Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Digital Correction for Disk File Errors

IP.com Disclosure Number: IPCOM000099369D
Original Publication Date: 1990-Jan-01
Included in the Prior Art Database: 2005-Mar-14
Document File: 2 page(s) / 84K

Publishing Venue

IBM

Related People

Elliott, PJ: AUTHOR [+2]

Abstract

Digital correction in a disk file position error signal (PES) is described in U.S. Patent 4,412,156. The present article extends the above correction method by using a self-clocking code on the servo surface to record the corrections to enable reading during settle. In addition, odd and even track corrections are separately recorded and code is included for rarely occurring uncorrectable samples.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Digital Correction for Disk File Errors

       Digital correction in a disk file position error signal
(PES) is described in U.S. Patent 4,412,156.  The present article
extends the above correction method by using a self-clocking code on
the servo surface to record the corrections to enable reading during
settle.  In addition, odd and even track corrections are separately
recorded and code is included for rarely occurring uncorrectable
samples.

      After a servotrack has been written, the servo information
(PES) does not always appear at the correct position when it is read
back.  This is due to two main reasons:

      1. The head and disk are vibrating while the track is being
written.

      2. The disk is not perfectly uniform magnetically, and small
imperfections cause apparent shifts in the servo track center.

      This means that data tracks deviate from their true center
lines in a pseudo-random way as they progress around the revolution.
This adversely affects the WWTMR (Write-to-Write Track
Mis-Registration) of the file.  If the PES could be improved, the
WWTMR of the file would be reduced, and hence the track density could
be increased.

      This can be achieved as described in US Patent 4,412,165 in
which a digital field is written ahead of the servo field to indicate
by how much the servo information is in error. The present article
extends the idea to a file with a two-track-wide dedicated servo
head, using a self-clocking code to encode the digital correction.

      In a particular file the requirements are a resolution of 1/128
track and a range of approx  1/8 track.  This means that 5 digital
bits are needed.  These are written in a 6-bit field which includes a
parity bit.  The code used includes additional error detection to
eliminate the possiblity of misreading it (*).

      Five bits can encode values from 힟 to -16/128. The code
-16/128 is used to denote an uncorrectable sample, i.e., on...