Semicon West: TechXSPOT reprises test workshops
Rick Nelson, Chief Editor -- Test & Measurement World, 7/25/2007 11:01:00 AM
San Francisco, CA. Combining six presentations from the Semiconductor Wafer Test and Burn-in and Test Socket workshops, a special TechXPOT session at Semicon West covered topics such as socket alignment and contact cleaning.
Kicking off the 2.5-hr. session, which began at 10:30 a.m. Wednesday July 18, David Shia, a mechanical engineer at Intel, described a Monte Carlo-based package-to-socket alignment assessment methodology.
Monte Carlo (MC), he said, overcomes limitations of alternative worst case (WC) and root-sum-square (RSS) methods. WC, he said, is unrealistically conservative, driving up costs or suggesting designs that are impossible to manufacture. RSS is more realistic, he said, but it relies on underlying assumptions of characteristics such as linearity. He cited power as an example. An RSS investigation of power would entail multiplication of voltage and current—a nonlinear operation that would make RSS results inaccurate.
MC overcomes such problems, he said. He described MC as a nondeterministic computational algorithm that makes use of random numbers. He said it’s useful for a variety of applications, ranging from business risk assessment to alignment accuracy assessment. He introduced the MC concept by describing how to use it to measure the area of a circle and went on to describe in detail how he applied it to assess five misalignment mechanisms. Those five are, on the socket side, shift of hole true position (TP) and shift of pin TP within a hole, and on the package side, shift of package within socket guide, shift of overall pad pattern, and pad-to-pad variation.
He presented a case study for a fine-pitch spring-probe socket for an LGA package. WC analysis yielded an unrealistic statistical Cpk value of 0.33. RSS analysis yielded an unrealistically high Cpk of 1.36. MC, he said, resulted in a realistic and workable Cpk of 1.16.
Byron Gibbs of Texas Instruments followed Shia. He described automatic contact-cleaning engineering study applied to package test. “Contact resistance is the enemy of test,” he said, and he developed what he called a surrogate cleaning device to fight that enemy.
His cleaning device is similar in size and shape to a device under test but has a cleaning medium applied to it. He turned to a handler vendor to provide a handler that contains a repository for the surrogate cleaning device. Experiments with manual and automated handler operation showed that the cleaning device improved first-pass yield, reduced rescreening rates, and improved equipment utilization. He concluded by noting that if you have 20 testers and increase utilization by 5%, you essentially get a free test cell.
Talkback
Related Content
Related Content
Sponsored Links
TMW Resource Center
Browse Resources by Type:
- Optimize Test Coverage with Top 10 Design Hints from JTAG Technologies
JTAG Technologies | Design Guide
VIEW NOW
Agilent Technologies | Web Event
VIEW NOW
Averna | Web Event
VIEW NOW
View More
