Expanding the use of synthetic instruments.

Dedicated test and measurement instruments offer many advantages. They tend to be easily programmed and optimized for the target application.

To facilitate their integration into laboratory environments or automatic test systems, these instruments typically are housed in a single rack-mountable VXI or PXI card cage and controlled via standard interface buses. Functionally, they execute tests and measurements using specifically designed hardware.

Due to their singular nature, dedicated instruments have definite limitations. Making multiple dedicated instruments work together in an organized and logical way can be somewhat of an engineering challenge. If the specs of an application change, or if an engineer needs to address the application in an unconventional manner, they offer little or no flexibility.

[FIGURE 1 OMITTED]

Test and measurement solutions that use multiple dedicated instruments can get unwieldy. They often result in a kludge of wires and test probes and require a lot more bench space. Such solutions usually are too cumbersome, difficult to document, and inconsistent relative to performance and repeatability.

If this vision suggests the use of virtual instruments, it is for good reason. Virtual instruments contain components that, when functionally designed in software by the end user, can perform simple but effective stimulus and measurement functions. A virtual instrument is like having a kit to build something but without instructions. The creativity is left to the user.

What Is a Synthetic Instrument?

Synthetic instruments (SIs) occupy the middle ground between dedicated and virtual. Pioneered by the NAVAIR Synthetic Instrument Working Group (SIWG), SIs contain basic building blocks that can be configured in software by the instrument vendor, a solution provider, or a user.

These building blocks may comprise a single package or be distributed across multiple packages and multiple test bus architectures. The blocks are used to address a specific application and, unlike dedicated instruments such as a synthetic spectrum analyzer, can be configured and reconfigured to address multiple roles.

Although most commonly associated with RF and microwave test applications, SIs can be applied to a broader range of applications from baseband analog test to serial communications buses. Two key benefits--greater flexibility and smaller footprint--enable test-system developers to produce more adaptable testers while reducing ownership costs and lowering the risk of obsolescence. The resulting solutions meet the Department of Defense (DoD) goals that originally motivated development of the instruments.

An SI addresses a closely defined set of application criteria, but the parameters of those criteria may not have been perfectly understood when the instrument was built.

Take the example of a serial bus tester implemented as an SI. When someone comes up with a design variation on bus rules or takes liberties with the specification in connection with a particular kind of test, we can modify the implementation details via software, which generally is impossible with a conventional instrument. With a hardware universal asynchronous receiver transmitter (UART), what you see is what you get. Its capabilities are fixed. By contrast, an SI can be reprogrammed.

[FIGURE 2 OMITTED]

For example, consider RF test. A downconverter converts RF frequencies to intermediate frequencies at, perhaps, 100 MHz. High-speed digitizers process the resulting waveform, and software- or hardware-based algorithms perform the corresponding mathematical analysis. An arbitrary waveform generator provides signals in the multi-megahertz region, from which they go to an up-converter and become RF.

Highly programmable downconverters, digitizers, arbitrary waveform generators, and upconverters all live in the same box. Instead of dedicated hardware as with a spectrum analyzer, the hardware is much more flexible. You can take advantage of recent developments in arbitrary waveform generators and digitizers as well as the increased computer power, digital signal processers (DSPs), and FPGAs that permit you to do high-speed math more easily.

Typically, an engineer designs an SI to solve some particular class of problem, such as RF test. The same hardware could provide several such RF/microwave solutions.

Traditional functions are spectrum analysis and signal generators. Higher-level functions would be a software-defined radio tester or an ECM threat generator. A single synthetic RF instrument can replace not only multiple traditional functions but also multiple application-specific higher-level functions.

Synthetic Serial-Bus Tester

While RF/microwave test has been the primary focus for synthetic instrumentation efforts, there also is rich experience in other areas. Teradyne's Bi-410 Synthetic Bus Test Instrument, for example, is used in a number of high-profile military and aerospace automatic test systems. By virtue of its synthetic architecture, the Bi-410 can replace multiple dedicated instruments that independently test the many protocols in older equipment.

In this case, the instrument is composed of several configurable building blocks. Rather than being dedicated to a single bus protocol, the blocks perform operations pertaining to sending and receiving bits, words, and frames associated with a variety of protocols. All of the detail attributes such as data encoding scheme, word size, parity, and frequency can be automatically controlled in software.

Serial buses, like the widely deployed MIL-STD-1553, come in many variations. Some are the result of early implementations that took place while the standard was evolving. Other variations are the result of deliberate departures from the specification to address project-specific requirements.

One intriguing alternative is a lower speed configuration in place of the normal 1-Mb/s rate. Another is the use of 4-bit word sizes in lieu of the standard 16.

SIs are built to support the standard and handle different encoding schemes, baud rates, voltage levels, and parity screens. Some of these variations were not anticipated when the standard was written. Synthetic instrumentation enables testers to adapt to these differences without hardware changes.

Many established aircraft utilize 1553. It's a proven interface that's been around for 25 years, and it will continue as a potent force for a long time to come. The most modern platform, the F-35 Joint Strategic Force (JSF), uses 1553 for critical control, coexisting next to the latest high-speed serial buses such as Fibre Channel and Fire Wire.

Test systems that must support the full range of aircraft requirements will experience several 1553 variations. Only flexible synthetic instrumentation is up to the challenge.

Traditional, rigid, standards-driven instruments typically cannot accommodate engineer tweaking in implementation of those standards. There is no better testament to this synthetic instrumentation than its acceptance by the Agile Rapid Global Combat Support (ARGCS) Program, a DoD project to demonstrate a single tester capable of testing legacy programs that span three decades of change from the U.S. Army, Navy, Marines, and Air Force as well as addressing new test development.

The ARGCS system incorporates many new test technologies to achieve a rapidly deployable, interoperable support capability (Figure 1). It features a diagnostic system capable of testing and maintaining analog, digital, and RF assemblies at all levels of the product cycle. Synthetic instrumentation is critical to the success of ARGCS in addressing such a divergent set of test requirements.

Building an Analog Synthetic Application Solution

How does all of this apply to end users? Take, for example, Teradyne's Ai-760 Analog Test Instrument and a signal transform test application that must verify the functionality of a line replaceable unit (LRU) used in critical avionics applications (Figure 2). The LRU is tested for expected changes in amplitude, frequency, and phase shifts between the input and output signals. Both stimulus and measurement components are needed, including an arbitrary waveform generator, two digitizers, and algorithms that check the automatic test markup language (ATML) instrument description and perform waveform analysis.

This application requires only a handful of the components the instrument actually contains. However, if the LRU changes, the instrument is easily reconfigured.

Figure 3 is a control unit tester, also for avionics applications. It needs three arbitrary waveform generators and three digitizers working in parallel. The Ai-760 has that plus a central DMM and a high-speed DSO.

[FIGURE 3 OMITTED]

Although strictly speaking this represents a general-purpose collection of instrument functions rather than an SI, it is representative of a new generation of instruments that you can mold to provide specific synthetic solutions. Like a Tinker Toy constructed from individual, yet multipurpose, components, the resulting solutions are flexible but application specific.

The growing popularity of synthetic instrumentation will mean that manufacturers will deploy fewer instruments having narrow missions. Distortion analyzers offer a very specific measurement and result. A high-speed digitizer can give you distortion information just as easily, but it also can support a wide range of other activities, including ones that designers have yet to consider.

About the Authors