By: Gary St. Onge
Vacuum and mechanical test fixtures are still an integral part of the process for most printed circuit (PC) board manufacturers. The typical boards manufactured today have an intermix of through-hole technology (THT) and surface-mounted technology (SMT) components. This shift in technology has brought with it a change in the fixturing approach and new concerns in some cases.
The trends and issues impacting fixturing can be separated into two basic categories: technical and service. Each of these areas contains challenges for builders and users of custom fixtures.
Access - Designing for testability which translates into access to probe targets has, in most cases, been integrated into the SMT design process. Special SMT probes have been on the market for more than 10 years as vendors have made significant improvements in pointing accuracy and electrical integrity. In some cases, however, targets are not available and, as miniaturization continues, many of the present probing methods probably will prove inadequate.
Special test modules (or translator modules) which recently entered the market provide both the necessary miniaturization and access for testing without test pads. These modules presently are available for J-lead and gull-wing components with lead spacings as small as .025". The module translates the miniaturized pattern to a .100" grid where standard probes can be used.
In the case of J-lead components, electrical contact between the module and the plastic-leaded chip carrier (PLCC) is through a wiping motion of the module during placement. For gull-wing components, the module contains small cantilevered pins which provide electrical contact. It is anticipated that this technology will be applicable for spacings down to .010".
Top-side access on fixtures will become more common. Challenges include making contact with small test targets now on two sides, access for hand probing, allowances for cooling and positional accuracy.
Many of the old-style dual-sided test fixtures completely enclosed the UUT in a vacuum which casued the unit to heat up when the board was powered up for an extended debug period. Some users allowed air to leak across the board in an effort to facilitate cooling. This has been shown to cause ESD-type failures. Most fixture companies now offer fixtures which allow for forced or convective cooling.
Accuracy - Lack of accuracy, as a result of tolerance build-up, can lead to false test results. The trend toward smaller test targets serves to increase the frequency of fixture-induced test failures.
For this reason, fixture builders migrate toward designs which minimize the tolerances between tooling holes and test probes. The test-probe manufacturers have improved pointing accuracy through process control and special probe sockets which strive to control the mounting angle. In the near future, greater emphasis will be placed on controlling tolerance stack-ups in the fixture assemblies.
Antistatic Fixtures - The trend to test higher-speed and lower-power devices has led to a proportional reduction in ESD tolerance.
Exposing a PC board with static-sensitive components to ESD during the test process generally can cause total component failure. Electrical overstress (EOS), on the other hand, can induce more insidious problems such as parametric degradation or reduced life expectancy (often referred to as infant mortality).
PC boards with EOS damage often pass functional test and are shipped as good finished products, which later prove to be marginal in operation and/or short-lived. A recent study1 showed that 55% of the field failures for a particular product were attributed to EOS. This can result in excessive warranty service and general reliability problems.
Fixtures now are available which are static-safe and provide protection to devices from ESD and EOS. It is envisioned that, in the near future, all test fixtures will be of the static-safe variety.
Board Flex - Fixture suppliers are being challenged to minimize the magnitude of board flexing which occurs when the UUT is drawn down against the bed of nails. Again, this is driven by SMT board designs which, due to component mounting, are much more susceptible to damage from board deflection.
When board flexing occurs, it is possible to create cracks in solder joints or the components if the fixture was constructed improperly. The old techniques used for gasketing and supporting the UUT need to be revised in order to minimize stresses on the board.
Recently, some methods have been developed which utilize computer modeling to effectively locate supports. These numerical methods consider the location and forces associated with test probes, the supports and the atmospheric-pressure differential. Using this data, board stress can be minimized effectively.
CAD/Tester Link - The use of computer-aided design (CAD) data in building test fixtures is on the increase. This is partially due to the drop in CAD hardware cost and their availability on personal computers.
The main reasons, however, center around accuracy, turnaround time and cost. When the CAD data is available, many of the manual data-entry tasks can be eliminated. This has the effect of reducing fixture development time with a proportional savings typically passed on to the user of the fixture.
The accuracy gains include both probe placement and wiring which can yield further savings by minimizing rework. The errors associated with any manual process can be eliminated by using the same data from which the board was created.
Service - The service requirements center around quality and delivery time. In many cases, a timely entry into a market is critical to the success of a product. With the trend toward shorter product life cycles and the SMT process facilitating rapid product transitions, the test area also is required to respond more rapidly to changes.
This means bringing a new fixture and associated test program on-line in a shorter time frame. To accomplish this, a fixture supplier needs to deliver a quality product on time with an absolute minimum (if not zero) number of errors. Presently, a two- to three- week delivery is adequate, but one to two weeks soon may be the norm.
Even this short delivery cycle will not satisfy the needs of users if the fixture arrives with errors. These errors, which can range from a misplaced wire to improper gasketing, can disrupt the manufacturing environment. Fixture suppliers will have to continue to improve their quaity and delivery service in order to meet the manufacturing challenge.
Gary St. Onge is the Vice President, Test Fixture Group, for Everett Charles Technologies.
Article printed in Evaluation Engineering, December 1988.
©1998-2003 Everett Charles Technologies. All rights reserved. Everett Charles Technologies, 700 E. Harrison Ave., Pomona, CA, 91767, U.S.A.