Browse Prior Art Database

Adaptive Base Casting Dynamics Estimator with Compensation for Base Casting Motion in Direct-Access Storage Devices

IP.com Disclosure Number: IPCOM000117931D
Original Publication Date: 1996-Jul-01
Included in the Prior Art Database: 2005-Mar-31
Document File: 8 page(s) / 262K

Publishing Venue

IBM

Related People

Chen, JC: AUTHOR [+4]

Abstract

In many disk files, digital head positioning systems are used both to move the read/write elements from track to track and to keep the elements close to the track center once they arrive on track. Typically, the plant model used in the digital servo includes only the dynamics of the power amplifier and Voice Coil Motor (VCM) (1 and 2). In the past, this was adequate because the force-to-mass ratios were sufficiently small. Fig. 1 illustrates that peak acceleration of the Head/Disk Assembly (HDA) base casting has increased an order of magnitude, with a concomitant increase in base casting motion.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 36% of the total text.

Adaptive Base Casting Dynamics Estimator with Compensation for Base
Casting Motion in Direct-Access Storage Devices

      In many disk files, digital head positioning systems are used
both to move the read/write elements from track to track and to keep
the elements close to the track center once they arrive on track.
Typically, the plant model used in the digital servo includes only
the dynamics of the power amplifier and Voice Coil Motor (VCM) (1 and
2).  In the past, this was adequate because the force-to-mass ratios
were sufficiently small.  Fig. 1 illustrates that peak acceleration
of the Head/Disk Assembly (HDA) base casting has increased an order
of magnitude, with a concomitant increase in base casting motion.

      To reduce space and power requirements for the Direct Access
Storage Device (DASD), the HDA mass has decreased.  The number of
tracks per meter has increased, thus reducing the tolerable Track
MisRegistration (TMR) in absolute terms.  Concurrently, there is a
desire to maintain (or reduce) access times.  Since the actuator mass
has not experienced the same decline as the HDA mass, the forces
required to maintain access time have remained relatively constant.
Additionally, control of the actuator dynamics has not advanced
enough that the servo bandwidth can be arbitrarily high enough to
result in an error transfer function that can attenuate the resulting
Position Error Signal (PES) sufficiently.  These factors combined
result in excessive TMR after an access operation for lower mass HDAs
when used with a "typical" digital servo architecture.

      One solution that has been successfully demonstrated is to
attach a sensor (accelerometer) to the base casting and, after signal
conditioning, use the resulting signal to modify the "normal" servo
compensation signal, thus accounting for base casting dynamics, as
shown in Fig. 2.  The disadvantages of this approach as compared to
Adaptive Base Casting Dynamics Estimator (ABCDE) include:
  1.  additional cost for the sensor, A/D converter, signal
       conditioning circuitry, cables, connectors and mounting
       hardware; and
  2.  lower reliability due to both the sensor and the supporting
       hardware.

      Implementation of the accelerometer solution demonstrated that
it was possible to achieve acceptable TMR if the base casting
dynamics were known.  A derivation of the required correction signal
is shown in  Fig. 3.

      The basis for the ABCDE system was the recognition that the
components of the HDA that contribute to base casting motion could be
modeled in the servo state estimator.  State variable feedback is
then applied to minimize the servo head position error in both
"seeking" and "track following" modes.  Also, it was recognized that
the parameters crucial to the success of the ABCDE experienced a wide
variation with time because of various effects.  Adaptive techniques
are used to track the parameter v...