摘要:
A multirate sampled closed-loop digital servo system controls read/write head positioning in a magnetic disc storage device. Using a multirate estimator, positional data embedded in the storage medium is sampled once per sector, and inter-sector positional information is predicted. At least two control signals are provided per sample. The control signals are outputted at a multiple of the measurement sample rate, e.g., twice the sample rate. The estimator estimates head position and velocity, system and DC bias, and accounts for estimator calculation time delay, in the system and in the control updates. Estimated states are based on the measurement, control value, last control value, and last estimated states. Input information used by the estimator includes the digitized sampled position error signal (PES), the Gray code value, the present digital control value, and the last digital control value. At every sample, the present invention calculates the control values based upon estimated read/write head position and velocity, and the absolute position of the target track. Both present and last control values are used for the estimator to compensate for computational delay and for additional system phase lags. To minimize computational delay, the algorithm implementing the estimator and controller requires but one multiplication and one addition are necessary to calculate the control value. Once the heads are moved close to the desired target tracks, the servo operates closed-loop in track-follow mode.
摘要:
A method of producing a position error signal includes storing a normal position error value generated from a normal servo field and a quadrature position error value generated from a quadrature servo field, where the normal servo field is ninety degrees out of phase from the quadrature servo field. A position error numerator is created by arithmetically combining the normal position error value and the quadrature position error value. A position error denominator is created based on the normal position error value and the quadrature position error value. The position error signal is produced by dividing the position error numerator by the position error denominator.