Abstract: In a disk drive having a dual stage actuator, a method to improve head switching time by aligning a target head with a target track before a head switch occurs by using a second stage actuator attached to the target head to align the target head with the new track. The radial offset required to align the target head with the target track is either calibrated at startup or is measured dynamically during normal disk operations. A dual stage actuator with either a milli-actuator second stage or a micro-actuator second stage, positions a read/write head over a disk surface. The disk drive has a radial offset table for storage of radial offsets for head switches.

Abstract: A method to adaptively reduce error introduced by the coil current rise time. A disk drive with a moving arm actuator driven by a voice coil motor is controlled by periodically modeling the coil current rise time. An adjustment to a commanded current value is based on the model of the coil current rise time and a magnitude change in commanded value. The model of the coil current is based on a linear approximation to the exponential rise time of the coil current. The commanded value is adjusted to compensate for the error in the control effort. For small commanded value differences no compensation is made for the coil current rise time. For medium command value differences a slope intercept equation is used to compute the compensated commanded value. For large commanded value differences half of an estimate of the maximum largest physical current change is used. The adjustment is made by the estimator in a position error signal feedback control system.

Abstract: An initialization for runout cancellation for a hard disk drive such as a Small Computer System Interface (SCSI) hard disk drive. The hard disk drive at a servo sample rate updates a set of state variables. At an index wedge the state variables are stored to a memory. Upon a command to perform a head switch the hard disk reads the previously stored state variables from the memory. The previously stored state variables are used to adjust the repeatable runout based on the location of a head relative to an index wedge.