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

Charge-Constrained Even Run-Length Code

IP.com Disclosure Number: IPCOM000111786D
Original Publication Date: 1994-Mar-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 2 page(s) / 69K

Publishing Venue

IBM

Related People

Ashley, J: AUTHOR

Abstract

Even run-length constraint codes [1] have recently been applied to a direct overwrite method of magneto-optic recording [3, 4]. When using this direct overwrite method of recording, the time between two successive magnetic transitions (each transition representing the symbol 1) must be an odd multiple of half the period of a sinusoidally oscillating magnetic field. This imposes the constraint that an even number of 0's (each 0 represented by the absence of a magnetic transition) must be written between successive 1's.

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

Charge-Constrained Even Run-Length Code

      Even run-length constraint codes [1]  have recently been
applied to a direct overwrite method of magneto-optic recording [3,
4].  When using this direct overwrite method of recording, the time
between two successive magnetic transitions (each transition
representing the symbol 1) must be an odd multiple of half the period
of a sinusoidally oscillating magnetic field.  This imposes the
constraint that an even number of 0's (each 0 represented by the
absence of a magnetic transition) must be written between successive
1's.

      When data is read from the disk, the current through the read
head inductive coil alternates in sign between successive 1's, and a
1 is detected when this current crosses a certain detection
threshold.  The readback signal is dc-coupled: in effect, a low
frequency component is added to the readback signal in order that a
long term average direct current of 0 is maintained.  As a
consequence, the 0-crossing threshold wanders from its optimal
position half-way between the current values before and after a
transition.  The slope of the signal at 0-crossings can be
significantly less than the maximum slope of the transitions.  Thus,
if the detector determines the occurrence of a 1 by using the value 0
as a fixed detection threshold, it will be more sensitive to noise
than a detector that somehow compensates for this effect.

      An alternative approach is to keep the simple 0-crossing
detector, but use a code having a spectral null at DC as well as
having even 0 run lengths.  The amplitude of the low frequency
component of the readback signal will be much lower, thereby easing
or obviating the task of adjusting the detection threshold.

      Our code achieves a spectral null at DC by imposing upper and
lower bounds on the amount of charge accumulated by the direct
current component of the readback signal [2].

      Disclosed herein is...