Next generation test systems
19 March 2008
Test managers and engineers use automated test systems in applications from design validation to end-of-line production test.
This is all with the goal of ensuring the quality and reliability of a product before it reaches the end customer.
Automated test systems can perform simple pass or fail tests or a whole range of detailed product characterisation measurements. As electronic products become more complex (the communications and semiconductor industries continue to drive the convergence of video, audio, wireless communication and internet connectivity into single products), test system designs must be flexible enough to support the variety of tests needed to accommodate different product models. They must also be scalable to deliver a larger number of test points, as new measurement functionality is required.
The ever-increasing demand for the latest product or technology, combined with the challenge of being the first to market, puts pressure on design and test engineering teams to shorten their product development cycles. To be successful, engineering teams need to develop new test strategies to decrease test time and improve efficiencies when moving from design to production.
Despite rapidly increasing functionality, the cost to manufacture the latest devices continues to fall. However, increased functionality often leads to a more expensive and time-consuming test process. This is a challenge for test engineering departments because they must develop test strategies that reduce cost per unit by increasing the throughput of their test systems, by reducing maintenance and upgrade costs, and by lowering the required capital investment.
To meet these challenges, test managers and engineers are being forced to abandon traditional test design strategies based on traditional box instruments or ‘big iron’ propriety ATE systems. Standalone instruments lack the flexibility needed, as the software processing and user interface are defined by the supplier and can only be changed by firmware updates. They also lack other capabilities, such as data streaming and synchronisation because they are designed to work primarily as standalone instruments and not for integrated system use. Propriety ATE systems, such as highly integrated testers for chip production, provide the performance that is needed, but are costprohibitive and can leave engineering teams vulnerable to obsolescence.
In response to these trends, test managers and engineers are now implementing modular, software-defined test architectures based on widely adopted industry standards to provide faster test throughput, greater system re-use and increased longevity.
Designing next-gen test systems
Many companies already employ a modular, software-defined test system strategy and have proven the return on their investment. For example, Microsoft developed the test system for the Xbox 360 controllers based on NI LabVIEW and PXI modular instruments that resulted in a test system that operates twice as fast as their previous generation. Sanmina-SCI builds FDA-approved medical device test systems based on NI TestStand and PXI that exceeds their requirements to test over 83,000 devices per week, and exceeds their yield production requirements by 95 per cent.
Some companies have written their own test executives and spent valuable engineering resources to develop test management software from the ground up. This often results in reduced productivity and, over time, resources can be tied up on software maintenance. To maximise productivity, engineering teams should use commercially available test management software, such as NI TestStand, to reduce the development of operations that are common for each device. By using this software, engineers can focus their development efforts on the operations that are different for each device.
Application development software
The application development environment (ADE), such as National Instruments LabVIEW or LabWindows/CVI, plays a critical role in test system architectures. With these tools, test system developers can communicate with a variety of instrumentation, integrate measurements, display information, and connect with other applications. The ADEs used to develop test and measurement applications must provide ease-of-use, compiled performance, integration of a diverse set of I/O, and programming flexibility to meet the requirements for a range of applications.
Measurement and control services play an important role by providing connectivity to various hardware assets in the system, system configuration and diagnostic tools. For example, NI Measurement & Automation Explorer (MAX) automatically detects hardware assets, including data acquisition and signal conditioning hardware; GPIB, USB, and LAN-controlled instruments; PXI systems; VXI devices; and modular instrumentation. This allows developers to configure them all in one place. Integrated diagnostic tests ensure that devices function properly and test panels provide a quick way to check the functionality of the hardware before developers begin programming. Measurement and control services should also provide integration with the application development software layer through API (application programming interfaces) so developers can easily programme the instruments.
An important aspect of the computing platform used is the ability to connect and communicate to the wide variety of instruments in a test system. There are several different instrumentation buses available for standalone and modular instruments including GPIB, USB, LAN, PCI, PXI, PCI Express and PXI Express. These buses have differing strengths, making some more suitable for certain applications than others. For example, GPIB has the widest adoption and availability of instrumentation, whereas USB provides availability, connectivity and high throughput.
Whilst LAN is suited to distributed systems, PXI Express or PCI Express provides the highest performance.
Measurement and device I/O
Fundamentally, there are two types of instrumentation architectures today; traditional and virtual. Both have measurement hardware, a chassis, a power supply, a bus, a processor, an OS, and a user interface.
The most obvious difference from a hardware standpoint is how the components are packaged. A traditional, or standalone, instrument puts all of the components in the same box for every discrete instrument. The supplier defines the measurement functionality, analysis, displays, and control of the instruments.
By contrast, modular, software-defined virtual instruments incorporate generalpurpose measurement hardware that helps users go beyond the standard capabilities and define their own measurements and user interfaces in software. With a modular approach, engineers can define the test system measurement functionality and build systems that scale to meet future demands.
When designing next generation test systems, it is important to incorporate strategies that increase system flexibility, deliver higher measurement quality, increase throughput performance, lower test system costs, and expand longevity.
Modular, software-defined automated test systems overcome the shortfalls of past solutions based on standalone instrumentation or cost-prohibitive propriety ATE systems. A modular hardware platform based on widely adopted industry standards, such as PXI, allows engineers to develop scalable test systems that tightly integrate functionality from a variety of instrumentation suppliers communicating over the most appropriate buses.
In addition, it also allows engineering teams to integrate current equipment investments, lowering the initial cost of implementation. Along with software-defined measurements that make use of the latest PC technologies (such as multicore processors and PCI Express), nextgeneration test systems can significantly improve throughput performance and scale to meet the demands of test systems for many different electronic products.
TRISTAN JONES is technical marketing engineer, National Instruments UK and Ireland.
Contact Details and Archive...
Related Articles...
Most Viewed Articles...