Disk drive adapting equalizer relative to bit error rate of sequence detector
View Patent ↗A disk drive is disclosed comprising a disk and a head actuated radially over the disk, wherein the head generates a read signal. A sampling device samples the read signal to generate a sequence of read signal samples, and an equalizer comprising a plurality of coefficients, equalizes the read signal samples to generate a sequence of equalized samples. A sequence detector detects an estimated data sequence from the equalized samples, wherein the sequence detector operates according to a target response comprising a plurality of target values. Control circuitry adapts the equalizer coefficients by computing error values in response to a difference between expected samples and the equalized samples, computing a gradient in response to a correlation of the read signal samples with the error values, and adjusting at least one of the equalizer coefficients in response to the gradient.
1. A disk drive comprising:
a disk;
a head actuated radially over the disk, the head operable to generate a read signal;
a sampling device operable to sample the read signal to generate a sequence of read signal samples;
an equalizer comprising a plurality of coefficients, the equalizer operable to equalize the read signal samples to generate a sequence of equalized samples;
a sequence detector for detecting an estimated data sequence from the equalized samples, wherein the sequence detector operates according to a target response comprising a plurality of target values; and
control circuitry operable to adapt the equalizer coefficients by:
computing error values in response to a difference between expected samples and the equalized samples;
computing a gradient in response to a correlation of the read signal samples with the error values; and
adjusting at least one of the equalizer coefficients in response to the gradient.
2. The disk drive as recited in claim 1 , wherein the control circuitry is operable to compute the gradient by scaling the correlation by at least one of the target values of the target response.
3. The disk drive as recited in claim 1 , wherein the control circuitry is operable to compute the gradient in response to an error event of the sequence detector.
4. The disk drive as recited in claim 3 , wherein the error event comprises a ‘+’ error event.
5. The disk drive as recited in claim 4 , wherein the control circuitry is operable to compute the gradient according to:
Δ
wm
=
∑
i
=
0
2
∑
j
=
0
2
g
i
g
j
(
e
k
-
i
x
k
-
m
-
j
+
e
k
-
j
x
k
-
m
-
i
)
where:
the equalizer comprises three coefficients w 0 -w 2 ;
the target response comprises three target values g 0 -g 2 ;
Δ wm is a gradient for the w m coefficient;
e k are the error values; and
x is a read signal sample.
6. The disk drive as recited in claim 4 , wherein the control circuitry is operable to compute the gradient according to:
Δ
wm
=
∑
i
=
0
2
∑
j
=
0
2
g
i
g
j
(
sign
(
e
k
-
i
)
x
k
-
m
-
j
+
sign
(
e
k
-
j
)
x
k
-
m
-
i
)
where:
the equalizer comprises three coefficients w 0 -w 2 ;
the target response comprises three target values g 0 -g 2 ;
Δ wm is a gradient for the w m coefficient;
e k are the error values; and
x is a read signal sample.
7. The disk drive as recited in claim 3 , wherein the control circuitry is operable to compute the gradient in response to the error event convolved with the target values of the target response.
8. The disk drive as recited in claim 7 , wherein the control circuitry is operable to compute the gradient by minimizing:
e y iT R z e y i
where:
e y i represents the error event convolved with the target values of the target response; and
R z is a noise correlation matrix.
9. A method of operating a disk drive, the disk drive comprising a disk, a head actuated radially over the disk, the head operable to generate a read signal, a sampling device operable to sample the read signal to generate a sequence of read signal samples, an equalizer comprising a plurality of coefficients, the equalizer operable to equalize the read signal samples to generate a sequence of equalized samples, and a sequence detector for detecting an estimated data sequence from the equalized samples, wherein the sequence detector operates according to a target response comprising a plurality of target values, the method comprising:
computing error values in response to a difference between expected samples and the equalized samples;
computing a gradient in response to a correlation of the read signal samples with the error values; and
adjusting at least one of the equalizer coefficients in response to the gradient.
10. The method as recited in claim 9 , wherein computing the gradient comprises scaling the correlation by at least one of the target values of the target response.
11. The method as recited in claim 9 , wherein the gradient is computed in response to an error event of the sequence detector.
12. The method as recited in claim 11 , wherein the error event comprises a ‘+’ error event.
13. The method as recited in claim 12 , wherein computing the gradient comprises minimizing:
e y iT R z e y i
where:
e y i represents the error event convolved with the target values of the target response; and
R z is a noise correlation matrix.
14. The method as recited in claim 12 , wherein the gradient is computed according to:
Δ
wm
=
∑
i
=
0
2
∑
j
=
0
2
g
i
g
j
(
e
k
-
i
x
k
-
m
-
j
+
e
k
-
j
x
k
-
m
-
i
)
where:
the equalizer comprises three coefficients w 0 -w 2 ;
the target response comprises three target values g 0 -g 2 ;
Δ wm is a gradient for the w m coefficient;
e k are the error values; and
x is a read signal sample.
15. The method as recited in claim 11 , wherein the gradient is computed in response to the error event convolved with the target values of the target response.
16. The method as recited in claim 12 , wherein the gradient is computed according to:
Δ
wm
=
∑
i
=
0
2
∑
j
=
0
2
g
i
g
j
(
sign
(
e
k
-
i
)
x
k
-
m
-
j
+
sign
(
e
k
-
j
)
x
k
-
m
-
i
)
where:
the equalizer comprises three coefficients w 0 -w 2 ;
the target response comprises three target values g 0 -g 2 ;
Δ wm is a gradient for the w m coefficient;
e k are the error values; and
x is a read signal sample.