Reply To: Aligning Modeled and Measured Eye Diagrams in Channel Analysis

[Dan Dvorscak]

When you extracted the various models, were they all at nominal conditions, or are you simulating across manufacturing process corners? Even if you just consider the effects of manufacturing on characteristic impedances, a single eye diagram simulation with all models at their nominal impedance is going to be something of a best case scenario. You may want to consider taking a design of experiments approach where you have multiple versions of each model  to account for that process cornering and set up a series of sweeps to include all of the potetnial scenarios.