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

Selection Algorithm for Head Takeoff and Landing Zones

IP.com Disclosure Number: IPCOM000115907D
Original Publication Date: 1995-Jul-01
Included in the Prior Art Database: 2005-Mar-30
Document File: 2 page(s) / 66K

Publishing Venue

IBM

Related People

Gillis, DR: AUTHOR [+2]

Abstract

This invention is an algorithm that combines component performance data with the unique customer usage pattern and cumulative file history to select the optimum location to land the heads during shutdown of the file. The selected landing position has the minimum inferred failure rate for stiction and Head/Disk Impact (HDI) due to wearout. Implementation will result in greatly enhanced Start/Stop (S/S) life, and reduced rest stiction failure rates.

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

Selection Algorithm for Head Takeoff and Landing Zones

      This invention is an algorithm that combines component
performance data with the unique customer usage pattern and
cumulative file history to select the optimum location to land the
heads during shutdown of the file.  The selected landing position has
the minimum inferred failure rate for stiction and Head/Disk Impact
(HDI) due to wearout.  Implementation will result in greatly enhanced
Start/Stop (S/S) life, and reduced rest stiction failure rates.

      This invention is an adaptive landing zone selection algorithm
to be implemented in DASD microcode resident in the HDA interface
electronics.

      The implementation of a basic landing zone  selection algorithm
requires a rudimentary microprocessor, some SRAM, and a clock.  The
addition of stiction measurement circuitry (i.e., a motor current
peak monitor) allows adaptive selection, in addition to a possible
"service request" function that will detect impending stiction
failures.

      Inputs required are one bit signaling a start, one bit
signaling a stop, and a digital input corresponding to the level of
stiction if the adaptive feature is selected.  The outputs are the
track position of the next landing cycle (and the SRQ bit if the
monitor feature is selected).

      The algorithm is designed to optimize the trade between wear
and rest stiction-- the lowest-wear CSS track is the most at-risk for
high rest stiction and vice-versa.  First a simulated lube vs track
distribution is then corrected for the number of cumulative  S/S
cycles at each track position  multiplied by the depletion rate.
Then t...