Dynamic magnetic information storage or retrieval – General processing of a digital signal – Head amplifier circuit
Reexamination Certificate
2000-06-15
2002-12-24
Faber, Alan T. (Department: 2651)
Dynamic magnetic information storage or retrieval
General processing of a digital signal
Head amplifier circuit
C360S053000, C360S065000, C360S077020, C360S078040
Reexamination Certificate
active
06498694
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to read-channel signal processing in media storage systems. More particularly, the present invention relates to a method and system for improved read-channel equalization through integration of servo information.
2. Description of Related Art
Data/media storage systems such as hard disk drives (HDDs), compact disks (CDs and CD-ROMs), digital video disks (DVDs), magneto-optical disks, etc., read and write data using magnetic or optical transfer techniques. In the case of magnetic media the process of reading the data involves measuring the magnetic field generated by the recorded data. In the optical media case the process of reading the data involves measuring the optical reflection properties of the recorded data as the data is lit by a laser beam.
Regardless of the technology used to read the data, once the data is read the measured signal is converted into an electrical signal (read signal). The electrical/read signal is then processed to infer the value of the data that was read (detected symbols). This is referred to as read-channel signal processing.
FIG. 1
illustrates a prior art read channel signal processing procedure. Data is stored in the media storage system as binary data
110
. The data is accessed using either the magnetic or optical media processes. In
FIG. 1
, read-channel
120
represents the modulation process where the magnetic or optical flux is written on the medium and the magnetic or optical signal is measured. The read-channel
120
also represents the conversion of the measured magnetic or optical signal into a filtered and sampled electrical signal referred to herein as the read signal
130
.
Next, the read signal
130
is processed to infer the value of the data that was read i.e., detected symbols
150
. The read-channel processing scheme illustrated in
FIG. 1
for inferring the value of the data that was read to arrive at the detected symbols
150
is referred to as Partial Response Maximum Likelihood (PRML) detection
140
. PRML detection
140
incorporates a linear partial response equalizer
145
followed by a Viterbi detector
146
. Other prior art methods for read-channel processing include Decision Feedback Equalizer (DFE) and Finite Delay Tree Search (FDTS). Such read-channel processes work in similar fashion to the PRML processes and are not discussed in detail herein.
Such read-channel signal processing systems suffer from Inter Symbol Interference (ISI) and noise that makes it difficult for the system to detect each written symbol. Thus, the read-channel processing systems have a corresponding error rate associated with the fact that the detected symbols
150
do not always match the values of the binary data
110
that was written. To decrease the error rate associated with read channel signal processing systems, such systems use models of the read channel to model the dynamic relationship between the actual value of the recorded data and the corresponding measured value of the electrical signal (read signal). The accuracy of these models greatly influences the Bit Error Rate (BER) of the read process. The BER is the ratio of the number of data bits whose inferred value was incorrect over the total number of data bits that were read and processed.
FIG. 2
illustrates a model used in prior art read-channel processing systems. Read channel model (model)
200
typically is a linear model
240
with additive white noise
220
. The concepts and processes of modeling are known in the art and are not discussed in detail herein. Model
200
and the processes of model
200
would be used within the read channel
120
illustrated in FIG.
1
. In
FIG. 2
, model
200
represents the assumption that the read signal
230
is produced by passing the binary data
210
through a linear filter
240
and then adding white noise
220
.
In prior art media storage systems read-channel signal processing is performed separately from servo processing. Consequently, in prior art read-channel signal processing the only source of error considered is the additive white noise discussed above. By separating the read-channel signal processing from servo processing, the read-channel equalization does not take into account the servo error of the read-head (i.e., actuator head) during processing of the read data. The servo error is the error between the actual position of the head and the desired position of the head on the media storage device. As track pitch is reduced to increase density, the tracking (or position) errors become a significant portion of the track pitch. In systems with reduced track pitch, during real-time operation the actual value of the servo error fluctuates widely due to controller design limitations and external disturbances. The position of the actuator head during a read and/or write operation will affect the value of the data read and/or written.
The positioning and motion of the magnetic or optical heads that read and/or write the data are controlled using sophisticated feedforward and feedback control methods (control methods). The main objective of these control methods is to minimize the servo error and improve the data access time of the system. The servo error, as stated earlier, is the error between the actual position of the head and the desired position of the head on the media storage device. The data access time is the amount of time that passes from the moment the command for reading and/or writing the data is issued to the moment that the data is actually read and/or written.
FIG. 3
illustrates an example of servo error, or positioning error on a rotating media storage device, for example a disk drive. It should be noted that servo error occurs on other media storage devices and that the example of a rotating media storage device such as a disk drive is merely meant to be illustrative and not limiting. It should also be noted that the systems described herein exchange the information on the servo error or positioning error as signals and thus the discussions herein may refer to the servo error or positioning error as the servo error signal or positioning error signal (PES), respectively.
Illustrated in
FIG. 3
is an enlarged version of a “track”
310
on a storage disk. In approximately the center of track
310
is a dashed line
320
which represents the data stored on track
310
. Actuator head
330
is illustrated in
FIG. 3
as being located directly above the center of track
310
where the data
320
is stored and is also located on a servo wedge
350
.
A servo wedge is like a “marking” on a disk (usually placed there at the time of manufacture) that delineates position on a disk. For example, a servo wedge in a magnetic recording media has a stronger magnetic field than other regions of the magnetic disk so as to delineate position. In the magnetic disk example, when a system senses the stronger magnetic field (i.e., the actuator head passes over a servo wedge) the position of the actuator head can be measured (or “sampled”) by the system. Based upon the signal detected at the servo wedge the position of the actuator head within the track can also be determined. In other words, the system is able to detect if the actuator head is on the center of the track or is off the center of the track. If the actuator head is off center, the system will also be able to determine which direction from center and by approximately how much the actuator head is off center. When a servo wedge passes under the actuator head, this occurrence is referred to as a “servo burst”.
It should be noted that the only time the position of the actuator head can be measured is at the point of a servo burst, and such measurements are referred to herein as “sampled data”. Based upon the sampled data the system can determine the position error (PES) of the actuator head.
In
FIG. 3
, servo bursts are represented as squares having an “x” thereon. The actuator head during read and/or write does not necessarily remain directly above the stored data
320
but instead fluc
Blakely , Sokoloff, Taylor & Zafman LLP
Faber Alan T.
Voyan Technology
LandOfFree
Servo error integration in read-channel equalization does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Servo error integration in read-channel equalization, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Servo error integration in read-channel equalization will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2989512