Thermal interaction between electronic components

The thermal performance of electronic components is generally quantified in terms of thermal resistances, which are used by system designers to calculate the operational junction temperature, and hence the expected Mean Time Between Failure. However, such thermal resistance data alone is of limited use to designers, and a need exists to increase awareness of the applicability of such results to practical situations. In this study the standard industrial practice of measuring a component's forced-air junction-to-ambient thermal resistance was applied to both a single component mounted on the recommended test printed circuit board, and to components of the same type mounted in a regular array on a larger test board. Experimental results highlight the variation of component thermal resistance with air velocity, copper tracking distribution, component position on the test board, powered adjacent components, the presence of insulation on the non-component side and on an adjacent board, with the latter creating a channel flow. The most influential parameter was powered adjacent components, with reduced copper coverage and component positional changes on the multi-component board accounting for up to a 25% increase in thermal resistance when compared with the standard value. This approach represents a necessary first step towards the development of a more accurate design tool based on standard test methods.