Enhanced Linearity Range for Magneto Resistive Transducer Servo by Changing the Direction of the Bias Current

The present invention pertains to magnetic data storage devices and more particularly to servo systems for such devices that employ magneto resistive (MR) transducer heads.

In a magnetic disk drive, the positioning of the transducer head over data tracks in alignment with the track centerline is accomplished using a "servo loop" including a mechanical actuator with a feedback path utilizing servo patterns written at the time of manufacture on sector segments or surfaces of the disk pack reserved for servo information. The servo patterns, head parameters, disk parameters and detection methods are customized to the specific application to ensure a linear response across the track or portion of the track. Measurable characteristics of the functionality of the servo system are the linearity achieved, the range over which the linear region is maintained, and the repeatability with respect to the signal to noise ratio, defects, and variations in the head's and disk's physical dimensions and magnetic properties.

Recent advances in magnetic recording heads have incorporated a magneto resistive (MR) element to sense the flux from the media. The MR element is typically biased to provide linearity, polarity and maximum sensitivity for detecting transitions written on the disk. The effects of biasing the sensor in combination with the physical geometry of the sensor cause non-uniformity of the read sensitivity along the length of the sensor with greater amplitude sensitivity occurring near one end. The non-uniformity of the read sensitivity along the length of the sensor causes non-linearity and a reduction in the width of the linear region of the servo systems position error signal (PES) for an amplitude type detection method. The PES's linear region becomes progressively more curved and narrower as the sensor's response along the MR elements length becomes more asymmetrical.

Reversing the bias current in an MR sensor causes the read back sensitivity profile to be mirrored about the center of the sensor. The technique of the present invention improves the PES linearity and range by altering the MR sensor's readback sensitivity while reading the servo burst displaced to one side of the track centerline as compared to that used to read the burst displaced to the other side of the track centerline by reversing the bias current. In this manner the more uniform higher gain portion of sensor response may be used to sense each of the servo patterns irrespective of the side of track center toward which it is displaced.

The principal benefit attained from implementation of the technique is to improve the servo system PES linearity by providing a more uniform and wider linear region from an MR sensor with an asymmetrical or non-uniform response. However, employment of the invention also provides benefits including reduction of the servo time required to write servo patterns during device manufacture, simplification of the microcode used for head positioning and improvement in the ability to compromise between servo and data requirements for the MR stripe height and length dimensions.

The reduction in servo writer time is realized by allowing wider servo patterns to be utilized for a given pitch while maintaining good linearity. An example being a servo system usage of `half track` patterns instead of third track patterns to achieve linearity across the width of the track pitch. The use of wider servo patterns also provides as a collateral benefit, higher servo signal amplitude due to the presence of fewer erase bands. Simplification of servo positioning is achieved by reducing the number of PES curves required to construct a linear PES.

The improved ability to compromise between constraints placed on the MR head design for the data and servo is based on the servo system requirements for a head with wider read width (in relation to data requirements) to maintain good PES linearity. Data requirements are generally optimized to provide a written track of larger width than the MR sensor's read width to maintain good soft error rate performance with file track misregistration (TMR). The present solution effectively increases the functional width of the sensor by selectively using the higher gain, more uniform portion of the MR stripe sensitivity for sensing the servo patterns thereby widening the response of the MR stripe's sensitivity.

FIG. 1 is a schematic illustration of the dual burst servo pattern for determining track misregistration or sensor position during track follow or positioning.

FIG. 2 is a graph modeling MR sensor amplitude responses along the length of an MR sensor for such devices with varying asymmetries.

FIG. 3 is a graph modeling servo PES linearity for the uniform and increasingly asymmetrical MR response functions illustrated in FIG. 2.

FIG. 4 is a graph illustrating the effect of MR bias reversal on PES linearity modeled for the sensor responses of FIG. 2.

FIG. 5 is a graph illustrating the effect of MR bias reversal on PES linearity, wherein sensor gain with bias in one direction is 80% the sensor gain with the bias in the other direction.

FIG. 6 is similar to FIG. 5 wherein sensor gain is 60% in one direction.

FIG. 7 is a schematic elevation of a typical magnetic disk drive.

FIG. 8 is a schematic showing of an MR sensor in cooperation with tracks of a data disk employing a quadrature type servo system.

FIG. 9 is a block diagram of a circuit for implementing the bias reversal technique of the present invention.

FIG. 1 is a schematic showing of a servo system illustrating the magneto resistive (MR) sensor positioned in alignment with the centerline of a data track, which is centered between the amplitude bursts A and B which are laterally displace by a half track pitch. The position error signal (PES) representative of the offset of the head from the track centerline is obtained by calculating (A-B)/(A+B) as the sensor moves along the track where A and B correspond to the amplitude of the read back signal from the A and B bursts respectively.

FIG. 2 illustrates the MR sensor responses having increasingly asymmetrical amplitude readback characteristics relative to the position along the MR stripe. The example modeled uses a unity track pitch and an MR sensor length of 2/3 the track pitch. The solid line curve 2A is the response curve where the varied response portion of the sensor is 10% of each side of the read width and produces an amplitude profile similar to that which occurs when an inductive read head is used. The other curves: 2B, 2C, 2D and 2E show the progressively increasing asymmetrical response profiles of sensor devices with 30%, 50%, 70% and 90% of the left side of the amplitude being varied. At a reference level of 50% of the peak response, the distance along line R from the curve at the left side of each profile to the common curve at the right side of the profile shows the reduction of the PES linear response region with the increase of the varied signal amplitude which occurs as a function of the shape of the sensor. The varied response shown by the family of curves is a result of the length to height ratio of the MR stripe. When the length to height ratio is higher, the uniform response portion at a given reference value or percentage of the amplitude response is greater. This characteristic cannot be optimized to the ideal condition since a lower length to height ratio yields a higher signal amplitude for a constant current density. Thus the designer is required to make a value judgement or trade off with respect to uniformity of response and amplitude of signal. The curve 2C represents a typical response curve which balances the competing requirements.

The graph of FIG. 3 plots the PES against the unity pitch. The ideal symmetrical signal would produce PES linearity substantially across the width of the sensor and the increasing asymmetry progressively reduces the linear response region, the substantially straight line inclined curve portion. The linear portion of curves 3A through 3E is effectively the function of the progressively reduced length of the 50% reference line R in FIG. 2. As the sensor moves over the A and B bursts the PES would be accurate if the ideal linear response was received as indicated by curve 3A which passes through 0 PES when the sensor is aligned with the track centerline. However, the asymmetry of actual MR sensors, as indicated by the progressively increasing asymmetry of curves 3B through 3E, cause the response curve to be displaced as a result of the degree of asymmetry caused by the MR sensor varied amplitude response characteristic. The response asymmetry cannot be eliminated in the MR sensor design since a trade-off is required between signal amplitude and the degree of asymmetry of the response for the given design and manufacturing tolerances. This variation of signal amplitude across the sensor width is not as highly significant when reading data, but becomes a substantial factor in the servo function during either access or track follow when signal amplitudes at the right and left sides of the MR sensor are compared to generate a PES.

By reversing the electrical bias on the MR sensor with respect to the A and B bursts of the servo, a reversal or mirror image of the asymmetrical MR sensor response curve is attained so that in reading the A and B bursts of FIG. 1 the asymmetries of the response are substantially cancelled. As shown in FIG. 4, this results in a PES of zero at the position of sensor alignment with the track centerline with the residual effects of the non-linear signals (curves 4C through 4E) relegated to a modest distortion between the PES and the displacement of the sensor from the track centerline.

A further effect of the PES signal amplitudes experienced during the use of a reverse bias sensor is that the maximum amplitude of the MR response using a bias current in one direction differs from the maximum amplitude of the MR response when the sensor has bias current applied in the opposite direction. This phenomena is illustrated by the curves (5A through 5E) of FIG. 5 where one amplitude is 80% of the other and in FIG. 6 (curves 6A through 6E) where one amplitude is 60% of the other. Comparison of FIGS. 4, 5 and 6 shows that the variations in sensor gains results in changes to the measured track center position and linearity. If the gain values remained the same, the effect would be a displacement of all track centerlines a uniform distance so that a nominal track centerline would be established and repeated without affecting the accuracy of reading or writing data. However, if changes in gain occur, the sensed track centerline moves radially and introduces all the associated problems, such as, overwriting adjoining tracks and increasing read errors as the read element is misaligned with the data or portions of the data have been erased during the writing of adjacent tracks. The impact of this effect can be considerably diminished through multiple methods. The sensor could be designed to have improved uniformity of gain for either direction of bias current, a normalization field could be provided as part of the servo pattern or reserved area of the disk to be used to normalize the head output for each bias direction, or the bias current could be varied based upon a previous characterization to compensate for the gain change. The normalization field would provide the benefit of compensating for gain changes caused by individual current cycles (bias dependent instabilities) which is recognized to directly cause an increase to TMR with this type of system.

A typical magnetic disk drive is shown in FIG. 7. A disk drive unit 12 includes a stack 16 of disks 18, each of which has at least one magnetic data storage surface 20, with the disk stack being rotated in unison by a spindle motor mounted concentrically within spindle 21. Transducer heads 22 are supported on arm-suspension assemblies 24 which are ganged together for simultaneous pivotal movement about an actuator support spindle 26. One arm assembly 24 includes an extension 28 on which is supported a voice coil 30. A voice coil motor 32 drives the voice coil 30 to move the heads 22 from one concentric track to another concentric track on the confronting data surface 20 or to cause the transducer head to be aligned with a selected data track centerline during read or write operations. The various components of the disk file are controlled in operation by signals generated by control unit 34 through signal and control lines 36. 37 and 38 respectively connected to the heads 22, the spindle motor and voice coil 30.

FIG. 8 schematically illustrates a sequence of data tracks (which are concentric circular tracks on the disk surface) that include a servo header portion 39 followed successively by A and B servo bursts: C and D quadrature servo bursts and the data sector portion. The transducer head MR sensor 42 has a length of about 0.6 the track pitch P; is aligned with the track centerline to read and write data and includes the bias leads 44.

The servo control, in accordance with the invention, is practiced by a portion of the control circuitry as schematically illustrated in FIG. 9. The sensor reads the servo header 39 which indicates that servo bursts are to follow and identifies the track type as odd or even. The servo positioning electronics 45 activates switch 46 to connect one of the bias current sources 47, 48 to the MR sensor 42 in accordance with the side of the track at which the A burst will be encountered and following the A burst, reverses the bias current direction by switching to the other of the current sources 47, 48 for the period during which the sensor passes the B burst. The signal from sensor 42, amplified by preamp 49, is connected to the servo positioning electronics 45 which compares the servo bursts and either pulses the voice coil in the direction indicated by the TMR value or refrains from transmitting a control pulse when the sensor is aligned with the track centerline or is within a predetermined minimum TMR value.

While the invention has been shown and described with respect to a preferred embodiment thereof, it will be understood by those skilled in the art that changes in the form and detail may be made in this embodiment without departing from the scope and teaching of the invention. Accordingly, the apparatus and method herein disclosed are to be considered as merely illustrative, and the invention is to be limited only as specified in the claims.