 System-Level JTAG: An application whose time has come
By Alan Sguigna
Vice President of Sales and Marketing
ASSET InterTech
Since its very definition, boundary scan (IEEE 1149.1/JTAG) has been on an upward trajectory. First it had to be embedded at the chip level. Then, as the technology was deployed, it moved up to become a mainstay of board test. And now we are in the midst of the next step upward as JTAG methods are implemented at the system level.
 This type of system-wide implementation of boundary scan has been referred to by several different names, including system JTAG test, embedded boundary scan, embedded JTAG test and others. At the end of the day, what it’s called doesn’t really matter. What matters most for you is the rapid ROI it has to offer and how you can take advantage of it.
The difficulty pinning a name on system-level JTAG stems from the fact that its implementations come in so many flavors. Some manufacturers are looking for a more efficient way to reconfigure programmable logic devices or to reprogram flash memories. Others want to be able to ensure high system availability through remote monitoring and troubleshooting. Still others see it as a way to reduce high field support and warranty costs.
Adding to the potential confusion is the myriad of ways that system-level JTAG can be deployed. For example, certain implementations will involve integrating a boundary-scan engine or JTAG executive kernel in the system, but others will rely on an external JTAG system connected to the system-under-test or system-to-be-programmed. Some deployments will be done exclusively for in-system reconfiguration of PLDs, CPLDs and FPGAs, while others will be intended solely for test purposes.
There is one thing that all of these implementations have in common. They all show that the reach of boundary scan has been extended beyond a single circuit board to include the entire system. Consequently, and for the sake of simplicity, we have settled on calling it system-level JTAG.
System-level ROI
What’s motivating this interest in system-level JTAG is readily understandable. It’s cost-savings or that ultimate arbiter of the marketplace, a company’s bottom line. A few examples might best illustrate this point.
Many manufacturers struggle with the so-called “no fault found” (NFF) predicament. In this scenario, a large number of high-availability systems have been installed in the field for several years. As is the case with every electronics system ever introduced, malfunctions develop, system performance may degrade over time, and faults and system failures take place. For example, zinc or tin whiskers may over time cause short circuits in operational equipment like servers or routers. To return a particular system to service quickly with as little down-time as possible, on-site technicians routinely remove one or more circuit boards that are suspected of causing the problem and replace them with new assemblies. The questionable boards are returned to a repair depot where they are tested. As it turns out, the subsequent testing of the circuit boards often determines that a high percentage of them are classified as “no-fault-found” and they are returned to service. The cost of the entire process to the manufacturer and to those who deploy the equipment can be staggering.
 System-level JTAG methods can help reduce NFF costs on several fronts. As part of a burn-in suite of tests before the system is shipped, system-level JTAG can identify intermittent faults that could worsen later in the field and lead to an NFF situation. Or, a field service organization could take advantage of system-level JTAG through remote access. That is, in response to an alarm condition, a central monitoring center could access embedded system-level JTAG facilities by using the Internet. Then, ScanWorks’ diagnostics capabilities could isolate the faulty assembly, device or pin. In this way, a remedy might be devised online. At the very least, a service technician could be dispatched with a specific corrective action plan in hand, eliminating the guesswork that all too often is typical of on-site troubleshooting.
Another cost-effective deployment of system-level JTAG might only involve in-system reconfiguration of logic devices. Many companies rely on their field service organizations to manually perform firmware updates in the field. The cost of labor alone for service technicians to go on-site to reconfigure firmware often causes delays in the roll-out of incremental upgrades which could improve system performance. With the remote access of system-level JTAG, even fairly minor firmware upgrades can be deployed immediately without the staggering costs of a manual, labor-intensive roll-out.
First Step: Plan Ahead
To cash in on the many benefits of system-level JTAG an organization must plan ahead. Company-specific cost studies will quantify the cost-savings. The intended applications for system-level JTAG – whether it will be used solely for test or just for logic reconfiguration or for both – must be defined and fully understood. And system design and architectural issues must be considered to ensure that the hardware and software resources needed by system-level JTAG are designed into the system. Another article in this issue of Connect, “Planning for System-Level JTAG”, goes into greater detail on many of these considerations.
The critical point to remember is that the sooner the process begins, the faster your company can begin reaping the benefits of system-level JTAG. Let us know how we can help. |
