Improved Compact Thermal Model for Over-Molded Packages

Epoxy molding compound (EMC) is the most commonly used encapsulating material for the semiconductor packages. Thermal conductivity of EMC is one of the key factors to package thermal resistances, especially for junction-to-case thermal resistance
$(\theta_{\text{JC}})$
of over-molded packages. Through the measured data, the change of thermal conductivity of EMC from 25
^º
C to 150ºC is about 10%. This difference affects the heat dissipation of electronics and results in the changes of package thermal resistances. In system-level thermal analysis, semiconductor packages are usually simplified as compact thermal models (CTMs) for saving the computer calculation time. Common CTMs are two-resistor compact thermal model (2R CTM) and DELPHI CTM. In the commercial thermal simulation software, the thermal resistance in 2R CTM or DELPHI CTM is not a temperature-related variable, but a constant. That means these CTMs cannot present the effect of thermal conductivity change under different temperatures. To consider the effect of temperature-dependent thermal conductivity to CTM, this work proposes a novel CTM called improved CTM and the procedure of constructing this CTM from DTM is described. The improved CTM proposed a this work combines DELPHI CTM with a block with EMC thermal property used to present the effect of EMC thermal conductivity change under different temperatures. An over-molded package is selected as the test vehicle. Junction-to-ambient thermal resistance
$(\theta_{\text{JA}})$
and junction-to-case thermal resistance
$(\theta_{\text{JC}})$
for this test vehicle are measured under different temperature conditions and the measured data are used for the simulation validation of DTM. After that, DELPHI CTM and the improved CTM are extracted from the validated DTM. In this work, the comparison between DTM, DELPHI CTM and the improved CTM are accomplished. The results show the thermal resistances predicted by the improved CTM have better agreement with those predicted by DTM, and this CTM can accurately present the change of thermal conductivity of EMC at different temperature conditions.