Bed-of-nail fixture users and suppliers constantly face new obstacles ranging from shrinking device and board geometries to requirements for dual or overhead access and increased flux residue from the soldering process. But they continue devising new solutions, since these fixtures still afford the most transparent access to IC and component leads–which simplifies test program preparation and fault diagnosis.
Shrinking geometries with tight lead spacing and smaller test-access areas now require better alignment provisions and better probe-pointing precision. High density also requires that probe diameters be smaller. But conversely, maintaining needed pointing accuracy demands greater probe rigidity.
The phase out of chlorofluorocarbons for board cleaning has introduced another set of problems. While well-controlled fluxing and soldering processes potentially leave very little residue, in practice, the test engineer is often confronted with no-clean boards that are not readily testable. These boards may be coated with layers of contaminants ranging in texture from hard and brittle to soft and gummy.
To address these problems, new probes have been developed and the basic fixture construction and actuation methods are undergoing fundamental changes. Fixturing solutions are also emerging for such extremely dense and fragile assemblies as multichip modules (MCMs) and Portable Computer Memory Card International Association (PCMCIA) cards.
Probes for No-Clean Boards
Two basic probe implementations have been devised to provide reliable contact with designated test targets on no-clean boards. One uses mechanical rotation with accompanying scraping action; the other relies on high force concentration to penetrate undesirable residue or contaminants.
The engineers at the Contact Products Division of Everett Charles Technologies (ECT) developed Twister(TM) probes with a helix rotating shaft captured in the probe receptacle to ensure precise alignment. Interconnect Devices, Inc. (IDI) offers rotating spring contact probes for 0.100(“), 0.075(“)and 0.050(“) center-to-center spacing. “The drill-like action of these rotating probes penetrates tough contaminates at a moderate (4 to 7 oz) spring force,” said Ed Schifman, President of IDI.
At ECT’s Ostby Barton Division, engineers implemented probe designs featuring stainless steel tips to cut through oxides, solder residue and other contaminants. The tips are gold-plated to provide superior electrical contact resistance and minimize surface oxides.
IDI also offers steel plunger tips on 0.100(“), 0.075(“) and 0.050(“) centers with a variety of plating options. “Steel plungers are substantially harder than the industry-standard beryllium-copper plungers,” said Mr. Schifman. “This increased material hardness provides a longer tip and edge life, prolonging the probe’s capability to successfully penetrate flux and contaminates in a no-clean environment.”
The degree, or ease, with which a probe can penetrate contaminates depends upon the spring force, plunger-tip design and the contaminate. IDI addresses spring force and plunger tip design by offering a variety of styles and options, and application guidance in the Fluxbuster(TM) section of its handbook.1
The tip style must be physically stable on the surface being contacted. For example, although a sharp chisel point may be ideal for a hole or pad, using it for a through-hole component lead would result in glancing and side loading.2
Selecting point styles is a somewhat subjective process. Experienced test engineers will often disagree on the most appropriate point for a given contact surface.
However, tests and field use have shown a particular group of point styles to be best suited for specific applications, said Arra Yeghiayan, Applications Engineer at QA Technology:
(o) For flat pads, chisels (such as QA point styles O3, 43, 53 and 63), triads (08 and 18), and spears (31 and 41) are recommended.
(o) For leads, crowns (34 and 24) and sharp triads (08) are recommended.
(o) For holes, chisels (03, 43, 53 and 63) and blades (51 and 61) are recommended.2
Force, Area and Contact Pressure
The ultimate degree of penetration achieved by nonrotating probes is a function not only of probe force but also of the contact area, which is affected by probe geometry and the condition of the probe tip. A tip which is blunt, either by design or because it has become worn or flattened during use, will make contact over a larger area than a sharp tip, resulting in lower contact pressure and reduced capability to penetrate contamination layers.
As seen in Figure 1, a probe with a low spring force and a relatively sharp tip could provide higher contact pressure than one with high spring force and a worn tip. Table 1 lists the contact pressures for various spring forces and contact diameters.
The contact pressures listed in Table 1 are significantly higher than the yield strength of solder, and will cause the solder surface to deform. As a sharp point initially bears against a solder pad, the solder will yield, the area will increase, and the contact pressure will drop until the pressure reaches the yield strength of the solder.2
Impact on Fixture Structures
The use of high and extra-high spring-force probes certainly improves contact reliability but simultaneously impacts fixture force requirements, structure and actuation methodologies. The force applied to the UUT becomes the critical factor. This total force is determined by multiplying the total probe count by the spring force of the selected probe.
If a vacuum-actuated fixture is to be used, assess the adequacy of the vacuum source, taking into account potential leakage problems due to the high force requirements. But even if vacuum power is adequate, modifying an existing fixture for use with no-clean process products may not be just a matter of replacing low-force by high-force spring probe–especially for high probe counts and small boards.
To provide effective probe contact, fixture designers have devised a variety of solutions. One method uses a vacuum-assist lid to increase the effective area pushing down on the PCB, since lbs of force = lbs/in.2 vacuum pressure times in.2 effective area. The second method, usually applied when an odd-shaped PCB is difficult to gasket or when PCBs have open vias, is a mechanical hold-down gate which, in effect, encloses the PCB and becomes part of the vacuum chamber.3,4
ECT has addressed the issue of no-clean mandates in two ways: by incorporating continuing refinements in the SuperKit(TM) vacuum fixturing system and by developing an entirely new pneumatic fixture system for applications in which vacuum fixturing can’t be used.
The new ECT all-purpose pneumatic fixture converts the linear motion of air-actuated ramp rails into uniform vertical motion on four corner-mounted bearing pins.
“Synchronous movement of the bearing pins maintains precise alignment of the probe plates and UUT when the fixture is engaged,” said Gary St. Onge, Director of Fixture Operations at ECT. “Since the fixture is pneumatically operated, adequate controlled pressure is always there to overcome the resistance of even highly populated probe fields.”
Some of the advantages pneumatic (or mechanical) fixturing has over vacuum fixturing were enumerated by Brian Lane, Sales and Technical Support Manager at CheckSum:
(o) Less UUT flex during pressure application; flexure could cause SMD joint failures.
(o) Less static build-up from vacuum flowing through open vias.
(o) Capability to test boards with many leaks, such as open vias.
(o) Capability to support higher probe densities; vacuum fixturing has a theoretical maximum of about 40 probes/square inch.
(o) Longer probe life from less contamination buildup.
“A specific problem of no-clean boards is the flux residue which can exacerbate deterioration of the gasketing material,” added Mr. Lane. “CheckSum’s mechanical and pneumatic fixture systems, without gasketing, solve this problem.”
Pneumatic fixtures can be used for top-side, bottom-side, dual-access and dual-stage testing. They can also provide the top access needed for implementing the new opens testing technologies exemplified by Hewlett-Packard’s TestJet, GenRad’s Opens Express and Teradyne’s Wavescan.
Regardless of whether vacuum or pneumatic actuation is used, today’s smaller test targets demand higher fixturing accuracies. For high spring-force probes especially, probe and board flexure must be minimal and high registration and pointing accuracies are essential. ECT, for instance, uses quad linear bearings in the SuperKit vacuum fixtures to provide the registration necessary to target the probes that penetrate contaminants.
“When the targets on a PCB are smaller than design for testability guidelines dictate, repetitive accuracy must be a prime consideration,” said Lisa MacMaster, Marketing Manager at TTI Testron. “Most fixture companies can ensure the accuracy of repeatably accessing targets as small as 25 mils by using guided probes. But the trend is moving toward even smaller target sizes. TTI Testron has a patent-pending solution, ULTRALIGN(R), which accesses targets with consistent repeatability at 0.015(“).”
MCMs and PCMCIA Cards
While being physically much smaller, MCMs and PCMCIA cards pose similar and, in many respects, greater test challenges than PCBs. Just like PCBs, they house components and interconnections on a substrate. These may contain shorts and opens and are potential candidates for in-circuit testing. As modular entities, they may be subjected to comprehensive at-speed functional device tests. Either test situation entails special mechanical access and–in the latter case–signal fidelity considerations.
“Some probes can approach the center-line spacing of an MCM,” said Ms. MacMaster. “But alignment of the MCM can present a problem because the MCM is typically a ceramic substrate and reliable mechanical reference points may not exist on the MCM for aligning it with the interconnect device. The fixture supplier must then design a clamping mechanism that acts as the reference designator for target access accuracy or provide means for optical alignment.
“PCMCIA boards can be as thin as 0.015(“), greatly increasing deflection concerns,” Ms. MacMaster continued. “Fixturing solutions for these may involve replacing board stops by Gl0 plates which have the board layout configuration routed into them. The G10 plates are placed on both sides of the board and keep it flat during test.”
For performing functional tests of PCMCIA cards, some test engineers choose to access the connectors with spring contacts rather than their mating half, noted Mr. Schifman. Advantages include ease of automation and contact reliability.
To facilitate such implementations, IDI developed the S-0-JS spring probe with a radiused tip, 0.016(“) dia, for a slip-fit with PCMCIA sockets. For MCMs, IDI offers a series of probes, including the QUAD 00 probe/receptacle that mounts on 0.020(“) centers, the MicroSeries, and the PENTA 0 probe for test access on 0.010(“) centers.
References
1. Third Edition Catalog and Source Book, IDI Interconnect Devices, Inc., January 1995.
2. “Probe Selection for No-Clean Flux Applications,” QA Technology Co., Inc.
3. Geary, G., and Smith, J., “Adapting Test Fixtures for No-Clean PCBs,” EE Evaluation Engineering, August 1994, pp. 79-81.
4. Geary, G., and Smith, J., “Adapting Test Fixtures for No-Clean PCBs,” Proceedings of NEPCON West, 1995.
These companies provided information for this feature:
Automated Test Engineering, Inc. (408) 435-8555
CheckSum, Inc. (206) 435-5510
Everett Charles Technologies (909) 625-9357
IDI Interconnect Devices, Inc. (913) 342-5544
QA Technology Co., Inc. (603) 926-1193
TTI Testron, Inc. (401) 766-9100
Copyright 1995 Nelson Publishing Inc.
March 1995