Applicability of JESD51-14 to clip-bonded, discrete power devices

Junction to case thermal resistance (R
th-JC
) is one of the most important metrics used to assess thermal performance of semiconductor components. During development of modern automotive ECUs the value of R
th-JC
from datasheets has been traditionally used to compare thermal performance of active electronic components and perform an informed selection among available choices. Economic impact of the right or wrong R
th-JC
value is high, therefore it’s important to know it as precisely as possible. JESD51-14 norm describes the most widely used measurement method for R
th-JC
together with a procedure for evaluating the experimental data. This paper presents some of the shortcomings of the method when applied to modern, clip-based power MOSFET packages used in current automotive ECU designs. Systematic bias in evaluation method and sources of random test setup variations are investigated to estimate the accuracy of resultant values. Conclusions about the applicability of JESD51-14 method to clip-based and other packages are formulated along with few suggestions for improvement.