Design News is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


Articles from 2004 In June

Exclusive Hands-on Review Display Kits

Instead of using discrete LEDs as indicators in a design, you might consider opting for a display that can present anything from basic alphanumeric information to full-color graphic images. Several kits now make it easier to connect different types of displays to a PC and run them through tests. To find out what these kits can do, I obtained several, ran the demo software, and exercised them independently. All kits came with serial-port connections that attached to my Windows 98SE lab computer. Most embedded-computer boards provide serial ports, so interfacing them to a display module should be a snap. Software and command sets, though, present challenges. Except for a few kits, most lacked any sample code written in C, Visual Basic, or other languages to help users determine what they would need to include in an application to make a display work.

Display Kits
Product Best Features Ease of Use Applicability Limitations Comments
Parallax G12023 Graphic LCD $109 Good display for its size. Quick setup. Simple interface accepts either TTL or RS-232C voltages. Helpful demo programs. Commands are easy to interpret and use. The BS2 computer provides an excellent way to put the display through a series of tests. Good in systems that need only a few lines of characters and simple monochrome graphics. Excellent in portable equipment. Graphics limited by the pixel count, but that's inherent in any small display. Almost any serial output-port can control this display. The Parallax BS2 computer simplified testing.
Cybernetic Micro Systems CYB-308 LED Display Kit 8 chip $245 Bright display for up-close applications. Easy setup, but difficult to figure out how to use commands to control the display. If an application needs this type of display, but not a serial interface, choose something else. This type of display will work if you need only characters and a few special symbols. No graphics. Smart display modules are expensive. The manual glosses over or ignores display-module features. If you want to use smart display modules, buy them from Agilent Technologies and do the integration yourself.
Noritake GUD 3000 ($288), GUD 7000 ($175) Graphic Development Kits Crisp, clear, easy-to-read display. Large characters and a good user interface. I/O ports on the GUD 3000 display provide unexpected flexibility. Easy to use and control. Top contenders if you need bright monochrome displays of text and graphics. A good match for combining graphics and characters. Viewports provide small areas you can control independently. Once in a while the user interface caused errors that forced a software restart. That limitation relates to the supplied software, not to the display. The manual on the accompanying CD-ROM suffers from poor organization and lack of an index.
Cybernetic Micro Systems CYB-003 Complete Starter Kit $595 Good use of high-level commands. Kit accepts several display types. An array of predefined viewports sets up separate regions for text or graphics. Easy setup using software that shows examples of kit options. The manual can seem overwhelming, though. An excellent way to control a large LCD without having to develop much software. Supplier offers a free limited-capability version of its CyberCom software. Users must buy the complete package at extra cost. I liked the CyberCom software and the availability of I/O lines on the module. Sample code seems dated, though. It runs on an old 8051 CPU.
Amulet Technologies Easy GUI 5.7" Starter Kit $399 HTML coding simplified development of display data. Combining graphics and a touch screen makes for a powerful display system. Setup goes smoothly and test displays come up right away. Using the HTML complier takes getting used to, but error messages help find and correct any coding problems. A great package for applications that must combine graphics and accept user input through a touch screen. The innovative use of HTML coding lets developers quickly set up complex displays and controls. Web-page design may take getting used to, but software provided does most of the work. Could cause a steep learning curve for someone without some exposure to HTML. Lots of sample code and examples will make this display system a hit. An HTML reference book will help users who have not designed Web pages. ezLCD Evaluation Kit $199 Full-color LCD will suit many needs. Supplied software has limited use. I wrote a Visual Basic program to exercise display commands and options. A nice display, but your software will have to keep track of many details. The small set of commands can add overhead to your application software. The display-controller firmware should offer more commands. A nice display, but the kit seems more a work in progress than a finished unit ready for engineers to evaluate. The company supports a discussion group on Yahoo.

G12032 120x32 Graphic LCD&?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />&o:p>&/o:p>

Summary: A good basic display that combines graphics and characters. The module's simple interface makes it easy to control with an embedded system. &o:p>&/o:p>


Analysis: This backlit display provides a 120 X 32-pixel area (2.3 X 0.6 inches) that will suit applications in handheld equipment, industrial controls, and similar devices. The display uses ASCII commands and its serial input accepts either ±10-V RS-232C or 5-V TTL signals. When first powered, the LCD runs through a visual self-test to show it works.&o:p>&/o:p>

Although the G12032 module will run directly from a PC's serial port, my tests used a BasicStamp BS2 computer module, also provided by Parallax The BS2 computer came with several demo programs from which I extracted code for my own experiments. After developing command sequences on the BS2 it would take little time to move them to another embedded controller.

Descriptions of the control commands seem complete and they include a helpful table of ASCII control codes and the corresponding keystrokes that create them. The display offers four or two lines of text with either 10 or 20 characters per line, depending on the font size you choose. The display responds to the carriage-return (CTRL-M) and line-feed (CTRL-J) characters to position text on a following line.

Graphics instructions use "escape sequences," such as "ESC P, x, y" that plots a pixel at point x, y. This instruction, and a few others, let me plot signal waveforms on the display. At first, I couldn't get the display to plot properly. The table of graphic commands shows the sequence of instructions, but it leaves out details of how to transmit those commands. The introduction to the commands includes them, but they should get re-emphasized in the command summaries, too.

The display's onboard memory stores fonts and as many as 15 full-screen displays. The ability to quickly update the display with stored images saves time. If an application requires several standard messages and graphics, you can set them up in preprogrammed screens. Then, updating the display requires only a simple screen-change command (CTRL E n), not a complete download of new data.&o:p>&/o:p>



CYB-08 Kit and CY308 chip $245&o:p>&/o:p>

HDSP-2112 display, $60 (1)&o:p>&/o:p>

CyberCom $75 (CD-ROM)&o:p>&/o:p>

Summary: The display modules provided with this kit work well on their own, so why complicate a design with an extra controller chip?&o:p>&/o:p>


Analysis: The kit arrived with the CY308 LED controller chip and four HDSP-21xx 5x7 dot-matrix LED display modules. The four display modules stack end to end to provide a line of 32 characters. Even without a filter, people can easily read the 0.2-inch-high characters. (Agilent Technologies sells the display modules in red, orange, yellow, or green versions.)

I had designed an interface using the same "smart" display modules, so I wanted to find out how the added controls provided by the CY308 chip would enhance the display's characteristics. After loading and running the CyberCom software, I clicked on the "LED" tab to configure the board for serial communications. Then, instructions guided me to set a display mode, send characters to the display and shift them right or left. Then I hit a brick wall.

The paper manual provides no examples of how to use the display's many commands, and for the most part, the explanations of control commands provide little assistance. The CTRL-H command, for example, should backspace over a displayed line of characters, much like a backspace key. During tests, though, the CTRL-H command simply cleared an entire line. Several changes to the display's operating mode yielded no suitable results. The manual should explain what a command does, how to use the command, and why and how to use it.

On their own, the HDSP-21xx smart display modules can dim, blank, or flash characters, attributes the kit's paper manual fails to cover in any detail. The modules also provide for 16 custom characters. Based on my experience with these display modules, I recommend you skip the CYB-308 kit and use just the smart modules along with firmware in your embedded controller. Each module's microprocessor-compatible interface simplifies control of many such modules in a system.&o:p>&/o:p>


GUD-3000 Graphic Development Kit, $288&o:p>&/o:p>

GUD-7000 Graphic Development Kit, $175&o:p>&/o:p>

Summary: A nice combination of graphics and characters on bright light-generating displays. The high-contrast output makes information easy to read under many lighting conditions. &o:p>&/o:p>


Analysis: Noritake sent me two displays, a GUD-3000 (256x64) and a GUD-7000 (140x32), along with a CD-ROM that included information files and programs that let me quickly display text and graphics. A test program's graphical user interface (GUI) transmitted character strings, custom characters, and graphics to the display. Several test patterns come on the CD-ROM.&o:p>&/o:p>

The GUI let me set up viewports, adjust scrolling speeds, observe character strings sent to a module, and so on. Many of the GUI controls seem intuitive, but complete instructions come on the CD-ROM. I particularly liked the capability to quickly design custom characters and to record strings of commands for later use. Once in a while the GUI causes errors that forced a software restart.&o:p>&/o:p>

To control the displays independent of the GUI, I sent them ASCII characters and control codes from HyperTerminal, a simple terminal-emulation program that comes with Windows. The data sheet for each display type includes information about hexadecimal codes and string formats that simplify establishing fonts, defining cursors, downloading characters, setting scrolling speeds, and so on.&o:p>&/o:p>

The GUD-3000 display includes a 4-bit input port and an 8-bit I/O port. These ports can control local devices, such as a keyboard, indicators, or other digital devices. The host processor transmits ASCII characters to a display module to control the ports. I didn't have time to test this capability.&o:p>&/o:p>

The sample C-language code on Noritake's CD-ROM will run on a Hitachi H8/325 microprocessor and control a display module using a parallel interface. Even though that interface doesn't match the one shipped with the kits, users can glean information about ASCII control sequences from the listings.&o:p>&/o:p>


CYB-003 Complete Starter Kit&o:p>&/o:p>

Summary: The powerful high-level commands and thorough documentation supplied with the Cybernetic Micro Systems LCD kit make it an excellent choice.&o:p>&/o:p>


Analysis: The kit I received included a CYB-003 controller board and an Optrex 240x60 monochrome display. The CYB-003 board will work with LCDs from suppliers such as AND, Optrex, Toshiba, Densitron, and Sharp, as noted in the documentation.&o:p>&/o:p>

Upon applying power, the controller cycled through a series of screens to demonstrate its capabilities. Removing a 40-pin demo IC let my PC control the display through a standard serial port. The package comes with paper manuals for the kit and for the embedded CY325 controller IC.&o:p>&/o:p>

During setup, the CD-ROM displayed board images that let me choose the proper controller so software installation could proceed on its own. The installed CyberCom program tests the serial connection to the controller and then runs a demo script. A CyberCom window displays the script's contents so I could see the transmitted commands and comments used in the demo. Two additional windows show commands and data I had CyberCom send to the display and what the display has transmitted to the lab PC. Another window let me type in commands and data that got sent to the display.&o:p>&/o:p>

Instead of running additional demo routines, I jumped into the manual and tried some of the 20 sophisticated commands. It took only a few minutes to learn how to initialize the display and send it some text, which the display automatically scrolls and wraps. All command strings sent to the controller start with a CTRL-C (control mode) and end with a CTRL-D (display mode), and the commands have a logical format. The command B x1,y1,x2,y2, for example, draws a box between diagonally opposite corners.&o:p>&/o:p>

The heart of the kit, the CY325 chip--which the company will supply separately--includes many built-in capabilities such as 250 predefined viewports developers can use for graphics or text. All viewports operate independently, so developers can plot pixels, draw lines, set font sizes, and so on, as needed. The simple command structure and sophisticated operations make the CYB-003 a great way to control a display.&o:p>&/o:p>



Easy GUI Starter Kit&o:p>&/o:p>

STK-GT570 &o:p>&/o:p>

Summary: The ease of setting up a display using HTML coding combined with a touch screen for user interactions make this display a top choice.&o:p>&/o:p>


Analysis: The company sent me a 360x240 pixel monochrome display (5.7 in.) connected to a board that provides a combination LCD-controller and microprocessor IC. The IC produces displays from information that includes hypertext markup language (HTML) tags used to create standard Web pages. By saving "web pages," the system takes most of the display-management burden off a host processor. Think of the display as a simplified browser displaying pages saved in local flash memory.&o:p>&/o:p>

After compiling a demonstration "Hello World" file using the Amulet HTML Compiler, a mouse click sent it to the display's flash memory. Any errors in HTML files get caught by the compiler. The compiler cannot accept all HTML commands, though, so documentation explains acceptable commands and their syntax.&o:p>&/o:p>

HTML coding for the Amulet system takes advantage of the display's touch screen. Like a mouse click on a Web page, a tap on any displayed "hot links" causes an action such as a jump to a new page or a serial transmission back to the host computer. The hot-link feature alone makes the display attractive in applications that require user interaction.&o:p>&/o:p>

I decided to create my own "Web page" that would contain a hot link to a second page. Amulet recommends developers use the Namo WebEditor (, and a 30-day trial version comes on the CD-ROM. WebEditor let me drag and drop elements on my page and I could view the HTML code, too. After I corrected several errors the compiler detected--I had to read the manual--my pages appeared on the display. Some display layouts may require a bit of hand tweaking.&o:p>&/o:p>

Sample files show working HTML code you can easily adapt. I copied several lines of touch-screen HTML code to use on my pages. With some knowledge of HTML, a developer can quickly have a display set up for testing. The capability to interact with a user through hotlinks enhances the value of this display system. I regret not having time to fully explore the Amulet system's many features and capabilities.&o:p>&/o:p>



ezLCD Evaluation Kit&o:p>&/o:p>

ezLCD-001-EDK &o:p>&/o:p>

Summary: A nice, small color display, but the dearth of commands and a work-in-progress feeling may deter all-but-the-most-determined users.&o:p>&/o:p>


Analysis: The ezLCD-001-EDK kit from contains a 240 X 160 pixel color LCD (2.7-inch) attached to a small interface board. The included CD-ROM provides clear start-up instructions and also explains how to program the board with the latest firmware, also on the CD-ROM or available from the company's Web site.&o:p>&/o:p>

The supplied program, Av232, provides a control panel that let me draw lines, set the display to a chosen color, and transfer several test bitmaps to the display. The bitmaps look good, and I think the display would work well in many applications. Unfortunately, the software doesn't do much more, and the kit seems like a work in progress rather than a full-featured package. &o:p>&/o:p>

The documentation explains 13 commands ranging from backlight control to placing lines and characters on the screen. Setting the display to a color, for example, requires the command string "$ byte !" where "byte" determines the color. Unfortunately, the software doesn't send those values to the display! At first I used HyperTerminal, and then I wrote a short Visual Basic program to control the display. The lack of a rich command set means your application code must track the positions of characters and other display characteristics.&o:p>&/o:p>

I liked the ezLCD kit's display, but the small command set and the lack of a tool to test commands may deter all but seasoned developers. If the vendor upgrades its software, the kit may appeal to a wider audience.&o:p>&/o:p>


Test Software:&o:p>&/o:p>

GO HYPER&o:p>&/o:p>


Microsoft provides the HyperTerminal terminal-emulation software with most versions of Windows, at least through version 98SE. Early versions of Windows installed HyperTerminal automatically, but later versions may require manual installation from a Windows distribution disk. To see if you have HyperTerminal look in the Communications portion of Windows' Accessories menu. The software lets you transmit and receive ASCII information through a PC's standard serial port at various bit rates. The free HyperTerminal Private Edition (HTPE) from Hilgraeve (Monroe, MI; works with Windows 95, 98, and NT 4.0. HTPE offers more capabilities than Microsoft's version.&o:p>&/o:p>


In several tests, I used Visual Basic 6.0 code to transmit strings to a display kit. You can find the short listing at Change the strings and characters as needed. Ensure the MSComm control exists in VB, and drag and drop it into your VB project to use my code. (You may need to load the VB MSComm control.) I set up the VB program to transmit characters embedded in the code, but you may decide to put display-control codes in a window and then transmit them to a display. I embedded codes in the VB software so I could easily modify and repeat sequences without much retyping. For more information on serial communication, refer to "Serial Port Complete," by Jan Axelson, Lakeview Research, Madison, WI. You can download serial-port code from Jan's Web site.&o:p>&/o:p>


You also can find the Basic Stamp code for a simple square-wave and triangle-wave pattern used with the Parallax G12032 120x32 Graphic LCD at Web page. The code runs on a BS2 computer from Parallax.--JT&o:p>&/o:p>


Defibrillator saves lives

Defibrillator saves lives

You've all seen the dramatic scenes on TV or in movies: An emergency medical technician or emergency-room specialist gets a cardiac-arrest case, and time is precious. The specialist quickly readies the victim, fires up the defibrillator unit, slaps on the paddles, and yells "clear." A few commands of "clear" and some jolts later, the patient comes back to life or, after a longer wait and more applied shocks, a doctor declares the patient dead.

The key factors in the outcome of this familiar scene are the defibrillating shock, along with the time period that elapses before applying it. The prospects of a successful outcome for the patient drop dramatically with every minute after the cardiac arrest occurs. Yet, defibrillator units are relatively costly and require considerable training to use. These units do more than administer a fixed, high-voltage zap. Instead, they must deliver a controlled, complex waveform into the variable impedance across the patient. This waveform differs from individual to individual, and the defibrillators must deliver these jolts while monitoring the faint electrocardiogram signals on the patient's skin. The sequence and system settings must be just right to be effective: too few joules or the incorrect waveform give an ineffective jolt; too many joules produce an overwhelming jolt that may seriously harm the patient.

To address these problems, Royal Philips Electronics began to sell its HeartStart home defibrillator to consumers in November 2002 after receiving Food and Drug Administration approval. Consumers must have a doctor's prescription to purchase the device. The company designed this unit, which took about two years from project initiation to engineering release to complete, with the key goals of adherence to FDA and mandatory medical protocols; ease of use by untrained users, or "lay responders"; and low cost. The fully self-contained, battery-powered unit measures 6.6321.8321.8 cm, weighs 2.1 kg with batteries, and sells for a retail price of $1,995.

A Shock to the System

Many U.S. patents cover the system design embedded in the HeartStart unit. One patent discusses the waveform it uses and the algorithm that controls this waveform; the other focuses on the circuitry design and the self-test algorithms. Although you might assume that the Philips unit uses a high-end processor, perhaps adding a DSP, the company chose a different approach. Instead, the unit uses a Motorola HC16 processor and an HC08.

The operational sequence begins before a user applies a jolt. The unit has a voice that tells the probably panicked lay responder what to do. The unit has no written instructions and displays no front-panel messages. The responder applies special pads to the patient, the unit senses that they are applied, and an automated sequence takes over. The system first reads the patient's ECG (electrocardiogram) to determine if a shock is appropriate treatment, and it simultaneously determines the impedance of the chest by injecting a small ac current and measuring the resulting voltage drop. With this information and its determination that the patient requires a defibrillation shock, the unit is ready to go.

The defibrillation waveform itself has total energy of about 150J, delivering about 1,800-V peak value and 20-A peak current; nominal load of the patient is 90-V. (Note that the voltage and current levels, energy, and protocols differ widely between internal and external defibrillators and pacemakers; these numbers apply only to external units).

Power for the unit comes entirely from a series-connected set of nine standard CR123 lithium batteries, such as those that cameras use, supplying a nominal 9V with 4.2-Ahr capacity. For ease of replacement, they are available in a custom battery pack rather than individual batteries, which users could insert incorrectly. The battery pack has enough reserve to deliver 90 typical shocks and provides standby power for four years.

The unit delivers the shock waveform once, and the defibrillator monitors the sourced waveform and then measures the patient's ECG to determine whether the jolt was successful. The American Heart Association protocol recommends that a defibrillator provide as many as three consecutive shocks, and the operator should then perform CPR (cardiopulmonary resuscitation) if the patient does not respond. Users of this unit, however, may not know or may be unskilled at CPR, so the unit uses a calm voice to coach the user through the CPR sequence. It also keeps prompting the responder to administer jolts until the patient is revived or the battery runs out. The defibrillator delivers a waveform that must meet stringent requirements and is critical to increasing the likelihood of success.

The design uses standard, off-the-shelf components, including a custom standard-cell ASIC that provides various circuit functions and supervision of disparate parts of the system. The designers of the unit simulated the performance of the ASIC simulating the critical 1.6 seconds of its actual operating time. The simulation took about 11 days on a high-end PC. Other than the simulation tool, the designers used no advanced EDA tools. The primary tools were standard board-layout software and numerous physical prototypes, which tested many packaging- and cost-related concepts. A key to minimizing cost was to reduce the number of internal pc boards, despite the mix of conventional and high-voltage signals.

Despite the waveform's complexity, the designers chose to use neither digital-waveform-synthesis techniques nor a supercapacitor, sometimes chosen for high-energy, high-rate-discharge applications. Instead, they chose a 100-mF film capacitor. The circuit controls the discharge through SCRs (silicon-controlled rectifiers) and an IGBT (insulated-gate bipolar transistor) with special attention to the di/dt parameter. The discharge-control circuitry generates the positive-going waveform, turns off, and then generates the negative-going waveform.

The designers also did not digitally synthesize the soothing voice that guides the untrained user. They instead used the voice of Peter Thomas, the reassuring narrator of the Public Broadcasting System Nova series. The defibrillator stores the voice and re-creates his speech using standard ADPCM (adaptive-differential-pulse-code-modulation).

In addition, testing the unit was critical to the design team, which faced the issue of how to know whether the design works to the mandated medical protocols and specifics. The designers needed to ensure that the unit, which can sit unused for long periods, will still perform properly.

To address these issues, the team used industry-standard patient simulators from Symbio and Dynatech-Nevada, now Fluke Biomedical. For the ongoing system check, the unit performs a daily self-test, as well as a series of tests when a user replaces the battery pack. These tests check battery condition, internal circuitry, the overall waveform-generation and -delivery subsystem, and calibration of key components and signals.

The self-contained HeartStart also has some communications capability. It records the first 15 minutes of the ECG and key parameters of the incident, and can transmit to a PC. Philips provides event-review software for viewing, analyzing, and managing the data from the defibrillator.

In Control: The architecture of the unit relies on one custom ASIC (a) to implement a low-level data-acquisition and analysis system working with a high-voltage, carefully controlled signal source (b).
Web Resources
Check out the links below for more info
Royal Philips Electronics:
HeartStart home defibrillator:

Self-Cleaning Sensors

Self-Cleaning Sensors

Who says sensors are easily contaminated? Craig Grimes has found a material that enables hydrogen sensors to rid themselves of impurities.

PRESENT POSITION: Associate Professor, electrical engineering/materials science, Pennsylvania State University

DEGREES: B.S., M.S. in physics, Penn State; Ph.D. in electrical engineering, University of Texas, Austin

HOW YOU DESCRIBE YOUR RESEARCH WHEN AT COCKTAIL PARTIES: There's a truth that the more sensitive a sensor is, the quicker it gets dirty. But photochemical properties of these sensors used to detect hydrogen can degrade any contaminants that fall onto them. The premise is that you can stick a sensor in the real world and it will continue to work.

WHAT MAKES THESE HYDROGEN SENSORS "SELF-CLEANING"? They are made of titanium nanotubes, lined up beside each other. Since the photocatalytic properties are so large with this nanotube geometry, the tube-to-tube contact creates amazing absorption characteristics that go across the spectrum from UV to visible light; the material absorbs the light, generating an electron-hole pair, which reduces contaminants.

ARE THERE ANY CONTAMINANTS THE SENSORS DON'T WORK WITH? Yes, certain types of salts will degrade the photochemical properties, so they don't work in the ocean or with WD-40.

WHAT'S THE BENEFIT OF SELF-CLEANING HYDROGEN SENSORS? First, from a cost perspective, self-cleaning sensors mean that you don't have to replace the sensors as often, if at all. Second, burning hydrogen results in emission of only water, which is perfectly clean. Once we have a useful supply of hydrogen, we can begin to wean ourselves off of our oil dependence.

WHEN WILL THEY BE COMMERCIALLY AVAILABLE? The self-cleaning hydrogen sensors have been licensed to SentechBiomed (; we're now focused on the photochemical properties.

To view images of the titanium nanotubes, go to Contact Grimes at [email protected].

Antimatter: Possible way to treat Cancer?

Antimatter: Possible way to treat Cancer?

Colliding Worlds: Standing inside a Rube-Goldberg-like particle accelerator at CERN, physicist Michael Doser explaisn to journalists how his team hopes to employe antimatter to destroy cancer cells. Located in Geneva, CERN is world famous center for nuclear research.

In a big pit at CERN, the European Institute for Nuclear Research in Geneva, a particle accelerator churns out antimatter that has the potential to treat cancer. A California company, PBar Labs, is sponsoring experiments to determine if antimatter can zap tumors in hamster cells. If this therapy works, plans are to treat human cancers. Unlike conventional radiotherapy, antimatter can be precisely calibrated to destroy only cancer cells. And as antimatter doesn't travel far in air, patients wouldn't need much shielding. The accelerator's proton source is a heavy target like tungsten. In the accelerator, these protons collide. Each million-proton collision generates one proton-antiproton pair. Physicists separate the antiprotons from the rest of the soup and slow them down in an antiproton decelerator shaped like a 100m-diameter doughnut swathed in foil and duct tape. The antimatter particles annihilate when they encounter matter, thus they can only be stored suspended in a vacuum. Antiprotons can be precisely targeted so that only in the last millimeter of their path (i.e., in the tumor) they slow down, interact with an atom and are annihilated. "The body is basically empty space," says Michael Doser, a physicist on the project who looks as if he were sent straight over from Hollywood casting. "All matter is illusion. To an antiproton, the body is just an array of atoms. Atoms consist of a tiny nucleus surrounded by a huge cloud of electrons. Antiprotons interact only with the nucleus. If you randomly kick a football through a field that only contains one other football, it is unlikely to hit anything, Doser explains.

Corrosion no more

Corrosion no more

The problem of localized corrosion is about to get a harder punch now that engineers at the Southwest Research Institute have developed the Multielectrode Array Sensor System (MASS). MASS uses miniature electrodes to detect where localized corrosion is occurring in everyday structures such as bridges and airplanes. With earlier detection, problems caused by localized corrosion-one of the most common ways for engineering components to fail-can be fixed easily and promptly. For info visit

RFID Extends to Medical Applications

RFID Extends to Medical Applications

Engineers creating tests for biowarfare threats such as anthrax wanted to put more data on sample collection containers than bar codes could provide. They turned to a technology long used in packaging and materials handling that's starting to gain acceptance in the medical and life sciences field-RFID (radio frequency identification).

"RFID has a lot of features that bar codes don't have," says Jim Whelan, president of Alexeter Technologies, a Wheeling, IL maker of systems that test for anthrax, ricin, and other hazardous materials. The company currently makes field units for emergency teams, but is developing a PDA-based unit designed for labs. In both markets, data management and traceability are key factors driving the need for RFID tags.

Foremost among RFID benefits over bar codes is memory capability, which lets engineers store data on a tag. Even the basic 128 bytes on chips made by Maxell Corp. of America are enough for the first members in the PDA line. Alexeter stores 14 parameters in about a third of the total memory, using just the write-only registers. "We don't want anyone to be able to write over data," Whelan says.

Another critical factor is small size.

The Maxell tags fit on the 1- x 3-inch disposable units used to collect suspicious powders. "We were able to put the tags on our existing cartridge without altering the molds, which was very important to us since re-tooling is very expensive," Whelan adds.

The RFID chip measures only 2.5 mm square including the antenna. Some chips require antennae that are nearly the size of a credit card, but Maxell's antenna is housed on the IC. "The antenna goes around 64 times. It's electro-formed on top of the silicon, not etched," says Rumi Kitatate, marketing manager for new product development at Maxell, headquartered in Fairlawn, NJ.

Maxell is also developing its own medical product-test tubes with embedded RFID tags. The tags make it simple to track each sample in a study. "There are 96 test tubes in a tray, they're very close together, and each needs to be read independently," Kitatate says.

By putting tags on the bottom of the tubes, a reader scans them when the rack is placed atop the reader. Kitatate notes that the readers are much less expensive than those for 2D bar codes now used for this type of task. "Those readers cost $3,000 to $5,000. Ours will be about half that," he says. The reader can be as small as a credit card and can attach to PCs using USB connections.

Chips will cost under $1 each, Kitatate adds. Even though the RFID tags will have a fair amount of memory, speed is well within most lab requirements. "It takes less than a second to read 128 bytes," he notes.

Elsewhere in the medical and life sciences market, researchers are exploring a number of programs that use RFID tags. For example, Intel Research Seattle in Washington is working on a program that puts tags on medical bottles and other items in homes of elderly and ailing people who want to live at home. The tags make it possible to determine whether medications are taken and other routine tasks are being performed, saving money by avoiding the cost of moving the person to a nursing home for monitoring.

The disposable devices measure 1x3 inches. Maxell's integral RFID tag and antenna, top left, is only 2.5 mm square. Its antenna is elctro-formed right on top of the IC silicon, bottom. RFID market growth, right, is being fueled by applications outside traditional materials handling, security, and automobile immobilization uses.

Web Source
RFID Resources:

Alexeter Technology
Maxell Corp. of America
Sun Microsystems/Capgemini

Use Grows as Size Shrinks

Use Grows as Size Shrinks

Automotive engineers continue to use electronics to add features and functions, driving growth for processor suppliers. While much focus is on high-end devices, 8-bit chips continue to see solid acceptance.

The automotive market is by far the largest user of 4- and 8-bit processors, which will continue to outship their high-end brethren, according to Dataquest Gartner ( In 2006, 8-bit shipments (with a fraction of 4-bit devices) will hit $3.9 billion, compared to $3.5 and $1.9 billion for 16- and 32-bit microcontrollers, respectively.

Even in demanding applications such as airbags, where processing must happen in milliseconds, their low cost and high power continues to earn them design spots. "We're using 8-bit microcontrollers in a variety of restraint products including airbag systems," said Thomas Stierle of Siemens VDO Automotive's Restraint Systems group in Auburn Hills, MI (

The growing number of airbags is among many applications that are keeping 8-bit volumes high.

"Approximately 50 percent of the processors in vehicles are 8-bit chips, it's still a growing area," says Willie Fitzgerald, automotive products marketing director at Chandler, AZ-based Microchip Technology (

Market watchers say there's still plenty of potential for small 8-bit microcontrollers to convert mechanical control functions by using electronic features. "There's always a place for mechatronics, adding a processor to sensors or actuators. As price comes down, that certainly becomes more attractive," says Paul Hansen of the Hansen Report on Automotive Electronics (

Microchip recently shrank an 8-bit processor down to a 6-pin SOT-23 package that it says is the smallest packaged CPU (see DN 06.07.04; Aimed at tier 1 automotive suppliers such as Bosch, Siemens VDO, and Delphi, the PIC10F holds 512 words in flash and 24 bytes of SRAM.

"The primary role is inside the car, either as an instrument or to create an atmosphere," says Fanie Duvenhage, product marketing manager for Microchip's Security, Microcontroller and Technology Development Division. The microprocessor creates various frequencies for chimes and other alerts, while lighting controls handle slow ramping and dimming, providing more capability than electromechanical controls, he adds.

Another potential application is to replace a watchdog timer, giving designers more flexibility and capability, Microchip's Fitzgerald says. The chip has a built-in timer, so programmability for extra functions can be added without increasing size. In these and other applications, the chip can replace a number of discrete components.

With the chip's price as low as $0.49 in quantity, "The cost is often comparable to a few logic gates and transistors," says Duvenhage.

In addition to its small size and price, marketers tout the simplicity of using a processor instead of components. "Engineers can design this in faster than they can find the perfect combination of discrete parts," Duvenhage says.

Still Growing: Eight-bit controller use continues solid growth in automotive applications, spearheaded by burgeoning airbag systems.

LEDs Help Save Firefighters

LEDs Help Save Firefighters

Because a new federal safety regulation says that firefighters must have a clear view of how much air is left in their self-contained breathing apparatus (SCBA), LEDs are being used to illuminate the indicator. The LEDs are easy to see through smoke and also function as a warning system to other firefighters. When operating normally, the LEDs give off a blue-green light; if assistance is necessary, the LEDs give off a yellow color. The series of LEDs selected for this task use a low amount of power, which means fewer battery replacements. For more information, go to the following website:

ColdFire Adds Interface Options, Encryption

ColdFire Adds Interface Options, Encryption

Freescale Semiconductor, the operation that's being spun-out from Motorola soon, is nearly doubling the number of products in its ColdFire line. The mainstays of the 20 new microprocessor devices are the MCF547x and MCF548x, higher-end parts that have more communication capabilities than previous family members, as well as data encryption and improved timing control.

The latest members of the ColdFire family all have CAN, Ethernet and USB on-chip, as well as a new CPU core. That core, the V4e, runs at up to 266 MHz for the 547x line, with a peak of 200 MHz for the 548x (

In addition to such clock rates, measured on another scale with speeds up to 410 Dhrystones, the line has two CAN 2.0B modules and one or two 10/100 Ethernet controllers. Those are the two mainstay networks in factory automation, medical instrumentation, and the important HVAC and building control field, which is the largest industrial segment for ColdFire devices.

Crestron Electronics Inc., a Rockleigh, NJ maker of control systems for lighting and HVAC control makers, picked the line for a forthcoming upgrade of its Isys touch sensitive control systems ( "Having multiple Ethernet, USB, and serial ports in a low-cost processor simplifies the design," says Fred Bargetzi, Crestron's vice president of technology.

The company also notes that data encryption is becoming increasingly important in a variety of fields. Designers want encryption even at the node level. For example at a college, encryption could thwart a mischievious hacker who might want to >>change an air conditioning node to heat, says John Sansing, standard products operation manager at Freescale's Transportation and Standard Products Group in Austin, TX.

Freescale is also upgrading its V2 core, adding 100 MHz to take performance up to 166 MHz. On the MCF523x, that core is integrated with an enhanced timing processing unit, which Sansing dubs "the mother of all timers. It's non-intrusive to the CPU, so the CPU can do other things." Most other timers require much interaction with the CPU, he adds.

Connections: The USB, fast Ethernet, and CAN links (optional modules, upper left) make the latest ColdFire products simpler to intergrate.

Rivaling a Fruit Fly's Brief Existence

Rivaling a Fruit Fly's Brief Existence

Galil Motion Control, a maker of microprocessor-based motion controllers, received an odd call from one of its suppliers recently. The caller wanted to know whether or not Galil had any spares of a particular electronic component on hand. Until then, Lisa Wade, VP of sales and marketing, had no inkling that the supplier was no longer planning to offer the component, which is used in a four-axis amplifier board that Galil makes.

"As soon as we got the call, we started to search for a substitute, but one didn't exist," says Wade. Instead, Galil engineers had to modify the board and change the spec. "This was a worst-case scenario," notes Wade. "Sometimes you get a year's notice, sometimes you get no warning."

The threat of component obsolescence is a growing problem for design engineers, particularly for companies like Galil, whose products may have life cycles of a decade or more. For many companies, the component they specify for their product today may be discontinued in 18 or 24 months. No big deal if you're designing cell phones, which have their own quick product life. But it's a whole other story for end products that have staying power.

One significant change today is that electronic components have become ubiquitous. They're now in cars, medical devices, durable appliances, and military tools-products that typically have a much longer life cycle, five to ten years, as opposed to cell phones or DVDs that share the lifespan of 18 to 24 months of the electronic parts inside. The problem is made even more difficult when OEMs hurry their design process by using existing designs for new products, thus creating products with components that are already aging.

"OEM design engineers have to be more careful when they're designing products," says Cathy Whittaker, VP of Arrow Alliance, a unit of Arrow Electronics Inc. that focuses on global customers and contract manufacturers. "They have to be knowledgeable about the life cycles of the components." She notes that the medical and military industries try to get around this challenge by choosing parts with multiple sources, the hope being that if a component goes obsolete from one supplier, another supplier will still produce it. So apprehensive is the military about supplies drying up that when forced to use a single source, it purchases a 10-year supply of the component.

In may ways, electronic distributors who have become accustomed to managing tens of thousands of part changes and end-of-life notices, are stepping up to the plate to help engineers manage the problem. Distributors generally get information on component changes directly from the suppliers themselves. "When we receive notice, we upload the information into our database within 24 hours," says Arrow's Whittaker. Once the distributor logs the information, a notice goes out to all its customers. "We identify what customers have purchased the component over the past year, and send a notice out to them all," says Whittaker.

In some cases, distributors know that a part will change or become obsolete while the product is still being produced. And again, distributors will try and send this information along to their customers as quickly as possible. Newark InOne, for example, provides customers with advance substitution notices. "We let people know in advance of changes so they can make a last-time buy or switch to an alternative and ensure an uninterrupted source of supply," says Steve Hopkins, VP of product management at Newark InOne.

Avnet Inc. also keeps customers alerted to part changes ahead of time. "We predict the end-of-life of parts," says Glenn Bassett, VP of Avnet's Premier Services Group, a unit that monitors supply chain issues. "You can tell when a part might go out of life by part class." Avnet offers a service that alerts its customers to end-of-life information even on parts that the company doesn't distribute. Arrow provides a similar service to its customers. Both companies maintain significant databases on components and offer access to the information as fee-based service.

Even the best distributors don't catch all end-of-life notices. "We receive good information from most of our suppliers, but there are some that are notoriously bad about providing information," says Bassett. He declined to specify exactly which suppliers are the offenders.

So what to do when the parts you need do go obsolete? If you absolutely, positively can't spec in a replacement, there are several sources for still locating parts that are no longer made. When components go obsolete, remaining inventory is flushed out to the secondary distributors that trade in overstock and out-of-use parts. Companies such as PartMiner Inc. and US-Bid Inc. scour the globe for out-of-use parts. "When obsolete parts are unavailable through traditional channels, many engineers come to PartMiner," says President Steve Codispodo. "We source parts through our network of 6,000 suppliers." PartMiner also offers subscriptions to its CAPS component database that cross-references components, helping engineers find replacements. And as a last resort, engineers could always try calling their customers-as was the case with Galil!

A Gazillion Notices: Part changes and discontinued components hit a high mark during the late days of the electronic industry recession in 2003. Though large numbers of componenets were either changed or discontinued, the number is actually a small percentage of all overall parts, less than 2/10ths of a percent.