PREDICTING END-OF-LIFE PERFORMANCE OF MICROELECTRONICS IN-SPACE

The prediction of degradation in the key parameters of microelectronics under mixed radiation environments is an essential skill of the space radiation effects community. This article acknowledges some criticisms of the level of that skill and its application to space missions. We describe some of the methods used, especially in ionization effects as they affect MOS devices, which is where a lot of the complexity is found. We also demonstrate that bulk damage can add confusion when mixed with ionization effects. Because mission times in space are very long compared with characteristic times for ''annealing'', the effects of low dose rates are discussed. Because the radiation doses encountered in space are ''moderate'' compared to some military fields, it is shown that test methods for these two fields cannot always be identical. Attention is directed to difficulties of and concerns in test and prediction routines and it is pointed out that complex variability in the radiation response of integrated circuits can usually be understood by reference to the great complexity of structure, physics and chemistry inherent in these devices. The main object is acknowledged: predictions which will allow Radiation Design Margins to be chosen, and to be maintained in practice in orbit, without gross overdesign.