|
|
|
|
|
Strain During Press Fit Connector Insertion in PCB Assembly
Production Floor |
|
Authored By:Shane Lewis, Ph.D., Asfaw Bekele, Tushar Tike Sanmina San Jose, CA, USA Vineeth Bastin Nordson-Dage Concord, CA, USA SummaryThe wide use of press fit connector technology has made the assembly process the intersection of fine pitch, high I/OIC packaging, high density interconnect, and PCB materials performance. Press fit is a process where connectors with pins are inserted through the Plated Through Hole (PTH) of a Printed Circuit Board substrate to establish mechanical and electrical contact instead of the traditional soldering process. The clearance between the PTH and the connector pin diameters are intentionally made close to zero so that intimate contact is made between the connector pin and the PTH after insertion. A controlled force is used to push the connector through the PTH. One of the consequences of such a process is damage to adjacent solder joints due to strain generated during the press fit insertion process. Experimental investigation was conducted to characterize the impact of strain generated during connector press fit. Tri-axial rosettes were attached to adjacent solder joints located at 0 mm, 5 mm and 10 mm away from the press fit site. Three PCB assemblies, one for each location were used for the study. The rosettes were placed at 4 corners of adjacent solder joints to capture all the strains. Strain gauge measurements were taken as the connector is pressed. The experimental results revealed that locations closer to the insertion site experienced the highest level of strain and the strain progressively decreased further from the insertion location. Furthermore, advanced x-ray inspection and computer tomography were demonstrated as a technique to evaluate and characterize failure sites. ConclusionsThe lack of reliability data to make a proper correlation between experimental data and design rule guidelines hinder us without validation. Micro-strains are a unit-less metric, and this makes the experimental data mostly dependent on the PCB dimensions, especially thickness. Although press fit connectors are a mature technology, the IO densities are increasing and the materials and form factors are changing. PCB materials are also changing, and tradeoffs sometimes have to be made in PCB design between the ideal construction and electrical characteristics of the system being constructed. Experiment illustrates the need for a comprehensive set of design rules and best practices for press fit and connector technology considering the need for high speed and high signal density on tightly populated, mixed PCBA technology using high performance PCB materials. We investigated a medium thickness PCB, with a 60 pin, rectangular press fit connector. The connector was an interference type pin collapsible eye, and measured the strain at increasing distances orthogonally from the corners. In defense of good ideas and half-baked in-vivo experimental execution, the setup has a number of flaws, including dimensional asymmetry and that it was not replicated. However, the ultimate intent was to establish a "trailhead" for using press fit (and "new" or other connectors) as designs and pin densities are outpacing practical working knowledge, at the expense of manufacturing costs. (See our "kickoff" road map of design rule guidelines in figure 18.) How our current test setup matches real world production usage of the connector is that the press machine is programmed without help of equipment or component suppliers ( a resources typically available in proportion to sales volume.) the die (press) design is exactly from the manufacturer spec sheet, it was purchased from local machine shop. In this test, the fixture and support die are matched to manufacturer recommendation, using low cost materials, and we used a single PCB and press fit connector for every test run. In a high volume high yield environment, the process wrinkles may hay been ironed out and some "special sauce" may have been added to the recipe. Test was executed in factory floor settings (an advantage of strain gauge systems, they are designed for field-use.) Each run was a cost (as in the real world) where a PCB, connector and strain gauges are consumed. Removal and reinstallation is another topic for consideration also! Strains in this configuration were within the limits of IPC 9704A. That is interpreted as "a BGA corner could be placed (within our framework) as close as possible to this connector, on this board, and the strain from the pressfit installation process would not cause a reliability concern." It would be a sweeping overstatement to claim this in a design rule or or to even claim that the BGA would survive if it were moved off the orthogonal test path. It illustrates the time and resources to make any claim of reliability and poses a larger question how granular do we get? Initially Published in the SMTA Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|