Browse Prior Art Database

Algorithm for Track Capture and Following on Disk Files

IP.com Disclosure Number: IPCOM000088326D
Original Publication Date: 1977-May-01
Included in the Prior Art Database: 2005-Mar-04
Document File: 3 page(s) / 36K

Publishing Venue

IBM

Related People

Craft, DJ: AUTHOR

Abstract

A track capture and following algorithm suitable for use on files which do not have available a continuous position error signal is described. The tracks on such files are normally divided into data sectors separated by short lengths of a special type of coding called a `servo burst'. This coding is written at the time the disk is manufactured, and interlocks are normally provided to prevent a write operation on the file from ever erasing any part of the servo bursts. When a data head mounted in this case on a rotary arm actuator passes over a servo burst, the signal from the data head can, when passed through suitable circuitry, generate a position error signal which is proportional to the distance the data head is offset from the center of the data track at that point in time.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 45% of the total text.

Page 1 of 3

Algorithm for Track Capture and Following on Disk Files

A track capture and following algorithm suitable for use on files which do not have available a continuous position error signal is described. The tracks on such files are normally divided into data sectors separated by short lengths of a special type of coding called a `servo burst'. This coding is written at the time the disk is manufactured, and interlocks are normally provided to prevent a write operation on the file from ever erasing any part of the servo bursts. When a data head mounted in this case on a rotary arm actuator passes over a servo burst, the signal from the data head can, when passed through suitable circuitry, generate a position error signal which is proportional to the distance the data head is offset from the center of the data track at that point in time. Conventional practice is to smooth out the sampled position error signal by means of a sample- and-hold circuit followed by a filter, the output then being fed to a servo amplifier which drives the actuator coil with current, thus generating a torque on the actuator.

The algorithm proposed is radically different in that as each position error becomes available it is used to calculate an actuator current which will, if applied and held constant over the next sector, result in the optimum trajectory for the actuator during that sector. The actuator is assumed to obey simple Newtonian mechanics with friction being neglected. The basic equations for the algorithm are as follows:

(Image Omitted)

where /./theta n-(1) is the previously-sampled rate of change of position error signal, /./Theta n is the currently-sampled rate of change of position error signal, theta n is the currently-sampled position error signal, In-(1) is the old value of actuator current drive active for the last t, In is the new value to be determined and held constant for the next t, K is the actuator torque-constant, j is the actuator inertia, and t is the inter-sector time. On sampling Theta n;theta n equation (i) is used first to calculate the external equivalent actuator current I(0) active for the previous period t. This value is then used as a correction factor in equation (ii) to produce In which is applied for the next period t. The operation of the algorithm is best understood by first considering the track capture situation in the absence of any external disturbing torque. Under these circumstances the equation for I(0) can for the moment be disregarded and I(0) made equal to zero in the equations for I(1) and I(2).

Consider the situation where the first servo burst in the figure has just passed under the data head. At this time the position error signal is available directly derived from the servo burst, and the rate of change of the position error signal is also available derived by differentiating the position error signal. For simplicity, both the position error signal and its first derivative are expressed in radians and ...