These days, it’s the rare professional who does not carry pocket-size information appliances. Pagers, cell phones, and palm-size organizers are revolutionizing the way we compute and communicate.

Ever since the PC was introduced, the test and measurement market has ridden the wave of technical advances in PCs and peripherals. These advances were, in large part, driven by the needs of business users. Nonetheless, the way test and measurement systems are now developed and deployed has undeniably changed because of the PC, and all companies involved in test and measurement have been affected. Just consider how fixed-function benchtop instruments now attempt to emulate the standard PC user interface (e.g., the HP/ Agilent Infinium oscilloscopes).

The parallels with the nascent handheld computer market are intriguing. In this burgeoning market, large companies are vying for the attention of the business world, bringing an ever-expanding array of technologies to the business professional’s and the field worker’s shirt pocket. Not only do we have fixed-function information appliances (e.g., pagers and cell phones), but we now have a variety of choices in programmable handheld computers, with such companies as Microsoft, Palm, and Symbian vying for the title of standard-bearer. Devices based on Windows CE and Palm OS are becoming more functional, with more features crammed into smaller packages and with longer battery life.

Yet, current handheld and portable instruments are largely fixed-function devices designed to do one or two things well. As generic handheld computers become more ubiquitous and more powerful, it seems only logical that economic and technological pressures will force developers to start basing handheld instruments on these mass market devices and their brethren.

Should the manufacturers of handheld multimeters, network analyzers, and oscilloscopes worry? Will the handheld test and measurement market follow the lead of the benchtop test and measurement market, foisting instrumentation functionality onto mass-market handhelds? Or will the unique needs of handheld test and measurement users limit the encroachment?

A Growing Market

The explosion in handheld computers is undeniable. Research firm Dataquest estimates that the handheld computer market will grow to $7.2 billion by 2003, with an installed base of 32.5 million units. That’s up from about 8.2 million units in 1998 and represents year-over-year growth in excess of 30%.

A typical handheld computer contains a general-purpose CPU, DRAM, RS-232 interfaces, and a graphical LCD with touchscreen. Some also include USB ports, keyboards, and compact flash or PCMCIA sockets.

Present handhelds are controlled by modern operating system software, such as Palm OS and Windows CE, with application programming interfaces that allow third-party developers to create their own application software. Metrowerks’ CodeWarrior and Microsoft’s Visual C++ and Visual Basic are examples of popular tools that are used to develop software applications for handheld computers.

Future handhelds will have sophisticated communications (e.g., Bluetooth, wireless Ethernet, and cellular), and if previous technical developments are any indication, they will provide more computational performance in smaller form factors and with better battery life.

Fixed in Function?

Although not as explosive as the general-purpose handheld computer market, the handheld instrumentation market has seen its share of innovations. Historically, the fixed functionality of a handheld multimeter, oscilloscope, or other instrument provided value to the manufacturer and the end user because it allowed lower cost, more straightforward design, and better performance. The influence of the PC and the increasingly powerful tools for developing so-called applied computing devices (e.g., Internet appliances) are changing this equation though, and handheld instruments are providing more features.

Consider Fluke’s ScopeMeter line of two-channel handheld oscilloscopes. The 190 Series offers a suite of measurements, in clud ing oscilloscope features, such as 200 MHz analog bandwidth, 8 bit vertical resolution, and external triggering; a full-function multimeter, with AC/DC voltage, resistance, and current measurements; and temperature measurements. With sophisticated data recording and trending capabilities, PC and printer interfaces, battery life of 4 hr., and a device weight of 4 lb., the ScopeMeter is the Swiss army knife of handheld instruments.

The huge market in telecommunications is forcing handheld telecom test equipment vendors to push the bounds of functionality and flexibility even further. TTC’s TTC 2000 Test Pad and Sunrise Telecom’s Sunset xDSL platforms are modular handheld platforms that can accept a variety of company-designed plug-and-play modules to perform various telecom tests (e.g., SONET, ADSL, and wireless).

Yet, despite movement toward PC-like ideas and philosophies, all these platforms are still proprietary. Most handheld instrument developers have not totally embraced the concepts of open hardware/software platforms and interfaces and third-party solutions, concepts that are taken for granted by developers and users of PC-based instruments.

Palm OS—Mass-Market Leader Spreads Its Wings

With more than 60% of the market (according to research firm IDC), Palm Connected Organizers (PCO) are the most popular handheld computers on the market. PCOs are redefining the way we use computers. Interestingly, the phenomenon that is Palm is in large part a result of what Palm did not put into the PCO. Instead of incorporating powerful CPUs, lots of memory, large screens, and all sorts of whiz-bang peripherals, Palm has kept it simple. PCOs are based on Motorola’s Dragonball microcontroller (a 68000 variant), Photo 1. Tangent Systems’ Versid MT-20s plugs directly into the Palm’s sync port and supports temperature measurement with two thermocouple input channels. It uses its own AAA battery and does not reduce the Palm’s battery life. The MT-40s (inset) connects via a cable and has four input channels. (Photo courtesy of Tangent Systems, Inc.) Photo 2. Datastick Systems’ MyCorder DAS-1206 connects directly to the Palm sync port and supports various sensor inputs through six analog input channels and one trigger input, with an internal sampling rate of 512 sps. The DAS-1206 also comes as a Springboard module for the Handspring Visor. (Photo courtesy of Datastick Systems, Inc.) running at <50 MHz, lacking a memory-management unit, and using a 16 bit external data bus. As with many microcontrollers, the Dragonball contains many onchip peripherals (e.g., an LCD controller, serial ports, and general-purpose I/O pins). So, while the PCO is by no means a powerful computer, it can be powered by AAA batteries for weeks at a time. By keeping things simple and using small batteries, the PCOs are inexpensive, very small, and convenient for users. In short, PCOs are good at the tasks for which they are designed (e.g., contact management and personal scheduling). With more than 6 million users and 50,000 developers (according to Palm, Inc.), PCOs have become a true platform for third-party software and hardware development, and test and measure ment peripherals have started to appear. Such companies as Tangent Systems (see Photo 1), Datastick Systems (see Photo 2), and Imagiworks have all developed peripherals that plug into the PCO’s sync port and allow low-speed acquisition of sensor data (e.g., voltage, current, temperature, and light). Furthermore, Palm has licensed the Palm OS to other manufacturers, such as Symbol Technologies and Handspring. Symbol has incorporated bar-code scanning into its devices, and Handspring’s Visor has the Springboard slot, a plug-and-play peripheral interface that has true possibilities for test and measurement. Unlike PCMCIA slots, the Springboard slot is open-faced, which allows modules to be as thick as necessary. As a result, modules can contain AAA batteries and strong connectors (see Photo 3). Furthermore, Springboard modules will hold driver software in their flash memory, so every Visor can automatically install, load, and run the appropriate driver (and application) software—no more fumbling with installing software from CDs. Although ideal for low-end test and measurement applications, the simplicity of the PCO may make it inadequate for more demanding jobs. High-speed data acquisition and execution of sophisticated application software on a system designed around the low-end Dragonball processor is questionable. But Palm, Inc., seems to be addressing these shortcomings. Published reports indicate that the Palm OS is being ported to ARM’s higher performance CPUs. It will be interesting to see how Palm OS grows and whether the expansion will be seen as having a multiplying effect on its popularity or as a dilution of its core functionality. With more than 60% of the market (according to research firm IDC), Palm Connected Organizers (PCO) are the most popular handheld computers on the market. PCOs are redefining the way we use computers. Interestingly, the phenomenon that is Palm is in large part a result of what Palm did not put into the PCO. Instead of incorporating powerful CPUs, lots of memory, large screens, and all sorts of whiz-bang peripherals, Palm has kept it simple. PCOs are based on Motorola’s Dragonball microcontroller (a 68000 variant), running at <50 MHz, lacking a memory-management unit, and using a 16 bit external data bus. As with many microcontrollers, the Dragonball contains many onchip peripherals (e.g., an LCD controller, serial ports, and general-purpose I/O pins). So, while the PCO is by no means a powerful computer, it can be powered by AAA batteries for weeks at a time. By keeping things simple and using small batteries, the PCOs are inexpensive, very small, and convenient for users. In short, PCOs are good at the tasks for which they are designed (e.g., contact management and personal scheduling). Photo 3. Handspring’s Visor is based on Palm OS and incorporates the innovative Springboard slot. Although it uses a standard PCMCIA connector, the Springboard slot is open faced, allowing modules to incorporate important test and measurement features, such as internal batteries, tall components, and large form-factor external connectors. (Photo courtesy of Handspring, Inc.) With more than 6 million users and 50,000 developers (according to Palm, Inc.), PCOs have become a true platform for third-party software and hardware development, and test and measure ment peripherals have started to appear. Such companies as Tangent Systems (see Photo 1), Datastick Systems (see Photo 2), and Imagiworks have all developed peripherals that plug into the PCO’s sync port and allow low-speed acquisition of sensor data (e.g., voltage, current, temperature, and light). Furthermore, Palm has licensed the Palm OS to other manufacturers, such as Symbol Technologies and Handspring. Symbol has incorporated bar-code scanning into its devices, and Handspring’s Visor has the Springboard slot, a plug-and-play peripheral interface that has true possibilities for test and measurement. Unlike PCMCIA slots, the Springboard slot is open-faced, which allows modules to be as thick as necessary. As a result, modules can contain AAA batteries and strong connectors (see Photo 3). Furthermore, Springboard modules will hold driver software in their flash memory, so every Visor can automatically install, load, and run the appropriate driver (and application) software—no more fumbling with installing software from CDs. Although ideal for low-end test and measurement applications, the simplicity of the PCO may make it inadequate for more demanding jobs. High-speed data acquisition and execution of sophisticated application software on a system designed around the low-end Dragonball processor is questionable. But Palm, Inc., seems to be addressing these shortcomings. Published reports indicate that the Palm OS is being ported to ARM’s higher performance CPUs. It will be interesting to see how Palm OS grows and whether the expansion will be seen as having a multiplying effect on its popularity or as a dilution of its core functionality.

Handheld and Benchtop— Different Challenges

So what will it take to allow handheld computers to be used for test and measurement in the way that PCs are? Although the parallels with PC-based instruments can be instructive, the differences between the two classes of devices are significant. Today’s handheld computers have limitations that must be rectified if programmable computer-based handheld instruments are to become viable.

Handheld computers are designed to be incredibly small compared with today’s PCs. While field use necessitates such miniaturization, it also presents challenges to designers of handheld instruments. There are many ways to analyze these issues, and I usually group them loosely into several categories:

Ruggedness. Handheld instruments are often used in harsh field environments, potentially much harsher than the conditions in which your typical benchtop instrument operates. The portable must be able to withstand high and low temperatures, dusty and wet atmospheric conditions, and shock and vibration. Most of today’s mass-market handheld computers are not designed for such operating conditions.

Signal Conditioning. An instrument is useless if it can’t adequately condition the analog sensor signals it acquires. Fixed-function instruments usually have built-in signal conditioning. On PC-based programmable instruments, signal conditioning is integrated in bulky cables, on plug-in boards, or in a separate box. For programmable handheld instruments, these implementations are often unacceptable (unless you have extra hands or wish to carry a lab bench into the field).

Connectors and Strain Relief. As in signal conditioning, connectivity in programmable handheld instruments will be an issue because connector panels and the connectors that they house aren’t likely to be integrated in the unit. Yet, unlike in the lab, field use of a handheld instrument practically necessitates adequate strain relief because cables and connectors are likely to undergo significant stresses.

Data Acquisition (DA) Peripherals. High-performance peripherals, such as those based on the PCI bus, are too large and too power hungry for the handheld world. For that matter, although PCMCIA is a popular form factor for handheld computers, many DA PCMCIA cards draw more current than the handheld computers that host them. In a domain where users expect their handheld devices to last for days or weeks on a couple of standard batteries, devices are not designed to power a DA card for one 8 hr. shift, let alone a few days at a time.

CPU Architecture. To reduce power consumption (and cost), handheld computers use CPUs that are more highly integrated, lack floating-point coprocessors and general-purpose DMA capabilities, and have slower clock speeds than desktop computer CPUs. Lower performance limits DA sampling rates and data analysis. Also, while desktop computers are limited to just a few CPU architectures, handheld devices are powered by a variety of incompatible CPUs from such vendors as Intel, AMD, MIPs, NEC, Motorola, Philips, and Hitachi. The propagation of architectures makes software development more complicated.

Memory Footprint. Handheld computers typically store programs and data in flash memory and battery-backed DRAM. DRAM is limited to 32 MB or less, and hard disk drives are rarely practical. So the amount of data that can be logged is limited. Fur thermore, sophisticated instrumentation and data analysis software can’t assume an essentially unlimited supply of memory.

Operating System. The hardware requirements of handheld computers necessitate operating systems designed for the unique needs of handheld computer users. As with CPUs, several largely incompatible approaches exist, including Microsoft’s Windows CE, Palm’s Palm OS, and Symbian’s EPOC (see the sidebars “Palm OS—Mass-Market Leader Spreads Its Wings,” “Windows CE—Lots of Choices,” and “Other Operating Systems Increase Options,”).

Limitations related to hardware and form factor are likely to be overcome in vertical-market devices designed on a core similar to mass-market handhelds but with features and extensions tailored for test and measurement applications. But the great variety of hardware features (such as widely varying display resolutions and color depths) in the handheld market only serves to make the designer’s job more difficult.

From a software standpoint, the main consequence of the proliferation of CPU/operating system architectures and the reduced memory footprint is that software drivers and development tools favored by PC-based instrument developers will not work without significant modification. For example, ComputerBoards must port its software drivers to Windows CE if it wants its PCMCIA cards to work with Windows CE devices. National Instruments must port LabVIEW, and HP/Agilent must do the same for HP VEE.

The difficulty in porting software drivers and development tools from the staid Pentium-based, large-memory footprint PC world to the memory-constrained, multiple-choice CPU and operating system architecture of handheld instruments is not to be underestimated.

As Chad Chesney, Data Acquisition Strategic Marketing Manager at National Instruments puts it, “There are a variety of barriers that must be overcome before user-defined handheld instrumentation solutions will be widely available. Rugged form factors, greater than 8 hr. battery life, and the easy integration of signal conditioning and connectivity will all be gating factors for mass adoption of handheld computers for measurement applications. Perhaps the most significant barrier is the ability for most engineers and scientists to easily develop, install, and execute their software applications on an embedded processor.”

Despite these issues, Chesney is upbeat. “Advances in these areas are already under way. It is just a matter of time before off-the-shelf data acquisition devices and application software, such as LabVIEW, are commonly used to create user-defined measurement solutions on handheld computer platforms,” he says.

The Future in Your Shirt Pocket?

Despite the shortcomings of today’s handheld devices, test and measurement handheld computers and peripherals based on Palm OS and Windows CE are starting to appear on the market. The huge handheld computer market and the handheld test and measurement market are colliding because the manufacturers of the new devices have solved at least some of the problems mentioned earlier.

Other Operating Systems Increase Options

Although Palm OS and Windows CE are duking it out in a well-publicized race for handheld market dominance, other options are to be considered as well. Symbian (a consortium founded by handheld organizer maker Psion, as well as several cellular giants, such as Ericsson, Nokia, and Motorola) has its EPOC operating system, which is targeted at smart cellular phones and handheld communicators that can browse the Web. In Europe, where cellular phones and personal organizers are even more present than in the U.S., EPOC is a popular option. Linux is also a force that may have to be reckoned with in this marketplace. RTLinux and uClinux are examples of ports of Linux to embedded CPU platforms (e.g., ARM, PowerPC, and MIPs). These CPU architectures are the ideal targets of handheld computer designers, and Linux is also receiving a lot of interest from developers of embedded systems, such as Internet appliances and set-top boxes.

As the handheld computer market continues to grow, the opportunities for programmable handheld test and measurement instruments will become more irresistible. As vendors of fixed-function handheld instruments and PC-based instrument hardware and software look to increase their revenue streams, it’s hard to argue against moving into the test and measurement handheld marketplace.

Windows CE—Lots of Choices

Handheld computing is one of the few areas in which Microsoft operating systems are not market leaders, at least from a market share standpoint. Although slimmed down dramatically and with many different vendors providing products, Windows CE still maintains significant compatibility with its desktop brethren. Yet, it looks bloated when compared with the simplicity of Palm OS. As a result, most Windows CE handheld devices are positioned at the high end, with such nifty features as MP3 and video playback capabilities, compact flash, and PCMCIA slots. The new PocketPC devices that are now coming to market may change this, but the jury is still out. But in vertical markets, such as industrial computing, Windows CE has made great inroads. This is due in large part to the fact that manufacturers in these lower volume niches are very concerned about compatibility and up front engineering costs and because they view Windows CE as part of a well-known family of operating systems for which there is a ready supply of hardware peripherals, software packages, and engineering talent. The fact that Windows CE devices typically have higher performance CPUs and larger memory footprints also helps. Industrial handhelds from such companies as Itronix and DVP (see Photo 4) have all the features of the typical off-the-shelf handheld computer but are more ruggedly packaged, with longer battery life and more peripheral interfaces. Because it is similar to desktop Windows, handheld computers based on Windows CE are a natural extension for developers of test and measurement peripherals and software, and support for the many test and measurement PCMCIA and USB peripherals is starting to appear. Still, long-term acceptance of Windows CE as a valid platform for handheld test and measurement will depend on the efforts of such companies as National Instruments, ComputerBoards, and Quatech. The conundrum of how to leverage this growing market is well recognized. As Kevin Kline, National Sales Manager at Quatech puts it, “With the popularity of notebook computers for field use, we see handheld devices, such as those based on Windows CE, as a natural extension of programmable field instrumentation and data loggers. Moving forward, we are evaluating Windows CE and the possibility of porting our software drivers to that platform, but the plethora of devices and architectures available for Windows CE and the fact that the operating system itself is evolving rapidly combine to make the task challenging.”

Still, just as PCs have not replaced fixed-function benchtop instruments, it’s fair to say that handheld computers will not totally replace fixed-function handheld instruments—at least not in the near term. But don’t be surprised if the next generation of handheld instruments that you design or use are built on an off-the-shelf handheld computer platform and third-party DA hardware and software tools.

A Sampling of Companies with Useful Products for Handheld T&M

Company Name Web Site Phone Number ComputerBoards, Inc. www.computerboards.com 508-946-5100 Datastick Systems, Inc. www.datastick.com 408-615-5774 DVP. Inc. www.omnimeter.com 301-670-9282 Fluke Corp. www.fluke.com 425-347-6100 Imagiworks, Inc. www.imagiworks.com 877-373-0300 Itronix Corp. www.itronix.com 800-441-1309 Handspring, Inc. www.handspring.com 888-565-9393 Hewlett-Packard (Agilent) www.tm.agilent.com 800-452-4844 Microsoft Corp. www.microsoft.com/windowsce 425-882-8080 National Instruments Corp. www.ni.com 512-794-0100 Palm, Inc. www.palm.com 800-881-7256 Psion PLC www.psionusa.com 978-371-0310 RT-Control, Inc. www.rt-control.com 416-621-0125 Quatech, Inc. www.quatech.com 800-553-1170 Sunrise Telecom, Inc. www.sunrisetelecom.com 408-363-8000 Symbian, Inc. www.symbian.com 650-598-4747 Symbol Technologies, Inc. www.symbol.com 800-722-6234 Tangent Systems, Inc. www.versid.com 800-992-7577 TTC www.ttc.com 800-638-2049

Home | Contact Us | Advertise © 2009 Questex Media Group, Inc.. All rights reserved. Reproduction in whole or in part is prohibited. Please send any technical comments or questions to our webmaster.