch.

Radio System Design

I/O Module

I/O Module

48 Ch.

48 Ch.

60 Ch.

For the harshest industrial and mobile applications, MEN Micro offers more solutions, including: ■

Core™ 2 Duo T7400, 2.16 GHz

■

Core™ Duo T2500, 2 GHz

■

Analog Input

Analog Output

Digital I/O

■

■

High Speed Serial (HSS)

HSS

HSS

■

Pentium® M 760/Celeron M 373, 1.2 GHz Side cards - UART, multimedia, USB and more!

Compatible 3U Intel® family with scalable computing performance Long-term availability- up to 15 years

Available I/O Modules Embedded Controller Power Supply

Dual GbE

Multi-Chassis Sync

Figure 3

Ext. Trigger

Ext. Clock Ext. Sync

1U acoustic server.

DAC Test ADC Test

DAC Test

RoHS MEN Micro, Inc. COMPLIANT 24 North Main Street 2002/95/EC Ambler, PA 19002 Tel: 215.542.9575 E-mail: sales@menmicro.com

www.menmicro.com

June 2007

23

SolutionsEngineering

PC/104-CPU Board 500MHz, fanless, high functionality, inexpensive rugged design

MGT

G4DSP-XD3

_ AMD Geode™ LX800 Processor with 500MHz _ up to 1 GByte DDR-DRAM SODIMM _ 4x USB V2.0, 2x COM, LPT1, FD, E-IDE, LPT1, KB, MS _ LAN Ethernet 10/100Base-T _ integrated 2D graphic controller CRT bis 1920 x 1440, 85Hz 24Bit Digital Video Interface (TFT) _ Operating Temperature 0°C up to +60°C _ Option: CompactFlash DVI- and LVDS-Converter Cablekit

DIGITAL-LOGIC offers a large variety of Embedded Computers in PC/104, EPIC, EBX, 3.5", smartModule and other form factors.

Further information:

www.digitallogic.com

24 Untitled-1 1

June 2007

6/14/07 12:53:33 PM

G4DSP-XD3

G4DSP-XD3

StarFabric G4DSP-XD3

G4DSP-XD3

V4DSP

V4DSP

V4DSP

ICS 8 5 5 1

ICS 8 5 5 1

ICS 8 5 5 1

CH 1&2

CH 3&4

CH 5&6

Figure 4

MICROSPACE MSM800SEL

StarFabric

StarFabric

G4DSP-XD3

Beam forming application example.

MHz by three dual channel, 12-bit ADC mezzanine cards, each connected to a dual FPGA host processor card via an XMC connector. Each FPGA processor card is locally connected to two quad DSP processor cards via 2.5 Gbit/s StarFabric links and to other FPGA processor cards via 2.5 Gbit/s MGT links.

Although the RocketIO lanes of the XMC interface are capable of transporting raw ADC samples at 400 Mbytes/s per channel to each of the dual FPGA processors, the backplane serial fabric is not capable of supporting this data rate. One solution is to design the system architecture so that the data rate is progressively reduced to a level that is compatible with the backplane signaling rate. In this example, this approach was feasible by using the onboard FPGA resources of the ADC mezzanine card and the dual FPGA card to preprocess the data. However, this technique may not be viable in higher data rate applications.

Future Developments

The complexity inherent in processing sensor-acquired data has led to the increasing use of highly scalable, very high-performance multiprocessor systems. However, developing sophisticated applications for such environments can be enormously challenging, and this has resulted in advanced software tools coming to market that are designed to accelerate the development of complex signal processing applications, providing the ability to reconfigure or scale the system to meet future application demands and delivering seamless integration between the single board computer, signal processing and sensor I/O. In hardware, it seems likely that the VITA 46/VPX standard will become widespread as the architecture of choice

for high-performance back-end sensor processing because of the substantially superior throughput of which it is capable when compared with VMEbus-based systems.

VPX provides backward compatibility with VME and PMC/XMC products and features a new, 7-row, high-speed connector with defined zones for VMEbus, serial fabric, XMC differential, PMC I/O and local I/O signals with the option of 3U and 6U form-factor boards. The 6U VPX connector provides for up to four ports of serial fabric per board, each port consisting of four 2.5 Gbit/s bidirectional links, and additional zones are reserved for up to two XMC cards to connect directly with the backplane.

The VPX implementation of the beam forming application described previously would offer a 2.6x increase in serial fabric throughput that, coupled with the use of a mesh-connected topology between identical processor boards and the XMC mezzanine cards, would deliver a level of performance well beyond that of which early signal processing systems were capable.

Advances in technology and the demands of the consumer and military markets have brought about a revolution in system design methodology. The back-end management and processing of sensor-derived data has become steadily more sophisticated and more demanding with advances in hardware and software enabling the development of new applications—and new applications spurring the development of new generations of hardware, software and system design. It is a trend that looks very likely to continue.

GE Fanuc Embedded Systems Towcester, UK. +44 (0) 1327 359444. [www.gefanucembedded.com].

Photographer: Remco Frank

What Do These Embedded Applications Have In Common?

Each embedded application is unique. That’s why VersaLogic offers an extensive line of embedded products with choices of size, performance, and features. That’s also why VersaLogic works closely with customers to tailor standard products to meet their needs. Product customization is available in quantities as low as 100 pieces!

More importantly, embedded customers have many common needs. Meticulous quality standards, dedicated customer support, and reliable product delivery are just a few that VersaLogic provides. That’s why customers from a wide range of industries rely on VersaLogic for their embedded computer needs.

Could you use uncommonly-good support? Call today to see if your unique application is a good match with VersaLogic’s unique approach to service and support. We may have more in common than you think!

Call 800-824-3163. Visit www.VersaLogic.com.

Rated a “Platinum Vendor”

by Venture Development Corporation (2002-2006) for excellence in customer service.

(800) 824-3163 • info@VersaLogic.com

IndustryInsight Embedded Choices for Medical Systems

Multicore Blade Servers Help Advance 3D Ultrasound Technology A new type of 3D ultrasound imaging technology uses blade servers and dual-core processors to provide increased precision, reduced invasiveness and lower cost in many medical applications.

by F rank Setinsek TechniScan Medical Systems

d

exploration er your goal peak directly al page, the t resource. chnology, and products

T

he medical imaging industry is continBecause of recent advances in comuously in need of new, innovative tools putational power and in 3D visualizafor the diagnosis and management of tion techniques, the use of 3D ultrasound disease. Faster, more accurate diagnosis can procedures is rapidly progressing and adhelp keep ever-spiraling health costs down dressing the need for increased precision, and improve overall patient care. Moreover, reduced invasiveness and lower costs in a larger percentage of patients may be more some cases. Historically, 3D ultrasound willing to undergo critical early detection was used primarily in the field of fetal improcedures if they can be made more com- aging, or scanning babiesâ&#x20AC;&#x2122; faces. Today, 3D fortable and less time-consuming. ultrasound technology is being considered Traditional ultrasound technology is for a number of new applications, including mpanies providing solutions now an inexpensive, compact and highly flex- vascular imaging, obstetrics, cancer detecoration into products, technologies and companies. Whether your goal is to research the latest ible imaging modality used tion and even intraoperative procedures. plication Engineer, or jump to a company's technical page,to the produce goal of Get Connected is to put you of type internal anatomy. It uses sound vice you require images for whatever of technology, nies and products you arethat searching waves travelfor.through water and bounce The 3D UltraSound CT Imaging off of tissues, organs, bones and other ana- System tomic substances to create a mirror image.

TechniScan Medical Systems, a deHowever, these ultrasound images can be veloper of ultrasound imaging technolquite fuzzy, which makes it difficult to de- ogy, is seeking to make an impact on tect small and subtle tissue variations, such breast imaging with a new kind of ultraas some early forms of cancer. In addition, sound technology that uses both speed traditional ultrasound is highly operator- of sound and attenuation measurements dependent and can be a time-consuming to develop a 3D-type image. The goal of procedure, especially when performing the UltraSound CT Imaging System is to whole-breast examinations. eventually develop a scanning system that differentially characterizes normal, benign and malignant tissue (Figure 1). By Get Connected with companies mentioned in this article. characterizing the ultrasound properties www.rtcmagazine.com/getconnected of these types of tissue, the UltraSound

End of Article

26

June 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

CT System has the potential to help physicians decide on appropriate clinical management, perhaps without the need for invasive follow-up procedures.

Conventional ultrasound measures the echo of sound waves as they reflect off of tissue to produce images. In contrast, the UltraSound CT system uses transmission ultrasound to produce two unique sets of images: one based on the speed of sound and one based on the attenuation of sound.

Once the patient is properly positioned on the ultrasound system, the scanning procedure itself is operator-independent. The images produced from these two different measurements can be displayed in many orientations for review by a radiologist (Figure 2).

System Technology Challenges

This application requires immense computational power to quickly process the huge amounts of data. A typical scan of each breast can generate about 30 Gbytes of raw data, or a total of 60 Gbytes per patient. Transforming that data into an image requires a highly compute-intensive algorithm. The final 3D image takes up only about 60 Mbytes, but a great deal

IndustryInsight

of computing effort is required to create that image.

Early development of this system focused on producing the highest quality images to obtain the most accurate data. However, after improving the image quality and ensuring the accuracy of its results, there were some serious performance hurdles to overcome. The first scans required nearly 20 minutes per 2 mm tissue layer. This was eventually brought down to about 12 seconds per layer, or 10 minutes per breast, through the use of a redesigned data acquisition system that captured data 100 times faster than the previous design.

Although TechniScan’s own optimization efforts made it possible to reduce the time needed for the patient to lie still for the scan to less than 10 minutes, the company was unable to reduce the amount of time required to compute and produce the 3D images needed for review. The current system incorporates a seven-node Linux cluster built with seven Kontron CP6011 SBCs. Each SBC features a 1.8 GHz Intel Pentium M processor with 2 Gbytes of memory, connected by a Fibre Channel host adapter to a 2 Tbyte RAID data storage system.

This system’s image processing time was about five hours per breast, obviously too slow. The company therefore asked Kontron and Intel to work together to help reduce the time needed to create an image.

TechniScan employed Intel’s compilers and optimization tools to obtain the best possible scan-to-image conversion time using this platform, about 2.5 hours. But even with a seven-node distributed architecture and optimized software, processing was still too slow.

It was determined that the system was limited by the computing capabilities of its processors. Among health care provider customers, the expectation is to review images with the patient immediately after the exam whenever possible. Because image processing begins at the start of the exam, the initial goal is therefore to produce complete images in less than 60 minutes.

Figure 1

The ultimate goal of TechniScan’s UltraSound CT Imaging System is to differentially characterize normal, benign and malignant tissue to help physicians decide on appropriate clinical management.

Figure 2

Image output based on a new kind of ultrasound technology depends on both speed of sound and attenuation measurements to develop a 3Dtype image.

Achieving Increased Performance with Multicore Architecture

To substantially decrease image processing performance time, TechniScan’s engineers began working on changes to their algorithm that would further speed calculations. They also turned to Intel software tools to help reduce calculation time. Once they reached the limits of those tools, the company looked to Kontron and Intel to find a more powerful processing platform to achieve its goal.

After analyzing the design challenges, a multicore platform was selected to deliver the needed performance. Intel’s 65 nm process technology makes it pos-

sible to integrate two cores in one physical package. As a result, the Intel Core Duo processor delivers nearly twice the performance as the previous Intel Pentium M 756, while consuming about the same amount of power.

According to Kontron’s benchmark data using a single-threaded software application, the company’s boards using the new Intel dual-core processors show an increase in floating-point performance of over 96.5%, higher integer performance of over 89.3% and double the 3D mark.

To meet the application’s requirement, it was apparent that a generic blade solution would not suffice.

Kontron built the CP6012 SBC using a June 2007

27

IndustryInsight

Advantages of StandardsBased Blade Solutions

Figure 3

With an array of seven Kontron CP601264 CompactPCI boards based on the Intel Core Duo T7400, TechniScan is moving toward its computing goal of delivering ultrasound images in less than 60 minutes from the beginning of the exam.

single Intel Core Duo processor T2500 at 2 GHz with 2 Gbytes of system memory. The boards support a configurable 64-bit, 66 MHz, hot-swap CompactPCI interface. With hot-swap and Intelligent Platform Management Interface support, the CP6012 meets the highest demands for the management of highavailability applications. The board is also designed for stability and packaged in a rugged format.

1 28 Untitled-2June 2007

The use of this SBC increased the number of total processing cores from 7 to 14, providing the necessary boost in performance. Initial results are encouraging. They show that, with the next-generation CP6012 board (the CP601264)—based on the Intel Core Duo T7400 processor at 2.16 GHz with up to 4 Gbytes of memory—TechniScan should be able to achieve its initial goal of delivering images in less than 60 minutes (Figure 3).

Blade server computing is transforming the server industry with a host of advantages in terms of design, functionality and total cost of ownership. By separating CPU and memory from other components—such as cabling, power supply, network connectivity and cooling systems—blade servers can significantly reduce massive enterprise server architectures into highly compact and dense form-factors.

Consequently, a number of blade servers can be combined into a single, powerful computing resource. Unlike their traditional rack counterparts, blade servers allow administrators to reduce power, cooling and space requirements and provide a real advantage due to the ease and speed with which they can be deployed. They offer huge scalability in a small space and flexibility in the type of blades that can be used within a single chassis, from low-cost uniprocessor blades to high-performance multiprocessor blades.

5/30/07 3:46:22 PM

IndustryInsight

Standardization is among the top issues that blade suppliers must address. Many companies are reluctant to install proprietary architectures into their corporate systems because they fear limited choices for add-ons and technological dead-ends. This was a key consideration for the 3D UltraSound CT Imaging System, since TechniScan was concerned about investing in many different form-factors or getting locked into a proprietary architecture with a single source.

A standards-based architecture like PICMG 2.16 eases the process of product integration and maximizes a company’s competitive advantage to meet its time-to-market window. PICMG 2.16, or CompactPCI Packet Switching Backplane, is an extension of the PICMG 2.x family of specifications that implements a packet-based switching architecture based on Ethernet on top of CompactPCI. This enables elements in a chassis to be considered “network elements” as opposed to the master/slave structure in the traditional, bus-based CompactPCI architecture.

The PICMG 2.16 specification was developed to extend the life of existing CompactPCI systems by combining the inherent robustness and reliability of CompactPCI with packet-switched Ethernet fabrics. The PICMG 2.16 platform evolved from the confluence of new IPbased communication applications, the growing popularity of CompactPCI and the fact that IP Ethernet switching has become the dominant LAN topology in the enterprise marketplace. Its necessity, however, was driven by the ever-present need for bandwidth and time-to-market pressures, both inherent limitations of the PCI bus.

Although PICMG 2.16 was created to conquer the limitations of the PCI bus, this new architecture is designed to complement existing CompactPCI systems, not replace them, thereby extending the life of the rugged and familiar CompactPCI architecture.

System architects developing around PICMG 2.16 can design highbandwidth, exceptionally scalable and distributed systems capable of offer-

ing high availability and fault-tolerant features not possible with existing busbased architectures.

On the Horizon

TechniScan recently completed a local phase of clinical trials for the 3D Ultrasound CT Imaging System. The company is continuing to work with Kontron and Intel to ensure that system performance will stand up to the rigors of the daily environment in which it will be used. These advancements will provide room on the imaging system’s platform to increase the size of the raw data files without affecting the time it takes to create an image. The Ultrasound CT Imaging System is still an investigational device and is awaiting clearance after FDA review.

TechniScan Medical Systems Salt Lake City, UT. (801) 521-0444. [www.technicalscanmedicalsystems.com].

June 2007

29

IndustryInsight Embedded Choices for Medical Systems

Choosing the Processor Brain to Control Heart, Lungs When developing a new medical device, system designers must weigh the trade-offs involved between building a custom processor-based control system vs. buying one off the shelf.

by P .J. Tanzillo National Instruments

d

exploration er your goal peak directly al page, the t resource. chnology, and products

T

he process of developing a medical device can seem like climbing a slippery slope. Even in the best cases when designers are dealing with sound, proven science, three fundamental and opposing pressures drive the development process.

First, quality is of the utmost concern because failure often leads to catastrophic consequences. On the other hand, reducing development time is equally critical to mpanies providing solutions now ensure that the new therapy or device can oration into products, technologies and companies. Whether your goal is to research the latest help andtotreat the largest number ofgoal people lication Engineer, or jump a company's technical page, the of Get Connected is to put you may also be key to establishing vice you require possible. for whateverIttype of technology, ies and productsanyou are searching early marketfor.position. Finally, in order for device companies to stay profitable, the cost of goods sold (COGS) and other economics must be considered.

To manage quality concerns, agencies such as the FDA help guide and enforce best practices for developing safe, reliable devices. The other two factors are up to the engineer, to work within those constraints and meet all of the requirements in the most timely and cost-effective manner.

End of Article Get Connected

Figure 1

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

30

June 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

Simplified block diagram view of the cardiopulmonary bypass (CPB) pump.

PCI-to -PMC

Technobox produces custom and off-the-shelf PMCs.

Serial

Ultra320

FPGA

async I/O, SATA, Ethernet, and IDE controllers. Plus a variety of adapters, extenders, and tools for PMC development or integration.

Adapters SCSI

solutions includes versatile FPGA-based digital I/O,

Quad Ethernet

Async

Our extensive range of

Make us your first stop for

SATA

PMC Solutions.

Reconfigurable

RoHS

CardBus

Digital I/O

Fast/Wide

SAS

PrPMC

PMC-to-PCI

IDE

www.technobox.com

CompactFlash

Your Source for PMCs, Adapters, and Tools.

RS-232/422/485 LVD/SE

PMC-X

PIM

Test Access & Bus Analysis

Carriers Extenders

140 Mount Holly Bypass â&#x20AC;&#x201C; Unit 1 Lumberton, NJ 08048 Tel. 609-267-8988 Fax 609-261-1011

IndustryInsight

Figure 2

The centrifugal pump uses rapidly rotated, concentric cones to propel blood forward by centrifugal force.

Once planning and development of device specifications are complete, one of the first engineering milestones is deciding which technology should implement the system’s primary controller.

This decision can make or break the product because it serves as the foundation upon which many other decisions are made. Not only must processor architectures, operating system capabilities and other components be considered, but also which portion of the system to design and which portion to buy off the shelf.

By designing and building a custom controller, the end solution can be completely specialized and costs optimized, but any design specification changes or oversights can cause lengthy and expensive delays. In addition, FDA validation and verification for custom hardware and software can be a grueling process, especially for smaller companies without prior experience.

By contrast, using an off-the-shelf platform can increase the COGS, and designers may end up paying for features that are neither necessary nor will be used in the design. Nevertheless, off-theshelf systems typically allow for a faster validation cycle and therefore faster time-to-market.

Determining System Requirements

First, designers should start by examining the device under development.

A cardiopulmonary bypass (CPB) pump, commonly called a heart and lung machine, can serve as an example for exam32

June 2007

ining the considerations when selecting a primary controller technology (Figure 1).

Medical professionals typically use these devices during open-heart surgery to oxygenate and circulate blood through the body, temporarily taking over the functionality of the heart and lungs.

Examining the system requirements that need to be met to achieve baseline functionality for a CPB pump, it becomes clear that device functionality can be broken down into two main phases: oxygenation and circulation.

Oxygenation is typically performed by a membrane oxygenator, which imitates a natural lung by interspersing a thin membrane of either microporous polypropylene or silicone rubber between the gas and blood phases. The controller acquires data from flowmeters and an oxygen analyzer to assess the current state of blood and algorithmically manage flow regulators, a gas blender, a gas filter and a moisture trap to control the ventilating gases.

Thus, this system requires multiple analog input and output channels.

The circulation pump is a simpler system, but involves many of the same I/O capabilities. Most modern CPB pumps employ centrifugal pumps, which consist of a series of nested, smooth plastic cones that, when rotated rapidly, propel blood by centrifugal force. An arterial flowmeter is required to determine forward blood flow since it can vary based on the resistance to the flow downstream.

Additional components of a typical system require even more analog, digital and other I/O capabilities. For example, to create a stand-alone user interface

that does not interfere with the system’s core control, a communications mechanism such as Ethernet or Serial RS-232 is needed. Furthermore, there are additional key electromechanical elements of the system, such as a heat exchanger to manage the temperature of the blood Once designers have an understanding of the problem at hand, a solution can be found in one of two ways: either buy an off-the-shelf deployment platform, or build one. It is crucial to examine both paths, understanding the risks and benefits associated with each.

Custom Embedded Design: The “Build” Approach

Before beginning development, designers must choose a processor technology. Five potential technologies could behave as the system’s primary controller: microcontrollers, microprocessors, DSPs, ASICs and FPGAs. Microcontrollers are extremely cost-effective and generally offer an integrated solution on a single chip, including I/O peripherals. However, they typically contain very small amounts of on-chip memory, so there is little room for complexity and expansion.

With microprocessors, clock rates are typically higher and there is almost always an external memory interface, so performance and expandability are generally not a concern. However, there are usually no on-chip analog peripherals, so these processors need to interface to external peripherals, making complex driver development a necessity for most applications.

IndustryInsight DSPs are specialized microprocessors with additional instructions that optimize certain mathematical functions such as multiply and accumulate. They are extremely useful for computationally heavy applications, but specialized knowledge is usually required to take advantage of this capability in software.

ASICs are designed for a specific application rather than general-purpose programmability. They are the most optimized solution, but their extremely expensive development and fabrication process is typically a deterrent to all except those

Figure 3

systems: a surgeon should be able to adjust the level of oxygen in the blood without affecting the flow rate. Therefore, an FPGA-based control scheme is the most appropriate solution since it allows truly parallel and independent control loops without threading.

However, for functions such as UI control and network connectivity, a microprocessor with an embedded operating system is the most common choice.

In these situations, developing a hybrid system is better than choosing the lesser of two evils (Figure 2).

The National Instruments CompactRIO platform is an example of a packaged embedded system.

with the highest-volume products.

An FPGA provides an interesting middle ground between custom ASIC design and off-the-shelf technology by providing a high level of specialization without high fabrication cost. Although they can be used for a variety of processing applications, complex FPGA design is uncommon because the programming paradigm of VHDL is foreign to most embedded software developers who are more comfortable with sequential programming in C. For the heart and lung machine, it may be best to use a combination of processor technologies to leverage the relative strengths of each. The oxygenator and blood pump are effectively independent

Once the processor technology has been decided, the next step is to develop the I/O circuitry. Because many of the signals in the CPB pump are analog in nature (flowmeters and motors), analogto-digital converters (ADCs) and digitalto-analog converters (DACs) are necessary, and corresponding software drivers need to be developed.

Off-the-Shelf Embedded System: The â&#x20AC;&#x153;Buyâ&#x20AC;? Approach

An alternative option is to purchase an off-the-shelf platform for developing the controller. Although this can cost significantly more than the cost of the components of the board, the developer can expect to reach the market much June 2007

33

IndustryInsight

more quickly. In addition, these systems typically have smoother paths for expansion, so addressing the inevitable feature creep that occurs after the first prototype is far less painful. As with processor technologies, there are a several deployment technologies to consider for use in a CPB pump.

Available in several form-factors, including Mini-ITX and PC/104, unpackaged embedded systems are typically the most cost-effective solution for off-the-shelf deployment. They come

Figure 4

By using a packaged embedded system rather than a custom design, the complexity of controller architecture can be significantly reduced.

with a variety of processor architectures, and each has a small amount of operating system and I/O support. However, the software development tools are almost never integrated and these systems typically still require verification of all regulatory certifications.

A packaged embedded system has the same components as an unpackaged embedded system, but it also delivers specifications for shock, vibration, operating temperature and environmental certifications. These systems are generally more expensive, but they often come with integrated software development tools and a more extensive set of integrated I/O options.

Industrial PCs leverage readily available PC technology, so they offer the most comprehensive options for development tools and I/O capabilities.

They also provide many of the same advantages of other packaged embedded systems regarding specifications and certifications. However, this capability comes at a cost, and these systems are 34

significantly more expensive than either of the other two options.

For the heart and lung machine, an off-the-shelf platform with a hybrid architecture similar to that of the custom solution is the best choice. In addition, since EMC/EMI regulations will become a factor in the hospital environment, developers should choose a packaged embedded system that has at least FCC Part 15 - Class A certification.

The National Instruments CompactRIO system meets these requirements by

June 2007

providing a real-time microprocessor, a programmable FPGA and a host of modules for a variety of analog and digital I/ O in a deployable form-factor (Figure 3).

Using this type of system can drastically reduce the complexity of the controller architecture (Figure 4).

Making the Decision

More often than not, technical capabilities are not the determining factor when deciding between build and buy.

Rather, it usually comes down to a simple financial analysis. If the return on investment of the engineering cost incurred in the development of the product can be justified by eventual profits, then the decision is a good one.

In order to make an educated decision, developers must first accurately estimate the cost of building their own custom solution. This is never as simple as it seems. Merely adding up the cost of the components on the board plus the hardware and software development time will result in a gross underestima-

tion of the total investment. Other, hidden costs need to be considered before the true cost of the job has been accurately assessed.

For example, manufacturing and inventory cost can typically account for an additional 20 to 30% of the COGS of the system. Also, on average about 30% of total software development time is spent on operating system, driver and middleware development, although choosing a packaged platform with integrated hardware and software can eliminate this need.

The developer must also account for other hidden costs, including environmental regulations, FDA validation, endof-life components and last-minute specification changes forcing design alterations and complete redesigns. Finally, the least tangible, but potentially the most impactful, is the opportunity cost of spending engineering time on designing this aspect of the system rather than other revenuegenerating projects.

Finally, for medical devices, the FDA validation effort for off-the-shelf components can be far less burdensome than a custom design, so long as due diligence is shown in choosing a vendor with a well-defined and well-followed product lifecycle development process.

If vendors with documented certifications like ISO-9001 Quality Management System are chosen, great strides can be made in easing the product validation effort.

When assessing technologies that are crucial to the success of a medical system product, developers should save their hard work for the systems that will either reach very high volumes, require an extremely specialized form-factor or have very stringent technical requirements, such as ultra-low power consumption. In areas where off-the-shelf technology can be used, it is best to let the suppliers take on the logistics and hidden costs so that developers can focus on differentiating technology that makes their products better.

National Instruments Austin, TX. (512) 683-0100. [www.ni.com].

GE Fanuc Embedded Systems

Flexibility like this deserves applause Introducing the industry’s most flexible 6U Compact PCI single board computers Of course, you could change your project. But wouldn’t it be better if you didn’t have to? If you could find the single board computer that precisely meets your needs?

Look no further. The new Intel®-based family of PICMG 2.1-compliant 6U CompactPCI SBCs from GE Fanuc Embedded Systems is characterized by outstanding flexibility in the range of options it offers. Market-leading performance is a given – and remarkably low power dissipation, coupled with optional ruggedization, means it can perform in the most challenging environments.

processor – coupled with ATI 2D and 3D graphics. Up to 4GB memory (soldered, in order to maximize robustness and reliability). Dual Gigabit Ethernet. Flash drive or UDMA hard disk – or both. Two RAID-capable SATA channels. Two PMC sites. Four USB 2.0 ports. There’s even a choice of operating system.

But it’s in its flexibility that it excels. Three choices of processor, including a state of the art Intel® Core™2 Duo

www.gefanucembedded.com

The CP11, CT11 and CR11 from GE Fanuc Embedded Systems. We designed them to be flexible so you don’t have to be.

CR11 6U Compact PCI single board computer with Intel® Core™2 Duo or Intel® Core™ Duo processors

© 2007 GE Fanuc Embedded Systems, Inc. All rights reserved.

exploration r your goal eak directly page, the resource. hnology, nd products

Executive Interview

“We strongly believe in VXS and it is rapidly gaining acceptance.”

RTC Interviews Fred Ruegg, President, Elma Americas RTC: Elma is well-known for its backplane and enclosure solutions, but what other markets do you serve? Also, what acquisitions have you made in the last few years?

Ruegg: Elma has five product divisions—Enclosures & Components, Backplanes, System Platforms, Cabinets, and Switches, Knobs & LEDs. In each of these divisions, we serve a wide range of industries, including Mil/Aero, Communications, Medical, Industrial, and more.

Elma Bustronic is our backplane operation in the Americas and we acquired Optima EPS for our cabinet division in 2004. Elma also recently acquired Mektron in the UK, a well-respected company for customized rugged enclosure solutions, among them many conduction-cooled designs.

area grew last year. We attribute this to slight growth in VME business and Elma gaining market share. Not every application needs the performance of high-speed fabrics, and the 2eSST/VME320 enhancements continue to keep VME64x a preferred solution. With its tightly coupled multiprocessing, vast vendor ecosystem and 20-year legacy with continuing advancements, VME will continue to hold its own. VME will remain strong in Mil/ Aero, particularly since that community likes to use a proven, well-understood technology, which presents less risk and superior longevity.

RTC: ATCA has been very much on the front burner for a number of companies yet it still eludes the kind of market penetration that many analysts have forecast. Each year analysts trot out RTC: In recent new projections saying that this will panies providing solutions now years we’ve seen a softening of VME across the board.

The be the year of ATCA. Now, halfway ation into products, technologies and companies. Whether your goal is to research the latest new products being throughis to2007 cation Engineer,number or jump to aof company's technical page, the goalintroof Get Connected put youthose predictions still ce you require for whatever typeshrunken of technology,dramatically and duced has haven’t been realized. First off, do you es and productsmost you areattention searching for.seems to be focused on see that ATCA has slowly been gainnewer VSO standards. However, we’ve ing momentum even though its growth also been made aware that legacy has not been explosive? Will the second VME—both VME 64 as well as the half of ’07 be the year that ATCA takes older version—is still alive and well. off? Why? Why not?

As a supplier of VME backplanes and Ruegg: ATCA has seen slow market boxes do you see the legacy VME busipenetration in its original core market— ness strong? Where, and what applicathe telecom central office. But, what is tions continue to use VME? exciting is the other applications where Ruegg: VME and especially VME64x we’ve seen growing ATCA interest. Elma are still quite strong. Our revenues in this was the first to offer AC power versions for ATCA system platforms. Many companies are interested in AC-powered ATCA Get Connected with companies mentioned in this article. for networking, scientific and even some www.rtcmagazine.com/getconnected military (communications-based) applica-

End of Article

36

June 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

tions. We have also recognized that there is a growing opportunity in IMS (IP Multimedia Subsystems) for ATCA. Interest has taken time in this area, but it seems to be growing significantly. We also feel that pricing points and customer comfort with the architecture has been improving in the central office. Therefore, Elma has invested in a second-generation car-

“There is a growing opportunity in IMS (IP Multimedia Subsystems) for ATCA.” rier-grade platform. It was specifically designed to offer superior cooling, full NEBS compliance and full redundancy for high availability (99.999% uptime).

RTC: In a similar vein, the military is becoming increasingly network-centric and as such, is requiring more networkoriented hardware. We’ve heard of at least a handful of ATCA systems and subsystems already going to military prime contractors and for development of military systems. As a supplier of hardware to the military, have you witnessed an increase in activity of ATCA in the military? For what applications and what kind of equipment?

Ruegg: You are correct in identifying Elma as a key supplier of COTS (defi-

Pump Up Your ARM ® Powered Design!

October 2-4 2007 Santa Clara Convention Center

ARM Developers’ Conference and Design Pavilion

Strengthen your skills and speed your time-to-market Only the ARM Developers’ Conference offers: �

� �

�

Over 90 track sessions providing a complete end to end design tutorial for leveraging ARM IP in advanced embedded applications Combined tutorials with Portable Design Conference Design Centers and exhibitions from leading ARM licensees and Connected Community members offer a full complement of workshops and presentations Forums and analyst presentations on industry trends

Plus: through the combination of track sessions, presentations and company design centers you will be fully immersed in leading strategies and methodologies for building complex designs with the ARM architecture.

And Finally-- Conference Delegates who pre-register before September 10 receive the conference proceedings on a 4 GB iPod Nano to take home!

Register early and save!

www.arm.com/developersconference

Co-Located with the Portable Design Conference & Exhibition

ExecutiveInterview

nition as per “Perry Memo”) chassis to the military and as such we have wide exposure to the varied architectures being considered and deployed. To your specific question about ATCA, yes, Elma has been participating in a number of rugged, COTS based, ATCA opportunities.

Elma’s experience may not be reflective of the market as a whole, however. Elma’s MIL-STD experience and modular approach to rugged MIL-qualified, chassis design combined with our ATCA strategy to support lower volume, special need applications has put us in a unique position in this growing niche. There is an inter-

est for ATCA in military applications, however, to this point the vendor base has focused on more monolithic, volume solutions. Successful suppliers of ATCA to the military will need to adapt their products and processes to a more sophisticated, flexible model.

RTC: MicroTCA has been getting more than its share of attention for a technology that has resulted in a $0 billion dollar market. I was going to ask when you believe MicroTCA will reach market maturity, however that might be prefaced by do you think MicroTCA will 38

June 2007

be a major factor in standard systems?

And for the second half of that question, When? And a third part of the question is what applications will it go into?

Ruegg: MicroTCA is still in its early stages. There are not yet a lot of products out in the market and the interoperability sessions are still in process (the MCH software development seems behind schedule). Most of the activity in MicroTCA is in initial product development and prototyping stages. The costs are still high for mass acceptance. This is why Elma developed a unique modular extrusion-based chassis platform in addi-

tion to the standard sheet metal solution.

It allows a wide range of chassis configurations and customization, while keeping development costs low, lead times short and even small quantities affordable. We believe in MicroTCA and have heavily invested in the architecture. The company offers by far the widest selection of MicroTCA chassis in the industry, including subrack sizes in 4U, 5U, 6U, 7U and 8U heights and a Portable Development Chassis in 4U width. We also have a unique liquid-cooled MicroTCA cabinet solution and will be launching MicroTCA handles and front panels soon.

RTC: In both RTC and its sister publication, PKG, which looks at all phases of electronic packaging, we’ve looked at developments in ruggedizing MicroTCA. According to PICMG president, Joe Pavlat, there are two efforts under way, one looking at extended temperature and outdoor ruggedness, and the other to meet shock and vibration demands as well as extended temperature. Do you have your thoughts rooted in one or the other approach? What do you believe will be the key to the acceptance of one, the other, or both?

Ruegg: Contrary to some advertisements, a real rugged MicroTCA does not exist yet.

Sure, one can put a MicroTCA card cage and backplane in a rugged enclosure and call it “Rugged MicroTCA.” The official PICMG Rugged MicroTCA committee has been formed only recently, with broad industry participation. They are discussing critical elements of the design. First, the committee needs to agree on what we are going to define as “rugged.” Second, there needs to be an agreed-upon test methodology. Third, the connector solution needs to be reviewed closely to see if it will live up to rugged COTS criteria. Only then can we develop a true rugged MicroTCA solution. We believe there will be different levels of ruggedness for different applications (particularly for Mil/Aero). As a leader in rugged COTS solutions, Elma will definitely contribute to this effort.

RTC: Switched fabric technology appears to be something that’s happening. Several variants from ATCA and AMC, which we just discussed above, to other approaches such as VITA 46 and VITA 41 are beginning to emerge as is CompactPCI Express. Do you see any of these emerging as critical technologies over the next several years? Which one (s)? How will the emergence of hybrid and non-VME products impact the backplane and packaging business?

Ruegg: At Bus & Board this year, Elma was proclaimed “King of the VXS Hill” as the company with the largest VITA 41 offering in the industry. We strongly believe in VXS and it is rapidly gaining acceptance. Many of the applications where VME64x may not be enough have moved to VXS. Since it is backward

ExecutiveInterview

compatible, it is often an ideal solution.

We also feel that VPX (VITA 46) will be popular for high-end applications, particularly where space is limited.

Elma has a Portable Development Chassis with a 5-slot 6U backplane and will likely have our 3U backplane version completed by the time this interview is printed. VPX 3U is particularly interesting for UAV and conduction-cooled ATR applications. CompactPCI Express interest was slow, but has picked up a bit recently. The emergence of these new technologies brings new challenges in power, cooling, backplane signal integrity and more. It will be increasingly important for customers to work with a leader in system platform solutions for technologies like VXS and VPX, etc. due to these complexities.

partners on a LFT solution for VITA 48 (VPX-REDI). LFT dissipates a lot of heat, with much fewer of the issues of mist/spray cooling. One of our modular MicroTCA cabinets is available with a liquid heat-exchange solution. Due to the cost impact,

RTC: Cooling is becoming an increasingly important issue as semiconductor components dissipate more and more heat. And, new packaging and assembly technology lets components be placed closer together. Traditional fans and turbines have trouble keeping up causing many companies to look to more exotic approaches such as liquid or spray cooling. Such approaches have been around for more than a decade, yet there are no real standardized approaches such as a standard placement for intake and exhaust for liquids.

There has been some talk, but very little action. Do you believe a standard— or set of standards—can be developed for mist/liquid cooling? Which one or both? Why, why not?

Ruegg: Cooling is definitely becoming more challenging. But, the creative and technological solutions are becoming better too. We are cooling AdvancedTCA with convection (forced air)-cooled thermal management solutions at 200W/slot and even up to 250W/slot. Elma has inhouse thermal simulation and testing equipment, including an airflow chamber, Flomerics thermal software, environmental chamber and much more. Conductioncooling is also a popular cooling option, particularly where airflow is limited (high altitudes, environmentally sealed, etc.)

Liquid flow-through (LFT) is becoming more common. We’ve been working with

RTC: How important is offering integration services for companies that do electronic packaging? Are higher levels of integration an increasing trend?

Ruegg: We feel integration is increasingly important and has been a focus for Elma for the last couple of years. In the past, we would provide a turnkey enclosure solution with the backplane, fans, power and other options—everything but the plug-in cards and the software.

But, some companies are looking to outsource more so that they can focus on their core competencies. So now, Elma offers full integration—including services such as software loading, system testing, FCC/UL/NEBS certification, logistics management and more. It is our philosophy to offer the customer the level of electronic packaging they request starting from components over system platforms all the way to fully integrated systems.

Ruegg: About 70% of Elma’s business is what we call “modified standards”— customized versions of our standard architecture chassis platforms like VME, cPCI, ATCA, etc. These are the best of both worlds. They are based on a stan-

“Successful suppliers of ATCA to the military will need to adapt their products and processes to a more sophisticated, flexible model.”

however, we foresee liquid cooling being limited to niche projects only.

RTC: Standards-based products have been a mainstay in the embedded computer industry almost since its inception. However, as applications call for smaller, more flexible solutions, do you think the industry is going to move in the direction of more custom products?

Will this be a volume-based decision?

What will happen to the smaller-volume applications that have depended on the standards?

dard, which is more cost-effective, built on a proven platform, and have a key component ecosystem. On the other hand, they facilitate customization and allow differentiation between products.

We feel customization is indeed increasing, however, we are committed to maintaining a basic standard component base and design principles that will allow cost-effective solutions for the low volume customer as well.

RTC: Talk of silicon-based lasers coming down in price to the point where they can reside on a chip as easily as transistors brings to mind the possibility of developing optical backplanes in the not-too-distant future. Intel already claims process developments that will bring the size and cost down to the point where this will be possible. Are you aware of any active efforts in developing optical backplanes? Do you think they can and will be a reality? When and how?

Ruegg: Even with lower costs, there are technical issues such as alignment and blind mating, sharp angles and bends, long haul and switching issues, ruggedness, etc. There is a saying in our industry that mainstream optical backplanes have been just “a couple years away” for the past 15 years!

Elma Americas Fremont, CA. (510) 490-7388. [www.elmabustronic.com].

June 2007

39

portabledesign conference & exhibition

santa clara october 3rd - 4th 2007 c o n v e n t i o n

c e n t e r

Connect with the leaders in portable design Donâ&#x20AC;&#x2122;t miss the show with the strongest signal strength in the industry

Unmatched Performance A Wide Range This event is

Interact with industry leaders in special panel discussions, presentations and exhibitions

registration

Superior Networking

opportunities at our receptions and Birds of a Feather events

portabledesignconference.com

sponsored by

Send a Clear Signal

to your boss that this is the one show you want to attend

d

Great Connections

!

of topics will be presented in over a dozen tracks at the conference, including two industry keynote addresses

AR M Co an Dev -loc el at co d m De op ed e wi e ex fo sig rs â&#x20AC;&#x2122; C th th pe r o n rie ne Pa on e nc ev vi f e en lio ere bo t n- n ce th an -

produced by Portable Design magazine, the largest and most influential publication in the US handheld and portable industry

Connecting

the

Brains:

Connecting Embedded Devices Moves Toward an All-IP World

AutomationandControl

Processor Control Modules Get Smaller, More Connected

IN A NETWORKED WORLD

The Guts That Make it Go:

AutomationandControl in a Networked World

The Guts That Make It Go: Processor Control Modules Get Smaller, More Connected An explosion of small form-factors in CPU modules and the growth of networking everything together in automation and control means that the processor modules used in control nodes and small devices have a lot more built-in connectivity, as well as lower power and cost. They are also expanding beyond the x86-based world to include processors based on non-PC technologies such as FPGAs and DSPs.

by Ann R. Thryft, Senior Editor

A

dvances in chip integration—primarily more functions squeezed into smaller CPUs—and plummeting hardware costs have resulted in an expansion of the types and sizes of small form-factor boards used as processor control modules in automation and control. An increasingly wide variety of small modules, often PCbased, with configurable I/O and network interfaces are finding their way into automated devices. The ubiquitous x86-based PC technology is found in a wide range of standard form-factors—including PC/104, ETX, EPIC, COM Express and PMC—as well as a huge number of custom board sizes.

But FPGAs are beginning to show up and, at the higher levels of control, DSPs are increasingly used.

The types of devices being automated are changing too, as control has moved outside the industrial context and into the everyday world and these devices are increasingly connected to the Internet. With such a huge range of possible designs, OEMs want as close to a “perfect-fit” solution as

42

June 2007

possible for each system—in terms of the application, time-to-market and total lifecycle costs—which is one major reason that there are so many small form-factors.

Small Form-Factors Proliferate

Designers are getting more specific about the amounts of electrical power and processing power required for the performance they need as well as application-specific I/O, physical board size, ruggedization and extended temperature operation, according to Bob Burckle, vice president of WinSystems. “Additionally, long-term availability of the hardware and initial engineering design support are key factors evaluated in order to make the decision as to what boards to buy.”

Other fundamental shifts lie behind the rise of multiple small form-factors in networked processor control boards. These include the fact that the cost of small boards is dropping rapidly, and a greater need for automation in applications where it either wasn’t needed before or less was required.

At the low end of the market are very small modules where PLC applica-

tions are moving up into the embedded space, says VersaLogic CEO Len Crane.

“At some point, you can’t stretch a PLC far enough. Those folks need lower-cost, lower-performance hardware.”

In some classic automation areas programmable automation controllers (PACs) are being substituted for older PLC technology. Inside the PAC is a smaller, low-power, x86 CPU from either Intel or AMD, which can be packed into smaller form-factors, says ADLINK Technology America’s general manager Pei Chien Lee. “So the PAC is a PC inside, but it can connect to DeviceNET and CANbus networks.”

The need for greater automation has helped shift many OEMs toward the “buy” end of the make vs. buy decision spectrum. This shift has already occurred with motherboards for two separate but related reasons: the “buy” option is cheaper at the same time that the “make” option has become more expensive.

Processor modules may not be far behind on this curve, since the economics of make vs. buy has moved toward buying

AutomationandControl

Pico-ITX Mainboard Takes on Networked Transportation

Figure 1

WinSystems’ LBC-GX500, a highly integrated SBC designed for machine-to-machine connectivity, features a wide variety of wired and wireless options. These include 802.11 wireless Ethernet, GSM/GPRS cellular modem, CDMA cellular modem, ZigBee wireless RF module, 10/100 wired Ethernet port, global-compliant dial-up modem, six USB ports and six COM channels.

small processor modules that are swapped out to create a next-generation system.

“At the board level this is very compelling whether you have large or small volumes,” says Steve Cooper, CEO of One Stop Systems. “The same trend applies in bigger systems, such as those based on CompactPCI Express. Bigger, compact custom systems are also going to standard off-the-shelf modules.”

But the variety of small form-factors, especially those running PC code, may have exceeded the limits of reasonableness. System design engineers want a stable platform with vendor support that will still be there a few years out.

“That’s a real problem with all of these formats, except for PC/104,” says Jeff Milrod, CEO of BittWare. “The newest standard isn’t necessarily the best one.

I’m going to invest in ‘old’ platforms like VME and PCI because I know they work. New technologies plus new formats equals risk squared. Take one or the other.”

12 Power Consumption (watts)

by Derek Lin and Gaynor de Wit, VIA Technologies

10

Reliably tracking hundreds or 8 thousands of long-haul delivery trucks is essential for distribu6 tion transportation operators. By 4 accessing wideband communica2 tion, in-cab computers can record routes driven and provide up-to0 3DMark2001 PassMark Samsung M.C.C. Idle the-minute delivery status and SE Windows Winstone Diagnostic 2004 cargo information. Likewise, police v2.07b patrol cars require a reliable in-veFigure Since minimal power draw from an inhicle computing platform to quickly vehicle computer is vital, the VIA EPIA PX Pico-ITX mainboard requires less than cross-check data during vehicle in11 watts in standard benchmark tests. spection or at a crime site.

Although the benefits of invehicle computers are clear, the challenges of providing transportation platforms are considerable. In addition to performance and connectivity, low power draw and heat production are essential, as are high reliability and compact size. A ruggedized case is essential to withstand the motion of bumpy or off-road driving, and the ability to build a fanless system with no moving parts is of considerable value.

Addressing these needs, the world’s tiniest full-featured x86 mainboard is aimed at next-generation small footprint embedded computers in networked control applications.

Based on the new Pico-ITX form-factor, which is optimized for power efficiency and space savings, VIA Technologies’ EPIA PX mainboard is smaller even than PC/104 at 10 cm x 7.2 cm. It gives system developers a standardized, ultra-compact, highly integrated platform that can be utilized across multiple embedded PC, system and appliance designs.

The VIA EPIA PX is powered by the 1 GHz VIA C7 CPU. Connectivity includes USB, LAN, WLAN and 3G (via USB), as well as ATA or SATA storage and a range of display support including LCD, VGA or LVDS/DVI. Additionally, the VIA PadLock Security Engine is a suite of security tools that, when enabled, offers on-the-fly AES encryption of up to 25 Gbits/s.

Minimal power draw and heat production are essential to in-vehicle platforms (Figure 1). The low power draw and effective heat dissipation of the VIA EPIA PX allows the design of small systems that can be passively cooled, sealed in dustproof chassis and stowed away in 1 DIN dashboards.

The Pico-ITX form-factor’s compactness is made possible by silicon power efficiency, which makes x86-based systems powered by car batteries a feasible design option.

Thermal design power (TDP), also known as maximum power use, is only 9 watts for the 1 GHz VIA C7 and 12W for the 1.5 GHz version. Combined with an average operating power of less than 1W for each version, this means that VIA CPUs are more power efficient than current x86 processor rivals. TDPs of the nearest Intel offerings range from 21W to as high as 84W for the Intel Celeron M. From automotive PCs embedded within dashboards and in-flight entertainment systems to industrial automation systems and portable devices, the Pico-ITX form-factor enables the design of full x86 computing devices in previously impractical locations and applications.

VIA Technologies, Fremont, CA. (510) 683-3300. [www.viatech.com].

June 2007

43

AutomationandControl

Logic Modules Deliver Compact, Flexible, Speedy Solution

by Guy Marom, Advanced Knowledge Associates Designers of systems used in networked real-time automation and control applications need three things: the ability to get to market simply and cost-effectively, a compact solution, and the ability to upgrade systems as required without major redesign efforts. Moreover, they need to know that their systems will be capable of adapting to and scaling out of component obsolescence issues later in their lifecycles.

A new approach that addresses all three issues centers on smart logic systems-on-module (SOMs). These building blocks are miniaturized, reconfigurable, fully self-contained, high-performance SOMs based on industry-standard programmable logic platforms that include processor cores and all necessary I/O and peripheral circuitry. They are designed to replace complex system elements that are often proprietary and therefore inflexible, bulky, costly, subject to obsolescence issues, inefficient in terms of power consumption and low in performance.

For example, many times whole mother/daughterboard configurations can be replaced by a single board with one SOM. In contrast to other form-factors such as PC/104-based devices, SOMs are smaller and more inherently rugged, and combine all the processing power and operating system elements needed to get a system up and running very quickly.

System-on-module suppliers such as Advanced Knowledge Associates (AKA) remain platform- and technology-agnostic, preferring to choose the most appropriate solution for the task. Technologies include Xilinx or Altera programmable logic chips, ARM or PowerPC MPUs and Linux, VxWorks or other embedded operating systems. Size and parts counts are also reduced. For example, AKA’s modules typically measure 2 in. x 2 in., yet contain over 200 components, including clocks, power, reset circuitry, temperature and passives.

Upgrading a SOM-based system is easy: designers can simply use a faster module with greater functionality. Nor is obsolescence an issue. Although single components frequently become obsolete, SOM makers can ensure continuity of supply by a simple internal redesign, which in most cases will not have an impact on parts qualification.

One SOM well suited to the demands of real-time control and automation systems is AKA’s LM125. Based on a high-density FPGA fabric and employing a Xilinx µBlaze CPU core, it incorporates 256 Mbytes of SDRAM and 512 Mbytes of flash to handle multiple boot images for logic and software applications, while supporting a wide variety of interfaces, including USB, MIL-STD 1553, RS-232, SPI, GPIO and I2C (Figure 1).

44

June 2007

External Logging & Terminal Interface Sensor

Sensor

Sensor

Ethernet

CAN External Control

1553

Actuator

Figure

LM125

Serial data link

ARINC 429

External Control

Actuator

Systems-on-module (SOMs) such as Advanced Knowledge Associates’ LM125 can provide compact, easily upgraded solutions with lots of I/O in networked, real-time control, bridging and industrial automation applications.

The module’s complete software stack includes operating system, industry standard boot-loader and built-in diagnostics for screening and fault tolerance support. One LM125 can be used as the main control subsystem for diverse applications in realtime networked control, such as motion control, protocol bridging, data acquisition and measurement.

The LM125’s modular architecture facilitates the simple integration of user logic and custom peripheral sets. Onboard programmable clock generation, voltage regulation and power monitoring further ease system integration. Because it is based on programmable logic technology from the world’s leading FPGA manufacturer, the Xilinx Spartan IIE, the LM125 enables great system flexibility, as designs can be reprogrammed to include new functionality.

Next-generation SOMs remove the time-consuming task of getting multiple elements such as processors, operating systems, interfaces, peripherals, clocking and power to work together, and are currently providing industry’s most compact and integrated system solutions.

AutomationandControl

More Connected Means More I/O

So how is networking affecting board design? Primarily in the increase in I/O on CPUs. “We’re starting to see more CPUs with more connectivity,” says Jonathan Miller, president of Diamond Systems. “For example, GPS is now being built in, similar to the way Ethernet is now included in all CPUs.” It’s become incredibly cheap to embed an Ethernet controller in small devices, such as device servers, or even photocopiers and soft drink dispensing machines. Processor modules are coming with a wide variety of industrial I/O, such as CAN, Profibus and Device Net, as well as ZigBee. For example, WinSystems’ LBC-GX500 SBC has a wide variety of wired and wireless connectivity options (Figure 1).

In addition, there is widely available PC-based I/O such as USB, which is replacing many older interfaces.

USB’s advantages are similar to those that made ISA popular for so long.

“It’s fairly simple, straightforward to use, inexpensive, readily available, familiar to most users and meets the needs of a niche market that doesn’t need the speed of PCI,” says Micro/sys president Susan Wooley. As time goes on, more MCUs will implement USB, and chipsets will have a higher USB port count.

There is also a shift from parallel to serial interfaces for control I/O. “Although very high-speed onboard interfaces by default end up being parallel, such as PCI Express, when you go offboard those that used to be parallel are moving toward serial, for example, from parallel ports to USB and from SCSI to SATA,” says Wooley.

Some simple expansion needs are not served well by a full-sized PC/104 add-on board. The cost of that board, along with the required stacking connectors, can make adding just a few digital or analog channels to a system fairly expensive for an OEM, says Crane. VersaLogic’s SPX expansion format was designed to address these issues (Figure 2).

Software Design Burden Grows

In addition, getting a handle on system performance has become increasingly difficult because a processor’s real speed can be hidden by OS bloat. As a result, customers are specifying two to three times as much Figure 2 The SPX I/O interface from VersaLogic, shown here on the Python EBX SBC, was designed to address performance as simple I/O expansion needs not served well by a full they actually need, PC/104 add-on board. The small board has a lowsays John McKcost interconnect system, is fast enough for most own, president of real-world signals, and uses a simple electrical Octagon Systems. interface that allows boards to be manufactured “If you’re running economically and allows users to design their own an industrial conI/O devices. trol program under Linux or Windows, you don’t really BittWare, Concord, NH. (603) 226-0404. know what the response time is. Be- [www.bittware.com]. cause time-to-market is killing everyone, it’s difficult for customers to find Diamond Systems, Mountain View, CA. (650) 810-2500. the time to optimize these programs.”

As any system designer knows, soft- [www.diamondsystems.com]. ware is becoming an increasingly imporKontron America, Poway, CA. tant part of the design burden. In some (858) 677-0877. [www.us.kontron.com]. cases, that design burden can reach 100%, says Miller.

Micro/sys, Montrose, CA. In general, at least 80% of design (818) 244-4600. [www.embeddedsys.com]. time is spent on software, since that’s where a system’s value-add lies. Al- Octagon Systems, Westminster, CO. though different products may use the (303) 430-1500. same system hardware components, [www.octagonsystems.com]. “it’s how the system is configured plus the application software that is installed One Stop Systems, Escondido, CA. (760) 745-9883. that can make it a medical blood ana[www.onestopsystems.com]. lyzer, a bomb sniffer, a red-light camera detection system, oil pipeline con- VersaLogic, Eugene, OR. troller or even a toll road tag reader,” (541) 485-8575. [www.versalogic.com]. says Burckle.

ADLINK Technology America, Irvine, CA. (949) 727-2077. [www.adlinktech.com].

Advanced Knowledge Associates, Santa Clara, CA. (408) 431-0735. [www.advancedknowledgeassociates.com].

VIA Technologies, Fremont, CA. (510) 683-3300. [www.viatech.com].

WinSystems, Arlington, TX. (817) 274-7553. [www.winsystems.com].

June 2007

45

AutomationandControl

Modular Touchscreen Workstations Improve Efficiency, Upgrades

by Dale Szymborski, Kontron Mobile Rugged Division In many different sizes and configurations, displays are used the factory’s equipment ROI. in industrial applications to help manage production. In some critiThe Kontron Stingray modular workstation is a three-piece cal factory environments—where equipment is exposed to various system consisting of display, workstation and PSU, which meets elements, temperatures, shock and vibration—they can quickly all of these criteria. A failed or obsolete component can be easily wear out, requiring frequent replacements or upgrades. swapped out, without using tools. The unit’s stainless steel front In demanding industrial environments it can be difficult to bezel seal assures protection and survives mandatory factory maintain maximum factor y floor uptime Customer Door when equipment must be replaced often. In many cases, entire systems 90 must be replaced in order to upgrade 30 Mounting Platform 50 Computer/Power Supply only one failed component. Beyond the 80 50 Toolless Knob associated material costs there are 70 Display Assembly (60-SUBP-0048) often additional labor costs, since specialized tools and trained professionals may be required to complete complicated implementations.

To minimize operational downtime, TBD equipment must be maintained as efficiently as possible. System modularity 20 TBD therefore becomes a key feature in this environment. The ideal system replaceToolless Handle ment would be a modular touchscreen workstation unit with display, workstation and power supply unit (PSU) modComputer Assembly (60-SUBP-0050) ules, which operates reliably at elevated temperatures and can be serviced inplant without the need for specialized Power Supply Assembly (60-SUBP-0049) tools (Figure 1).

40 If any element—the display, work10 station or PSU—can be assembled Figure To achieve maximum uptime, automation facilities can replace only and maintained on the floor by factory failing components at a minimum of expense with a modular touchscreen employees, rather than outsourcing the workstation unit such as the Kontron Stingray. work to a team of technical professionals, costs can be kept down and uptime maximized. Modularity would also allow the system’s components hose-downs. Custom expander plates let a display fit into nonto be utilized independently—as a touch display only, a computer standard, pre-existing enclosure openings. host only or as a complete workstation—to conserve power and Low-wattage displays, highly efficient industrial-grade power minimize unnecessary wear on equipment. The modular worksta- supplies and Intel’s LV Pentium M processors reduce heat protion would also feature a small footprint to facilitate incorporation duction. The separation of CPU, power supply and display allows in space-constrained existing housings. maximum airflow and minimizes trapped heat. Each modular comEnvironmental hazards must also be taken into account, ponent is contained in its own perforated, aircraft aluminum houssince a system used within a factory may be exposed to high ing, providing maximum rigidity and optimal heat dissipation. temperatures and required to operate non-stop for long periods of time. The utilization of low-power components would allow the sys- Kontron America, Poway, CA. (858) 677-0877. tem to generate less heat, thus improving reliability and ultimately [www.us.kontron.com].

46

June 2007

Industrial SBC Supports Wired and Wireless Communications WinSystems’ LBC-GX500 is a highly integrated, single board computer (SBC) designed for machine-to-machine connectivity with a wide variety of wired and wireless options. It provides an open and powerful platform for management of geographically distributed machinery.

• AMD GX500@1W processor • PC-compatible: supports Linux and Windows® XP embedded and other popular RTOS • Operates from -40º to +85º C with no fan • 10/100 Mbps Intel Ethernet controller • 10 COM ports and 6 USB ports • Socket support for 56kbps POTS modem, GPRS/CDMA cellular modem, ZigBee and 802.11a/b/g wireless modules • 48 bi-directional TTL digital I/O lines • Flat panel and CRT supported • Onboard AT keyboard, PS/2 mouse, LPT, FDC, and UDMA disk controllers • Type I and II CompactFlash cards supported • PC/104 and PC/104 Plus expansion • Optional 12-bit A/D converter, 8 SE/4 DI • Optional Trimble GPS receiver • EBX-size: 5.75” x 8.00” (147mm x 203mm) • Long-term product availability • Quick Start kits for software development Off-the-shelf delivery, knowledgeable technical support, long-term availability and the right price makes WinSystems’ LBC-GX500 the SBC of choice for your application.

Call 817-274-7553 or Visit www.winsystems.com Ask about our 30-day product evaluation!

715 Stadium Drive • Arlington, Texas 76011 Phone 817-274-7553 • FAX 817-548-1358 E-mail: info@winsystems.com

TM

AutomationandControl in a Networked World

Connecting the Brains: Connecting Embedded Devices Moves Toward an All-IP World Many technologies have been developed for networking embedded devices, creating a world of specifications that has solved a big number of connectivity problems. These subnets are increasingly interacting with Ethernet and hence the Internet. Now, embedded devices all the way to individual sensors are becoming full-fledged IP citizens.

by T om Williams Editor-in-Chief

T

here was a time when the notion of automation and control called to mind the networked factory floor with automated machines running and passing data among each other and to HMI monitoring systems that are connected to servers from which relevant data is made available to management such as sales and purchasing. This is still a very big part of the realm of networked automation and control, but it is being complemented by a much larger world of networked autonomous and semi-autonomous devices that act in the public sector as well as industry and also in the consumer world, and that ultimately connect to the Internet.

A large number of industrial protocols are hard at work today keeping factory operations humming. These include Profibus, DeviceNET, Controller Area Network (CAN), LONworks and a num-

48

June 2007

ber of PLC protocols to name a few. There have been successful efforts to integrate these clusters to work within an enterprise Ethernet from where it is an easy and natural step to connect with the full Internet.

That has been done by using gateways that translate between the IP network and the addressing schemes of the various network protocolsâ&#x20AC;&#x201D;an approach that has and is working well.

However, with the expansion of the world of networked devices, we are seeing more varied applications, a greater need for easy expansion and configuration of connected devices and the push to directly address ever smaller elements down to individual sensors and actuators. The common ground, increasingly, is Internet Protocol addressing that allows all nodes to dwell in the world of IP. They may have different link schemes and physical inter-

connects, but the common world is turning out to be IP. This brings dual challenges. The first involves adapting the traditional Ethernet to address some of the needs of embedded devices such as with determinism issues.

The second involves getting IP/Ethernet connectivity into new designs and into existing equipment with a minimum of cost and a maximum of reliability. One fact helping to meet these challenges is the increasing speed of Ethernet, now commonly at 100 Mbits/s and moving toward Gigabit Ethernet, which offers significantly more bandwidth than traditional field buses.

Ethernet Powerlink, for example, is a real-time Ethernet protocol that adopts the CANopen object dictionary for communicating with real-time elements in network domains that are specifically separated from the non-real-time Ether-

AutomationandControl

net but communicate with it over bridges between real-time and non-real-time domains. The packets are separated into periods for isochronous and asynchronous data communication. Real-time data such as cyclic control loops are transferred in the first period of a packet cycle—the isochronous period—while non-real-time data comes later in the same packet, in the asynchronous period. Each node in a given real-time domain is granted access in every cycle to maintain determinism.

Another variant, called Ethernet for Control Automation Technology (EtherCAT), repeatedly sends a single packet to all nodes in the domain. That packet is divided into frames of bits that are assigned to each individual node, or slave, and sent around regularly to the slaves by the EtherCAT master. At each node, data is read from and written to the bit positions assigned to that node. With a given number of slaves in the system, the timing of each successive packet reaching any given node is known as is the round-trip time for the packet making possible very accurate time synchronization in a distributed system (Figure 1).

Since EtherCAT, as implemented for example by Sysgo, utilizes a single Ethernet packet for any given domain, the total number of bits available to the connected slave nodes is limited to the maximum size of an Ethernet packet. Of course, the various frames assigned to different slaves can vary within that upper limit. This makes EtherCAT an attractive choice where the amount of data that must be shared is limited to such things as variables and state information.

Another Ethernet variant that offers more flexible use of data at the ap-

plication level is the Data Distribution Service (DDS), which is now a standard defined by the Object Management Group (OMG) and offered as a product by several companies such as PrismTech and Real-Time Innovations. DDS is known as a “publish/subscribe” paradigm in which an individual node “publishes” or makes data available that may be of use to other nodes and each node “subscribes” to that data in which it is interested. According to Rajve Joshi of Real-Time Innovations, this middleware technology, “frees the application programmer of these devices from having to worry about the communication details.” Rather, it allows him to work with data that is meaningful at the application level.

DDS also incorporates a quality of service feature that not only sees to it that each node receives the data it needs, but also guarantees timing relationships. For example, node A may be publishing status information every 10 microseconds. Node B subscribes, but only requires status updates every 100 microseconds while Node C subscribes for updates every 500 microseconds. Each node has a time window to receive its data. The middleware receives and acknowledges when the data is received in time by each node and is alerted if it is not, at which point it can take appropriate action. In addition, a subscriber node can specify the order in which it needs to receive updates. For example, “turn on the pump” might need to happen before, “set the pressure valve.” DDS can work over a variety of transport mechanisms including backplanes, switched fabrics and other network technologies.

Implementing real-time control and communication over Ethernet has always

Kontron HMIs...

V Panel Express Scalable display sizes: 12.1 to 17 inches v Performance: up to Intel® Core™ Duo processor v Fanless cooling & IP 65 front v

www.kontron.com

Untitled-4 1

1-888-294-4558 - info@us.kontron.com

June 2007

49

6/14/07 2:25:03 PM

AutomationandControl

Ethernet HDR FH EH Figure 1

WC CRC

In an EtherCAT network, each embedded node reads from and writes to a bit field assigned to it within a single Ethernet packet. The packet is sent around the network with deterministic periodicity, which enables the synchronization of nodes.

had to contend with the fact that it is a “carrier sense, multiple access/collision detect” (CSMA/CD) technology that is inherently nondeterministic. Simply using straight Ethernet for real-time control has always required modifications that also often separate the modified Ethernet domain from that of the enterprise by using bridges and routers. This has proven to be of real value as evidenced by the acceptance of these variations and of others.

A parallel development to distributed networked automation is the proliferation of small form-factor boards that have quite powerful processors that can be directed at demanding real-time tasks locally. While there is still a need to synchronize events within the network, more critical issues can be dealt with by a single processor at a local node while less timecritical data is shared over the network.

In addition, the networks of embedded devices are constantly being expanded, reconfigured and redistributed. Think of things like autonomous metering, monitoring the power usage of devices as they are installed in various locations or managing a fleet of vehicles and the

50

Data

June 2007

cargo they are carrying. This makes the use of standard networking technologies very attractive because system developers need to focus on the needs of their applications, not worry about the details of network communications.

Thus the tendency to let distributed nodes handle the hard real-time issues locally and communicate needed data in what may be real time, but is somewhat less demanding or “softer” real time. This may give up something in terms of network-wide determinism, but it offers the advantage of using unmodified Ethernet, which can communicate directly with IP and hence be integrated with the Internet for worldwide access down to the individual sensor or actuator if so desired. Today, the majority of traffic on the Internet is already between machines more than it is between people.

Of course, it can be said that almost every CPU board made has at least one Ethernet interface. However, on most of these boards, the main processor is responsible for executing the network protocols and managing network communications in addition to running the actual application code. A number of companies

are now offering modular solutions that let board designers integrate an embedded Web server onto their small form-factor devices at a modest cost, allowing them to communicate over a standard IP network (see sidebar “Connecting Equipment to the Internet”). These include Digi, Connect One and Lantronix.

The Connect One solution bases its technology on its iChip, and Lantronix solutions are based on a processor it has developed called the DSTni for executing the TCP/IP protocols and handling the networking, while the Digi Connect ME family uses its own ARM7TDMI processor. All three also offer WiFi and LAN connectivity via 802.11b/g and 10/100BaseT Ethernet. Solutions are available at different levels of integration, although the exact mix of offerings differs among manufacturers.

Basically, a designer can (given sufficient volumes) choose to integrate at the chip level, the module level or board level.

It is also possible to Web-enable legacy equipment with external device servers that connect to an existing device’s serial port requiring no modification of the

AutomationandControl

ZigBee Mesh for Industrial Automation

by Eugene Fodor, Rabbit Semiconductor Only a couple years ago, Bluetooth was a technology that sounded intriguing, but had not yet really emerged. Today, people ever ywhere have donned Lieutenant Uhura style mobile phone ear gear—à la Star Trek—to free their hands as they talk to business associates, friends and loved ones. The rapid emergence of ZigBee offers the embedded market the same intrigue and will deliver as quickly on the science fiction of the past.

ZigBee offers inexpensive, low-power, secure wireless solutions that can operate in all the different ISM bands specified in the 802.15.4 standard, but what’s really exciting is the ability to take advantage of mesh networking. Now devices are not only wireless, but they can easily become par t of a living network that can

are as Unique...

Figure

In a mesh network, individual nodes can fail and connectivity can still be maintained as the devices find an alternate route. Here the signal originally passed through the blackened-out node, but when it failed, automatically rerouted itself to the lamp.

dynamically change and reform as necessar y, lowering the cost of finding and fixing failed network devices. This feature of mesh reduces cost in deployment because the network heals itself around failed network nodes. That healing ability means that the automation and control of environments like factor y floors and building controls are less prone to bottleneck points of failure. When checking the health of devices, the nodes that are alive will appear providing a highly reliable picture of the entire application space. Only the nodes that have failed will be out of reach, and therefore easy to identify and find. This proper ty allows control systems to be managed with a higher degree of accuracy.

Embedded developers can leverage today to build these future embedded applications by using technology like core modules that exemplify the control industr y’s direction of future applications by combining an I/O-rich microprocessor with a ZigBee radio.

Micro Client

Scalable display sizes: 7 to 15 inches v Compact Flash & CAN bus on Board v Fanless cooling & IP 65 front v

www.kontron.com 1-888-294-4558 - info@us.kontron.com

Untitled-4 1

June 2007

51

6/14/07 2:26:52 PM

AutomationandControl

Connecting Equipment to the Internet

by Daryl R. Miller, Lantronix Attaching your equipment to the Internet makes vital information available anywhere in the world, empowers novel applications and opens up new revenue streams.

While there are a number of ways to add connectivity, utilizing device servers can be the most cost-effective and quickest solution. These powerful little products bundle an operating system, full TCP/IP stack, security protocols, Web server and complete tunneling application to transport data across the Web. (Tunneling is a means by which the data from your device is “packaged” and put on the network for transport and then “unpackaged” at the other end).

Device ser vers are easy to connect with your equipment and offer several inter faces including: RS-232, RS-422, RS-485, SPI, USB and even individual digital I/O lines (GPIO).

A variety of options exist for connecting your equipment to the Internet via hardwired Ethernet (802.3) and wireless (802.11b/g). Some device ser vers will even bridge between wired and wireless should you already have a wired Ethernet connection and want to add wireless to your product. With the help of a device ser ver, users can easily enable the communication between a host computer and their equipment or from one piece of equipment to another over the Internet or Ethernet network using tunneling.

Wireless is a hot technology that has broad applicability. It’s a real challenge for most equipment manufacturers to implement since the technology is largely targeted at PCs. Using a wireless device server you can quickly reap the rewards of wireless connectivity including mobility, flexibility and the ability to connect to previously out of reach equipment. And with the latest security standards like WPA2, 802.11i and AES (Advanced Encryption Standard), you can rest assured your wireless data is just as safe from prying eyes as wired data.

Most countries regulate wireless transmitting equipment even when it is on a public frequency. These tests and certifications can be quite expensive. Make sure you implement a device server or module that already has certifications—you will save money, time and a lot of headaches.

Regardless of the connectivity option selected, device networking provides the ability to remotely manage, monitor, diagnose and control equipment from anywhere, at anytime. The result is maximized efficiency and increased information awareness providing manufacturers lucrative opportunities to serve a wide range of industries.

application other than to direct external communication to that port. Digi with its Connect ME family and Lantronix with its XPort and WiPort products, offer very small integrated Web server modules that include processor, a choice of CPU interfaces, network protocols and flash memory for storing Web pages and an RJ45 connector. They also offer pin-compatible wireless versions.

There are numerous options for the designer in terms of the configurations and the connectivity scenarios that can

52

June 2007

be set up depending on the application.

The common theme, however, is that the application processor only needs to send commands and data to the server module in order to communicate over the IP network, be it an Ethernet LAN or the Internet or both. In addition, Lantronix has recently introduced its XPort Direct, an embedded device gateway that does not have internal Web pages (Figure 2). This actually expands the number of options available to the developer at low cost.

In one scenario, it might be desirable

Figure 2

The Xport Direct from Lantronix offers Ethernet connectivity without internal Web pages, an option for developers who simply want to move data to and from an embedded device for use and analysis on a central server.

to directly access an embedded Web page for maintenance diagnostics. In addition to initiating communications, the application can also constantly update the internal Web page in real time so that a user can get current data on the status of the application or alternatively set parameters and issue commands to control or reconfigure the device. Alternatively, the application can initiate communications for alarms and alerts when a critical parameter gets out of range. Internal Web pages can have links to external pages anywhere on the Internet if so desired.

Another scenario might call for the aggregation of data from some large number of devices that are widely separated geographically. The Lantronix XPort Direct was designed for, among others, applications that simply need to get data back to some central point such as a Windows Server application like a database or spreadsheet, or a central Web page.

For example, in a scenario where there are a thousand freezers all reporting their power usage, they can all simply periodically send the data. The data can then be aggregated, processed and presented to whatever number of users who choose to log into the central server. If the data is

AutomationandControl

Application Profile

ZigBee Stack

IEEE 802.15.4

Figure 3

Compliant Platform

& as Diverse…

A ZigBee-compliant platform consists of a ZigBee stack layered on top of the 802.15.4 link layer. To create a ZigBee product, the developer adds an application that conforms to one of the commercial or proprietary application profiles that let the product interact with other products designed to that profile.

put into something like an SQL database, then it can be available to any number of different applications on the server or on user workstations. This also allows collaboration for things like troubleshooting without depending on the limited Web services of a single embedded module.

Of course, such Ethernet and Internet capabilities also lend themselves to autonomous interaction between embedded networked devices—a capability not precluded by other networking technologies, but made more available by virtue of its ability to be an IP citizen. The expression “machine to machine” (M2M) is a somewhat indistinct concept that at least covers the notion of independent modules communicating with each other in some distributed system where a good deal of the interaction takes place without human intervention. At some point or points, of course, there is a user interface that often deals with the system from a high point in the communication hierarchy.

A simplified example of M2M would be a fleet management system where trucks

are fitted with computers and GPS systems.

The management can track the locations of all the trucks and modules at the terminals, automatically register which cargo is loaded and offloaded at each location as well as what loads are at each terminal and are destined for which delivery points. The automatic coordination of this data lets the system decide which truck can most efficiently carry which cargo and can dynamically update the routes of the trucks. A great deal of the data exchanged between modules on trucks, in terminals and RFID data from cargo is exchanged autonomously. User intervention comes at a higher level where a human operator can see the “big picture.”

Security is obviously an important issue when dealing with networked devices.

There are a number of solutions coming online, including separation kernels that “wall off” the kernel from other operations and other OSs. Various forms of encryption are also implemented. The fact that these Ethernet modules are based on a separate microprocessor than the CPU also goes toward maintaining a firewall

Barracuda

Versatile Configuration – Intel® Pentium® processor Workstation or Display-only v 15-inch Full Sunlight-Readable v Corrosion-proof NEMA-4, IP 65 aluminum closure v

www.kontron.com

Untitled-4 1

1-888-294-4558 - info@us.kontron.com

June 2007

53

6/14/07 2:29:45 PM

AutomationandControl

between the network and the application, limiting the commands that can be passed between them. In fact, Amit Resh of Connect One says, “The OS we use is not open source and that by itself creates another layer of security compared to the Open Source operating systems. The architecture, which is two separated CPUs with separate and different memory allocations, makes it even more secure, while the communication between the two CPUs is via the AT+i command set.” The instruction set of the Connect One iChip processor is also not publicly available.

Figure 4

The Rabbit Semiconductor RCM4510 module allows OEMs to enable a wide range of devices with ZigBee connectivity.

IP-based networking between widely separated devices allows a great deal of flexibility and can be carried out by both wired and wireless network connectivity.

Coming in at a lower level but soon fitting into the IP universe is a growing generation of wireless mesh networks. These can consist of both sensors and actuators—controlled devices—but their main characteristic is that data can take multiple paths through the nodes in the mesh to reach its destination. This multi-hop approach allows the use of very low-power devices.

The basis for these networks is the IEEE 802.15.4 specification, which is a lowpower, low data rate wireless network.

The most widespread implementation based on 802.15.4 is ZigBe, named after the zigzagging of bees (see sidebar “ZigBee Mesh for Industrial Automation”).

A ZigBee platform consists of a ZigBee

54

stack on top of the 802.15.4 radio. An application is built upon the ZigBee platform. ZigBee defines what are known as “application profiles,” which define how devices interact without interfering with other applications. There can be publicand manufacturer-specific profiles. Thus a manufacturer can produce a line of devices that works together in an application that is sold as a product. Alternatively, multiple manufacturers can produce devices that can collaborate in a particular application and the user can buy from a range of choices (Figure 3).

June 2007

For example, the ZigBee Alliance— consisting of over 200 members—has recently announced the new public Telecom Profile, which will support the incorporation of mobile phones in a ZigBee network that is capable of secure mobile payment (wave the phone in front of a scanner and enter a PIN), health care monitoring, peerto-peer small data sharing and other location-based services and features. Other application areas include building control and monitoring, automatic meter reading, industrial plant monitoring and wireless sensor networks.

ZigBee nodes do not have IP addresses, so a gateway is needed for access from IP networks. Such a gateway was recently introduced by Meshnetics, which also supplies 802.15.4 ZigBee OEM modules for integration on manufacturers’ devices. The ZigBit Ethernet Gateway acts

as a bridge between the IP network and the ZigBee network and also allows the user to set up, manage and reconfigure the ZigBee network. Multiple gateways can connect to the LAN and can be accessed from a user server or over the Internet.

To enable OEM devices with ZigBee connectivity, Rabbit Semiconductor also offers pin-compatible RabbitCore modules, one for 802.11 WiFi and another, the RCM4510W, featuring ZigBee/802.15.4 connectivity with the XBee Series 2 ZigBee module from MaxStream and the Rabbit 4000 processor running at 29.49 MHz. Features include up to 49 GPIO lines shared with 6 serial ports and 4 channels of analog inputs, hardware DMA, quadrature decoders, PWM, and up to four levels of alternate pin functions (Figure 4).

There are other small wireless networks, including wireless sensor networks other than ZigBee that are built on top of the 802.15.4 specification. One of these is offered as a development kit, or “primer pack” from Arch Rock. The Primer Pack/IP has implemented the Internet Engineering Task force (IETF) proposed 6LoPAN standard for IPv6 over 802.15.4 and makes it possible to assign an actual IP address to any given wireless element in the mesh network. So everything is addressable via IP like any other actor or device on the Internet or IP LAN. According to Arch Rock’s Brian Bohlig, “All we had to do was figure out how IP is carried over 802.15.4, then you can use an IP address rather than an address that was created by these devices.”

This represents a closure in the sense that there have been and are many networking schemes, such as Profibus, CAN, DeviceNet, token ring, etc., where devices can communicate with each other at this link layer. All these schemes also have a way to communicate with an IP network in service of an application where a gateway translates between the addressing scheme of the particular link layer and the IP network. Now we are starting to see a homogenous IP environment emerge where the richness of IPv6

AutomationandControl

can assign IP addresses to the smallest devices making these subnets also full citizens of the IP family.

Interestingly, this may start to happen with USB as well. Micro/sys recently introduced a specification called Stackable USB that will allow USB to be used as a low-end I/O bus among embedded modules based on a redesigned connector (Figure 5). While initially implemented on PC/104 modules, Stackable USB can potentially connect among a number of small form-factor modules including EPIC, EBX, COM Express and mini-ITX to name a few.

John Carbone of Express Logic notes that, “USB is coming on strong in embedded,” and that it can be very efficient when a vendor makes the effort to integrate it with an RTOS, saving the user the trouble of doing the integration and also reducing the footprint. USB 2.0 now offers a raw data rate of 480 Mbits/s and is morphing into an “On the Go” version that can function as either a host or a device, plus a Certified Wireless USB is on the way that will have the obvious advantage of wireless connectivity. In addition, there are efforts underway to bring USB into the TCP/IP community as well and extend the IP world via IPv6 to USB devices in the near future.

Figure 5

The Stackable USB specification uses a redesigned connector (the bottom connector is shown here) that lets modules be stacked without wires and for more rugged applications than standard USB.

as your application needs.

Lantronix Irvine, CA. (949) 453-3990. [www.lantronix.com].

Meshnetics Bellevue, WA. (909) 512-6374. [www.meshnetics.com].

PrismTech Burlington, MA. (781) 270-1177. [www.prismtechnologies.com].

Arch Rock San Francisco, CA. (415) 692-0828. [www.archrock.com].

Rabbit Semiconductor Davis, CA. (530) 757-8400. [www.rabbitsemiconductor.com].

Connect One San Jose, CA. (408) 572-5675. [www.connectone.com].

Real-Time Innovations Santa Clara, CA. (408) 200-4700. [www.rti.com].

Digi International Minnetonka, MN. (952) 912-3444. [www.digi.com].

Sysgo Klein-Winterheim, Germany. +49 6136 9948-0. [www.sysgo.com].

Express Logic San Diego, CA. (858) 613-6640. [www.rtos.com].

ZigBee Alliance [www.zigbee.org].

Stingray v v v

Modular Design Tool-Less Repair Capability Panelmount HMI Workstation Solution

www.kontron.com 1-888-294-4558 - info@us.kontron.com

Untitled-4 1

June 2007

55

6/14/07 2:30:09 PM

Software&Development Tools Embedded Data Management

Embedded Software Development Needs a More Automated, Test-Driven Approach Software is increasingly complex, more is being reused and different components come from widely different sources. A unified and automated testing strategy can greatly improve quality and production schedules if applied throughout the development process.

nd

er exploration ther your goal speak directly cal page, the ht resource. echnology, s and products

by M ark Underseth S2 Technologies

E

mbedded software development has evolved into a largescale, globally distributed endeavor, posing significant loration into products, technologies and companies. Whether your goal is to research the latest engineering management challenges. Embedded projects pplication Engineer, or jump to a company's technical page, the goal of Get Connected is to put you involvetypehuge teams of developers, outsourcers, third-party rvice you requirenow for whatever of technology, software technology vendors, chipset partners and even open anies and products you are searching for. source. However, software development methods and practices are largely the same as ten years ago, especially in the areas of integration and testing. As a result, companies are struggling with the challenge of managing, integrating and verifying a vast array of components from many different sources. As a shortterm fix, software managers add more engineers and resources but with limited effectiveness and at very high cost. Often, they still end up delivering software releases late and with compromised quality.

The growing complexity of embedded software development requires a new more reliable and scalable approachâ&#x20AC;&#x201D;one that adopts developer testing practices and implements auGetearly Connected tomated software verification prevent with companies mentioned intothis article. and detect more defects www.rtcmagazine.com/getconnected sooner. The objective is to have all developers and integrators create reusable tests that can be shared and automated throughout

mpanies providing solutions now

Growing Functionality and Product Derivatives

Diversity of Embedded Software

Increasing Complexity, Costs and Time-to-Market

Behavioral Issues and 3rd Party Applications

End of Article

Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

n.

56

June 2007

Reliance on Global Software Supply Chain

Figure 1

Increasing market demands have created new complexities and challenges for embedded software development.

REAL-TIME & EMBEDDED COMPUTING CONFERENCE NEW TRAINING WORKSHOPS & SEMINARS Register Today at www.rtecc.com

Enter a World of Embedded Computing Solutions Attend open-door technical seminars and workshops especially designed for those developing computer systems and time-critical applications. Get ahead with sessions on Multi-Core, Embedded Linux, VME, PCI Express, ATCA, FPGA, Java, RTOS, SwitchFabric Interconnects, Windows, Wireless Connectivity, and much more.

Meet the Experts Exhibits arranged in a unique setting to talk face-to-face with technical experts. Table-top exhibits make it easy to compare technologies, ask probing questions and discover insights that will make a big difference in your embedded computing world. Join us for this complimentary event!

Be sure to enter the drawing on-site for an iPod Video

Join Us In...

January

February

March

April

Copenhagen, Denmark Santa Clara, CA

Melbourne, FL

Huntsville, AL Atlanta, GA Phoenix, AZ Albuquerque, NM

Boston, MA Bethesda, MD

May

June

July

Milan, Rome Italy Dallas, Austin, Houston, TX

Chicago, IL Minneapolis, MN

August Longmont, CO Colorado Springs, CO

September

October

November

December

Beijing, Shanghai, China Ottawa, ON Montreal, PQ Shenzhen, Xi’an, China San Diego, Long Beach, CA

Tyson’s Corner, VA

Detroit, MI Toronto, ON Vancouver, BC

Seattle, WA Portland, OR

www.rtecc.com

Software&DevelopmentTools

the development cycle. This strategy replaces ad hoc testing with continuous automated testing to ensure the on-time delivery of fully tested and properly functioning products (Figure 1).

Outsourcers Open Source The market for sophisticated embedded software-based OEM Developers Chipset Partners 3rd Party Vendors products is exploding. As the industry evolves to support an ever growing range of capabilities and features, the underlying microprocessors become smaller and faster, and the software content just keeps multiplying.

A few years ago, a typical embedded application included a few thousand lines of monolithic code developed by a few developers at a single location. Today’s embedded application may incorporate millions of lines of code developed by more than a Stylus Applications 3rd Party Applications Finger Applications hundred developers. It has become a complex software platform comprising a huge number of software components brought toCore Net UI PIM gether from various sources and locations.

OpenMoko Application Framework In the past, the embedded software industry was driven by matchbox GTK+2 technical competence, meaning that smart technical guys someudev blueZ dbus GSM GPS kdrive 7 llbx11 how always came through. There has never really been a push to Linux Core Services Linux User Interface invest in processes and technologies to address this rapid increase Linux 2.6 Kernel & Device Drives in complexity. Now, in the race to beat the competition, product developers and manufacturers face greater time-to-market presMobile Handset Hardware sures and tighter product release schedules. Yet they have more (Target Board) (PC) code to manage, limited ability to properly test it, and less time to find and fix problems. This is a recipe for disaster.

Figure 2 Embedded applications have become complex It’s very apparent that the current approaches and technolosoftware platforms comprising a huge number gies that were sufficient in the last decade are no longer adequate. of software components brought together from various sources and locations.

Software methods, practices and technology simply have not evolved in response to these new complex integration issues.

Possibly the most alarming observation is the relative time and resources devoted to quality assurance (QA) or product testing. In many companies, time spent on software coding or implementation is relatively short, while integration activities can take twice as long. However, product testing efforts are truly daunting, taking 5-10 times as long as implementation, while staffed with very large teams that continue to grow.

At many companies, integration testing is merely a “smoke test” or “sanity test” to confirm a viable software build by manually executing a rudimentary set of tests. Even when integration testing is more extensive, the test coverage is limited by the time-consuming nature of manual testing. Often the first time all embedded software components are extensively tested as an integrated whole is during QA or production testing.

As a result, QA engineers usually uncover large volumes of defects. A hiatus ensues as managers redirect developers from other work Figure 3 To enable developers to exercise their code as they write it, an to isolate, characterize and debug large numbers embedded software verification platform, like STRIDETM for example, of critical and serious defects. Engineers try in could invoke APIs and simulate other dependent code, all through a vain to salvage a release schedule but they know GUI on the host, and without writing test code. the odds are stacked against them. By catching

The New Complexities of Embedded Software

Open Embedded

x11 Applications

Web Browser IM Book Reader Terminal (others)

Contacts Messages Application Manager Search (others)

Dialer Main Menu Media Player Clocks (others)

x86 SDK

58

June 2007

Software&DevelopmentTools

defects late, developers are fixing bugs when they’re the most difficult, time-consuming and expensive to resolve. With so many critical and serious defects, software managers inevitably not only miss their delivery schedule, but also find it difficult to predict new delivery dates. Worse yet, they cannot be sure that the code they ultimately release is high quality and free from costly or dangerous errors (Figure 2).

Mass Storage Modules for VMEbus and CompactPCI®

Objectives for a New Verification Strategy

Clearly the embedded industry is overdue for a more scalable and effective software testing and integration strategy. Any effort to improve software integration and verification should address several key objectives.

First, manual testing by developers or integrators must be minimized. Manual testing is too tedious, time-consuming, erPMC CompactFlash Module ror-prone, and is not a good use of valuable engineers. Also, for Two Type I/ Type II CF Sockets companies building multiple products concurrently from the same code base, engineers do not have access to all hardwaresoftware permutations. Secondly, the process must achieve phase See the full line of Mass Storage Products at containment to prevent or catch defects early, before QA. Imporwww.RedRockTech.com tant metrics of a new process will be significantly fewer defects or call Toll-Free: 800-808-7837 escaping into QA, and reducing the time and resources required Red Rock Technologies, Inc. 480-483-3777 in product testing.

Next, processes and infrastructure must be implemented that are truly scalable and capable of supporting integration and testing of many components for multiple product lines concurrently.

And finally, improving the predictability of releases requires ear- edrock_04.indd form is to assist developers in creating reusable, automated tests 1 2/2/07 1:21:52 PM lier visibility into the software health of releases. Managers must quickly and easily by providing specialized tools and techniques. be able to pinpoint trouble spots, and make decisions based on For example, the verification platform might enable developers real data or metrics. to quickly break dependencies by simulating missing code with a GUI, simple scripts or C/C++ code. Or perhaps the verification An Automated, Test-Driven Strategy platform would support recording and playback to automate a To achieve these objectives, the mantra needs to be “Au- series of manual test operations. tomate, Automate, Automate.” Then, reuse and automate the An embedded software verification platform also provides tests whenever code has changed or at various integration points value at this stage by enabling developers to validate their tests throughout the development cycle. This strategy sounds concep- before code is available. For API-level testing, a software veritually simple but it does nonetheless require a process change in- fication platform can execute and verify tests by simulating or volving adoption of new methods, implementation of automated modeling APIs through simple scripts, C/C++ code, or a GUI. infrastructure, and a change in mentality regarding the impor- This provides developers with a very simple, quick means to extance of developer testing. Let’s examine this strategy more ecute tests and feed them canned responses to validate them. For closely, looking at certain test-driven software practices and the example, if code-under-test depends on the return values of anrole of an embedded software verification platform. other application interface, the developer can dynamically mockDue to the complexity of today’s embedded software, devel- up the desired return values (Figure 3). opers need to actively participate in the verification of their code.

This new approach succeeds only if developers and integraAs the creators and architects of software components, only de- tors create and deliver automated tests that are reusable by anyvelopers truly understand the inner workings of their code. Fol- one, and if the tests can be aggregated and automated in largelowing a best practice from Test-Driven Development (TDD), de- scale testing. To do this requires that a common test framework velopers would ideally create tests up front before implementing and test guidelines be used by all software developers, integracode. Developing tests in advance has the added benefit of flesh- tors and testers. ing out your design, and especially aids in designing for testabilThe first key is the software verification platform, which ity. By thinking about testing up front, you’ll take into account serves as the common test framework that supports, manages and and implement facilities to access internal APIs, data structures automates tests from all of the developers and integrators. With or other information that aid in testing. the diversity of embedded software components, this means that Thus, the first role of an embedded software verification plat- the test framework should ideally be flexible enough to support June 2007

59

Software&DevelopmentTools

Typical Development Process

Write Code & Fix Defects

Manual Unit Testing

Defects Reported By Product Test Check In Code

Automated Unit & Integration Testing

Check In Code & Test Assets

Build Application

Automated Execution of Test Assets

Engineering Releases

Product Test

Fully Automated

An embedded software development process can be dramatically transformed by leveraging automated tests created by developers and integrators.

various testing strategies. Depending on the type of embedded software component, certain approaches may be more suitable.

“Hard” real-time code may require tests written in native code to be directly built into the target; “soft” real-time applications can be exercised remotely from the host, possibly using a scripting language; network protocols might have internal state machines that should be verified through white-box techniques; data-centric APIs may require facilities to efficiently enter complex data.

Secondly, the development team must conform to a level of uniformity when it creates tests. For example, guidelines might require that tests be written to be self-contained, i.e., not dependent on the preceding execution of other tests. Standard entry and exit criteria would guarantee that tests enter and leave the target in a consistent, known state, enabling tests to be executed in any sequence. All tests would leverage the same error handling and recovery mechanisms. Internal policies and conventions would establish naming conventions, archiving and maintenance policies, and the standard languages for implementing tests.

Continually Leveraging Automated Testing

Once everyone has made the effort to deliver automated tests, the development team begins to derive major benefits from applying them. There are many opportunities to leverage the automated tests, virtually for free. For example, developers can perform regression testing of their own code whenever they modify their software. Tests can also be re-executed at “integration points,” or whenever developers integrate their code with other software components.

The developers’ tests can be aggregated and automatically executed with an automated complete target build, at regularly scheduled intervals. This practice is referred to as “continuous integration” and is very effective for achieving and maintaining

60

Product Test

Automated Activity

Automated Report Generation

Developer & Integrator Activities

Figure 4

Engineering Releases

Manual Activity

Development Process With Unified Verification Strategy

Write Code & Fix Defects

Manual Integration Testing

Build Application

June 2007

software stability because defects are uncovered earlier and frequently, providing continual visibility and metrics into the software’s overall health. The complete collection of developers’ and integrators’ tests can then also be executed by the QA team to complement their black box testing.

In these cases, the role of the software verification platform is to provide a framework for management, reporting and automation by aggregating, organizing, controlling and executing tests, and then collecting, analyzing and displaying results.

Implementing this unified verification approach will indeed transform the software development process with significant new attributes. Tests from all developers—even geographically widely separated—will be managed and automated from a single common framework. Anyone will be able to reuse and execute any test for any component at any time. A growing portfolio of developers’ tests can be automated for regression or verification at various integration points or whenever code is modified. This will enable metrics for software health and completeness to be collected much earlier in the development cycle.

Even if this strategy is applied incrementally or selectively to certain development teams, the impact can be dramatic. A greater volume of defects can be caught and prevented before product test resulting in shorter cycles and fewer resources required in product test. This, along with increased visibility and predictability into software health and delivery schedules, will lead to a higher quality product, on time. These are all tangible benefits that can transform a company’s embedded software engineering into a true competitive advantage (Figure 4).

S2 Technologies Cardiff, CA. (760) 635-2345. [www.s2technologies.com].

Products&Technology Session Border Controllers Provide Flexibility

A growing number of telecommunications carriers seek to lower their cost of entry but retain the ability to scale VoIP services. At the same time, smaller service providers need the functionality, reliability and performance of large session border controllers in a smaller form-factor. Two session border controller platforms from AudioCodes fulfill both of these needs with a great degree of flexibility and functionality and enable service providers to offer, and peer to, multiple VoIP services on a single platform.

The low-cost, 1U small-footprint nCite 1000 is targeted toward smaller VoIP service providers that require a platform scaling from 100 to 4,000 concurrent sessions. It addresses both peering and hosted VoIP applications and services. The full-featured, IMS-ready, nCite 4000 is targeted toward carriers with networks ranging from smaller, rapidly growing network build-outs, to larger, more established networks. It delivers from 21,000 to 84,000 simultaneous sessions while incurring less than 50 microseconds of media latency.

Features of the nCite 1000 and 4000 include virtual session border controller partitioning that lets service providers offer carrier peering, hosted FW/NAT traversal and SIP/H.323 interworking on the same platform, as well as support for ENUM, DNS, CDRs and privacy headers, TCP/UDP/IPSec and TLS support, and seamless transcoding support leveraging AudioCodes’ IPmedia media servers.

AudioCodes, Plano, Texas. (214) 291-1000. [www.audiocodes.com].

Family of PXI Express High-Speed Instruments and 18-Slot Chassis

Ethernet Switch Boasts Two 10 Gig Ports

A new PICMG 2.16 embedded Ethernet switch features 24 10/100/1000 Mbit/s switch ports, two 10 Gbit/s uplink ports and support for IPv6 routing. The CPC6620 from Performance Technologies is the latest addition to its line of switches and is a key component of its line of high-availability Advanced Managed PlatformT solutions. Available in ruggedized and conformal-coated versions with fiber optic 10 Gbit/s uplinks, the CPC6620 can be configured to monitor network status and to continuously check its own health through real-time integrity tests. In the event of system or network failure, data can be automatically re-routed to an alternate path.

Performance Technologies’ complete line of highavailability Advanced Managed Platforms provide an alternative to existing commercial and proprietary systems. Configurations include 1 Gbit/s or 10/100 Ethernet switches, comprehensive remote shelf management, high-performance x86 and PowerPC compute elements accommodating Linux, SolarisT or Windows operating systems, and HA middleware.

Options include application processors, a wide range of networking I/O products and communications protocols, and NexusWare, the Company’s CGL 3.2-registered and POSIX-compliant Linux distribution and development environment.

Performance Technologies, Rochester, NY. (585) 256-0200. [www.pt.com].

PPC 8641-Based VME SBC Boasts XMC, CompactFlash Sites

A new family of modular instruments from National Instruments includes the NI PXIe-5122 100 Msample/s, 100 MHz dual-channel digitizer and the NI PXIe-6537 and NI PXIe-6536 50 MHz and 25 MHz 32-channel digital I/O modules. The NI PXIe-1065 18-slot chassis offers up to 1 Gbyte/s per-slot dedicated bandwidth and a combination of both PXI and PXI Express slots. The new PXI Express products are suited for applications such as signal intelligence, spectral monitoring, semiconductor chip characterization and video test.

The new instruments offer high-speed data recording and playback using the latest industry-standard bus, PCI Express, which delivers up to 1 Gbyte/s of bidirectional streaming throughput on a x4 PCI Express link. With the NI PXIe-5122 digitizer (based on a x4 PXI Express link), for example, engineers can capture high-speed analog signals on two simultaneous channels at 100 Msamples/s with 14-bit resolution and stream the signals across the PCI Express bus at the digitizer’s full rate of 400 Mbytes/s to the PC memory and/or hard disk. The new chassis includes a combination of PXI and PXI Express hybrid slots to accept a mix of both existing PXI modules and high-bandwidth PXI Express modules. All of the new products integrate with a variety of software including the NI LabVIEW graphical development environment, LabVIEW SignalExpress interactive measurement software and NI TestStand test management software.

New military systems as well as upgrades to older legacy systems will benefit from a new VME64x 6U SBC based on the PowerPC Power Architecture. The SVME/ DMV-184 from Curtiss-Wright Controls Embedded Computing combines the dual 1.0 GHz core Freescale 8641 PPC with 2 Gbytes of onboard DDR2 SDRAM via dual 64-bit DDR2 memory controllers.

The general-purpose 2eSST “VME320”capable SBC has two PMC sites (one with VITA 42.3 XMC capability), one Interface Personality Module (IPM) site, one CompactFlash site for onboard mass storage and up to three Gigabit Ethernet ports. The XMCcapable mezzanine site has an 8-lane PCIe link to the 8641 for multiGbyte/s performance. The board is compliant with Curtiss-Wright’s Continuum Software Architecture (CSA), delivering optimal interoperability with the company’s latest SBC and DSP boards and easing development and technology insertion over long life-cycle programs.

The SVME/DMV-184 is available in both air-cooled (SVME) and conduction-cooled (DMV) configurations. A factory-installed IPM can be configured with multi-function RS-232/422/485 serial ports, MILSTD-1553, SCSI, Serial ATA, with LVTTL and differential discretes.

Software support includes Curtiss-Wright’s standard CSA firmware, CSA VxWorks Board Support Package and Driver Suite, MIL-STD1553 software driver and Continuum Vector signal processing library.

National Instruments, Austin, TX. (512) 683-8411. [www.ni.com].

Curtiss-Wright Controls Embedded Computing, Leesburg, VA. (703) 779-7800. [www.cwcembedded.com].

62

June 2007

COM Express Eco-Design with Core 2 Duo and Intel 965GM Chipset

Built around the Intel Core 2 Duo processor, the new ETXexpress Computer-on-Module from Kontron runs at up to 2.2 GHz with Intel’s latest 965GM Express chipset. The module features up to 800 MHz FSB. The clock speed can be dynamically reduced, as required by the operating workload, offering an improved fit for a range of embedded applications including those that are temperature- and power-sensitive. Additionally, the processor can to go into a deeper sleep state for enhanced power efficiency. The chipset logic recognizes whether data is stored on the processor’s on-chip cache memory or in RAM so the chipset does not have to wake up the processor to check the on-chip cache memory. The inclusion of onboard TPM 1.2 ensures a fit for applications that call for enhanced security options.

The ETXexpress-MC will offer support for GEN4 graphics. It contains Intel’s Direct-X9-compliant Graphics Media Accelerator (GMA) X3100 for piping video data through the pixel shaders. The integrated Display Power Saving Technology (DPST) 3.0 saves up to an additional 400 mW of power by switching between progressive and interlaced display modes. The ETXexpress-MC supports fast dual-channel memory up to 4 Gbytes via two 533 MHz or 667 MHZ DDR2 SO-DIMM sockets positioned on the top side of the module. It has 5x PCI Express x1 lanes, 3x SerialATA ports and 8x USB 2.0 ports along with Gigabit Ethernet.

Kontron America, Poway, CA. (858) 677-0877. [www.kontron.com].

Pentium M/Celeron M Control Board Features Six GbE Ports

A high-performance control board designed for IDS/IPS, firewall, VPN gateway and Unified Threat Management (UTM) applications features six GbE or 10/100 LAN ports. The MB-06067 from WIN Enterprises uses an Intel Pentium M or Celeron M processor with a 400 MHz FSB, an Intel 852GM express chipset and a ICH4 I/O controller. It supports system memory of up to 1 Gbyte with one 266 MHz DDR memory socket.

Both CompactFlash and memory can be replaced. Utilizing either the Intel 82541PI or Intel 82551ER Ethernet controller, the new board supports six GbE copper or 10/100 LAN ports with optional bypass function on two ports. Two serial ports, one parallel port, three USB 2.0 ports, an E-IDE connector and a CompactFlash type II socket complete the interfaces supported by the board.

Connected with technology and M The MB-A6067 control Get board with Intel Pentium M/Celeron companies providing solutions now socket type and six GbE copper (Intel 82541PI) ports with two ports Connected is a new resource for furtherThe exploration bypass is priced at $395 for Get a single unit, processor not included. into products, technologies and companies.

Whether MB-B6067 control board with Intel Celeron M 600 MHz/512 Kbyteyour goal is to research the latest datasheet from a company, speak directly cache and six 10/100 LAN (Intel 82551ER) ports with two ports bypass with an Application Engineer, or jump to a company's technical page, the costs $356 for a single goal ofunit.

Get Connected is to put you in touch with the right resource.

Ad Index

level ofMA. service you require for whatever type of technology, WIN Enterprises,Whichever N. Andover, (978) 688-2000.

Get Connected will help you connect with the companies and products [www.win-ent.com]. you are searching for.

www.rtcmagazine.com/getconnected

VXS Board Packs Six A/D Channels in a Single Slot

System designers developing advanced signal generation solutions in radar, electronic warfare and mobile communications applications need high-density processing solutions. A 6U VXS payload card from TEK Microsystems combines six independent channels of 16-bit, 160 Msample/s A/D conversion with FPGA-based DSP processing technology in a single slot along with a single D/A conversion output channel.

The Tarvos VXS implements the new LTC2209 A/D converter from Linear Technologies Corporation, providing a high-performing, lownoise, low-jitter A/D conversion solution.

The board’s A/D converters are linked into a Xilinx FPGA equipped with an advanced DDR SDRAM memory architecture with a capacity of up to 5 Gbytes on a single card. The FPGA supports high-speed off-board communications through two front-panel high-speed serial ports or eight high-speed serial links over the VXS standard P0 connector onto the backplane, using protocols such as Gigabit Ethernet, VITA 55/ Aurora, Serial RapidIO and Serial FPDP, programmable via firmware.

The Tarvos VXS is designed to support multichannel applications requiring a high degree of synchronization and minimal skew between channels.

Triggering waveform outputs to an accuracy of one sample period are fully supported and the cards have been designed to be scaled into fully synchronized multi-board applications if very high channel counts are needed. Pricing starts at $19,900 for single unit quantities.

TEK Microsystems, Chelmsford MA. (978) 244-9200. [www.tekmicro.com].

CompactPCI SBC/PMC Carrier Card Dissipates 5W

Get with technology and companies providing The latest member of GEConnected Fanuc Embedded Systems’ PowerPact 3 Connected is aand new multifunction resource for further exploration family, the IMPCC2 3U Get CompactPCI SBC PMC car- into products a company, rier card, dissipates onlydatasheet 5 watts,from a low among speak SBCs.directly with an Application Engineer, or in touch with the right resource. Whichever level of service you require for Based on the low-power 603e PowerQUICC Get Connected will help you connect with the companies and products yo processor, the IMPCC2 enables the dewww.rtcmagazine.com/getconnected velopment of small, lightweight systems. It features two 10/100 BaseT Ethernet ports, two USB 2.0 ports, four high-speed serial channels, optional VGA video and up to two Gbytes of flash memory module storage. It can be optionally configured as a stand-alone or peripheral PowerPC processor using the embedded 266 MHz 603e core in its PowerQUICC (8270) communications processor.

Available in five ruggedization levels, the board can be deployed either stand-alone or used in conjunction with other PowerPact 3 boards.

The IMPCC2 provides support for a wide range of GE Fanuc EmbedGet Connected with companies and ded Systems andfeatured third-party products in thisPMCs section.and is available with comprehensive operatingwww.rtcmagazine.com/getconnected system and Deployed Test support. Pricing starts at $2,500.

E

Products

GE Fanuc Embedded Systems, Charlottesville, VA. (800) 322-3616. [www.gefanucembedded.com].

Get Connected with companies and products featured in this section. www.rtcmagazine.com/getconnected

June 2007

63

Products&Technology

5U Five-Slot AdvancedTCA Chassis with Shelf Manager

An AdvancedTCA chassis includes a shelf manager and dual hot-swappable I2C controlled fan trays. The aTCA-8505 from Adlink Technology also features a push-pull design for a right-to-left transverse of airflow and a more efficient cooling system. In order to accelerate the transition speed within the blades, the chassis backplane supports a full-mesh transition module. All these features make the aTCA-8505 suitable for network security appliance providers who demand high performance and remote control functionality.

The aTCA-8505 has three slots for nodes, two slots for fabrics, and dual shelf manager slots to meet redundancy requirements. The shelf manager provides a UART and two LAN ports for external connectivity. Using off-board I2C buses, external FRUs and intelligent fan board data can be easily accessed to monitor environment status and adjust fan power to all the blades inside the chassis. All Adlink AdvancedTCA CPU blades and switch blades support both base and fabric interfaces. The transition mode of the base interface is dual star and the fabric interface is full mesh. The aTCA-8505 is priced starting at $3,200 with volume discounts available.

ADLINK Technology America, Irvine, CA. (866) 423-5465. [www.adlinktech.com].

Low-Power PCIe Switches Target Control Planes and Consumer Devices

An eight-port, eight-lane and a five-port, five-lane switch for PCI Express have been introduced by PLX Technology. The two new devices also bring consumption levels down to 0.8W, and are offered in the smallest package at 15 mm x 15 mm—suitable for designing into today’s space-constrained PCB real estate. These lowcost switches represent PCIe standards penetrating the price-sensitive embedded and consumer markets.

The PLX ExpressLane PEX 8509 (eight lanes, eight ports, 1.2W) and PEX 8505 (five lanes, five ports, 0.8W) switches offer PCIe connectivity attributes garnered from PLX’s third-generation architecture. This architecture features the company’s proprietary Cut-Through design that lowers latency (118ns for the PEX8509, 138ns for the PEX 8505) for maximum switch throughput, three integrated on-chip Hot-Plug controllers, true peer-to-peer functionality, and flexible and configurable ports.

While perfect, the PEX 8509 and PEX 8505 are equally suitable for control planes in the communications and networking markets as well as for notebook and docking stations, industrial-control systems, medical imaging, embedded systems and consumer markets, such as set-top boxes, digital video recorders and multi-function printers. These switches comply with the PCI-SIG PCI Express Base Specification, Revision 1.1. Volume pricing will be $12 for the PEX 8509 and $8 for the PEX 8505.

PLX Technology, Sunnyvale, CA. (408) 774-9060. [www.plxtech.com].

Hazardous-Classified Product Displays Are Class 1 Zone 2 Certified

Designed to meet the operational needs and strict safety standards of harsh environments in the oil and gas industry, the “Barracuda” from Kontron is a rugged product family, class 1 Zone 2 certified and designed for maximum protection in hazardous-classified location environments from drilling and pipeline monitoring to chemical processing.

The Kontron Barracuda family, which includes workstation and display configurations, was designed to meet the tough safety and environmental standards needed by the world’s leading oil and gas service companies. The new product line continues the company’s leadership legacy for offshore and land-based oil and gas operations.

It features a corrosion-proof NEMA4 and IP65 high-strength aluminum enclosure to provide the highest system reliability. The Barracuda’s touch-screen display is suitable for outdoor use under direct sunlight. The high-bright technology also results in improved image quality by means of optimized transmission of display graphics The addition of a front panel on/off switch allows the backlight to be turned off and the touch-screen disabled while cleaning the unit, without turning off the complete system.

Able to withstand operational temperatures of 0 to 50°C and offering a -40°C option for use in low temperature extremes, the fully sealed unit withstands corrosive liquids, high shock and vibration. The Kontron Barracuda offers unsurpassed durability, reliability and safety.

Kontron, Poway, CA. (858) 677-0877. [www.kontron.com].

64

June 2007

PCI Express x4 Express Card for Laptops

A new PCI Express (PCIe) x4 Express Card from One Stop Systems enables laptops to operate with high-speed expansion capabilities at 2.5 Gbits/s to a x4 downstream device including an expansion chassis or storage system. The PCIe x4 cable downlink connects to all One Stop Systems’ expansion chassis for additional add-in board capacity.

The Express Card connects to a downstream device through a PCIe x4 cable.

Key features of the PCIe x4 Express Card include LVPECL Spread Spectrum reference clock buffer outputs, electrical isolation at cable connector, low power and a powered cable connector for cables requiring active equaliza- tion for additional distance.

The Express Card does not require software drivers and supports up to 7-meter passive and 25-meter active cables based on the PCIe Cable standard. In addition to also connecting to the PCI Express x4 cable, the PCIe x4 Express Card automatically downshifts to the x1 EC standard. The PCIe x4 Express Card (Part #: OSS-PCIe-HIB2-EC-x4) lists for $496 and is available immediately.

One Stop Systems, Escondido, CA. (760) 466-1664. [www.onestopsystems.com].

Flash Drive Water-Resistant to 200 Meters and Shock-Proof

An extremely rugged line of USB 2.0 flash drives is designed and engineered to be the industry’s toughest USB drive. The Flash Survivor from Corsair is water-resistant, CNC-milled aluminum encased, and shock-proof to safely store user’s information and files in the most demanding environments. Flash Survivor is immediately available in two variations: the GT 8GB and 4GB. Flash Survivor features water resistance to 200 meters (650 Feet) or 20 atmospheres and is sealed with an EPDM (ethylene propylene diene monomer) waterproof seal. The solid-state drive features triple-point protection against shock and impact because it is encased in CNC (Computer Numerical Control) milled aluminum (as found in aircraft part production), which ensures consistency in material quality, thereby guaranteeing the USB drive’s toughness. It features a sustained read/write performance up to 34 Mbytes/s and 28 Mbytes/s respectively.

The Flash Survivor family of USB drives is preloaded with a security application that allows users to create a hidden, password-protected partition on the drive. The password is encrypted with 256bit AES encryption, the most secure encryption algorithms available. The Corsair Flash Survivor GT 8GB and Flash Survivor 4GB are priced at $129.99 MSRP and $59.99 MSRP respectively.

Corsair, Fremont, CA. (510) 657-8747. [www.corsair.com].

GPS Receiver System Has 1 PPS Output

For a large variety of mobile and in-vehicle testing applications, GPS capability is extremely useful. In particular, mobile data acquisition and control systems need position, velocity and UTC time information. The DNA-GPS GPS receiver system from United Electronic Industries adds GPS technology to the company’s popular PowerDNA, UEILogger and UEIPAC data acquisition and control “cubes.”

The DNA-GPS provides location, velocity and UTC time and offers an optimal combination of high accuracy, low power, small size and ease of use. It is built on the Garmin GPS 16-HVS. This provides location information with a positional error of less than three meters in areas served by the WAAS (most of North America) and 15-meter accuracy worldwide. Steady-state velocity measurements have an accuracy of 0.1 knot.

Get Connected with technology and The GPS is also a usefulcompanies source of providing accurate solutions time andnow date information. The GPS 1 pulse/s (PPS) output is synchronized tofor UTC time Get Connected is a new resource further exploration into products, technologies within +/- 1 microsecond. This makes the GPS’s 1 and PPScompanies. signal anWhether excep-your goal is tosynchronizing research the latest datasheet from is a company, tionally accurate means of systems.

Price $495. speak directly

Ad Index

with an Application Engineer, or jump to a company's technical page, the

United Electronicgoal Industries, Canton,isMA. of Get Connected to put(781) you in 821-2890. touch with the right resource. [www.ueidaq.com].

Whichever level of service you require for whatever type of technology,

Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected Serial Communications Cards Serve Legacy Products

Optically Isolated USB to RS-232/422/485 Adapters with Locking USB Connector

Offering designers the choice of one or two serial ports, serial devices can connect peripherals such as barcode scanners, serial displays and data acquisition modules to any USB port. Each DB9M serial port is in the SeaLink +IDIN and +2IDIN serial adapters and is configurable for RS-232, RS-422, or RS-485 via dip switches accessible through the case. Optical isolation protects the host computer from damaging voltage surges and ground loops commonly found in industrial and OEM applications. Each adapter includes Sealevel’s patent-pending SeaLatch locking USB connector design that prevents accidental cable disconnection.

For setup, install the included software and then connect the serial adapter to your USB port.

SeaLink serial ports appear as standard COM ports to the host system enabling compatibility with legacy software.

SeaLink serial adapters are USB bus powered eliminating the need for external power supplies. Status LEDs display transmit and receive activity, and SeaLink serial adapters support data rates to 921.6 Kbits/s. The serial adapters include a removable plastic clip that snaps onto a 35 mm DIN rail. The clip can also be attached with screws to a wall, under a counter, or to any other flat surface. Both adapters are available now and prices start at $139.

Sealevel Systems, Liberty, SC. (864) 843-4343. [www.sealevel.com].

A new line of industrial serial communication cards from Aaxeon Technologies includes PCI, PCI Express and PCMCIA cards. The family of products includes 2-, 4- and 8-port RS-232 Get Connected with technology and companies providing cards and 2-port RSGet Connected is a new resource for further exploration into product 422/485 cards for both the datasheet from a company, speak directly with an Application Engineer, o PCI and the PCI Express in touch with the right resource. Whichever level of service you require for bus. In addition, thereGet areConnected 1- and will help you connect with the companies and products y 2-port RS-232 PCMCIA cards. www.rtcmagazine.com/getconnected Pricing for the RS-232 universal PCI cards ranges from $39 for the MSC-102A 2-port card to $169 for the MSC-108A 8-port version. The MSC102B 2-port RS-422/485 card is priced at $99 and the MSC-102B-SI 2-port version with isolation and surge protection is priced at $199. Pricing for the RS-232 PCI Express cards ranges from $99 for the 2-port MSC-202A to $269 for the 8-port MSC-208A. The RS-422/485 PCI Express cards are priced at $149 for the 2-port MSC-202B and $239 for the MSC-202B-SI 2-port version with isolation and surge protection.

USB-to-serial cards are priced from $109 for the UTS-404A USB-to 4-port RS-232 industrial high-speed card to $469 for the UTS-408A-SI USB-to 8-port RS-232 industrial high-speed surge + Get Connected with companies and products featured in serial this section. isolation card.

PCMCIA cards are priced at $75 for the MSCwww.rtcmagazine.com/getconnected 301A 1-port RS-232 PCMCIA card and $99 for the MSC-302A 2-port RS-232 PCMCIA card.

E

Products

Aaxeon Technologies, Brea, CA. (714) 671-9000. [www.aaxeon.com].

Get Connected with companies and products featured in this section. www.rtcmagazine.com/getconnected

June 2007

65

is to research the latest datasheet from a company, speak directly with an Application Engineer, or jump to a company's technical page, the goal of Get Connected is to put you in touch with the right resource.

Whichever level of service you require for whatever type of technology, Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected

Advertiser Index Get Connected with technology and companies providing solutions now Get Connected is a new resource for further exploration into products, technologies and companies. Whether your goal is to research the latest datasheet from a company, speak directly with an Application Engineer, or jump to a company's technical page, the goal of Get Connected is to put you in touch with the right resource. Whichever level of service you require for whatever type of technology, Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected

Company

Page

Website

ACCES I/O Products..............................................................................................12..................................................................................................www.accesio.com Advanet Technologies............................................................................................28...........................................................................................www.advanettech.com

Products

End of Article

ARCOM.................................................................................................................29....................................................................................................www.arcom.com ARM Developersâ&#x20AC;&#x2122; Conference.................................................................................37.....................................................................www.arm.com/developersconference Digital Logic AG.....................................................................................................24............................................................................................. www.digitallogic.com

Get Connected with companies and Get Connected Embedded Planet..................................................................................................17.................................................................................... www.embeddedplanet.com products featured in this section.

with companies mentioned in this article.

www.rtcmagazine.com/getconnected Extreme Engineering Solutions, Inc........................................................................67...................................................................................................www.xes-inc.com www.rtcmagazine.com/getconnected GE Fanuc Embedded Systems..............................................................................2, 35......................................................................................................www.sbs.com Harting, Inc............................................................................................................9....................................................................................................www.harting.com

Get Connected with companies mentioned in this article.

Harting, Inc. EPT.....................................................................................................6..............................................................................www.harting.com, www.ept.com

Get Connected with companies and products featured in this section.

www.rtcmagazine.com/getconnected

www.rtcmagazine.com/getconnected Innovative Integration............................................................................................33........................................................................................www.innovative-dsp.com Intel......................................................................................................................15.......................................................................................................www.intel.com Kontron America...........................................................................................49, 51, 53, 55.........................................................................................www.kontron.com Logic Supply, Inc....................................................................................................4..............................................................................................www.logicsupply.com MEN Micro, Inc......................................................................................................23.............................................................................................. www.menmicro.com Micro Memory LLC.................................................................................................68......................................................................................... www.micromemory.com Microsoft Windows Embedded...............................................................................13..............................................................................www.microsoft.com/embedded One Stop Systems.................................................................................................61.................................................................................... www.onestopsystems.com Performance Technology........................................................................................21.......................................................................................................... www.pt.com Phoenix International..............................................................................................4................................................................................................. www.phenxint.com Portable Design Conference..................................................................................40...................................................................... www.portabledesignconference.com Real-Time & Embedded Computing Conference......................................................57......................................................................................................www.rtecc.com Red Rock Technologies, Inc...................................................................................59........................................................................................... www.redrocktech.com Swell Software......................................................................................................18.........................................................................................www.swellsoftware.com Technobox.............................................................................................................31..............................................................................................www.technobox.com Ultimate Solutions.................................................................................................19..................................................................................................... www.ultsol.com VersaLogic Corporation..........................................................................................25.............................................................................................. www.versalogic.com WinSystems..........................................................................................................47............................................................................................www.winsystems.com

RTC (Issn#1092-1524) magazine is published monthly at 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673. Periodical postage paid at San Clemente and at additional mailing offices.

POSTMASTER: Send address changes to RTC, 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673.

66

June 2007

All the cards...

…for all the solutions Bringing P.A. Semi to PMC, Extreme Engineering Solutions introduces the XPedite8000, the highestperformance, lowest-power PrPMC (Processor PMC) solution available today. With the P.A. Semi PA6T-1682 integrated platform processor, the XPedite8000 offers: • Dual 2.0-GHz Power Architecture processor cores.

• Two independent DDR2 SDRAM channels for maximum bandwidth.

• Two front-panel Gigabit Ethernet ports.

• Two PTMC-compliant P14 Gigabit Ethernet ports.

• PCI /PCI-X PMC interface operating at up to 133 MHz. • Linux, VxWorks, and QNX support.

For customers looking for one vendor to provide the complete system solution, X-ES provides full component selection, operating system support and integration services.

www.xes-inc.com www.xes-inc.com

©2006 Extreme Engineering Solutions sales@xes-inc.com

©2007 Extreme Engineering Solutions

Ò

Á

MD<Û:8II@<IJÛ=FIÛJN@K:?Û=89I@:J £xÝäÊEÊ £ÈÝä UÊ/Ü Ê - iÊÓ6*ÇäÒÊ-ÞÃÌi «Ã UÊ/Ü ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ Ê6 {Ê-8xxÊ ÀÊ 8£Èä Á

Ê - iÊ «> Ê iÛ Vià UÊ/Ü Ê8 É* Êà ÌiÃ

UÊ£{Ê `i«i `i ÌÊ i ÀÞÊ>ÀÀ>ÞÃÊÜ Ì ÊÌ Ì>

Ê L> `Ü `Ì Ê vÊ ÛiÀÊ{äÊ ÉÃ

UÊ ÕÀÊi Li``i`Ê* ÜiÀ* ÒÊV «ÕÌiÀÃ UÊ « iÌiÊ L >À`Ê-iÀ > Ê,>« ` "Ê

Ê ÃÜ ÌV Êv>LÀ VÊV iVÌ Û ÌÞ

£xÝä

UÊÊÊÊÊÊÊÊÊÊ6 / Ê{£Ê ÀÊÊÊÊÊÊÊÊÊÊÊÊÊ6 / Ê{ÈÊv À >Ìà UÊ,Õ}}i`Ê ÀÊ i`Ê> `Ê `ÕVÌ

Ê i`ÊÛiÀÃ Ã

£ÈÝä

Ê >à ÞÊÌ iÊ ÃÌÊ «ÀiÃà ÛiÊL >À` ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊiÛiÀÊLÕ ÌÊv ÀÊ * Ê«À ViÃà }° ÊÊÊÊÊÊÊÊÊ+Õ ÌiÊ« Ãà L ÞÊÌ iÊ ÃÌÊ «ÀiÃà Ûi ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊiÛiÀÊLÕ ÌÊv ÀÊ6 ° Á

,"Ê ",9 x{äÊ6>ÃÃ>ÀÊ Ûi Õi]Ê >ÌÃÜ ÀÌ ]Ê > v À >Ê £Î££Ê1°-° ° ­1-®Ê/i Ên£nÊ nÊääÇäÊÊUÊÊ­1-®Ê >ÝÊn£nÊ nÊ{{x ÜÜÜ° VÀ i ÀÞ°V ÊÊUÊÊÃ> iÃJ VÀ i ÀÞ°V