Electrical testing provides valuable fault data for process controls and is the final determination of the product quality level reaching the customer and the consumer. Investment in electrical test equipment remains one of the more critical decisions made by PCB fabricators. This equipment must be capable of testing emerging PCB technologies at a sufficient rate sufficient to meet turnaround time and volume requirements. The critical part of the test system is the fixture - the tooling that connects the tester with the PCB. There is an axiom in test that says "a test system is only as good as its test fixture design." If reliable, repeatable mechanical contact with the PCB fails to occur, the testing process is compromised, regardless of the quality of the electronic measurement system involved.

The latest generation of PCB technology comprises quad flat packs (QFPs) with centers of 0.020” and less, ball grid arrays (BGAs) with centers of 0.050” and less, and other devices with small pad geometries. These components have dictated new fixture designs be developed to accommodate the new technology. Like the PCBs themselves, the test fixtures must be produced and implemented in shorter times and cost less than previous generations.

Test Pin Displacement
Longer pins have greater displacement - the distance a pin will tilt from the "on grid" point in the bottom plate of the fixture to the "on product" point in the top plate of the fixture as shown in Figure 1.

 

The greater displacement advantage of longer pins is key to solving density issues that occur when a device footprint on a PCB requires more grid points than are available immediately beneath the device. A longer pin with a greater deflection significantly enhances the ability to solve density problems by creating a larger area from which to draw grid points. In certain fixture designs, this capability virtually eliminates the need for expensive double-density grid electronics, a considerable and measurable asset. Pin length and displacement overshadow grid density in adding capability in testing fine center and high density PCBs.

Extreme pin displacements, however, introduce new challenges. As test pads decrease in size and increase in density, a long test pin with a headless tip feature has become essential in providing the required displacement, while ensuring the pointing accuracy necessary to repeatedly contact the smaller targets. Pointing accuracy, usually described as a radial dimension, is the pin’s ability to hit a given target. In a headless pin fixture design, the tip of the pin is retained in the top plate and protrudes slightly, if at all, to hit the board as shown in Figure 2.

Improving pointing accuracy in this fashion creates a problem in retaining the pin in the fixture. Methods had to be developed to capture the pins in the fixture. There was also a need for grid fixtures for the top of double-sided universal grid systems. Previous headed pin fixture systems could not be used in this application: when the fixture for the topside grid was flipped over, the pins would fall out. The headless, retained-pin grid fixture was the solution for improving pointing accuracy and providing a fixture for the top of double-sided universal grid systems.

Fixture Designs

Figure 3 shows headless pin fixturing methods in current use. Each style has its benefits and limitations.

Figure 3a represents a bent or curved pin design, the advantage being that the pin hits the target perpendicularly. The hole diameters in the top plate can be tightened to improve pointing accuracy because all the pins are passing through with little or no angle. Long pins allow for a greater degree of displacement, and retention methods allow top and bottom grid fixturing to be identical. Disadvantages are that hanging middle plates reduce the torsional stability of the fixture design, increasing chances of fixture distortion under compression and producing binding with false opens or loss of registration and/or pointing accuracy.

Thicker diameter pins do not bend easily, and bending or curving any pin can cause binding in the fixture. Pin binding that cannot be overcome produces false opens. For this fixture style, additional test system spring probe force is required in the grid system to overcome binding. The pins in the fixture that need no bending will transfer the additional spring force directly to the board feature, potentially marking or gouging it

In this topside pin load design, the crimp in the pin rests on the second plate and pins cannot be removed without disassembling the fixture. When all the pins have been loaded, the top plate is put on with some difficulty, because the top plate holes are the tightest in the fixture, and because the top and second plates are drilled to the same pattern. The bending of the pins therefore takes place as the top plate is mounted, a difficult task with high point-count and high density fixtures. Also, pin crimps close to the top plate, and in proximity under compression, can touch and cause false shorts. A final caveat is that this fixture design is compressible. When the system compresses the board on the fixture, the fixtures compress as well, producing a dynamic pin displacement, plate location, and a pin curving/bending situation that is difficult to compensate for in design and is likely to produce binding, false opens, loss of pointing accuracy, and shorting at the crimp point.

In Figure 3b, leaning, rather that bending the pins significantly reduces binding, and eliminates the need for additional spring force. The crimp is on the bottom of the fixture instead of the top, greatly reducing the possibility of shorting in the fixture. Loading from the bottom side significantly increases the loading rate, since the bottom holes are fairly loose, and are uniform due to the on-grid assignment. The fixture is rigid (noncompressible), with uniform plates and spacer locations providing a stable platform for the fixturing software to assign pins and lean or displace them.

The dual top plate allows for top plate rigidity and maximum pin deflection. The first top plate is thin to permit a tight hole diameter for accurate pin tip pointing without binding. The second top plate is thicker for top plate stability and flatness, with an oversized hole to avoid pin binding while providing guidance. Unfortunately, the first of the dual top plates is very thin, providing pointing accuracy, but is subject to material movement due to drilling, and the inherent reduced stability of thin materials. The squared profile of the dual top plate hole provides a catch point for pins that slip or “walk" down into the fixture.

The reduced pin length (3") of this design sacrifices displacement and therefore slightly limits density and fine pitch capability. Pins are not normally removable from this style fixture, but a version of the design is available with a sheet of drilled Mylar as the retaining plate for the crimp pin, similar to Figure 3c, with the Mylar sandwiched between standard polycarbonate plates, allowing pins to be from the bottom side without disassembly.

Figure 3c uses a grooved pin retained by a drilled Mylar sandwich. Pin loading rates are greatly improved by the non-oriented pins, grooved on both ends, which allows handfuls of pins to be loaded from the bottom side. A countersunk top plate permits tight top hole sizes for better pointing accuracy, while maintaining top plate flatness and thickness for stability. Unfortunately, pin motion under compression wears and distorts the Mylar over time, causing the pins to slip and fail. Multiple loads and reloads of Mylar-retained fixtures tend to further reduce retention life. Featured pins, such as crimp or grooved music wire pins, exhibit a weak point at the feature at reduced pin diameters, limiting ultra-fine center capabilities. Effective pin diameter limits are approximately 0.019 inches, at which point, material strength at the feature location under normal stress compression is insufficient.

The fixture style shown in Figure 3c incorporates additional plates for pin guiding throughout the fixture, most notably near the top plate. The plate under the top plate eases loading of especially fine or dense fixtures and reduces the opportunity for the pin to “walk out” of the hole during testing and catch under the top plate.

The style shown in Figure 3d represents some of today's most advanced features, including non-oriented, non-featured pin design for maximum pin load rates, minimum pin cost, maximum pin displacement and minimum pin diameters for fine center testing. After loading the pins, a drilled latex sheet is pressed over them to retain them. The latex is not sandwiched; it floats between the bottom plates, reducing wear on the latex and extending retention life. This fixture design utilizes a series of uniform, one-piece, molded perimeter support stacks progressively tiered to provide precise plate spacing and uniform fixture height. This accuracy is achieved due to the singular and consistent dimensional tolerance typical of a molded part, as opposed to the cumulative tolerance stack-up of conventional nylon spacers between each plate. The net result is a more solid fixture that maintains the flat parallel relationship of bottom to top plates under compression. This is essential for repeatable contact when testing fine pitch and high densities. Other benefits of molded spacers include approximately 50% reduced material cost and significant gains in fixture assembly time.

Additional economies can be realized on low and medium technology PCBs using the same fixture design by removing plates. Implementing a latex fixture style with longer, headless test pins can provide an effective method of overcoming the technology barriers that limit the electrical test function.

A thorough cost/benefit analysis is needed to determine the financial advantages of any particular style of test fixture. Reviewing a PCB fabricator's comparative fixturing can be quite revealing. Optimizing existing fixturing can eliminate the need for a costly investment in a double-density system while improving throughput and turnaround time.

Bibliography
"Bare Board Electrical Testing" Philip J. Hallee, The Printed Circuit Handbook, 4th Edition, Coombs ed. published by McGraw-Hill, New York.

Philip J. Hallee is regional sales manager with Everett Charles Technologies, Pomona Ca. Article printed in PCFab, June 1996