Zero forcing adaptive equalization in a disk drive read channel

Details Zero forcing adaptive equalization in a disk drive read channel Zero forcing adaptive equalization in a disk drive read channel

1999-07-12

2002-04-30

Hudspeth, David (Department: 2651)

Dynamic magnetic information storage or retrieval

General recording or reproducing

Specifics of equalizing

C375S350000

Reexamination Certificate

active

06381085

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention is related to the field of disk drive systems, and in particular, to disk drive systems and circuitry that use a zero forcing algorithm to produce the coefficients for the adaptive filter in the read channel.
2. Statement of the Problem
FIG. 1
depicts a conventional system that includes a host computer system
100
and a disk drive system
102
. The disk drive system
102
includes control circuitry
104
and disk device
106
. The disk device
106
stores data for the computer system
100
. To transfer this data from the disk device
106
to the computer system
100
, the disk device
106
transfers a signal
120
to the control circuitry
104
. The signal
120
is an analog representation of the data. The control circuitry
104
converts the signal
120
into a signal
126
for the computer system
100
. The signal
126
is a digital representation of the data and is suitable for processing by the computer system
100
. Thus, the control circuitry
104
converts an analog representation of the data into a digital representation of the data.
Those skilled in the art will appreciate that numerous conventional components of the disk drive system
102
are not depicted on
FIG. 1
for the purpose of clarity. For example, the disk device
106
typically includes disks on which data is written, heads to write/read the data to/from the disks, and motors that position heads and rotate the disks. The control circuitry
104
typically includes a controller, servo circuitry, and a read channel. The controller manages data transfers. The servo circuitry controls the motors to position the heads and rotate the disks. The read channel converts the analog signals from the disks into usable digital data. The read channel includes an adaptive filter
108
, a Least Mean Square (LMS) circuit
110
, and an adder
112
that are shown on FIG.
1
.
The adaptive filter
108
is a digital Finite Impulse Response (FIR) filter that receives an input signal
121
. The input signal
121
is a sampled version of the signal
120
. The adaptive filter
108
processes the samples in the input signal
121
to generate the output signal
122
. In particular, the adaptive filter
108
digitally alters pulses in the input signal
121
into a shape that is more suitable for processing by subsequent detector circuitry (not shown). The adaptive filter
108
continually improves its performance by adjusting internal coefficients in response to a coefficient signal
125
. It should be appreciated that improving the performance of the adaptive filter
108
will reduce data errors in the signal
126
.
The adder
112
receives a copy of the output signal
122
and an ideal signal
123
. The ideal signal
123
can be generated in numerous ways, such as using a slicer on the output signal
122
or by using a digital copy of the data stored on the disk device
106
. The adder
112
subtracts the output signal
122
from the ideal signal
123
to generate the error signal
124
. The adder
112
provides the error signal
124
to the LMS circuit
110
.
The LMS circuit
110
receives input signal
121
and the error signal
124
. The LMS circuit
110
applies an LMS algorithm to produce the coefficient signal
125
that alters the coefficients in the adaptive filter
108
. The LMS algorithm is:
C
K+1
=C
K
+&mgr;e
K
X
K
where:
C
K+1
=the new coefficient signal
125
C
K
=the old coefficient signal
125
&mgr;=the step size
e
K
=the error signal
124
=i
K
(ideal signal
123
)−y
K
(output signal i
22
); and
X
K
=the input signal
121
.
The upper case variables represent vectors that are comprised of scalar values that are represented by lower case variables. For a ten tap filter, the term e
K
X
K
can be represented by the following values: [e
K
x
K
, e
K
x
K−1
, e
K
x
K−2
, e
K
x
K−3
, e
K
x
K−4
, e
K
x
K−5
, e
K
x
K−6
, e
K
x
K−7
, e
K
x
K−8
, e
Kx
x
−9
]. A more economically efficient implementation replaces the term e
K
X
K
in the LMS algorithm with x
K
E
K
. For a ten tap filter, the term x
K
E
K
can be represented by the following values: [x
K
e
K
, x
K
e
K+1
, x
K
e
K+2
, x
K
e
K+3
, x
K
e
K+4
, x
K
e
K+5
, x
K
e
K+6
, x
K
e
K+7
, x
K
e
K+8
, x
K
e
K+9
].
Thus, the LMS circuit
110
improves the bit error rate performance of the disk drive system
102
by adjusting the coefficients in the adaptive filter
108
. Unfortunately, the bit error rate performance of the conventional disk drive system
102
suffers because the adaptive filter coefficients do not converge to a solution for optimum bit error rate performance. The convergence problem is derived from the fact that LMS circuit
110
adjusts the coefficients using a Mean Squared Error (MSE) driven process. Although MSE is a convenient metric that correlates with bit error rate, the correlation is not perfect. Thus, the convergence problem in the conventional disk drive
102
permits additional data errors to remain that prevent or slow the operation of the computer system
100
. The additional data errors also require more expensive disk drive components to compensate for the errors.
Given the enormous growth in the demand for higher capacity computer data storage, there is an acute need to continually improve the performance of disk drive systems. In particular, solutions are needed to reduce the problem of data errors in disk drive systems. These solutions will allow less expensive components to be used while maintaining or improving current error rates. The cost savings can be passed on to the consumer in the form of less expensive computer data storage.
SUMMARY OF THE SOLUTION
The invention solves the above problem by using a zero forcing algorithm to adjust the coefficients in the adaptive filter. Testing has demonstrated that systems using the zero forcing algorithm have better bit error rate performance than conventional systems using the LMS algorithm. Thus, the invention allows the read channel adaptive filter to converge to a solution closer to the minimum bit error rate than does LMS circuitry using an MSE driven process. Consequently, the problem of data errors in disk drive systems is reduced, so less expensive disk drive components may be used while maintaining or improving current bit error rates.
The invention includes disk drive circuitry, systems, and methods. The disk drive system comprises control circuitry and a disk device. The disk device stores data and transfers an analog signal representing the data. The control circuitry receives the analog signal, converts the analog signal into a digital signal, and transfers the digital signal. The control circuitry includes zero forcing circuitry and an adaptive filter. The zero forcing circuitry produces new coefficients for the adaptive filter.
In some examples of the invention, the control circuitry includes an analog-to-digital converter, adaptive filter, detector, decoder, and both zero forcing circuitry and LMS circuitry. The analog-to-digital converter receives and samples the analog signal to generate a sampled signal. The adaptive filter shapes the sampled signal based on coefficients to produce an equalized signal. The detector detects binary data from the equalized signal, and the decoder decodes the binary data to generate the digital signal. Either the zero forcing circuitry or the LMS circuitry may be selected to produce the coefficient signal that adjusts the coefficients in the adaptive filter.


REFERENCES:
patent: 4866647 (1989-09-01), Farrow
patent: 5313472 (1994-05-01), Hara
patent: 5467232 (1995-11-01), Ouchi et al.
patent: 5487085 (1996-01-01), Wong-Lam et al.
patent: 5563819 (1996-10-01), Nelson
patent: 5696639 (1997-12-01), Spurbeck et al.
patent: 5717619 (1998-02-01), Spurbeck et al.
patent: 5726818 (1998-03-01), Reed et al.
patent: 5744993 (1998-04-01), Sonntag
patent: 5754353 (1998-05-01), Behrens et al.
patent: 5760984 (1

Affiliated with

Also associated with

Rating

Zero forcing adaptive equalization in a disk drive read channel does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Zero forcing adaptive equalization in a disk drive read channel, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Zero forcing adaptive equalization in a disk drive read channel will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2895414