546
World Abstracts on Microelectronics and Reliability
RA system using process failure analysis for ICs. A. TAKAIDE and N. MANAGE. Proc. IEEE Reliab. Maintainab. Syrup., Philadelphia (Jan. 18-20, 1977) p. 1. This paper presents a discussion on the reliability assurance system of integrated circuits, which at present mainly comprises sampling stress testing and screening. These procedures, especially sampling stress testing, are shown to be ineffective in assuring highly reliable devices. Upon careful examination, we have found out that the failure mechanisms observed in actual use correlate well with failure mechanisms experienced in manufacturing process, And we will show in this paper that these correlations are used effectively to establish a new reliability assurance system. The repair or modification of pads, tracks and plated-thru holes on PCB's. Electron. Prod. p. 19 (April 1977). Modern double-sided printed circuit boards feature extremely dense, highly complex, tightly interwoven networks of slender circuit tracks and component mounting pads. As a result, a critical problem facing the Electronics Industry today is how to repair or modify an ultra-complex, exceptionally high cost circuit board without further damaging it, or even worse ... ruining it entirely.
Reducing solder shorts in wave soldering. R. H. RUHL, DANIEL NETZ and R. HECKMAN. Electron. Prod. p. 8 (April 1977). In an effort to further reduce the level of shorting, a study was undertaken to do the following: (1) review the mechanism causing solder shorts, (2) investigate the suitability of applying process control techniques to the wave soldering process, (3) explore the effect of board design changes.
Travelling-wave tube reliability estimates, life tests and space flight experience. VINCENT R. LALLI and CARLTON E. SPECK. Proc. IEEE Reliab. Maintainab. Symp., Philadelphia (Jan. 18-20, 1977) p. 12. An assessment of the probable failure modes of high-power traveling wave tubes (TWT's) intended for use in space is presented. The infant mortality, useful life, and wearout phase of the tubes life are considered. The performance of existing developmental tubes. flight experience, and sequential hardware testing are evaluated. The reliability history of TWT's in space applications is documented by considering (I) the generic parts of the tube in light of the manner in which their design and operation affect the ultimate reliability of the device, (2) the flight experience of medium-power tubes, and (3) the available life test data for existing space-qualified TWT's in addition to those of high-power devices, Based on generic failure rate data from existing literature, an analytical estimate of the MTBF of a specific highpower TWT used for the Communications Technology Satellite (CTS) is calculated. The same procedure is also applied to a tube characteristic of existing space-qualified devices in order that the validity of the analytical predictions may be compared with demonstrated MTBF. A failure effects analysis with criticality ranking is then presented for the CTS tube in order to uncover those components which most strongly affect its reliability, the critical design weaknesses of the tube, and its wearout modes. A sequential test is then described which determined early within the development program whether the tube could demonstrate with reasonable confidence a given MTBF range.
Dormancy effects on nonelectronics. DONALD F. COTTRELL. Proc. IEEE Reliab. Maintainab. Syrup., Philadelphia (Jan. I 8-20, 1977) p. 7. The results of a 12-month study program conducted by Martin .Marietta Corporation to collect and analyze reliability data for nonelectronic components in the dormant mode form the basis of this report. Data were collected from such sources as private contractors, government facilities, research institutions, and
educational institutions. The collected data wcrc ti~en grouped, analyzed, and statistically tested. Approximatcly 170 billion part-hours of dormant data covering 90 major part classes and subclasses were collected.
Optimum test samples for reliability improvement. EL~Lr BOYLE and LELAND STEWART. Proc. IEEE Reliah. ~/lai,tainab. Syrup., Philadelphia (Jan. 18 21), [977~ p. 24Y F o p mulas are derived and discussed which provide optimum sample sizes for discovering failure mechanisms. T~t o cases are considered: (A) The sample size for a single hardwmc componcm is determined to obtain an optimal bala~ce between expected improvement in reliability and test cosL (B) The sample sizes are determined for all hardware components which comprise a system such that the expected improvement in reliability of the s~stem is maximized for a Nven fixed total test budget Reliability evaluation of integrated circuits, t'roc. 5ym?. 1 EE tferiot-Watt University, Edinburgh (March 31, 1977) p. (~r Every manufacturer has a slightly different method of generating his reliability data. It is therefore difficult lor a user inexperienced in reliability calculations to make a valid comparison. Towards this end the concepts introduced earlier will be of great value. The following list should be verified before any conclusions ~q" vendor superiority are drawn. 1. Confidence limit. 2. Reject criteria ldegradation, data si~ect, t'ul~clional, catastrophic, specific mechanismsl. 3. Temperature of test. 4. Activation energy. 5. Distortion of failure rate due to ~ lo~ number oi device hours. 6. Biasing configurations. 7. Test monitoring (system failures can produce imprcssire results due to less stringent stress being applied to the device). Only if all these factors are considered can a u'uly ol!icctix c comparison be made.
Contamination in solder baths. DONALD A, I!L.t,ID]I. Electron. Prod. p. 16 (April 19771. tt is necessary to maintain a control of the level of contamination in :t solder bath within specific limits in order to avoid problems. Grainy or gritty solder joints can result in the development of cracks within the joint, if not immediately detectable, then after the product has been put into service. Contamination in the solder can result in changing the mobility or fluidity of the solder. If solder becomes sluggish, the drainage characteristics as the printed circuit board exits from the solder wave might result in icicles and bridges which will require additional touch-up, an expensive operation which can be minimized using high-purity solder when initially filling the machine and by maintaining the alloy within limits by undertaking the minor expense of a solder analysis on a routine, periodic, quality control basis.
The life cycle cost impacts of unsafe designs. RicnaaD [, W~:BER. Proc, I EEE Reliab. Maintainak Syrup.. Phihutelphia (Jan. 18-20, 1977)p. 120. Accident costs represent a significant part of the cost of ownership. If life cycle costs are to be controlled, weapon system designs must consider the impact that unsafe designs have on the cost of ownership. Accident costs cannot be brought under control by concentrating on reducing only the accident rate; the average cost of each accident must also be reduced. Each of the cost elements that go to make up the cost incurred by an accident must be explicitly used to make those design tradeoffs necessary to reduce life cycle cost. The Air Force's Directorate of Aerospace Safety has been active in developing an awareness among program and system managers of the life cycle accident cost concept and in developing