JP2001148168A - Disk recording and playback device - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To effectively suppress crosstalk noise between adjacent tracks without significantly changing the current system.
As a result, it is an object of the present invention to provide a disk recording / reproducing apparatus capable of realizing a high track density. An encoder included in a data channel of an HDD is disclosed. The encoder 11 suppresses the occurrence frequency of at least one or more suppression patterns among 2T to 4T, which are highly affected by the amount of crosstalk noise from adjacent tracks, for recording data to be recorded on the disk 1, The suppression pattern is converted into RLL coded data such that the frequency of occurrence of each of the suppression patterns is lower than the frequency of occurrence of each of all of the coded patterns except for that. The data channel 7 supplies the recording data encoded by the encoder 11 to the write head 31 via the write amplifier 61.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is applied to a disk recording / reproducing apparatus such as a hard disk drive, and more particularly to an encoding circuit and a decoding circuit for executing recording and encoding processing.

[0002]

2. Description of the Related Art In recent years, hard disk drives (HD)
In D), in order to increase the storage capacity, technology development for increasing the density of tracks (higher TPI) configured on a disk is being promoted. In order to increase the TPI, it is important to reduce crosstalk noise from adjacent tracks during data reproduction.

As a specific prior art, a proposal has been made to realize a narrow track pitch by controlling a recording current of a head in accordance with a recording signal pattern, thereby making the effective recording track width independent of the recording signal pattern constant. (See, for example, JP-A-11-144205,
Prior art). In addition, there is a proposal to provide a guard band member made of a non-magnetic material between adjacent recording tracks to reduce side fringes of recording magnetic domains, thereby narrowing the spacing (for example, Japanese Patent Application Laid-Open No. 9-97).
No. 419, a second prior art). Further, there is a proposal to reduce crosstalk noise by using two heads having a magnetic film formed obliquely to a recording track and having different azimuth angles (for example, Japanese Patent Application Laid-Open No. 7-320227). , A third prior art). Further, by providing a clock phase difference of 180 ° between the even-numbered and odd-numbered tracks, the magnetization reversal phase is shifted by 180 ° between adjacent tracks, so that the recording is performed.
There is also a proposal to reduce crosstalk noise between adjacent tracks (see, for example, JP-A-6-131614, which is a fourth prior art).

[0004]

As described above, high T
In order to achieve PI, it is necessary to reduce crosstalk noise from adjacent tracks on the disk. For this reason, the first to fourth prior arts as described above have been proposed. The first prior art utilizes the characteristic of magnetic recording that the effective recording track width becomes narrower as the recording density increases during magnetic recording. However, the data pattern to be actually recorded is a random pattern, and in order to alleviate the nonlinear transition point shift (NLTS) phenomenon that occurs during high-frequency recording, pre-recording compensation (precompensation) is dynamically performed on the random pattern. You need to control. For this reason, it is necessary to increase the speed of the circuit system. Further, since the pre-recording compensation amount and the recording current value are dependent on each other, it is necessary to correct both at appropriate and high speed with respect to the random pattern, and the control for that is complicated.

In the second prior art, a guard band of a non-magnetic layer is physically provided between tracks to reduce crosstalk noise. However, the two types of magnetic layers
In order to further increase TPI to form tracks,
Higher precision processing technology is required, and significant improvement over the current disk media is required. A third prior art utilizes azimuth recording. But,
There are various problems to be solved in actual HDD application, such as the need to maintain the azimuth angle between the head and the disk medium and between adjacent tracks during data recording and reproduction, and practical application is not easy. . The fourth prior art utilizes an adjacent phase difference, but will have a higher TPI in the future.
It is considered that clock control becomes difficult in realizing the realization.

[0006] In short, there are prior arts for reducing the crosstalk noise between adjacent tracks to increase the TPI, but they are all difficult to realize, for example, requiring a significant improvement over the current system. .

An object of the present invention is to provide a disk recording / reproducing apparatus capable of effectively suppressing crosstalk noise between adjacent tracks without largely changing the current system, and consequently realizing a high track density. Is to provide.

[0008]

According to the present invention, when data is recorded on a disk, the amount of crosstalk noise differs due to the difference in the recording signal pattern of each track, and the recording signal pattern is controlled. The present invention relates to a disk recording / reproducing apparatus having means for reducing crosstalk noise. That is, the disk recording / reproducing device of the present invention is connected to a head, and in a data signal processing means for processing each data signal corresponding to recording data to be recorded on the disk recording medium and reproduction data reproduced from the disk recording medium, An encoding unit for encoding the recording data into RLL encoded data, and a decoding unit for decoding the reproduced data are provided. The encoding means suppresses the occurrence frequency of the suppressed RLL encoded pattern in which the reproduction error rate due to the crosstalk noise is degraded, and suppresses the occurrence frequency of the suppressed RLL encoded pattern more than the occurrence frequency of all the patterns except the suppressed RLL encoded pattern. R so that the frequency of occurrence is small
Generate LL encoded data.

Specifically, when the encoding means generates the RLL encoded pattern from the recording data, the encoding means generates a specific encoded pattern (at least a 3T pattern) in which the amount of crosstalk noise between adjacent tracks relatively increases. ) Is generated such that the frequency of occurrence of all the patterns except for the suppressed RLL coded pattern is lower than the frequency of occurrence of all the patterns except the suppressed RLL coded pattern. Here, a specific coding pattern (3
T pattern) is, for example, “1” next to the information bit “1”.
Is included in the lowest frequency coding pattern (11T pattern) which becomes the information bit "1" next to 10 consecutive information bits "0" from the 1T pattern from the highest frequency coding pattern (1T pattern). Is
This is estimated in advance through experiments and the like.

[0010] With this configuration, the configuration of the encoder (decoder at the time of data reproduction) that executes recording and encoding processing when recording data on the disk is changed without changing the entire system of the disk recording / reproducing apparatus. Thus, crosstalk noise between adjacent tracks can be effectively reduced. Therefore, an error rate at the time of data reproduction can be improved with a highly feasible configuration, so that a higher track density of the disk can be improved and a large-capacity disk recording / reproducing apparatus can be provided.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. Here, the occurrence frequency of the suppression pattern whose error rate is degraded due to the crosstalk noise is suppressed, and the RLL encoded data is reduced so that the occurrence frequency of the suppression pattern becomes smaller than the occurrence frequency of all patterns except the suppression pattern. As an example of generation, it is assumed that the suppression pattern is completely deleted.

(System Configuration) FIG. 1 is a block diagram showing a main part of an HDD relating to the embodiment.

HDDs are roughly classified as shown in FIG. 1, a disk 1 as a recording medium, and a spindle motor (SP).
M) 2, a head assembly, a data signal processing system,
It has a disk controller (HDC) 8 and a servo system 9.

The head assembly includes a head slider 3 on which a read head 30 and a write head 31 are mounted,
An actuator 4 holding the slider 3 and a voice coil motor (V
CM) 5. The read head 30 is mounted on the disk 1
Is a read-only head for reading data from, and usually comprises an MR head (or GMR head). The write head 31 is an inductive head for magnetically recording data on each track on the disk 1 according to a recording current pattern supplied from a write amplifier 61 described later. The VCM 5 drives the actuator 4 in the radial direction of the disk 1 and seeks the slider 3 under the control of the servo system 9.

The HDC 8 has a function of constituting an interface between the HDD and the host system, receiving recording data (WD) from the host system, and transferring reproduction data (RD) to the host system. The servo system 9 includes a CPU constituting a main control device of the HDD, and mainly executes rotation control of the disk 1 and head positioning control.

The data signal processing system is roughly divided into a preamplifier circuit 6 and a data channel (read / write channel).
And 7. The preamplifier circuit 6 has a read amplifier 60 and a write amplifier 61. Read amplifier 6
0 amplifies the data signal read by the read head 30 and sends it to the data channel 7. The write amplifier 61 converts the recording encoded data supplied from the data channel 7 into a recording current pattern, and sends the recording current pattern to the write head 31.

The data channel 7 comprises a read channel system and a write channel system, and a PRML (partial response maximum likelihood) for processing recording and reproduction data signals.
hood) type signal processing circuit. Read channel system is A
GC amplifier 20 and low-pass filter (LPF) 21
, A digital PRML processing circuit 22, a decoder (decoding circuit) 23, and a descrambler (descrambler).
ler) 24. The AGC amplifier 20
(Automatic gain control) function to keep the level of the input data signal constant. The LPF 21 is a filter for removing noise.

The digital PRML processing circuit 22 has an A /
It is a digital signal processing circuit including a D converter, a digital waveform equalization circuit (digital filter), a Viterbi decoder, and the like. Decoder 23
Is a decoding circuit paired with the encoder 11 related to the embodiment, and decodes the encoded data encoded by the encoder 11 into the original recording data. The descrambler 24 is a light channel scrambler (s
and a function of returning the recording data randomized by the crawler 10 to the original recording data. Note that the arrangement relationship between the descrambler 24 and the decoder 23 may be reversed.

The light channel system includes a scrambler 10,
Encoder 11, precoder 1
2, and pre-recording compensation
n) A circuit (write pre-con) 13 is provided. The scrambler 10 is a circuit for randomizing the recording data WD from the HDC 8, and usually comprises an exclusive OR circuit.
The encoder 11 is a recording encoding circuit related to the embodiment, and converts recording data (randomized data) into recording encoding data of an encoding pattern capable of suppressing crosstalk noise between adjacent tracks, as described later. Convert.
The precoder 12 is a kind of decoder, and has a PR (partial r
esponse) It is a circuit for giving equalization and interference of inverse characteristics. The write pre-con 13 is a circuit that performs a compensation process for suppressing a phase shift at the time of data reproduction during recording.

(Operation of Data Channel) The operation and effect of this embodiment will be described below with reference to FIG. 1 and FIGS. 2 to 4.

The main components related to the embodiment are:
An encoder 11 included in the data channel 7. The operation of the write channel including the encoder 11 will be described with reference to the flowchart of FIG.

When the HDC 8 receives the recording data WD to be stored on the disk 1 from the host system, it sends it out to the data channel 7 (step S1). In the write channel of the data channel 7, the scrambler 10
A randomization process (randomization) is performed so that a biased bit pattern that easily causes a reproduction error does not occur (step S2). As described later, the encoder 11 performs an encoding process on the randomized recording data to generate encoded recording data (step S3). The encoded recording data is an RLL (run length li).
(coded) encoded data. The encoder 11
A block encoding process for converting a block of an arbitrary data sequence into a predetermined code sequence is executed. Further, the write channel sends the RLL encoded data subjected to the recording compensation processing to the write amplifier 61 via the precoder 12 and the write pre-con 13 (step S4). Light amplifier 61
Is converted into a recording current pattern and supplied to the write head 31 (step S5). By this write head 31, RLL encoded data is magnetically recorded on a designated track on the disk 1.

On the other hand, at the time of data reproduction, the read head 3
By reading 0, the RLL encoded data recorded on the designated track on the disk 1 is read. Read amplifier 60
Is a data signal (R) read by the read head 30.
LL encoded data) and sends it out to the read channel of the data channel 7. In the read channel, the signal is sent to the digital PRML processing circuit 22 via the AGC amplifier 20 and the LPF 21 as described above.

The digital PRML processing circuit 22 executes, for example, digital waveform equalization processing in accordance with the equalization class of EEPR4 and maximum likelihood decoding processing by a Viterbi decoder. The decoder 23 performs a decoding process (3 in this embodiment) which is a reverse process of the encoding process by the encoder 11 described above.
(Decoding processing for the reproduction data from which the T pattern has been deleted). Here, the decoder 23 may have a configuration including a Viterbi decoder. The descrambler 24 is a decoder 2
3 to return the data decoded to the original recording data and send it to the HDC 8 as reproduction data RD.

(Recording Encoding Process) Next, the recording encoding process by the encoder 11 of the embodiment will be specifically described. Note that the decoding process performed by the decoder 23 is the reverse process of the encoding process, and a description thereof will not be repeated.

In this embodiment, as shown in FIG. 3A, as the RLL encoded data, the period of the highest frequency (repetition period) at the time of recording is T, and an arbitrary periodic signal is nT (for example, the highest period). A periodic signal of 1/6 times the frequency is 6T
), The period coding pattern of the lowest frequency is referred to as an 11T pattern. Therefore, the cycle coding pattern of the highest frequency is a 1T pattern, and is hereinafter referred to as a 2T pattern, a 3T pattern, and a 4T pattern. 1T
The pattern is an encoding pattern that generates an information bit “1” and then becomes “1”. The 11T pattern is an encoding pattern that generates ten consecutive information bits “0” from information bits “1” and then generates information bits “1”. Similarly, each coding pattern of the 3T pattern and the 4T pattern is expressed as shown in FIG.

The encoder 11 of the embodiment executes an encoding process according to a code generation rule as shown in FIG.
Here, FIG. 4 is a code state transition diagram. Starting from state 1, the state shifts to state 2 when information bit “0” is generated, and to state 3 when information bit “0” is generated. Thereafter, the state shifts to the horizontal state in the same manner. However, when the information bit “1” is generated on the way, the state returns to the state 1.

The code generation rule which means in the code state transition diagram of FIG. 4 indicates a procedure for generating encoded data from which the 3T and 4T patterns have been deleted. The 3T and 4T patterns are sensitive to crosstalk noise between adjacent tracks when data is recorded on each track of the disk 1, and have an error rate of 1.5 to 2 digits during data reproduction, as described later. Also causes a phenomenon of deterioration. 3T and 4T
The pattern deletion processing is 3 counting from the left end of state 1.
The next state of each of the fourth and fourth bits is a bit "1"
Is not generated. Here, as described above, the 1T pattern is an encoding pattern of the highest frequency that generates an information bit “1” and then becomes “1”. The 11T pattern is an encoding pattern that generates ten consecutive information bits “0” from information bits “1” and then generates information bits “1”. This 11T pattern is to be applied to a design for increasing low frequency components so as to be as advantageous as possible in magnetic recording. This is because, when the recording density of the disk 1 is increased, PE (partial erasure), which is a typical non-linear phenomenon, and a thermal fluctuation phenomenon in which the magnetization once recorded disappears are generated as much as possible. Has the effect of not letting you do.

Hereinafter, the operation and effect of the encoding processing of the embodiment will be specifically described.

First, as shown in FIG. 6, a designated track (TR1) at the time of data reproduction is adjacent to an adjacent track (TR).
In 2), it is assumed that data 62 of a predetermined periodic signal (nT pattern) is continuously recorded in advance. Data consisting of a random pattern is recorded on the designated track (TR1). FIG. 5 shows the result of measuring an error rate (bit error rate: BER) using the read offset amount (arrow 60) as a parameter when reproducing data from the track (TR1) by the read head. The read offset is a state in which the position of the read head on the track (TR1) is displaced (offset) in the direction of the adjacent track (TR2) (track center 61).

In FIG. 5, the vertical axis represents the error rate (BER), and the horizontal axis represents n of the predetermined period signal recorded on the adjacent track.
The T pattern is shown. The parameter is a read offset amount (0 to 0.34). FIG. 5 is basic data indicating how much a margin of an error rate remains with respect to a track density (TPI) on a disk.

As is clear from the measurement results shown in FIG. 5, if the read offset amount is "0", that is, if the read head is completely in the on-track state, any periodic signal from the adjacent track (TR2) will not matter. No error rate degradation is observed for crosstalk noise (characteristic curve 100). On the other hand, as the read offset amount increases, the error rate deteriorates (characteristic curve 10).
1-104). In particular, it can be confirmed that a specific periodic signal (here, 3T pattern) significantly reduces the error rate. In short, it shows that there is a difference in the amount of crosstalk noise received by the read head in the track (TR1) depending on the type (nT pattern) of the periodic signal recorded on the adjacent track (TR2).

FIG. 7 is a diagram showing the relationship between the off-track amount for an adjacent track and the error rate (BER). FIG. 8 is a diagram showing the relationship between the cycle n of the recording signal of the adjacent track and the signal amplitude (amplitude power).

Here, the crosstalk noise is a leakage signal from a track other than the track to be recorded / reproduced, and the influence from the adjacent track is particularly maximized. The influence of the recording signal on the adjacent track is related to two factors: the frequency of occurrence of the magnetization transition of the adjacent track and the signal amplitude (amplitude power). As shown by a characteristic curve A in FIG. 8, the frequency of occurrence of the magnetization transition is inversely proportional to the period n of the recording signal of the adjacent track. As shown by the characteristic curve B, the signal amplitude decreases as the recording density (high TPI), that is, the signal period n decreases.

With the increase in the recording density (higher TPI) of the HDD, the maximum recording frequency is shifted to the higher frequency side, while the low frequency signal is contained as much as possible so as to be magnetically advantageous. Therefore, the signal amplitude power difference between the lowest frequency and the highest frequency of the recording data is increasing. Specifically, as shown in FIG.
The signal amplitude power X in the 1T pattern is much smaller than the signal amplitude power Y in the T or 12T pattern (characteristic curve B). Therefore, even when viewed as crosstalk noise, when a recording signal of a 6T pattern or a 12T pattern exists in an adjacent track with respect to a 1T pattern, crosstalk noise to the track becomes larger.

On the other hand, as described above, the frequency of magnetization transition occurs more frequently in the 1T pattern than in the signal recorded at a high frequency (see the characteristic curve A in FIG. 8). Therefore, in the crosstalk noise, two factors, the frequency of occurrence of magnetization transition and the signal amplitude power, both depend on the period n of the recording signal of the adjacent track (characteristic curve C). The characteristic curve C shown in FIG. 8 is expressed as a product of the above-described elements. From the characteristic curve C, it can be confirmed that the period n of the recording signal of the adjacent track at which the crosstalk noise is maximized exists. That is, as described above, the recording signal of the adjacent track at which the crosstalk noise is maximized has a period n
Is an encoded pattern of a 3T pattern in which is “3”.

As described above, according to the present embodiment, as a coding pattern of a recording signal in which crosstalk noise from an adjacent track becomes remarkable, a 3T pattern is centered.
Specific coding patterns, including 2T and 4T patterns, can also be envisioned. In other words, when the recording data is RLL-encoded, in order to suppress the deterioration of the reproduction error rate due to crosstalk noise, the RL that generates the specific pattern (at least the 3T pattern) is generated.
What is necessary is just to delete from an L encoding pattern. Therefore, the encoder 11 according to the embodiment generates RLL encoded data in which a 3T pattern (which may include a 2T pattern or a 4T pattern) in which the error rate degradation is remarkable is deleted according to the code generation rule shown in FIG. Therefore, it is possible to magnetically record the RLL encoded data which is hardly affected by the crosstalk noise from the adjacent track on the designated track on the disk 1. As a result, it is possible to achieve a narrow track in which adjacent tracks are close to each other, and realize a high track density (high TPI).

Here, as described above, the error rate deterioration amount due to the 3T pattern or the 4T pattern depends on the reproduction distance (read offset amount) to an adjacent track. In other words, the error rate deterioration amount increases as the track pitch, which is the distance between the track to be reproduced and the adjacent track, decreases. As a specific example, when the encoded data from which the 3T pattern and the 4T pattern of the embodiment are removed is used, the margin in the off-track direction for keeping the sixth power at the reproduction error rate is increased by 0.13 μm on one side. It becomes possible. Estimated on both sides of the track, this corresponds to an increase in margin of 0.26 um. Therefore, when the track pitch is constant, the positioning accuracy of the head is ± 0.13 u.
You will earn a margin of m. On the other hand, when the error rate is constant, the track pitch can be reduced by 0.26 μm. Therefore, when the same embodiment is applied to a conventional HDD, the track density is 18 kTP.
As I, the track density can be improved by about 18%.

(Modification) FIG. 9 is a diagram related to a modification of the embodiment. FIG. 9A shows a code generation rule for deleting only the 3T pattern from the generated encoded data. FIG. 3B shows a code generation rule for deleting only the 4T pattern from the encoded data to be generated.

In the same embodiment, the reason why the influence of the crosstalk noise on the 3T pattern and the 4T pattern (or the 2T pattern) becomes large has been described. However, the signal amplitude power ratio of the highest / lowest frequency and the magnetization transition It is also assumed that the periodic pattern, which is likely to be affected by crosstalk noise, slightly changes due to a change in the occurrence probability ratio or the like. Even in this case, the code generation rule for deleting the periodic pattern can be easily analogized.

From FIG. 7, it is predicted that when the track width is reduced by about 10%, the error rate (BER) is deteriorated on average by about -0.3 digits. Furthermore, if the track width is made narrow, the average BER is degraded, and it can be estimated that there is a recording pattern in which the amount of crosstalk noise is maximized. Here, the nT pattern dependence curve of the BER shows an extreme value, and it is not rare that the curve degrades by one digit from the average BER. Therefore,
By controlling the BER maximum pattern, the BER is improved, and the design of the track width according to the desired BER can be narrower than the current state.

[0042]

As described in detail above, according to the present invention, H
In a disk recording / reproducing apparatus such as a DD, when recording data is encoded into RLL encoded data, RLL encoded data is generated in which the frequency of occurrence of a code pattern highly influenced by crosstalk noise from adjacent tracks is suppressed. . Therefore, by recording the RLL coded data on the disc, it is possible to effectively suppress crosstalk noise from an adjacent track during data reproduction and improve a reproduction error rate. In other words, since the error rate between adjacent tracks can be relatively improved, it is possible to easily realize a higher track density.

[Brief description of the drawings]

FIG. 1 is an exemplary block diagram showing a main part of an HDD according to an embodiment of the present invention.

FIG. 2 is an exemplary flowchart for explaining the operation of the data channel according to the embodiment;

FIG. 3 is an exemplary conceptual diagram for explaining an encoding pattern related to a recording encoding process according to the embodiment;

FIG. 4 is a code state transition diagram for explaining a recording encoding process of the embodiment.

FIG. 5 is a diagram showing a relationship between an encoding pattern and an error rate in the embodiment.

FIG. 6 is a conceptual diagram showing a data recording / reproducing state for obtaining the measurement result.

FIG. 7 is a view showing a relationship between an off-track amount and an error rate in the embodiment.

FIG. 8 is an exemplary view showing a relationship between a signal period of an adjacent track and a signal amplitude in the embodiment.

FIG. 9 is a code state transition diagram for describing a recording encoding process related to a modification of the embodiment.

[Explanation of symbols]

 Reference Signs List 1 disk 2 spindle motor (SPM) 3 head slider 4 actuator 5 voice coil motor (VCM) 6 preamplifier circuit 7 data channel 8 disk controller (HDC) 9 servo system 10 scrambler 11 Encoder 12 Precoder 13 Prerecording compensation circuit (write precon) 20 AGC amplifier 21 Low-pass filter (LPF) 22 Digital PRML processing circuit 23 Decoder (decoding circuit) 24 Descrambler 30 Decoder Head 31 Write head 60 Read amplifier 61 Write amplifier

Claims (3)

[Claims] 1. A disk recording medium that is a data recording medium, a head that records and reproduces data on and from the disk recording medium, and a recording data to be recorded on the disk recording medium connected to the head and Means for processing each data signal corresponding to reproduced data reproduced from the disk recording medium, the encoding means for encoding the recorded data into RLL encoded data, and the decoding means for decoding the reproduced data And the encoding unit suppresses the occurrence frequency of the suppressed RLL encoded pattern in which the reproduction error rate is degraded due to the crosstalk noise, and suppresses the occurrence frequency more than the occurrence frequency of all the patterns except the suppressed RLL encoded pattern. R
A disk recording / reproducing apparatus, comprising: data signal processing means configured to generate the RLL encoded data so that the frequency of occurrence of the LL encoded pattern is reduced. 2. The method according to claim 1, wherein, when the encoding pattern having the highest frequency is a 1T pattern, a frequency of occurrence of at least one of 2T to 4T patterns whose reproduction error rate deteriorates from the recording data. RLL coded data of the RLL coded pattern is generated such that the frequency of occurrence of the suppressed RLL coded pattern is lower than the frequency of occurrence of each pattern except for the suppressed RLL coded pattern. 2. The disk recording / reproducing apparatus according to claim 1, wherein the decrypting unit is configured to decode the RLL encoded data reproduced from the disk recording medium by the head into the original recording data. 3. In order to suppress deterioration of a reproduction error rate due to crosstalk noise or the like, when the highest frequency encoding pattern is a 1T pattern, the frequency of occurrence of at least one of 2T to 4T patterns is determined. 2. The disk recording / reproducing apparatus according to claim 1, wherein the apparatus is configured to generate the RLL encoded data so as to suppress the frequency of occurrence of the 5T pattern or less.