Trends in Test part 3 – Easier access to FPGA-enabled instrumentation
16 April 2009
In the third of five instalments, National Instruments looks at trends in test. This month focuses on FPGA-enabled instrumentation.
We have previously looked at how software-defined instrumentation combined with off-the-shelf multicore technology is enabling us to push the boundaries of the systems we deploy. We will continue to explore the trends that are re-inventing how we create our test systems by now looking at Field Programmable Gate Array (FPGA) based instrumentation.
So what is an FPGA? At the highest level, FPGAs are re-programmable silicon chips. Using pre-built logic blocks and programmable routing resources, users can configure these chips to implement custom hardware functionality without having to pick-up a breadboard or soldering iron. The re-configurability of FPGAs mean they are able to instantly take on a brand new ‘personality’ simply by defining digital computing tasks in software and compiling them into a configuration file or bitstream.
Traditional software tools have made the programming of FPGAs a complex task, preventing the domain expert engineer or scientist from having access to this technology. It has been limited to those digital designers who have been trained in using a hardware description language (HDL). We are now starting to see new tools emerge that enable the programming of FPGAs without any background in hardware design. National Instruments’ LabVIEW is a high-level graphical development tool, which via the LabVIEW FPGA Module, allows you to directly program FPGAs. This gives domain experts the ability to quickly implement FPGA-based systems, meaning time can be focused on functional design and algorithm development; low level details such as how to connect the logic blocks on the FPGA to give the desired functionality does not need to become an issue.
The flexibility and rapid prototyping capabilities of an FPGA can substantially reduce costs in the face of increased time-to-market concerns. Let us take for example an automotive company designing a new engine control unit (ECU). Testing the ECU in isolation can only take us so far; ultimately the ECU needs to be tested within the context of the dynamic system it is controlling. What if the engine is still in development? What if a test failure could damage the engine? What if it is expensive to run the engine for the duration of the test cycle? Hardware-in-the-Loop (HIL) simulation addresses these challenges by providing a virtual or simulated system enabling the testing of an embedded controller, such as an ECU, in isolation. Minimising the number of system level tests creates a more efficient test process and therefore time and money can be saved. To perform HIL or any other type of sensor simulation, the system must operate deterministically and at high speeds to create an accurate representation of the real world system. Although HIL systems can be implemented without the use of an FPGA, the flexibility, ease of programming and performance of an FPGA bring significant benefits.
Reint Smit from Neopost Technologies built a HIL test system to develop complex mail-handling machines, "With this test system, we determine failures in embedded hardware and software at an early stage, which helps us significantly shorten the test time and time to market. It also helps reduce costs because we catch the errors before they are carried out through the entire development cycle," said Smit.
There is no reason why this idea of FPGA simulation in system level tests cannot be applied to component level test in the semiconductor industry. Rather than designing a custom ASIC to test something like a memory controller, FPGAs offer the ability to test an idea or concept and verify it in hardware without going through the fabrication process. The large initial investment in ASIC design is easy to justify if you are manufacturing thousands of chips per year, but if custom hardware functionality is required in one, ten or a hundred systems, this is scarcely the case.
It is also inevitable that our test systems have to evolve as the products we test iterate throughout their lifecycles. The costs of making incremental changes to FPGA designs are negligible when compared to the large expense of re-designing an ASIC, especially when using a high-level graphical approach to programming. The benefits do not stop there. We have seen how computer processor technology has progressed over recent years through the increase in speed and number of cores, and similar technology growth is true for FPGAs; and it is advancing fast. By using a modular approach to our test systems, through an industry standard like PXI, we are able to continually benefit from these rapid technology developments. We are able to upgrade an FPGA board in isolation and immediately take advantage of the higher performance; the only requirement would be a simple re-compile of the existing code.
As the problems we are working to solve become ever more complex, we will continue to see strong adoption of FPGA-based instrumentation. In next generation test systems, high-level system design tools, such as NI LabVIEW, will become ever more valuable as scientists and engineers are driven to achieve more, with smaller budgets and less time.
Richard Silley is Technical Marketing Engineer at National Instruments UK & Ireland
Contact Details and Archive...
Most Viewed Articles...