Graduation date: 2006
This thesis presents a Z-parameter based model to predict the substrate
noise coupling between two contacts in a heavily doped substrate for frequencies
less than 2 GHz. The empirical model is scalable with contact size and spacings
between the contacts and model parameters can be readily extracted from simu-
lated or measured data. The error is within acceptable limits and computational
costs associated with extraction of substrate parasitics is significantly reduced by
using this model compared to numerical techniques. An application of the model
to analyze the substrate noise coupling between a digital and analog block is also
demonstrated.