PCB stack-up design and optimization for next generation speeds

Abstract

An efficient empirical simulation tool was presented to predict the transmission line performance of coupled differential microstrip and strip lines. Based on the Artificial Neural Network (ANN) method, the predicted transmission line performances were accurate and were suitable for printed circuit board (PCB) stack-up design. Design examples were presented how this tool helping on differential strip line structure parameters design and predicting their performance range with the structure parameters variation.