Here is the abstract you requested from the MIL_2010 technical program page. This is the original abstract submitted by the author. Any changes to the technical content of the final manuscript published by IMAPS or the presentation that is given during the event is done by the author, not IMAPS.
|Printed Wiring Board Failure - A Detailed Analysis|
|Keywords: board, short, contamination|
|The testing of circuit card assemblies (CCAs) stops when anomalous electrical conditions are detected by systems put in place to monitor the testing. These systems are set with parameters that allow for automatic shutdown of tests in progress when a failure condition is detected. Such was the case when a test point that was set to see a nominal voltage between 34.151244 V and 36.661244 V detected a value of 8.682 x 10 -005 V; essentially a dead short. Extensive troubleshooting was performed to locate the cause of the test anomaly. The troubleshooting absolved all electronic devices within the affected circuit. The printed wiring board (PWB) itself was identified as the source of the electrical short. A bulge, approximately 250 mils in diameter, was observed on the surface of the PWB. The bulge was visually examined and photographed. Further non-destructive steps included documentation via X-ray radiography. A material with an image density similar to that of a nearby soldered pin was observed to be extending from that nearby pin and into the PWB base material. The section of the PWB containing the bulge was cut out using a low speed diamond saw. Further cutting of the section excluded all but an approximately 1 x 2 in PWB segment surrounding the anomaly. Higher resolution, multi-axis X-ray imaging followed by careful cross sectioning revealed a void in the area immediately outside of the plated though-hole (PTH). This void contained solder and some melted copper. In conclusion, the assembly suffered electrical shorts due to the intrusion of conductive materials into the dielectric layers of the PWB upon which it was built. The semi-spherical, smooth shape of the void suggests that it was pre-existent to the assembly manufacturing operation at Orbital. Evidence collected during the sectioning process suggests that there was a small gap in the dielectric at the edge of the outermost, or layer 12 solder pad, that allowed solder to flow into the void. Melted copper attests to the high temperature resulting from the electrical power available to the short. Crystallographic structures in the layer 10 copper indicate a heat flow direction radiating from the general area of the solder mass outward. The delamination of layer 12 and other damage artifacts pointed to an exhaust path for the gasses created during the thermal event. The presence of conductive materials most likely of an ionic or metallic filament nature, combined with time and heat as the unit had been powered for almost 78 hours before the electrical shorts manifested, resulted in the reported electrical shorts and expulsion of material from the PWB.|
|David Boyarsky, Engineer
Orbital Sciences Corporation