Q5 Interview - Peter van den Eijnden, JTAG Technologies

Peter van den Eijnden, managing director of JTAG Technologies, talks to Electronics Weekly about boundary-scanning - what value it adds, why the tools are perceived as being expensive, and the long-term stability of manufacturing test systems.

What value-added does boundary-scan (aka JTAG testing) bring to the manufacturing process?

Boundary-scan represents a low-cost, if not lowest-cost, means of verifying the soundness of the interconnects between [predominantly digital] components on PCBs and within systems. It therefore allows for the testing of nets within PCB structures that do not surface between BGA-packaged devices and which are therefore inaccessible via traditional ICTs (In-Circuit Testers). Boundary-scan technology can also be used to perform certain Functional Tests (FTs) on logic clusters within a design.

What's more, with the increasing availability of affordable design simulation and prototyping tools, engineers are sending designs to manufacture with greater levels of confidence (i.e. "correct-by-design") than ever before; so production FT need not be as exhaustive as in the past, and limited FT performed by boundary-scan will often suffice.

Boundary-scan is also used for in-system programming (ISP) of PLDs and memories such as parallel Flash and serial PROMs based on I2C, SPI or other serial interfaces.

Boundary-scan tools are perceived as expensive, why?

This perception, which tends to be taken by smaller companies rather than larger ones, is typically against boundary-scan [test] development tools, rather than the manufacturing, shop-floor test run-time systems.

As you are no doubt aware, many man-years of development work goes into providing high levels of both automation and accuracy, and these costs do have to be paid for. As mentioned, the larger organisations don't perceive boundary-scan as expensive, not because they have more money to spend but because they tend to appreciate that accurate tools with high levels of automation save their engineers a great deal of time.

Within SME organisations, the cost of man-hours is often looked at in a different way, causing them to look more at the cost of a tool rather than the man-hours it saves.

Does boundary-scan have a role to play in strengthening European Manufacturing?

Most certainly. The majority of outsourcing, for the manufacture of high volume products, to the Far East is done because of cheaper labour rates. However, boundary-scan reduces test time, therefore reducing overall manufacturing time.

If you take into account shipping delays and costs from say China, the cost savings don't stack up as well as they used to; especially on medium volume sophisticated products that may require a high degree of expertise to manage and maintain the manufacturing and test processes. Here, the 'expertise' that is often sent out from the West at high costs is not initially accounted for.

Already we are seeing Hungary, Romania and other Eastern European countries becoming viable high-volume manufacturing alternatives to the Far East. So I think it is just a matter of time before Central Europe can claw back the high volume work that went to the Far East.

What is your view on the long-term stability of manufacturing test solutions?

Process inspection tools, such as AOI and (perhaps to a lesser degree) X-ray will always have strong roles in volume manufacturing, despite the high costs of these tools.

As for ICT and FT, they will also play valuable roles in future manufacturing test, but I believe dependency on these test methodologies will decrease - and here I am talking about the percentage of board real-estate that will be tested using traditional ICT and FT. Boundary-scan testing will, I am convinced, be employed to greater degrees in manufacturing - either as standalone test tools or integrated into traditional ATE, for example. The ability for tools to perform statistical analysis will also provide valuable feedback when tuning the manufacturing process. The value in test is not only about pin-pointing the fault on a PCB to enable a repair but it is also there to assist the production engineer tune the manufacturing process.

Are new standards being developed, and how are boundary-scan tools likely to evolve?

The most recent standard is IEEE 1149.6, which allows for the testing of high speed, AC-coupled interconnect (such as LVDS), and tools are just becoming available which cater for "dot 6".

This year will most likely see IEEE 1149.7 emerge as a formal standard, and in the future we might see dot 7 compliant devices which may still have a standard TAP but also be accessible via a 2-pin interface. We would therefore be looking at new or adapted hardware and software.

Greater levels of automation will be par for the course and, whilst remaining a superb test solution for manufacturing, boundary-scan will push further into the "embedded systems" space. Indeed there are already mechanisms for embedding boundary-scan controllers into a system - with up- and down-links providing wired or wireless remote access to systems, boards and chips.

See also: Q5 - Interviews with electronics industry leaders
Read all the Electronics Weekly Q5 interviews. From ARM's chairman, Sir Robin Saxby, to touchscreen technology firm Zytronic's MD, Mark Cambridge, the business leaders share their particular insights on the UK electronics industry.
http://www.electronicsweekly.com/q5s