406
World Abstracts on Microelectronics and Reliability
resistors, although started with a small sample size, gives encouraging results. Large sample sizc (550-1000) and long test durations (800-1000 hr at each level) are recommended for life-tests. As no data is available for indigenous components regarding failure rates and deviation of parameters with time, better experiments can be designed based on this experience.
The corrosive activity of fluxes. A. BOCHENEK and B. WOZNIAKOWSKI. Microelectron. & Reliab, 14, 295 (1975). The corrosive effects of fluxes on conductive paths are one of the major factors limiting the reliability of electronic equipment. The process begins from the moment when flux touches the metal surface and lasts during the process of soldering, until the apparatus is washed. Every one can notice the effects of the corrosive action of flux residues when equipment is used. Obviously, the constructor and the technologist of assembly should know about magnitude of corrosive activity of particular types of fluxes in order to choose the right type, to design thickness and arrangement of conductive paths, and to control the assembly parameters, as well. High reliability photodiodes for space applications, M BRILMAN. Microelectron. & Reliab. 14, 305 (1975). This paper presents the techniques used to manufacture a high performance silicon photodetector having several precisely defined photosensitive areas. The principle of the device design is based on a PIN structure. Metallic layers are used to mask parasitic photosensitive areas. The metal layers are used in such a way that they do not degrade the electrical characteristics of the junction. Special care has been taken during manufacturing to assure the reliability of the product, which is to be used in a meteorological satellite. Problems were encountered in defining significant tests characterizing the reliability of the device under various environmental conditions. The reliability characterization is difficult due to the semiconductor and optical aspects of the structure. Tests were therefore imposed to specifically monitor the reliability of either the electrical or opticalparameters. The device was then required to withstand both sets of reliability tests. A comparative reliability evaluation of CMOS. A maturing technology. D. J. BURNS and V. C. KAPFER. Proc. Reliability and Maintainability Syrup. Washington D.C.p. 354, 28-30 January, 1975. A reliability evaluation program consisting of 180 C/MOS devices from 4 vendors has been in progress at RADC for about 2 years. The date codes for these devices range from 1969 to 1973. The testing program included storage temperature, thermal shock, moisture resistance w/bias, temperature cycling w/bias, and bias power and temperature step stress (BPATSS). The results are presented detailing the predominant failure modes and mechanisms which were induced by this stress testing program. Evaluation testing of integrated circuits. R. A. HOMAN and M. W. ROSSMAN. Proc. Reliability and Maintainability Syrup. Washington D.C., p. 372, 28-30 January, 1975. Evaluation/qualification test approaches applied to various types of integrated circuits are summarized. Linear devices as well as C-MOS and TTL digitals were evaluated. Among the evaluation techniques utilized were 3. C I R C U I T
AND
SYSTEMS
RELIABILITY,
Reliability case history of an airborne air data computer, T. MURATA. IEEE Trans. Reliab. R-24. (2) 98 1975. The life of an airborne Air Data computer and some of its
computerized circuit modeling and simulation for purposes of nominal, worst case, and sensitivity analyses. physical construction analysis, node voltage probing on both digital and linear devices, characterization by means of exploratory and design limit testing. Other tests performed were environmental, burn-in, life, specification limit, and functional tests. The data analysis approach, typical results and observations on the effectiveness of the various techniques are presented.
Reverse bias stresses on emitter-base junctions. R. F. HAYTHORNTHWAITE and R. E. THOMAS. Proc. 1975 Canadian SRE Reliability Syrup. Ottawa, p. 231, 10 May, 1975. Evidence is presented showing that the gain degradation mechanism in bipolar transistors subjected to either high temperature reverse bias (HTRB) stressing or emitter base avalanche (EBA) breakdown stressing is similar. The study was conducted to determine if the short EBA test could be substituted for the much longer HTRB test in predicting transistor stability under operating conditions. Reliability of interconnections on microcircuits. D. L. FRESH. Proc. Reliability and Maintainability Syrup. Washington D.C., p. 568, 28-30 January, 1975. Defects in metal interconnection patterns constitute a significant factor in the reliability of semiconductor microcircuits. Electrical opens, occurring in the vacuum-deposited metal films, have achieved considerable notoriety and have been the basis for reliability studies by numerous investigators. The application of the scanning electron microscope (SEM) is becoming relatively widespread in detecting these defects in metallization. Considerable work has recently been directed toward improving the SEM procedures and techniques, including the preparation of a test method in MIL-STD format. The current status of the SEM test method; its utility, suitability, and acceptance in detecting defects in metal interconnection patterns; and its extension and applicability to other types of defects in high reliability semiconductor microcircuits are discussed. On computer-oriented design of microstrip amplifiers with component tolerance assignment. E. MARAZZ1, V A MONACO, A. BRINI and V. SOLARO. Aha Frequenza XLIV (4), 174 (1975). Component parameter tolerances can cause performance degradation of some circuits during mass-production. Production yield estimates relative to both circuit response specifications and given component tolerances and techniques for increasing yield without reducing tolerances or introducing expensive adjustments are topics of interest in circuit design. These design aspects are illustrated using a 2 GHz thin-film microstrip reflection amplifier. Inspection and supervision of cable quality, during fabrication, using an auto-test unit TM02. M. CADE, P. FENOUILLET, A. CLAUSTRE and J. C. LEROY. Cables and Transmission, 29 (2) 213 1975. (In French). Following important work into statistical control process of urban cables at the factory sites, carried out by the Centre National d'Etudes des T616communications, the French Company S.A.T. has studied and developed an auto-test unit designed to test out cables during fabrication. After going over the principles underlying a verification of cable parameters, the expos6 sets out a table of auto-test measurements TM02, an instrument which proves to be very useful when assessing cable quality. MAINTENANCE
AND
REDUNDANCY
components have been observed in the field and in the laboratory. This paper treats 3 aspects of the reliability. Part 1 shows the relationship of predicted to actual MTBF