Content area
Abstract
Density limits in magnetic recording technology rely on the minimization of noise to obtain sufficient signal-to-noise ratio (SNR) in recording systems. A major degradation of SNR comes from off-track misregistration during head operation. With the ever-decreasing information bit width, these off-track problems, such as precise positioning of the head during read/write operations, have never before become so important.
Here, in this dissertation, several issues related to noise and off-track or track phenomena are studied both experimentally and theoretically. The mechanism of medium noise, the dominant noise source, is investigated. Analyses of noise modes and the impact of track edge noise on the noise composition is made. Intrinsic magnetization noise is studied by a deconvolution technique. A simple model reflecting the actual random granular structure of the recording media is developed to simulate the noise characteristics of realistic recording systems. With this simple model, a study of position error signal in both longitudinal and perpendicular recording systems is made. This model is further extended to write arbitrary length pattern to study off-track performance of a hypothetical 1Tbits/in2 recording system.
In addition to track misregistration, another source of SNR degradation is medium thermal decay after the recording. The evolution of medium noise components are studied during the course of thermal decay both experimentally and theoretically. SNR and the bit error rate (BER) are also calculated using a simple self-consistent model.
Beside medium noise, thermal magnetization fluctuation noise resulting from ambient thermal fluctuations in the GMR head is addressed. Theoretically, a calculation method based on a matrix representation and diagonalization technique is developed using a theory based on the idea of eigenmode. Experimentally, both low frequency (spinstand tester) and high frequency measurements are done to observe and characterize this type of head noise.
