Pitfalls when correlating TLP, HBM and MM testing

Abstract

Correlation between human body model (HBM), machine model (MM) and transmission line pulse (TLP) testing is still under discussion. In this paper, it is shown that the lack of correlation is due to misinterpretations in the test results, due to the fact that soft and hard failures are often confused. This is illustrated by ESD results of test structures with grounded-gate NMOSTs and field oxide devices, measured according to HBM, MM and by means of TLP testing. It is shown that for a valid comparison, it is essential to combine electrical measurements with a thorough physical failure analysis in order to establish the failure signature. Once the failure signature and the current path are taken into account, a good correlation is found.