From reliability(IC) point of view, (FI)CDM test method been used for decades to qualify IC robustness. And this test method is a way to separate ±charge in IC, which “NET” charge is zero.
And from the other point of view, CDM model is generated by + or - charge in IC, which “NET” charge is not zero.
For some case, which was diagnosed by CDM damage, and we try to replicate by (FI)CDM tester, somehow it can’t.
So I am wondering that what’s going on:
Anyone have experience on this?
In reality, a charged IC and its discharge(CDM) is more complicated than (FI)CDM?
Hi Bruce, while I can’t answer all of your questions, it is often the case that the damage caused in the real world cannot be exactly reproduced in the lab with a CDM test. A “real world” event does not have the same characteristics as what it experienced in the lab.
I have had events where I found a CDM like discharge in the process but was causing the HBM diodes to fail. The CDM and HBM results are helpful when trying to figure out where the failures could occur but the failures don’t always correspond to the lab results.