This paper derives a formula that shows a direct relationship between the noise spectrum and the magnetization configuration of a thin-film medium. This formula was obtained by Fourier transforming the down-track autocorrelation function of jMx-Mz, where Mx and Mz are the longitudinal and vertical components of magnetization. The formula takes into account not only the variance of the magnetization deviation, which is generally used to estimate media noise, but also the correlation between deviations in magnetization. This lets us calculate the noise power spectrum as well as the signal power, once the magnetization configuration in a medium is obtained, which means that the magnetic microstructures such as vortex, zig-zag domain, etc., are automatically included in noise calculation. To investigate the relationships between the noise properties and the magnetic properties of media with various anisotropy fields (Hk), axis orientation of easy magnetization, and stiffness constants (A), recorded magnetization configurations were simulated using the Landau–Lifshitz–Gilbert equation.1,2 The easy axis was oriented from uniaxial direction to random direction in plane. The simulated noise power spectra were maximized at a relatively long wavelength, which was varied by changing Hk, or the easy axis orientation. The noise power was strongly influenced by Hk and the axis orientation of magnetization. This confirms that preferential orientation along the longitudinal direction and a large Hk are effective for improving the S/N ratio. On the other hand, the stiffness constant less improved media noise than Hk and axis orientation, suggesting that the preferential orientation of the easy axis is more effective to reduce media noise than isolating magnetic grains.1
Read full abstract