Authored By:
David Lober, Mike Bixenman, D.B.A., Mark McMeen
Magnalytix
TN, USA
Summary
From the development of Surface Insulation Resistance (SIR) an emphasis has been placed on utilizing a few “standard” test patterns, namely the “Y” test pattern, and interdigitated “comb” patterns. These designs have historically lagged “leading edge” PCB design rules, and even typical PCB design rules. These patterns share a common feature, in that they are “spherical cows”, meaning that they are a highly simplified model of a complex phenomenon. The long, uniformly spaced, parallel electrode shape of the “Y” and “comb” patterns allows for the assumption of a uniform electric field. This means that the force imparted on an ion between the test patterns is independent of where it is located. However, this approximation is not valid when applied to modern components.
“Real” footprints of modern components have complex geometries which violate these assumptions. This may lead to certain areas underneath a component to experience higher electric fields, meaning ionic contamination is subjected to greater forces, leading to increased ElectroChemical Migration (ECM) risk. Newer SIR test patterns are starting to use real component footprints as the basis for their design.
In this paper, which is the first in a series of examining the relationship between SIR, ECM, and SIR test patterns, we examine the effect that various design factors of SIR test patterns has on the electric field on the surface of the PCB.
Conclusions
The goal of this paper is to start transitioning the industry from a spherical cow, to a cylindrical one. We are not attempting to create a comprehensive model of SIR, but we are attempting to improve the current model.
Initially Published in the SMTA Proceedings
|
