Side-reading free spin-valve read sensor

The feasibility of high track density recording in a perpendicular magnetic recording system has been investigated. The properties of the magnetic field of a single pole type (SPT) head having a narrow track width were studied by using a large scale model of a head and a medium. The y-field gradient of the SPT head near the track edge is steeper than the x-field gradient of a ring head, and it was confirmed by Bitter technique that the recorded track width acculately coincides with the width of the main pole of the SPT head. Perpendicular recording using the SPT head also has superior off-track characteristics and the crosstalk signal is hardly detected even at very narrow track pitches. From a magnetics point of view, no problems exist in narrowing the track width of the SPT head in perpendicular recording. Narrow track recording and reproducing with SPT head of only 6.5 μm in track width could be achieved with sufficient signal to noise ratio in perpendicular flexible disk systems.

As continued reducing the track width (TW) of magnetic read sensor for gaining areal density in hard disk drive (HDD) industry, stripe height (SH) is needed to be reduced as keeping aspect ratio(TW/SH) for maintain resistance and signal constant. This paper is proposed to characterize the free layer behavior of tunneling magneto resistive (TMR) head with multiple SH by using ferro magnetic resonance analyzer (FMRA) that is able to analyze by layer-level state without applying external field. The results are found that, at shorter SH, effective stiffness field of free layer is increased which is based on decreasing FMR amplitude and increasing FMR frequency. As a result, magnetic read sensor is shown less sensitivity and signal to noise ratio (SNR) with shorter SH. The SH is a correlated function of the effective stiffness field in free layer that is referred to changed shape anisotropy field which has an effect to read ability.

In this article we present a quantitative micromagnetic modeling analysis of thermally activated magnetic noise in spin valves heads at deep submicron track widths. It is found that, for these spin valve heads, the thermal magnetic noise is at a magnitude that is comparable to that of the head’s Johnson noise. Reducing the sensor track width by keeping the same aspect ratio and film layer thickness will not necessarily yield an increase of thermal magnetic noise, which is attributed to the accompanying increase of the demagnetization field within the sensing layer. However, both the magnitude of noise and the noise spectra are strong functions of the sensor stripe height. An excessive stripe height will yield significantly higher thermal magnetic noise.

For the MR strips stabilized by the permanent magnet films, we numerically analyze the effects of the leakage field of the permanent magnet films on the bias angle in the MR films. Using the magnet films as the domain control films, magnetization rotation in the MR films at the track edges is suppressed by their leakage fields and will in turn decrease the effective track width. The decrease effective track width is thought to be due to the large leakage field especially at the track edges, so we examined some cases of decreasing the leakage field at track edges. Among the cases, the effective track width was most effectively enlarged when the distance between the magnet films was widened compared with the spacing of the leads. >

The data retrieval resolution of spin-valve reading head is a function of the spacing between the shield layer and the free layer of the spin-valve head - the head gap. Spin-valve head is of multi-layer structure. The layer thickness variation among different wafers leads to the variation of the gap length. Therefore it is important to develop a suitable methodology for testing and evaluating the gap length fluctuation not only during the head fabrication process but also at the magnetic integration level. This paper reports a novel method for the quantitative evaluation of the gap length fluctuation among different heads. The proposed method proved to be simple and easy for implementation.

Unshielded spin-valve sensors (6×2 μm2) with both synthetic ferrimagnet free (SF) and synthetic antiferromagnet pinned (SAF) layers are demonstrated, exhibiting high sensitivities, good thermal stability and reduced free layer effective magnetic thickness (teff). The free layer teff can be reduced to ∼10 Å while maintaining its physical thickness near 60 Å, without significant magnetoresistive signal loss (MR∼6%). After current biasing, the sensors show a linearized response, demonstrating a coherent switching of the synthetic free layer. Sensor sensitivity (S) increases with decreasing free layer effective magnetic thickness (S≈0.5%/Oe at teff=10 Å) as expected from the decrease of the free layer demagnetizing field. The coupling field (Hf) between the free and pinned layers increases with decreasing teff (Hf∝Ms2/(teffMefff)). The sensors are thermally stable upon anneals up to 300 °C.

This paper discusses the relationship between the effective write track widths of thin-film heads and the coercivities of thin-film disks. The off-track performances of the output amplitudes and media-noises were investigated by using Co-Cr-Ta sputtered thin-film disks with different coercivities. The effective write track width becomes larger with an increase in the disk coercivity if the gradient of the head write field is large. If the gradient of the head write field is small, the wavelength of the zigzag wall becomes short and its amplitude becomes large in the track width direction of the transition regions. At the track edges of the transition regions, the track width direction component of magnetization increases because of the bitpattern bridging between the neighbor transitions.

Magnetoresistive spin valve sensors based on the giant- (GMR) and tunnelling- (TMR) magnetoresisitve effect with a flux-closed vortex state free layer design are compared by means of sensitivity and low frequency noise. The vortex state free layer enables high saturation fields with negligible hysteresis, making it attractive for applications with a high dynamic range. The measured GMR devices comprise lower pink noise and better linearity in resistance but are less sensitive to external magnetic fields than TMR sensors. The results show a comparable detectivity at low frequencies and a better performance of the TMR minimum detectable field at frequencies in the white noise limit.

This paper describes the noise spectra in the GHz range of tunneling magnetoresistive (TMR) read heads as a function of both bias current and external magnetic field in order to understand the effect of spin-torque (ST) on the thermally excited magnetization fluctuation noise (thermal mag-noise) in the TMR heads. A strong noise spectral peak at around 3-4 GHz is either enhanced or reduced depending on both the bias current and the external magnetic field. This result demonstrates that the ST influences the thermal mag-noise originated from the free layer under the positive (or negative) bias current in case the magnetization of the free and reference layers are nearly anti-parallel (or parallel) to each other.

Narrow track recording in a double-layer perpendicular film medium is studied via micromagnetic modeling. A 3-D probe head field with a track width W=1.5 mu m is used for recording simulations. Recordings of multiple consecutive transitions are simulated. Edge transition patterns and edge overwrite properties are characterized. Simulated transition magnetization pattern at track edges turns towards to the head motion direction and extends to the next transition front. Edge overwrite is studied by simulating recording of a new track over a previously recorded track. It is found that the edges of the new track exhibit no separations with the overwritten track and the intertrack edge of newly recorded transitions is strongly modified by the previously written transitions.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

This study considers the analysis of the developing flow due to the mixing of two uniform incompressible fluid streams with different velocities. Using the method of matched asymptotic expansions, solutions valid in three regions of the free shear layer are obtained. The expansion process uses a small parameter based on the difference between free-stream velocities and, unlike previous studies, is applied to the complete Navier–Stokes equations. The controversy involving the choice of a proper third boundary condition for the free shear layer between parallel streams is avoided in the present approach. This is achieved by matching the downstream solutions, valid inside and outside the mixing layer, to the solution describing the flow in the initial mixing region near the start of mixing. Results of the analysis indicate that the transverse velocity, in the developing free shear layer between two uniform streams, approaches similarity at a much slower rate than the longitudinal velocity. The analysis also showed that, in the early stages of the mixing process between two uniform streams, the faster stream is initially displaced away from the shear layer. This initial flow displacement is not predicted by the downstream mixing-layer similarity solution.

The read performance of spin-valve heads with overlapping leads is studied by using a micromagnetic model. In the case of hard biased spin-valve heads without an overlapped region, the sensitivity (output per lead distance) falls significantly with decreasing lead distance because of the existence of dead zones at both ends of the track. However, heads with 0.5-μm-wide overlapped regions at the ends of the sensing film have high sensitivity even if the lead distance is reduced to 0.5μm. Their output for an effective track width of 1.0μm is 30% larger than that of heads without the overlapped region when the heat generation is the same. Their output has a smaller dependence on the longitudinal bias strength, indicating a larger margin for the longitudinal bias. Spin-valve heads with overlapping leads are promising for high-track-density recording.

A magnetoresistive (MR) read sensor design for magnetic recording has been proposed in an attempt to solve the magnetization distribution and thermal stability problems of nanometer-scale read sensors. In this design, the free layer has a perpendicular-to-plane easy axis and the reference layer has an in-plane easy axis. The giant-magnetoresistive film with this structure is demonstrated in which the free layer is a [CoFe∕Pt]n multilayer. The structure of the [CoFe∕Pt]n multilayer has to be optimized so that the magnetic properties of the free layer can meet the requirements of a MR read head.

Experiments were conducted in the nonequilibrium region of a free mixing layer with unequal freestream velocities. Four velocity ratios U(1)/U(2) of 0.32, 0.46, 0.74, and 0.96 were used in this investigation. The growth of the shear layer as well as the velocity adjustment in the near wake were examined. There was reasonable agreement between the measured mean velocity profiles and those computed using the K-epsilon turbulence model. Some periodic turbulence velocity fluctuations were observed in the mixing layer, but their frequency remained the same along the flow.

Spin-valve sensors of the type NiFe/Cu/Co have been designed for optimal biasing behavior and successfully incorporated into a gigabit-type shielded read head configuration with a read trackwidth of 2 /spl mu/m, a read gap of 0.25 /spl mu/m, and a MR sensor height of 1 /spl mu/m. The spin-valve sensor had a structure of 100 /spl Aring/ NiFe/25 /spl Aring/ Cu/22 /spl Aring/ Co/110 /spl Aring/ FeMn, and yielded a net spin-valve coefficient of /spl sim/3.5% at the completion of head processing. Uniform field testing of the read heads after wafer fabrication and lapping showed quiet and stable spin-valve response with near optimal bias performance. Recording tests of the read heads at a head-disk clearance of 1.5 /spl mu/m showed linear, non-saturated signal response on a media with an areal moment as high as 1.25 memu/cm/sup 2/, yielding reasonably symmetrical signals with peak-to-peak amplitudes ranging from /spl sim/750 /spl mu/V//spl mu/m to as high as /spl sim/1000 /spl mu/V//spl mu/m of read trackwidth. Linear density rolloffs and microtrack profiles have also been studied, and results showed behaviors closely agreeing with design targets for high density recording operations. >