Abstract
This paper describes a thermal measurement method to test circuit boards
in a production environment. For this work, the transient thermal
response of a known good component/board and a faulty component/board
was obtained. The impact of emissivity on the temperature recorded by
the thermal imaging system was calculated. The transient thermal
profile of a known good component/board was modified so that its
emissivity matched the emissivity of the component/board under test.
Modified transient thermal response of a known good component/board was
compared to its counterpart, with and without faults. Cumulative and
total absolute temperature variations were usod as parameters to
compare two profiles for a certain time interval. In all the cases
tested, these parameters--unlike simple temperatute differencing--
correctly provided a consistent criterion to accept or reject a
component/board.