Thermocouple attachment using epoxy in electronic system thermal measurements — A numerical experiment

Abstract

Thermocouples have been widely used in electronics thermal measurement. Although there are many ways to attach a thermocouple to an electronic component, epoxy especially rapid bonding epoxy material has been the favorable choice due to its convenience of use. However there remains lack of comprehensive understanding of the accuracy of the measurement results and what should be done to minimize the error introduced by the thermocouple and epoxy. In this study two parameters were introduced to describe the causes of error in thermocouple measurement using epoxy. A total of eight variables that may occur in the epoxy attachment were investigated based upon a numerical experiment setup, which consisted of a detailed replication of an actual thermocouple and a typical electronic component thermal model. Different combinations along with their measurement errors were provided for side by side comparison. The measurement error could be as high as 25 ∼ 40% in some cases and even for the best case scenario it was still above 4% in this study. The quick-dry epoxy is practically good enough for most electronic thermal measurements, but attentions must be paid to control several variables that can be commonly neglected in order to assure results are acceptable. The findings from this study can be applied by thermal engineers to achieve the best practice during thermal design and measurement.