How a grassroots organization can help get managers on board.

Every company always looks to improve their bottom line, but I would venture to say doing so becomes even more important in times of economic turmoil. In the past few years, companies have embraced defect detection; even prevention and process control have become givens. It is only lately, however, that design for test (DfT) and design for testability are becoming prevalent. Defective products not properly tested because they lacked DfT can cut into revenue generation through slower time-to-market, recalls, rework, extension of the warranty period and premature product termination.¹

As a concept, DfT is not new. I would argue, however, the financial upside of these implementations is only now really being acknowledged. For example, we’ve given lip service over and again to reducing returns, yet new examples of products coming out with huge warranty return rates continue to surface. By embracing DfT and testability concepts, these situations would be avoided or, at the very least, reduced.

Let’s explore cost, as it is an interesting factor. Let’s say the designer is designing a microprocessor with >1000 pins and choosing not to use boundary scan. Then the in-circuit test system will need a >1000 pin bed-of-nails fixture. Let’s say the boundary scan microprocessor was going to be $50 and the non-boundary scan was $30. The cost savings is $20 per processor, but where does the added complexity from using that >1000 pin fixture and the fixture cost break even? The added cost for the boundary scan processor will pay off (and then  some!), but decision-makers tend to get caught up with initial pricing at the expense of big-picture items such as return over time and return on investment. In fact, this is an issue that faces all test – regardless of quality of the design. Understanding the initial investment will pay off and then some is the concept every inspection and test supplier is trying to demonstrate all the time.

At this point, I suspect the test engineers are nodding and saying, “I know all this, but how do I convince management?" and in fact, that’s exactly what needs to be done to make DfT a success. Management teams must embrace the long-term gain (shorter time to market, less warranty return, less rework, etc.). This is one of the main goals of the Testability Management Action Group, or TMAG (tmag4dft.org). Its grassroots efforts are targeting management on all levels to get the message out to positively impact future products.

From a practical perspective, what can the average test engineer do? As with anything design-related, one of the biggest drivers needs to be collaboration across organizations. I can remember several examples from when I was a process development engineer where a design came my way for build/test and there were no fiducials on the board, thus eliminating any hope of a robust solder paste or optical inspection. Another example: If a product is designed such that it is to be tested by AXI only, and the designer throws it over the wall to the test group, which in turn expected an ICT implementation, there will be issues that cost money and time.

The SMTA (smta.org) has partnered with TMAG and is working on a refreshed Testability Guideline. Be sure you and your design engineers have a working copy. This will prevent simple mistakes from entering in product designs and will permit standardizations that will reduce design time over time. Within these guidelines you’ll find instructions for keep-outs, clearances, tolerances, carrier recommendations, and so on.

In general, be proactive. Partner with your design counterparts. Join TMAG and learn how to approach management to sign on the dotted line. Incorporating design for test and testability will bring things further upstream and will standardize on many aspects, thus making the implementation phase easier and more robust.

References

1. Louis Y. Ungar, “The Economics of Harm Prevention Through Design for Testability,” International Test Conference Proceedings, October 2008.
2. Thomas J. Coughlin, “Designing for Testability… The Technology, the Technique, and the Economics,” July 30, 1996 Web posting, http://members.aol.com/prpca/designof.htm.

Stacy Kalisz Johnson is Americas marketing development manager at Agilent (agilent.com); stacy_johnson@agilent.com.

Sidebar: Benefits of DfT²

• Reduces the time required to pass the design into manufacturing
• Lowers manufacturing cost
• Minimizes the design engineer’s involvement in production setup
• Improves cross-functional communication and cooperation between design, engineering and manufacturing
• Lowers both initial and lifecycle costs
• Decreases test times and virtually eliminates harrowing production delays
• Guarantees more efficient diagnosis and repair in the field
• Provides more accurate diagnostics to the part.

• Prev
• Next