Cirrus Logic Inc.

Seagate Technology Incorporated

Christopher Thomas Settje

Paul M Romano

Louis Supino

David L Schell

Timothy Francis Ellis

Reed David Hanson

Michael Edward Baum

Rodney Brittner

Khalil B Dizaji

Steven Alan Koldewyn

The present invention reduces the time required for coefficient convergence and thus improves time-to-ready. To that end, the present invention provides an initial learn&#x2014;preferably non-zero&#x2014;coefficient. The initial learn coefficient is based on a translation of timing information to position information. One aspect of the present invention uses address marks of a disc then translates them to phase position information. The present invention measures the times between address marks of adjacent servo sectors. Variations between those times is related is related to disc runout, particularly disc eccentricity. Also, the servo sectors are used to identify at which phase the maximum amplitude (or magnitude) of the timing variations occur. This phase information and maximum amplitude are used as variables in determining the initial learn coefficient. The present invention also utilizes a plant model to predetermine a variable. This variable is used in conjunction with the phase and maximum amplitude variables to determine the initial learn coefficient. Other aspects of the present invention use approximations or statistical values to determine the initial learn coefficient. For example, the maximum amplitude can be based on a clearance between an inner diameter of a disc and a spindle. The phase can be based on a value within a range of phases. The plant model can be changed to include assumptions, approximations or other factors.

Initial learn of adaptive feedforward coefficients

Using a 'reference' sensor, a reference position is detected in a frame of reference that is moving along a predetermined path relative to the reference sensor. The reference position is used to derive a position-dependent indicator of a position-dependent 'longitudinal' offset component (i.e. along the path). An output from the target sensor is monitored using a &#x201c;target window&#x201d; that is partially based on the position-dependent indicator.

Monitoring a sensor's output using a target window that is based on a longitudinal offset's position-dependent indicator

A discrete-time sliding mode controller (SMC) controls the motion of a read head actuated over a disk storage medium during tracking operations. The discret-time SMC comprises a linear signal generator for generating a linear control signal, and a sliding mode generator for generating a sliding mode signal. These signals are combined and applied to a voice coil motor (VCM) for positioning the read head over a particular data track recorded on the disk. The linear signal generator and the gain (c&Ggr;) &Dgr; are designed such that once the sliding mode variable &sgr; crosses the sliding line the first time, it will cross the sliding line in every successive sample period resulting in a zigzag motion about the sliding line wherein &sgr; changes sign at every sample period. The switching action and resulting chatter of the sliding mode controller are minimized by adjusting the width of the boundary layer through appropriate selection of predetermined constants &Dgr; and &lgr;. The boundary layer can be made arbitrarily small so as to better approximate and thereby attain the benefits of continuous time SMC. In addition, the required control effort is minimized while still providing adequate immunity to parameter variations and external load disturbances.

Disk storage system employing a discrete-time sliding mode controller for servo control

A discrete-time sliding mode controller (SMC) controls the motion of mechanical apparatus such as a read head in a disk storage system. The overall control effort is generated by combining a linear control effort with a discrete-time sliding mode control effort generated by switching between gains in order to drive the system's phase states toward a sliding line trajectory. An estimate of a reference is generated and added to the control effort, thereby providing an approximation of the derivative of the reference signal. Estimating the reference signal may be performed by various methods, depending on whether the reference is known, unknown but repeatable, or completely unknown and non-repeatable. A least-mean-square (LMS) algorithm is employed to estimate the reference signal by computing coefficients of a function which minimizes a particular system parameter. If the reference is an unknown, repeatable sinusoid, such as the runout signal in a disk storage device, the coefficients of a synthesized sinusoid may be determined by minimizing the position error phase state. Alternatively, the coefficients of the synthesized sinusoid may be determined by minimizing the sliding mode variable &sgr;,which defines the position of the system phase states relative to the sliding line. For references that are unknown and non-repeatable, an estimate of the derivative of the reference signal is computed, weighted, and added to weighted derivatives of the estimate, where the weights or coefficients of this function are also determined by minimizing the sliding mode variable &sgr;,.

Reference estimator in a discrete-time sliding mode controller

Adaptive discrete-time sliding mode controller

Growing community of inventors

Westminster, CO, United States of America

Christopher Thomas Settje