Dynamic magnetic information storage or retrieval – General processing of a digital signal – Head amplifier circuit
Reexamination Certificate
1999-02-03
2002-09-10
Faber, Alan T. (Department: 2651)
Dynamic magnetic information storage or retrieval
General processing of a digital signal
Head amplifier circuit
C360S065000, C360S067000, C360S031000, C360S053000, C360S066000
Reexamination Certificate
active
06449110
ABSTRACT:
FIELD OF INVENTION
The present invention relates to the recording and reproduction of binary data in magnetic disk storage systems for digital computers, particularly to a magnetic disk storage system employing a non-linear transducer (e.g., a magneto-resistive (MR) read head) adjusted to operate in a non-linear region but with higher gain, together with a non-linear correction circuit for attenuating the non-linearity in the read signal.
BACKGROUND OF THE INVENTION
Computer systems typically comprise a disk storage device, for example a magnetic or optical disk drive, which provide an inexpensive means to store large amounts of digital data in a non-volatile manner. The disk storage device is essentially a communication system where the storage medium (magnetic or optical), transducer, and read/write electronics constitute the communication channel. Similar to other communication channels, the digital data in storage devices is “transmitted” through the channel by modulating an analog signal. In magnetic disk storage systems, for example, the digital data modulates the current in an inductive write coil in order to write a sequence of magnetic transitions onto the surface of a magnetic disk in concentric, radially spaced tracks. And in optical disk storage systems, the digital data may modulate the intensity of a laser beam in order to write a series of “pits” onto the surface of an optical disk in tracks that spiral inward toward the center of the disk.
During a read operation, a transducer or read head is positioned in close proximity to the surface of the disk, and while the disk spins under the read head, the read head senses the alterations (magnetic or optical) representing the digital data. The read head generates an analog read signal comprising pulses induced by the surface alterations. In magnetic recording, for example, the read head comprises a sensor that is responsive to the changes in the magnetic flux caused by the magnetic transitions representing the digital data. The two main types of magnetic sensors employed in magnetic storage devices include the conventional inductive coil read head which is sensitive to the change in magnetic flux, and the more recent magneto-resistive (MR) read head comprising a resistive element which is sensitive to the strength or magnitude of the magnetic flux. Both sensors generate an analog read signal comprising pulses induced by the magnetic transitions, but the MR read head exhibits substantially higher sensitivity and noise immunity which is why they are displacing the older inductive coil type read heads.
As with other bandlimitted communication channels, the maximum capacity of a disk storage system is approximated by Shannon's equation for the capacity of an additive white Gaussian noise channel:
C
=
W
log
(
1
+
P
N
0
W
)
.
In the above equation, W is the channel bandwidth, N
0
is the noise power spectrum, and P is the signal power. The bandwidth W of a disk storage system is, for the most part, limited by the characteristics of the storage medium. Thus, once the storage medium is chosen, the maximum capacity of the storage system is essentially a function of the signal power P and the noise power N
0
(i.e., the signal-to-noise ratio or SNR). Certain characteristics of the storage medium also contribute to the noise power in the read signal, so designers generally choose the least expensive medium that will provide the highest bandwidth and SNR to attain maximum storage capacity.
In addition to innovations in the storage medium itself, attempts to increase storage capacity generally focus on improving the actual SNR through improvements to the transducer and drive electronics, as well as improving the effective SNR through the use of error correction codes (ECC), such as the Reed-Solomon ECC codes, and through the use of sophisticated signal processing techniques spawned by communication theory.
One such advancement in communication theory that has recently been applied to disk storage systems to achieve significant gains in storage capacity is partial response (PR) signaling with maximum likelihood (ML) sequence detection. Partial response signaling refers to a particular method for transmitting symbols represented as analog pulses through a communication medium. The benefit is that at the signaling instances (baud rate) there is no intersymbol interference (ISI) from other pulses except for a controlled amount from immediately adjacent, overlapping pulses. Allowing the pulses to overlap in a controlled manner leads to an increase in the symbol rate (linear recording density) without sacrificing performance in terms of SNR. Stated differently, a partial response signal provides an increase in the effective SNR by making more efficient use of the channel bandwidth.
Partial response channels are characterized by the polynomials
(1
−D
)(1
+D
)
n
where D represents a delay of one symbol period and n is an integer. For n=1, 2, 3, the partial response channels are referred to as PR
4
, EPR
4
and EEPR
4
, with their respective frequency responses shown in FIG.
1
A. The channel's dipulse response, the response to an isolated symbol, characterizes the transfer function of the system (the output for a given input). With a binary “1” bit modulating a positive dipulse response and a binary “0” bit modulating a negative dipulse response, the output of the channel is a linear combination of time shifted dipulse responses
y
(
t
)=&Sgr;
a
n
p
(
t−nT
)
where a
n
denotes the write current symbols +1 and −1 at time n and p(t) represents the channel's dipulse response shifted by nT (n symbol periods). The dipulse response for a PR
4
channel (1−D
2
) is shown as a solid line in FIG.
1
B. Notice that at the symbol instances (baud rate), the dipulse response is zero except at times t=0 and t=2. Thus, the linear combination of time shifted PR
4
dipulse responses will result in zero ISI at the symbol instances except where immediately adjacent pulses overlap.
It should be apparent that the linear combination of time shifted PR
4
dipulse responses will result in a channel output of +2, 0, or −2 at the symbol instances (with the dipulse samples normalized to +1, 0, −1) depending on the binary input sequence. The output of the channel can therefore be characterized as a state machine driven by the binary input sequence, and conversely, the input sequence can be estimated or demodulated by running the signal samples at the output of the channel through an “inverse” state machine. Because noise will obfuscate the signal samples, the inverse state machine is actually implemented as a trellis sequence detector which computes a most likely input sequence associated with the signal samples. The algorithm for selecting a most likely sequence through a trellis was invented by a man named Viterbi, and thus the algorithm is commonly referred to as the Viterbi algorithm.
The Viterbi algorithm for a PR
4
trellis sequence detector is understood from its state transition diagram shown in FIG.
2
A. Each state
2
is represented by the last two input symbols (in NRZ after preceding), and each branch from one state to another is labeled with the current input symbol in NRZ
4
and the corresponding sample value
6
it will produce during readback. The demodulation process of the PR
4
sequence detector is understood by representing the state transition diagram of
FIG. 2A
as a trellis diagram shown in FIG.
2
B. The trellis diagram represents a time sequence of sample values and the possible recorded input sequences that could have produced the sample sequence. For each possible input sequence, an error metric is computed relative to a difference between the sequence of expected sample values that would have been generated in a noiseless system and the actual sample values output by the channel. For instance, a Euclidean metric is computed as the accumulated square difference between the expected and actual sample values. The i
Bliss William G.
DeGroat Ronald D.
Cirrus Logic Inc.
Faber Alan T.
Shifrin Esq. Dan A.
LandOfFree
Optimizing operation of a disk storage system by increasing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Optimizing operation of a disk storage system by increasing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Optimizing operation of a disk storage system by increasing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2853528