Analysis of Capacitor non-idealities on PDN performance

Accurate modelling of power distribution network (PDN) design in high-speed systems is a critical challenge. The equivalent electrical circuit of decoupling capacitor (decaps) becomes ambiguous due to temperature, bias voltage and aging dependency. This paper presents a workflow to model the decaps by considering of the temperature and bias voltage as the uncertainty parameters for predicting the PDN impedance. A pareto-front analysis is performed to quantify the uncertain parameter sensitivities on the PDN impedance. A test PDN structure with multiple decaps were analyzed to quantify the variations in the impedance compared to the nominal case. Results show that variations in decap parameters can lead to a 25% overestimation in the overall PDN impedance.