ALBERT

Description: A flight printed wiring board (PWB) for a satellite project was observed to have a high incidence of measling. Other PWB's produced for the program by the same manufacturer did not exhibit the degree of measling as did the 'measle-prone' board. Measling susceptibility during hand soldering and measling effects on PWB insulation resistance were investigated for three production PWB's. Measling resistance was significantly different between the three boards: the 'worst' exhibited five times the number of measles as the 'best' board. 'Severe' measling (that which is likely to affect board reliability) did not exist on the 'best' board, even under extreme soldering conditions (399 degrees C for 12-15 sec.), whereas the 'worst' board showed an average of one 'severe' measle for every two pads under more normal soldering conditions (288-343 degrees C for 2-5 sec.). Both soldering time and temperature affected measling, with time having a slightly greater influence (2 percent versus 12 percent). Measling effects on PWB insulation resistance were inconclusive. These were evaluated by in situ resistance measurements on the same three boards at elevated temperature and humidity. The measured resistance for all three boards decreased for exposures greater than 50 degrees C and 50 percent relative humidity. The 'measle-prone' board showed a resistance decrease at only 25 degrees C and 50 percent relative humidity. However, no definitive difference was detected between measled and not-measled (control) samples. The boards evaluated were production boards, so the effect of interlayer traces connecting the plated-through holes was not controlled. It is likely the resistance measurements were over different volumes of PWB laminate, which would account for the widely varying resistances measured. Thermomechanical measurements on board laminate materials did not reveal any differences attributed to measling. Differences in glass transition temperature were significantly different when measured by DTA, but not by SDT. Laminate thermal expansion differences were significant for the 35 degrees C evaluations, but not for any higher temperatures.

Description: Tests were performed to determine if thermal shocking is destructive to glass-to-metal seal microelectronic packages and if thermal shock step stressing can compare package reliabilities. Thermal shocking was shown to be not destructive to highly reliable glass seals. Pin-pull tests used to compare the interfacial pin glass strengths showed no differences between thermal shocked and not-thermal shocked headers. A 'critical stress resistance temperature' was not exhibited by the 14 pin Dual In-line Package (DIP) headers evaluated. Headers manufactured in cryogenic nitrogen based and exothermically generated atmospheres showed differences in as-received leak rates, residual oxide depths and pin glass interfacial strengths; these were caused by the different manufacturing methods, in particular, by the chemically etched pins used by one manufacturer. Both header types passed thermal shock tests to temperature differentials of 646 C. The sensitivity of helium leak rate measurements was improved up to 70 percent by baking headers for two hours at 200 C after thermal shocking.