Marvell International Limited

Edward Ying

Pantas Sutardja

David Rutherford

Man Cheung

Luan Ton-That

Methods, systems, and apparatus include, in one aspect, a method including receiving from a controller a signal for controlling a device for rotating a machine-readable medium; and increasing a bandwidth of a transfer function corresponding to the controller by at least filtering the signal to compensate for a pole of the transfer function.

Calibrating spindle motor controllers

Methods, systems and computer program products for compensating repeatable timing variations associated with a spindle motor are described. Specifically, a repetitive error correction factor may be determined using a computational model which predicts timing variations. The correction factor can then be used to cancel the effect of the actual timing variations upon the spindle motor.

Repetitive error correction method for disk-drive spindle motor control systems

Methods, systems and computer program products for compensating repeatable run-out (RRO) are described. In some implementations, RRO may be compensated by predicting the RRO and subsequently subtracting the predicted RRO from the actual value of the RRO to cancel out the undesired effect caused by the RRO during operation. More specifically, RRO generated at each spindle motor rotation may be predicted by measuring the position errors associated with the RRO with respect to a current position of a read/write head. A computational model then may be utilized to determine one or more correction factors to compensate for the RRO.

Repetitive error correction method for disk-drive servo control systems

Methods, systems, and apparatus, including computer program products, are described for calibrating spindle motor controllers. In one aspect, a method is provided that includes receiving a compensator output from a spindle motor controller that compensates for variations in a velocity of an associated spindle motor, and filtering the compensator output to compensate for a spindle motor system pole and produce a high frequency roll-off.

Systems and techniques to calibrate a control loop include, in at least one implementation, a system including the control loop configured to generate a clock signal and lock the clock signal to timing marks detected on a machine readable medium, a repetitive error correction module configured to receive a predicted phase and a corrected phase error for the clock signal, generate a predicted repetitive phase disturbance using the predicted phase and the corrected phase error for the clock signal, and calibrate a phase error to compensate for variations in repetitive phase errors in the clock signal using the predicted repetitive phase disturbance; and a servo track generator configured to generate servo tracks using the clock signal.

Calibrating control loops

The present disclosure includes systems and techniques relating to writing ramp-tracks on a computer readable medium. In some implementations, a trajectory is identified across previously written tracks on a surface of a machine-readable medium for a read-write head to move along and build up velocity. Identifying the trajectory includes identifying one of the previously written tracks as a launch location from which to launch the read-write head. Identifying the trajectory also includes identifying a start writing location near a blank region of the surface of the machine-readable medium at which to turn on a write gate. A current is applied to the read-write head to cause the read-write head to move along the trajectory. In addition, the identified launch location is adjusted until a target velocity is obtained at the start writing location.

Method to write ramp-track

Methods, systems, and apparatus, including computer program products, are described for calibrating control loops, specifically phase-locked loops. In one aspect, an apparatus is provided that includes an oscillator model that generates a predicted phase based on an input, a first averaging submodule that generates an average predicted phase over a predetermined number of samples, and a first summing submodule that receives a first corrected phase error and generates a predicted repetitive phase disturbance using the first corrected phase error, the predicted phase, and the average predicted phase.

A phase error reduction system includes a control module, a phase-locked loop (PLL) module, and a harmonic removal module. The control module generates source timestamps for a plurality of synchronization marks in a source signal using a clock and generates a plurality of target tirnestamps. The PLL module determines phase errors between the source timestamps and the target timestamps and minimizes the phase errors. The harmonic removal module removes harmonics of the phase errors using a weighted moving average filter (MAF). The harmonic removal module comprises a repetitive feed forward (RFF) module that includes an amplifier the scales the phase errors, a delay buffer that generates RFF commands to reduce the phase errors, and a summing module. The MAF filters the RFF commands. The summing module provides sums of the phase errors scaled by the amplifier and the RFF commands filtered by the weighted MAF to the delay buffer.

Growing community of inventors

Berkeley, CA, United States of America

Edward Ying