Quantitative Analysis and Prediction of EMI From Spread Spectrum of Differential Random Signaling Based on IC Source Modeling

Abstract

This article proposes a quantitative analysis and prediction method for electromagnetic interference (EMI) from common mode (CM) noise in differential random signaling based on integrated circuit (IC) source modeling. To comprehend the EMI source, a spectrum of the CM noise due to nonideal differential random data was theoretically analyzed from the perspective of duty cycle distortion. Also, the effect of spread spectrum clocking (SSC) was derived and calculated as an EMI mitigation method. Next, to predict the EMI in high-speed differential signaling, an IC source model of a current mode logic driver was obtained from a Thevenin equivalent circuit with measurement-based CM source impedance and voltage. The IC source model was validated with the theoretical CM spectrum and was used for field simulation radiated from a simplified peripheral component interconnect express channel structure. The accuracy of the field simulation was estimated by comparing the simulated fields to the measured fields. Finally, to quantitatively estimate the degree of the EMI reduction using the SSC, spread field spectrum was obtained by applying the SSC calculation to the simulated fields and validated by comparing the spread field spectrum measured with an SSC generator.