Microelectronics and Reliability
Pergamon Press 1966. Vol. 5, pp. 85-96.
Printed in Great Britain
ABSTRACTS ON MICROELECTRONICS AND RELIABILITY
RELIABILITY OF COMPONENTS, TUBES AND TRANSISTORS
Graphics sampling plans for consumer acceptance of electronic components. L. DANZIOE~, Industrial Quality Control 21, No. 6, pp. 312-317 (December 1964). Sampling plans have previously been derived for testing a minimum mean life with a guaranteed consumer protection. These non-replacement plans are examined in a different light, expanded, graphed and compared to replacement plans. A table is derived for quickly approximating the operating characteristic curve for any choice of consumeroriented plan, thus enabling one to evaluate the protection that the plan affords the producer. All of the plans are based on the assumption that the distribution of times-to-failure is exponential. Controlling part dimensions during fabrication and heat treatment. J. A. FERRANTE,Metal Progress 87, No. 1, pp. 87-90 (January 1965). This article discusses production experience and practices of heat treating IBM component parts. Various problem parts are disclosed that typify methods of maintaining close tolerances and good parts after heat treatment. Heat treat cycles, equipment, straightening practices and metallurgical theories are examined to evaluate their respective roles. Experimental test data on the dimensional stability of maraging steels are also discussed. Cathode/heater-insulation failure in oxide-cathode valves. C. H. R. GENTRY,D. NEWSON and R. C. OLDFIELD,Proc. IEE 112, No. 8, p. 1501 (August 1965). Heater/cathode insulation in indirectly heated valves is considered as a two-part phenomenon: the comparatively slow deterioration during operational conditions followed by a rapid thermal breakdown, which is caused directly by the sudden passage of a large current between heater and cathode. Examination of heater/cathode failures in valves tested under rigidly controlled conditions has indicated some of the laws governing the deterioration of the insulation. The effects of various parameters such as temperature of operation, thickness of insulation and applied potential have been studied, and comparisons have also been made for different heater and cathode materials. The experimental evidence is discussed in an attempt to explain the failure mechanism involved, and the conclusion that electrolysis is the principal factor is supported by the results of a number of experiments carried out with specially designed electrode structures. A new approach to the attainment of the highest possible reliability in tantalum capacitors.
J. BURNHAM,IEEE Transactions on Component Parts, p. 21 (March 1965). The question of how to achieve the highest possible reliability for tantalum capacitors is discussed from both the theoretical and practical point of view. A method of achieving reliabilities of the order 95 per cent at a confident level of 95 per cent or better is described which involves an analytical treatment of the physics of the main mode of failure, all accounting for over 90 per cent of a failure; and a statistical model is derived which demonstrates this reliability on a highly accelerated test. The test is applicable on a 100 per cent basis to the test capacitors and, as it is non-destructive, it allows the attainment of a reliability figure for each individual unit. This also provides a method of attaining very high reliabilities at a low cost and the test can be carried out in 24 hours or less.
Application of the Eyring Model to capacitor aging data. H. S. ENDICOTT,B. D. HATCHand R. G. SOHMER, IEEE Transactions on Component Parts, p. 34 (March 1965). The Eyring Model has been used to derive the "power rule" for capacitors. Analytical models for evaluating progressive and step stress tests are presented. Various methods are discussed for determining the value of the exponent for the power rule. In particular, the relation between the exponent of the power rule and the ~ of the Weibull 85