Effect of Grain Size Distribution on the Performance of Perpendicular Recording Media

Micromagnetic simulations of perpendicular recording in hard disk storage media have been performed with model media of variable microstructural disorder. Simulations indicate that increasing disorder, either due to size and shape distribution or due to disordered packing, decreases signal and increases noise. The mechanism observed in the model is that, in a disordered microstructure, there is a distribution of magnetostatic and exchange coupling between grains that acts to create clusters of grains that act collectively. These clusters increase the auto-correlation function of the spatial distribution of magnetization that is a measure of the magnetic feature size. Consequently, the transition width between recorded bits increases and the position variation of the transition locations (jitter) increases, so that signal-to-noise ratio (SNR) falls. The results suggest that microstructurally ordered media will exhibit higher performance, and that such effects may ultimately demand the use of self-assembled or patterned media with regular packing and very narrow size distribution     

Noise due to particle distributions in metal-particle media

Noise is a concern when it comes to storing data because noise can cause a corrupted signal and, therefore, lost data. Micromagnetic simulation was used to gain insight and understanding of particulate media noise due to nonuniform particle distribution. These distributions result in a spatial fluctuation of the anisotropy field and orientation of the particles, which introduces randomness into the recording process. The effect of the randomness (i.e., noise) on the recording process was studied, as were effects of reduced media thickness and improved particle alignment on the noise level and record nonlinearity. Effect of the noise on write-equalized recordings was also investigated. The focus of the paper is on thin metal-particle media     

Media limitations on achievement of maximum detection window

The recording performance of a digital recording system is primarily dependent upon the isolated half pulse width (T50), signal amplitude, intersymbol interference (ISI), signal to noise ratio (SNR) and the horizontal detection window margin (Tm). These characteristics are related to the thickness and magnetic properties of the media, as well as the head/media interface. This paper addresses the relationship of the media characteristics in longitudinal digital magnetic recording for both particulate and thin film media to the overall recording system performance. The results are shown to proceed from further development of the quasi-optimum time containment filter analysis for media noise dominant and electronic (white) noise dominant recording channels [1], [2].     

Recording by an ideal single-pole head onto an array of Stoner-Wohlfarth magnetic particles

This paper presents a theory of recording by a probe head onto an array of individual magnetic particles. The theory predicts the area packing density of particles and shows that head-to-medium spacing is the prime factor in determining its value. The paper investigates the trade between attainable packing densities and variations in medium properties.     

Thermal effect limits in ultrahigh-density magnetic recording

In current longitudinal magnetic recording media, high areal density and low noise are achieved by statistical averaging over several hundred weakly coupled ferromagnetic grains per bit cell. Continued scaling to smaller bit and grain sizes, however, may prompt spontaneous magnetization reversal processes when the stored energy per particle starts competing with thermal energy, thereby limiting the achievable areal density. Charap et al. have predicted this to occur at about 40 Gbits/in². This paper discusses thermal effects in the framework of basic Arrhenius-Neel statistical switching models. It is emphasized that magnetization decay is intimately related to high-speed-switching phenomena. Thickness-, temperature- and bit-density dependent recording experiments reveal the onset of thermal decay at “stability ratios” (K_uV/K_BT)₀ ≃35 ± 2. The stability requirement is grain size dispersion dependent and shifts to about 60 for projected 40 Gbits/in ² conditions and ten-year storage times. Higher anisotropy and coercivity media with reduced grain sizes are logical extensions of the current technology until write field limitations are reached. Future advancements will rely on deviations from traditional scaling. Squarer bits may reduce destabilizing stray fields inside the bit transitions. Perpendicular recording may shift the onset of thermal effects to higher bit densities. Enhanced signal processing may allow signal retrieval with fewer grains per bit. Finally, single grain per bit recording may be envisioned in patterned media, with lithographically defined bits     

Model for demagnetization-induced noise in thin-film magneticrecording media

A linear, statistical model is described which predicts the power spectrum of measured noise in bulk-demagnetized (i.e. AC-erased) thin-film magnetic recording media. It is shown that the noise is the result of magnetic flux which is ascribed to erasure-induced transitions along the track length in the medium. The noise power spectrum for a rigid disk medium is shown to correspond to the power spectrum of Poisson-distributed induced transitions along the track length, while noise along the track width is sufficiently described in terms of a uniform, average magnetization with small variance. Experimental data from two thin-film disks are used with the model to estimate the Poisson parameter for each disk. It is demonstrated that AC-erased noise from particulate media can be considered as a limiting case of the Poisson model     

The average power density spectrum of the readback voltage from particulate media

An excellent report on the spectrum of readback voltages from magnetic media can be found in Ref. [1] (Thurlings) where the power density spectrum was developed from basic assumptions on the distributions of the particles in the media. In attempting to fit our experimental observations with the above model, it was found that the expressions suggested in [1] had to be extended in order to fit the variations in gain as a function of frequency of the written signal. The extensions include the formulation of a simple but adequate model for magnetic interaction among particles. This interaction model explains the dependence of noise on the frequency of the written signal.     

Micromagnetic predictions for signal and noise in barium ferriterecording media

A dynamic, micromagnetic recording process simulation is used to predict signal and noise for barium ferrite media. The model has a high spatial (particulate level) and temporal (≈10^-9 s) resolution. The theory includes head saturation, image charge, particle magnetostatic interactions, and particle anisotropy. Most input media and head characteristics are taken from direct measurements, such as electron microscopy or magnetometry, or from the literature. However, several variables, such as head-medium spacing, can only be estimated from the available data; errors in these estimates could affect the predicted numbers by several dBV. Overall, the calculations yield very high accuracy; both signal and bulk erase noise match experiment to within ≈1 dBV. This unusually close agreement for the bulk erase noise is obtained without the benefit of particle-particle interactions. This latter point places barium ferrite in interesting contrast to acicular particle media where interactions are believed to substantially reduce short-wavelength bulk erase noise     

Correlation-sensitive adaptive sequence detection

In high density magnetic recording, noise samples corresponding to adjacent signal samples are heavily correlated as a result of front-end equalizers, media noise, and signal nonlinearities combined with nonlinear filters to cancel them. This correlation significantly deteriorates the performance of detectors at high densities. In this paper, we propose a novel sequence detector that is correlation sensitive and adaptive to the nonstationary signal sample statistics. We derive the correlation-sensitive maximum likelihood detector. It can be used with any Viterbi-like receiver (e.g., partial response maximum likelihood, fixed delay tree search, multilevel decision feedback equalization) that relies on a tree/trellis structure. Our detector adjusts the metric computation to the noise correlation statistics. Because these statistics are nonstationary, we develop an adaptive algorithm that tracks the data correlation matrices. Simulation results are presented that show the applicability of the new correlation-sensitive adaptive sequence detector     

Magnetic properties of new (NP) hydrothermal particles

A new type of uniaxial particles for magnetic recording whose precursors are produced via an original hydrothermal process was recently introduced in the market. The characterizing feature of these materials is their extremely uniform size and ellipsoidal shape. Most of their properties can be ascribed to the lack of external (pores, dendrites, sharp edges) as well as internal (because of ellipsoidal shape) - self-demagnetizing sources, which suggests the name of "Non-Polar (NP)" particles. As a result of this morphology, they show unique magnetic properties. The magnetization reversal mechanism is not accounted for by any known mechanism as fanning or curling and the rotational hysteresis gives very sharp Wrvs. H curves with the lowest values for the integral ever measured for particulate media. The behaviour of coercivity versus packing fraction for Cobalt-surface-modified NP particles is described by a new "constricted magnetization" model. The reasons for these unique characteristics and the practical impact that these materials may have in the field of magnetic recording are discussed.     

Sources of noise in erasable optical disk data storage

Noise sources in the readback signal for phase-change and magneto-optical disks at red, green, and blue wavelengths are examined, and a simple model is presented to explain the observed noise spectra. For phase-change disks the media noise, which corresponds to ~0.4% fluctuation in the disk's amplitude reflection coefficient, is the limiting performance factor for the conventional detection scheme. In magneto-optical media the depolarization noise, whose fluctuations are ~0.05% of the disk's reflection coefficient, is the major contributor to the media noise in the differential detection scheme. In phase-change optical disks the main sources of noise are the roughness of the groove profiles and the graininess of the polycrystalline recording layer. In nongrooved regions of the disk the media noise measured with green light is found to be nearly the same as that obtained with the red light. In magneto-optical disks the scattering of light from the rough groove profiles, as well as media inhomogeneities, gives rise to depolarization. Measurements on nongrooved regions of a magneto-optical disk indicate that the media noise obtained with the green light is somewhat higher than that obtained with the red light.     

Magnetooptics, lasers, and memory systems

The field of magnetooptics is reviewed and the application of a magnetic memory system as a readout technique is discussed. A review and comparison of the fundamental magneto-optic effects and their utility in a system is presented. It is shown for a longitudinal Kerr readout system that laser and shot noise limit wide-band (1 MHz) signal-to-noise ratios to about 40 dB. Media noise problems are reviewed. The limitations to packing density are discussed, and it is concluded that packing densities greater than 10⁷bit/in²(including suitable guardbands) are practical. The various techniques for optico-thermal recording are surveyed. A discussion of related hardware components (such as optical modulators and lasers) is presented. It is concluded that a viable magnetooptic detection-laser beam memory system is practical. No suitable nonmechanical scanning system has yet been developed.     

Analogy between magnetic recording media noise and shot noise

By comparing the analyses of magnetic recording media noise and shot noise, this paper shows that a very close analogy exists between the two. It demonstrates that all the well-known results for both kinds of noise are caused by random processes that have equal probabilities in equal volumes.     

Characteristics and applications of metal tape

To manufacture low noise and high output tape, we have to choose excellent particles and also make its surface as smooth as possible. We developed a metal tape with higher output and higher S/N compared with those of conventional oxide tape by introducing new technologies of higher packing density and mirror surface finishing. This metal tape realized higher recording density which could be used in 8 mm Video, electronic still camera ("MAVICA"), magnetic contact printing system and various digital recording systems. This paper describes characteristics and applications of the metal tape.     

Noise characteristics of stacked CMOS active pixel sensor for charged particles

The noise characteristics of a stacked CMOS active pixel sensor (SCAPS) for incident charged particles have been analyzed under 4.5 keV Si⁺ ion irradiation. The source of SCAPS dark current was found to change from thermal to electron leakage with decreasing device temperature. Leakage current at charge integration part in a pixel has been reduced to 0.1 electrons s⁻¹ at 77 K. The incident ion signals are computed by subtracting reset frame values from each frame using a non-destructive readout operation. With increase of irradiated ions, the dominant noise source changed from read noise, and shot noise from the incident ions, to signal frame fixed-pattern noise from variations in sensitivity between pixels. Pixel read noise is equivalent to ten incident ions. The charge of an incident ion is converted to 1.5 electrons in the pixel capacitor. Shot noise corresponds to the statistical fluctuation of incident ions. Signal frame fixed-pattern noise is 0.7% of the signal. By comparing full well conditions to noise floor, a dynamic range of 80 dB is achieved. SCPAS is useful as a two-dimensional detector for microanalyses such as stigmatic secondary ion mass spectrometry.     

Basic properties of AC noise

The noise characteristics of ac demagnetized particulate media with special emphasis on correlation function analysis are discussed. The recording samples were specially prepared iron-oxide media coated on glass substrates by means of a spin coating technique. The media were nonoriented, which meant that the particles in the dried coating constituted a virgin state which is different from the ac erased state. The noise spectra for the virgin state and the ac demagnetized state appear to be equal for wavelengths longer than about 20 μm; for shorter wavelengths, the difference is about 2 dB. A correlation function measurement was made for the long wavelength region. The noise voltage was digitized and stored in a computer. After the measurements, the correlation between different noise voltages was computed. As a result, no correlation was found at all (ρ = 0) between virgin noise and ac noise. When the sample was ac demagnetized again, there was a well-defined correlation (ρ = 0.32) between successive ac noise levels. This can be explained on the basis of the ac demagnetization process.     

An investigation of high-frequency bias-induced tape noise

Bias-induced tape noise remains a major limitation of the SNR in audio magnetic tape recording systems. Defined as the increment in system noise incurred when the bias oscillator is turned on, the noise can originate from a number of different causes; namely, bias oscillator harmonic distortion, magnetized heads, the earth's magnetic field, and an intrinsic noise source. The latter noise source is our primary concern here. Such record system parameters as head-to-tape spacing, gap length, bias current, and bias frequency were investigated with regard to their influence on this intrinsic bias noise source. Two models of the mechanism of intrinsic bias noise are examined. The first, the "amplitude modulation model." proposes that bias noise is generated by amplitude modulation of the recorded bias signal by the physical and magnetic variations of the head-tape system. In this model, bias noise is merely the lower AM sidebands of the recorded bias signal. The second model relates bias noise to the interaction fields in erased tape. This model proposes that these fields behave similarly to normal recording fields and can be "re-recorded" on the tape at an enhanced level. The two proposed mechanisms are examined in the light of the experimental data. The amplitude modulation model is shown to agree with all the observed data with the exception of the existence of bias noise at bias wavelengths smaller than the particle size. The second model, which does not incorporate a wavelength dependency of bias noise, is in qualitative agreement with the observed data. Methods are discussed for reducing the bias noise without materially affecting the system performance.     

A particle packing model for sand–silt mixtures with the effect of dual-skeleton

The study of particle packing models for binary mixtures is important in the field of granular materials, from both theoretical and practical perspectives. A number of particle packing models have been developed for predicting packing density (or void ratio) of a binary mixture. However, the measured results and the predicted values do not always agree with each other, particularly in the range of fines content between 25 and 50%. It is postulated herein that the discrepancies between the measured results and the predicted values are primarily due to the incorrect assumptions used in the existing models. In the existing models, the packing density is determined from one of the following two assumed mechanisms of particle mixing: (1) the mixed packing has a dominant large-particle skeleton and the small particles fill the voids of the large-particle skeleton, or (2) the mixed packing has a dominant small-particle skeleton and the large particles are embedded in the small-particle skeleton. It is obvious that the first assumed mechanism is only applicable for mixtures with low fines content, whereas the second assumed mechanism is only applicable to mixtures with high fines content. Therefore, the predictions from existing models are unsuitable for mixtures with medium fines content, such as a mixture of fines content between 25 and 50%. In this study, a 3-D discrete element simulation is carried out to show that, for a mixture of medium fines content, the packing structure has a dual-skeleton, which is neither dominated by a large nor small-particle skeleton. Then, we postulate that, in the mixed packing, both mechanisms can take place: filling of small particles and embedment of large particles. The concepts of “dual-skeleton index” and “index size” are proposed to account for the interactive effects of filling and embedment. Based on this postulation, we develop an analytical method, which has the capability of predicting minimum void ratio for sand–silt mixtures with various fines contents. The developed model is then validated by the experimental results obtained from 16 types of sand–silt mixtures.     

Crucial points in perpendicular recording

Since the introduction of perpendicular recording on a floppy disc by IWASAKI in 1977 and its equivalent design on a rigid disc (SPH-like sensor + double-layer medium) in 1981, many tests have been carried out on different R/W sensors. For each test the main goal was the fci record or the improvement of the magnetic layer. Seen from the recording system point of view, the head and the medium are looked at as a unit through a specification, unchanging with increasing area density. For example, a minimum of 26 dB and 70 % must be achieved for the S/N ratio and the resolution respectively. By considering the noise of the best electronic channel (with a thin film head), and ignoring mechanical and medium noises, the output signal must be at least 250 μv pop. For a 50 Kfci application, however, a sensor does not yet exist. Using a ferrite head with a 1.2 μm gap length to write on FeTbGd, the level of the signal will not be high enough to be used. It is improved with a 0.6 μm gap head but then, the field doesn't allow us to write ! Such problems exist also with thin film heads or SPH like sensors on rigid discs. To improve the R/W process, the trend is to use a double layer medium e.g. CrCo/FeNi. The results show that this direction is not necessarily the best. For example, when erasing or over-writing with the head, some domains appear in the FeNi film which create noise from the track or its edge. Another example is the fact that the optimum parameters for a medium such as CrCo are not always compatible with the characteristics of the head (i.e. Hc, the thickness, the crystallographic orientation, the bit stability compared to the write field, the signal, the noise...).     

Recording performance of discrete track patterned media fabricated by focused ion beam

We report on the recording performance of discrete track patterned media fabricated by focused ion beam (FIB). We investigated performance over a small area by spinstand read/write testing. Discrete track patterned regions show smaller magnetic track width and better signal separation between adjacent tracks and therefore higher track density than that of nonpatterned continuous media as a result of reduced side fringe effect and edge noise. We found that, at a designed groove depth of 4-8 nm, the shallow FIB etched grooves already provide good isolation between adjacent tracks, indicating the superiority of ion beam induced modification of magnetic properties in film media over physical modification of disk surface topography. This has implications for discrete track recording and media fabrication.