The effect of thermal feedback and statistical technological dispersion of microcircuits parameters on their thermal modes

Abstract

Abstract During functioning of integrated circuit, the consumed power converts into heat, which makes high temperatures appear in it. Due to the strong dependence of the power on the temperature, its value changes as a result of self-heating, and the changed power, in turn, makes the microcircuit’s temperature change, thus closing the thermal feedback loop. Neglecting the thermal feedback in a microcircuit results in a significant discrepancy between the real temperature and its prognosticated values. The statistical technological dispersion of microcircuits electric and thermal parameters is the second most important factor that must be taken into account for adequately modeling thermal processes therein. Due to this dispersion, microcircuits’ real temperatures are not exactly known and determinate; on the contrary, they are stochastic, and their values are vary within some intervals. The statistical dispersion of a microcircuit’s parameters is conditioned by the technology of manufacturing and mounting the microcircuit in an electronic device, as well as by the stochastic parameters of ambient, such as the temperature, velocity and direction of flow of environment. In this article examines mathematical modeling the separate and simultaneous influence of the thermal feedback and statistical dispersion of the parameters on a microcircuit chip’s temperature. It has been shown that neglecting the said factors leads to the inadequate modeling of thermal and electrical regimes of microcircuits and to errors of design of electronic devices based on them.