Western Digital Technologies, Inc.

Western Digital Corporation

Thao P Trieu

Ara W Nazarian

Richard Koonwai Wong

Charles W Simmons

Edgar D Sheh

Disclosed is a disk drive having a plant with a moveable transducer and a servo controller that more precisely controls the transducer during short to medium seeks by producing a multi-rate feed-forward command effort signal. The servo controller includes a reference position generator that provides a reference position signal that varies as a function of time along a shaped position profile for moving the transducer from a start position to a target position in a smoothly controlled fashion, a feed-forward control path that receives the reference position signal and applies a feed-forward command effort signal at a feed-forward rate that is greater than the servo sampling rate to move the actuator more closely along the shaped position profile between servo samples; and a feedback control path that receives the reference position signal and the indicated position signal and applies a feedback command effort signal at the servo sampling rate based on a difference between the reference position signal and the indicated position signal to make minor corrections to error remaining after application of prior feed-forward command effort signals.

Disk drive with feed-forward control path that receives a reference position signal to apply a feed-forward command effort at a rate greater than a servo sampling rate

Repeatable runout cancellation in sectored servo disk drive positioning system

A method in a magnetic storage disk drive for independently learning and thereafter canceling repeatable runout that is written to the servo fields of each servo track during the servo writing process. The method does not require a servo writer because the disk drive may independently detect the runout without spinning out of control in response to such runout by operating its servo control loop in a low bandwidth mode so that it is less responsive to the higher frequency components of the runout, and by removing the effect of the DC component of the runout or selected harmonics on each revolution. The method takes several revolutions to 'learn' the servo wedges of a current track by initializing a wedge runout value for each servo wedge and an average uncorrected runout value and then successively developing an interim wedge runout estimate for each servo wedge by adding a fractional portion .lambda. of a raw position error signal to a fractional portion (1-.lambda.) of the wedge runout value, computing a corrected position error signal by subtracting the interim wedge runout estimate from the raw position error signal, and computing a new wedge runout value by subtracting the average uncorrected runout value from the interim wedge runout estimate and, subsequent to each revolution, updating the average uncorrected runout value.

Repeatable runout cancellation in sectored servo disk drive positioning

The invention is a method of calibrating a position error signal (PES) to overcome the problem of a measured displacement or PES that varies as a non-linear function of real displacement when the read transducer is positioned to either side of a null position where the PES equals zero. The PES is notably nonlinear while reading servo burst information with a Magneto-Resistive (MR) head which has nonlinear magnetic characteristics. The method uniquely calibrates the measured PES to provide a corrected PES without requiring any knowledge of the actual displacement from the position where the measured PES equals 0. The calibration according to this method, therefore, may be performed at anytime without need for real displacement information from a servowriter or any sort of special calibration track. The method involves positioning the MR head to a plurality of predetermined, measured displacements corresponding to partial track increments of real displacement of unknown magnitude on either side of the null position, measuring a PES gain value at each of the plurality of measured displacements, and then correcting subsequent measured displacements based on the plurality of measured displacements and corresponding PES gain values. The correcting step is preferably accomplished by using the measured displacements and corresponding PES gain values to establish coefficients of a correction function c(y) which substantially provides unity gain relative to real displacement over a limited range of measured displacements. The preferred correction function c(y) is a partial sum Fourier series.

Growing community of inventors

Redwood City, CA, United States of America

Thao P Trieu