Simulation of the effect of medium noise on servo position error signal detection in longitudinal recording

A medium model with random grain structure and desired grain size distribution is used to study the effect of medium noise on position error signal (PES) signal and sensitivity using 100 Gb/in/sup 2/ parameters in longitudinal media. Three PES detection schemes are compared: (1) average peak voltage; (2) integration of the absolute playback signal; and (3) maximization of the integration of the product of a phase-adjusted square wave with the playback signal. Scheme III is found to have the best output PES sensitivity for every case. The PES, PES sensitivity, and null point position of the RES signal along with their distributions are studied. It is found that the PES sensitivity has a nearly Gaussian distribution. In most cases, the null point position of the PES has a nonzero average and an asymmetric distribution. Media with different grain sizes and size distributions are investigated and compared. The grain size distribution has a small effect on the PES detection. Increasing the average grain size dramatically increases the standard deviation of the PES sensitivity while slightly decreasing the PES sensitivity. The simulation results also show that increasing the medium grain orientation ratio (OR) improves the PES sensitivity.