JP2010080002A - Data reading method, read channel, and storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To read data with high accuracy even if there is a flaw on a storage medium. <P>SOLUTION: When a SYNC mark is written in a flaw part of a storage medium and unable to be read, the length of a preamble is changed (e.g., shortened) to overwrite it with the changed preamble and the SYNC mark (steps S18 and S20). Thus, the SYNC mark is movable (overwrite) to a part of no flaw. When the data is read, the data is read by timing (interval) based on the length of the changed preamble (steps S22 to S30), and hence the data can be read irrespective of any positional change of the SYNC mark. By reading the data using the SYNC mark, the data can be read with high accuracy without being affected by jitters during reading of the SYNC mark. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a data reading method, a read channel, and a storage device, and more particularly, a data reading method when a sync mark arranged immediately before data cannot be read, a read channel suitable for carrying out the data reading method, and The present invention relates to a storage device including the read channel.

  Conventionally, a magnetic disk device is widely known as an information storage device capable of reading (reproducing) data from a storage medium. This magnetic disk device uses a magnetic disk as a storage medium, and records data on the magnetic disk (each track) in units of blocks constituted by a predetermined number of symbols, that is, in units of sectors. In this type of magnetic disk apparatus, a function for accurately detecting the leading symbol of data is indispensable. Therefore, a synchronization signal pattern (hereinafter referred to as “SYNC mark”) is written before the data. By detecting this SYNC mark, it is possible to correctly detect the leading symbol of data, thereby ensuring data synchronization.

  For this reason, when the SYNC mark cannot be detected accurately, data synchronization is not ensured. Therefore, when the read operation (read operation) is performed as it is, data with a timing shifted by several symbols, for example, is read.

  This is called a framing error, and the read error covers the entire sector, and correction using ECC (error correction code) is impossible. Therefore, detection of the SYNC mark (synchronization signal detection) is a very important factor for improving the performance of the magnetic disk device.

  Usually, the detection of the SYNC mark is performed by pattern-matching a predetermined SYNC mark with the read data and counting the number of matching symbols. When the number of matching symbols is equal to or greater than the predetermined number, a synchronization detection signal is generated. Output. However, if a defect due to a scratch or the like exists in a portion where the SYNC mark is written on the magnetic disk, the SYNC mark cannot be detected, and it becomes impossible to ensure data synchronization. (For example, see Patent Documents 1 to 3).

  Recently, in order to solve this problem, a dual SYNC mark function is provided in which two SYNC marks are provided in one sector, and when one SYNC mark cannot be detected, data is generated by detecting the other SYNC mark. Has been proposed as a function of RDC (read channel) (hereinafter, this technology is referred to as “conventional technology 1”).

  In addition, when detecting a SYNC mark composed of a bit pattern of a specific length, there is also proposed a technique that employs a function that allows a specific number of bit errors to be detected even if the entire SYNC mark cannot be read completely. (Hereinafter, this technique is referred to as “conventional technique 2”).

JP 2001-184806 A JP-A-6-111492 JP 2001-143406 A

  However, in the above prior art 1, it is generally rare that the SYNC mark is written on a defect portion on the magnetic disk. Therefore, since one SYNC mark is usually sufficient, always providing two SYNC marks in one sector may deteriorate the format efficiency by one SYNC mark and reduce the recording density. .

  Further, in the above-described prior art 2, when the allowable range of bit errors is increased (the number of allowed errors) is increased, there is a possibility that a portion that is not a SYNC mark is erroneously detected as a SYNC mark.

  On the other hand, as a technique for solving the above-described problem, a technique for reading data at each of a plurality of timings estimated to be in the vicinity of the start of data after the read gate is turned on has been proposed. However, there is a variation (jitter) in the timing when the read gate is turned on, and also when writing data to the magnetic disk, there is a variation (jitter) in the timing from when the write instruction is issued until the actual writing is performed. Therefore, the possibility that the data can be read accurately from the top of the data is very low.

  Accordingly, the present invention has been made in view of the above problems, and an object thereof is to provide a data reading method capable of reading data with high accuracy. It is another object of the present invention to provide a read channel capable of writing a sync mark at a position different from the scratched part by overwriting even when the sync mark is written in a damaged part on the storage medium. It is another object of the present invention to provide a storage device that can reproduce data with high accuracy.

  The data reading method described in the present specification is based on the preamble added before the sync mark when the sync mark added before the data cannot be read when reading the data recorded on the storage medium. An overwriting step of overwriting the storage medium with the changed preamble and sync mark, and after reading the overwritten sync mark, taking into account the changed preamble length And a reading step for reading the data.

  According to this, as a case where the sync mark cannot be read, it is assumed that the sync mark is written on a defect portion of the storage medium. However, in the overwriting process, the length of the preamble is changed. Thus, by overwriting the preamble and the sync mark, it is possible to move (overwrite) the sync mark to a portion having no scratch. Further, in the reading step, data can be read regardless of the position of the sync mark by reading the data in consideration of the changed preamble length. As a result, data can be read based on the sync mark, so that data can be read with high accuracy.

  The read channel described in the present specification outputs at least one of a preamble, a sync mark, and data to a preamplifier connected to a head that records data on the storage medium in order to record data on the storage medium. The output mode of the output unit is changed to the first mode for outputting the preamble, the sync mark, and the data in this order, and the length of the preamble is changed from the length in the first mode. And a switching unit that switches between a second mode that sequentially outputs a preamble and a sync mark.

  According to this, since the switching unit sets the output unit to the first mode, the output unit outputs the preamble, the sync mark, and the data in this order, so that normal data recording can be performed. By setting the output unit to the second mode, only the preamble and sync mark having different lengths from the first mode are output from the output unit, so that the preamble and sync mark having different lengths can be overwritten. it can. As a result, even when the sync mark is written on the scratched portion of the storage medium, the sync mark can be moved (overwritten) to another new location.

  A storage device described in this specification includes a preamplifier connected to a head that performs recording / reproduction of data to / from a storage medium, and a read channel described in this specification connected to the preamplifier. .

  According to this, even if a sync mark is written on a scratched part on the storage medium, it is provided on the storage medium because it has a read channel that can write the sync mark to a position different from the scratched part by overwriting. It is possible to reproduce the data with high accuracy by avoiding the influence of the scratches.

  The data reading method described in this specification has an effect that data can be read with high accuracy. In addition, the read channel described in this specification has an effect that the sync mark can be written at a position different from the scratched part by overwriting even when the sync mark is written in the damaged part on the storage medium. In addition, the storage device described in this specification has an effect that data can be reproduced with high accuracy.

<< First Embodiment >>
Hereinafter, a magnetic disk device 100 as a first embodiment of a storage device according to the present invention will be described in detail with reference to FIGS.

  FIG. 1 schematically shows the configuration of a magnetic disk device 100 according to the first embodiment. As shown in FIG. 1, the magnetic disk device 100 includes a printed circuit board 90 and a disk enclosure 80.

  The printed circuit board 90 includes a data buffer 12, a flash ROM (Read Only Memory) 14, a shock sensor 16, a servo controller 18, a clock generation unit 20, a read channel 22, and a hard disk controller 24. The disk enclosure 80 includes a preamplifier 30, a head 32, a voice coil motor 34, a spindle motor 36, and a magnetic disk 38 as a storage medium.

  The data buffer 12 is a storage unit that temporarily stores data from a host (not shown), and the flash ROM 14 is a storage unit that stores data such as programs used by the hard disk controller 24.

  The shock sensor 16 detects vibrations in the vertical and horizontal directions with respect to the magnetic disk device 100 and notifies the servo controller 18 of vibration information that is information of the detected vibrations. The servo controller 18 receives commands from the hard disk controller 24, The voice coil motor 34 and the spindle motor 36 are controlled based on vibration information and the like. Further, if the writing is in progress, the writing operation is stopped once, and the control for rewriting again after the vibration is recovered is entered.

  The clock generation unit 20 is a device that generates a clock and supplies the generated clock to the hard disk controller 24. The read channel 22 is a device that accepts access from the hard disk controller 24 or the preamplifier 30 and bridges data between the hard disk controller 24 and the preamplifier 30. The read channel 22 has a write system 22a shown in FIG. 3 and a read system 22b shown in FIG.

  The hard disk controller 24 performs overall control of the entire magnetic disk device 100.

  The preamplifier 30 transmits data input from the read channel 22 to the head 32 (recording element) and transmits data read by the head 32 (reproducing element) to the read channel 22.

  The head 32 has a recording element for writing data to the magnetic disk 38 and a reproducing element for reading and erasing the written data.

  The voice coil motor 34 drives a head stack assembly (HSA) 33 that holds the head 32 under the instruction of the servo controller 18 and positions the head 32 at a desired position on the magnetic disk 38. is there.

  The spindle motor 36 rotates the magnetic disk 38 at an appropriate rotation speed under the instruction of the servo controller 18.

  The magnetic disk 38 is a storage medium that stores data by changing the magnetization state of the applied magnetic material. Here, before the data to be written on the magnetic disk 38, as shown in FIG. 2, a mark (SYNC mark) used for finding the head of the data at the time of reproduction is written. In addition, a synchronization signal (preamble) of a PLL circuit used at the time of reproduction is written before the SYNC mark. Note that an ECC correction code, a postamble, and the like are written after the data, but the illustration thereof is omitted in FIG.

  Next, the write system (write system) 22a of the read channel 22 will be described. FIG. 3 is a block diagram of the write system 22 a of the read channel 22.

  As shown in FIG. 3, the write system 22a of the read channel 22 includes an NRZ interface 42, a scrambler 50, an encoder 44, a parallel-serial converter 46, a precoder 48, a multiplexer 56 as an output unit, A correction unit 58, a PECL driver 60, a handshake byte detector 62, a delay circuit 64, a data formatter 66 as a switching unit, and a multiplexer 68 are included.

  The NRZ interface 42 receives data (NRZ data) output from the hard disk controller 24, and the scrambler 50 is used for randomization to eliminate the regularity of the NRZ data.

  The encoder 44 limits the continuous length of “no magnetization reversal” or adds a parity bit for the post processor 86 (see FIG. 4) in order to prevent failure of the timing recovery 98 (see FIG. 4). . The parallel-serial conversion unit 46 converts NRZ parallel data into serial data. The precoder 48 is a kind of conversion circuit, and outputs data (write data) to be written to the magnetic disk 38 to the multiplexer 56.

  The multiplexer 56 outputs any one of write data, preamble, and SYNC mark output from the precoder 48 to the correction unit 58.

  The correction unit 58 shifts the inversion point in advance according to the data pattern at the time of writing in order to correct NLTS (Non Linear Transition Shift). The PECL driver 60 is a kind of digital switch, and can transmit signals at high speed.

  The handshake byte detection unit 62 detects Sync.Byte (also called handshake byte) sent from the hard disk controller 24 into the read channel 22 via the NRZ data.

  The delay circuit 64 delays the output of the detection result by the handshake byte detector 62 by a delay (delay time) set in the hard disk controller 24.

  The data formatter 66 outputs a preamble via the multiplexer 56 when the write gate is turned on in the hard disk controller 24. Further, when the handshake byte is detected, the SYNC mark is output by calculating backward from the timing when the head of the write data passes through the precoder 48. Then, the write data that has passed through the precoder 48 is output immediately after the output of the SYNC mark is completed.

  Next, the read system (read system) 22b of the read channel 22 will be described. 4 is a block diagram of the read system 22b of the read channel 22. As shown in FIG.

  As shown in FIG. 4, the read system 22b of the read channel 22 includes a VGA 72, an ASC 74, a CTF 76, an ADC 78, an FIR 82, a Viterbi decoder 84, a post processor 86, a decoder 88, and a data formatter. 92, and the NRZ interface 42 common to the write system 22a of FIG. The lead system 22b further includes an SM (SYNC mark) detector 94, an adaptation engine 96, a timing recovery 98, an ASC controller 102, an AGC controller 104, and a TBG generator 106.

  The VGA 72 and the ASC 74 constitute a feedback loop and control the equalized reproduction signal amplitude to an appropriate value. Along with the FIR 82, the CTF 76 equalizes the input waveform (reproduced waveform) to the target waveform (Target) and cuts off the high frequency. The FIR 82 equalizes the input waveform to the target waveform together with the CTF 76, and performs adaptation in response to changes in the input waveform. The Viterbi decoder 84 selects a plausible data series by using the relationship between data before and after.

  The post processor 86 detects an error in a specific unit bit (Code Word) using parity, selects several types of error patterns prepared in advance, and corrects them. The decoder 88 limits the continuous length of “no magnetization reversal” to prevent the timing recovery 98 from failing, and adds a parity bit for the post processor 86.

  The timing recovery 98 forms a feedback loop and controls the sample timing of the equalized reproduction signal to an appropriate value. The TBG generator 106 (reference clock generator) generates a reference clock for data. In the present embodiment, the TBG generator 106 generates a data transfer clock based on a signal input from an external fixed frequency oscillator.

  Next, a method for reading data recorded on the magnetic disk 38 in the first embodiment will be described.

  First, based on FIG.5 and FIG.2, the outline | summary of the data read-out method of this embodiment is demonstrated easily.

  Usually, as shown in FIG. 2, since the preamble and the SYNC mark are added to the data recorded on the magnetic disk 38, the preamble is read out by the head 32 to synchronize and detect the SYNC mark. (Detection by the SM detector 94 in FIG. 4) starts data reading. However, as shown in FIG. 5A, when the SYNC mark is written on the scratched portion (defect portion) DF on the magnetic disk 38, the SM detector 94 cannot read the SYNC mark. For this reason, in the state of FIG. 5A, the data reading start position cannot be grasped, and data cannot be read.

  In the present embodiment, as described above, when it is determined that the SYNC mark cannot be read, the length of the preamble is already written on the magnetic disk 38 as shown in FIG. 5B. It is assumed that the change is made shorter and the preamble and the SYNC mark after the change are overwritten on the magnetic disk 38. Note that an interval J shown in FIG. 5B indicates a deviation of the writing position due to jitter (Jitter) generated when the changed preamble is overwritten.

  By performing the overwriting as described above, the SYNC mark is arranged (written) while avoiding the scratched portion (defect portion) DF on the magnetic disk 38.

On the other hand, when reading data, unlike before overwriting, the data does not exist immediately after the SYNC mark. Therefore, as shown in FIG. 5C, a predetermined interval is provided after the SYNC mark is detected. Then, data reading is executed. In this case, the interval is
(Preamble length before change) − {(Preamble length after change) + (jitter amount)}
That is,
(Preamble shortening amount)-(jitter amount)
However, the exact value of the jitter amount is unknown.

Therefore, in this embodiment, as an interval,
(Preamble shortening amount)-(variable (γ))
And reading data until the data can be read accurately while changing the value of γ.

  Next, a specific execution method of the data reading method will be described with reference to the block diagrams of FIGS. 3 and 4 with a focus on the flowchart of FIG.

  When the read gate is turned on in the hard disk controller 24, the flowchart of FIG. 6 is started. In step S12, in the read system 22b of the read channel 22, the length of the preamble is set to “normal mode”, the interval is set to “0”, and the read mode is set to “data read mode”. These settings are normal data reading modes. Therefore, in step S14, the read system 22b of the read channel 22 of FIG. 4 reads out the data recorded on the magnetic disk 38 using a general method.

  Next, in step S16, the data formatter 92 (or hard disk controller 24) determines whether or not the data has been read. If the determination here is affirmed, the entire process of FIG. 6 ends. On the other hand, when judgment here is denied, it transfers to step S18. Note that when the determination in step S16 is negative, as shown in FIG. 5A, the SYNC mark is present in the flaw (DF) portion DF of the magnetic disk 38, so that the SYNC mark cannot be read. A case is assumed.

  In step S18, the data formatter 66 of FIG. 3 sets the length of the preamble to “shortening mode” and sets the writing mode to “overwrite mode”.

  Next, in step S20, the data formatter 66 controls the multiplexer 56 to output the preamble for a time shorter than usual and to output the SYNC mark after outputting the preamble. As a result, the preamble and the SYNC mark whose length is set short are overwritten on the magnetic disk 38 as shown in FIG. 5B. In this case, the SYNC mark is written at a position shifted from a defect on the magnetic disk 38.

  Next, in step S22, in the timing recovery 98 of FIG. 4, the interval (I) is set to the initial value (α), and the reading mode is set to “SYNC mark recovery mode”. As the initial value (α) of the interval (I), for example, it is possible to set the same value as the preamble length change amount.

  Next, in step S24, data is read using the read system 22b of the read channel 22 of FIG. In this case, in the read system 22b, after reading out the SYNC mark, reading of data is started after an interval (I).

  Next, in step S26, the data formatter 92 (or hard disk controller 24) determines whether or not the data has been read. If the determination here is affirmed, all the processes in FIG. 6 are terminated. On the other hand, if the determination is negative, the process proceeds to step S28.

  In the next step S28, the timing recovery 98 determines whether or not the interval (I) exceeds a preset maximum value (β). If the determination here is negative, the process proceeds to step S30, where the timing recovery 98 increases the interval (I) by γ and returns to step S24. Thereafter, steps S24, S26, and S28 are performed until data can be read (until the determination in step S26 is affirmed) or until the interval (I) exceeds the maximum value β (until the determination in step S28 is affirmed). repeat. If the determination in step S28 is affirmative, it is determined that data could not be read, and the process ends in error.

  In the first embodiment, when data input from a host (not shown) is written to the magnetic disk 38 (at the time of normal data writing), the length of the preamble is set to “normal mode” and writing is performed. Set the mode to “Normal mode”. As a result, a normal length preamble, SYNC mark, and data are sequentially written on the magnetic disk 38 (see FIG. 2). In the present embodiment, the normal mode corresponds to the “first mode”, and the overwrite mode in steps S18 and S20 described above corresponds to the “second mode”.

  As described in detail above, according to the first embodiment, the length of the preamble is changed when the SYNC mark cannot be read because the SYNC mark is written on a defect portion of the storage medium ( And the preamble and the SYNC mark are overwritten, so that the SYNC mark can be moved (overwritten) to a portion having no flaws. Further, when data is read, the data is read at a timing (interval) based on the changed preamble length, so that the data can be read regardless of the position change of the SYNC mark. By reading data using such a SYNC mark, it is possible to read data with high accuracy without being affected by jitter when reading the SYNC mark.

  Further, according to the first embodiment, the read operation is repeated until the data can be read while changing the read timing (interval) of the SYNC mark. Therefore, regardless of the magnitude of jitter at the time of writing, Reading can be performed with high accuracy.

<< Second Embodiment >>
Next, a second embodiment of the present invention will be described. Unlike the first embodiment, the second embodiment is characterized in that the preamble length is changed longer when the SYNC mark cannot be detected.

  First, an outline of the data reading method of the second embodiment will be briefly described with reference to FIG.

  In the second embodiment, as shown in FIG. 7A, when the SYNC mark is written in the flaw (defect) portion DF of the magnetic disk 38, as shown in FIG. Overwrite by changing the length of. As a result, the SYNC mark is arranged (written) while avoiding the scratched portion (defect portion) DF on the magnetic disk 38.

  However, when overwriting is performed by the above-described method, the data is eroded by the preamble or the SYNC mark, and therefore the data originally written cannot be read only by the data read immediately after the SYNC mark is detected.

  For this reason, in the present embodiment, as shown in FIG. 7D, from immediately after the read gate is turned ON in the hard disk controller 24 (immediately after receiving the handshake byte), as shown in FIG. Dummy data having a size corresponding to the eroded portion is output from the read channel 22 to the hard disk controller 24 in advance.

  Then, the hard disk controller 24 performs ECC correction calculation processing on the dummy data and the data read after detecting the SYNC mark so as to read the data originally written.

  Next, a specific execution method of the data reading method will be described with reference to the block diagrams of FIGS. 3 and 4, focusing on the flowchart of FIG.

  In the flowchart of FIG. 8, steps S12 to S16 are the same processes as those in the first embodiment. And when judgment in step S16 is denied, it transfers to step S118.

  In step S118, the data formatter 66 of FIG. 3 sets the length of the preamble to “decompression mode” and sets the write mode to “overwrite mode”.

  Next, in step S120, the data formatter 66 controls the multiplexer 56 to output a preamble longer than usual, and to output a SYNC mark after outputting the preamble. As a result, the preamble and the SYNC mark having a long length are overwritten on the magnetic disk 38 as shown in FIG. 7B. In this case, the SYNC mark is written (overwritten) at a position shifted from a scratched portion (defect) on the magnetic disk 38.

  Here, in practice, as shown in FIG. 3, the write system 22a of the read channel 22 is provided with a delay circuit 64, and thus the above processing is executed using the delay circuit 64. That is, the delay circuit 64 determines the time from when the handshake byte detector 62 detects a handshake byte (a signal for switching from preamble output to SYNC byte output) until the data formatter 66 is notified of the detection result. Delay (delay). In this case, the preamble is continuously output from the read channel 22 to the preamplifier after the write gate is turned on until the detection result is notified to the data formatter 66. Therefore, as shown in FIG. 9, it is possible to change the preamble length longer by the delay from the time when the handshake byte detection is completed (time point a) to the time when the SYNC byte recording starts (time point b). . In this case, even if the beginning of the data has already passed through the precoder 48, there is no problem because the output of the data is not selected in the multiplexer 56.

  Through such processing, a “New preamble” and a “New SYNC mark” as shown in FIG. 9 are recorded on the magnetic disk 38.

  The data formatter 66 turns off the RW signal of the preamplifier 30 immediately after outputting the SYNC mark after detecting the handshake byte (see FIG. 9). By doing so, it is possible to prevent the write data from being output to the preamplifier 30. Thereby, only the preamble and the SYNC mark can be overwritten on the magnetic disk 38.

  Returning to FIG. 8, in the next step S122, the initial value (κ) of the dummy data length is set in the data formatter 92 (see FIG. 7C), and the reading mode is set to “SYNC mark recovery mode”. . As the initial value (κ) of the dummy data length, for example, the same value as the preamble length change amount (expansion amount) can be set.

  In step S124, data is read using the read system 22b. In reading this data, after the read gate is turned on in the hard disk controller 24 as shown in FIG. 7 (d), it is prepared in advance from the time when the handshake byte is detected as shown in FIG. 7 (e). The dummy data having a length (κ) (for example, dummy data consisting of data “0000...”) Is transferred to the hard disk controller 24, and the data is read after the SYNC mark is detected and transferred to the hard disk controller 24. The transfer timing of these data is controlled by RRCLK shown in FIG. Further, these data become normal data through ECC correction calculation processing in the hard disk controller 24.

  Next, in step S126, the data formatter 92 (or the hard disk controller 24) determines whether or not the data has been read. If the determination here is affirmed, all the processes in FIG. 8 are terminated. On the other hand, when judgment here is denied, it transfers to step S128.

  In the next step S128, it is determined whether or not the dummy data length (κ) exceeds a preset maximum value (δ). If the determination is negative, the process proceeds to step S130, the dummy data length (κ) is increased by γ, and the process returns to step S124. Thereafter, steps S124, S126, until data can be read (until the determination in step S126 is affirmed), or until the dummy data length (κ) exceeds the maximum value δ (until the determination in step S128 is affirmed). The process / determination of S128 is repeated. If the determination in step S128 is affirmative, it is determined that data could not be read, and the process ends in error.

  Also in the second embodiment, when data input from a host (not shown) is written to the magnetic disk 38 (at the time of normal data writing), the length of the preamble is set to “normal mode”. Set the writing mode to "normal mode". As a result, a normal length preamble, SYNC mark, and data are sequentially written on the magnetic disk 38 (see FIG. 2). In the present embodiment, the normal mode corresponds to the “first mode”, and the overwrite mode in steps S118 and S120 described above corresponds to the “second mode”.

  As described above in detail, according to the second embodiment, the length of the preamble is changed when the SYNC mark is written on the defect portion of the storage medium and cannot be read. Since the preamble and the SYNC mark are overwritten (decompressed), it is possible to move (overwrite) the SYNC mark to a portion having no scratch. When data is read, dummy data having a length determined based on the data expansion is prepared in advance, and ECC correction calculation processing is performed together with the data read after the SYNC mark is detected. As described above, since data can be read using the SYNC mark, it is possible to read data with high accuracy without being affected by jitter at least when reading the SYNC mark.

  Further, according to the second embodiment, since the read operation is repeated until the data can be read while changing the data length of the dummy data, the data can be read accurately regardless of the magnitude of jitter at the time of writing. Is possible.

  In each of the above embodiments, the case where the SYNC mark is overwritten while completely avoiding the scratched part (defect) DF is illustrated, but the present invention is not limited to this, and the SYNC mark is a part of the scratched part ( It may cover a small area. Even in this case, it is possible to detect the SYNC mark and the data by adopting a method of allowing and detecting a specific number of bit errors as in the prior art 2.

  The above-described embodiment is an example of a preferred embodiment of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the scope of the present invention.

1 is a block diagram schematically showing a configuration of a magnetic disk device according to a first embodiment. FIG. It is a figure which shows roughly the data recorded on a magnetic disc, a SYNC mark, and a preamble. It is a block diagram which shows the structure of the write system of a read channel. It is a block diagram which shows the structure of the read system of a read channel. It is a figure for demonstrating the overwrite method of the SYNC mark and the reading method of data which concern on 1st Embodiment. It is a flowchart which shows the data reading method which concerns on 1st Embodiment. It is a figure for demonstrating the overwrite method of the SYNC mark and the reading method of data which concern on 2nd Embodiment. It is a flowchart which shows the data reading method which concerns on 2nd Embodiment. It is a figure for demonstrating the method to decompress | expand and output a preamble, and the method of overwriting only a preamble and a SYNC mark.

Explanation of symbols

22 Read channel 56 Multiplexer (output unit)
66 Data formatter (switching unit)
38 Magnetic disks (storage media)
100 Magnetic disk device (storage device)

Claims (10)

When reading the data recorded on the storage medium, if the sync mark added before the data cannot be read, the length of the preamble added before the sync mark is changed, and the post-change An overwriting step of overwriting the storage medium with a preamble and a sync mark of
A data reading method, comprising: a reading step of reading the overwritten sync mark and reading the data in consideration of the length of the preamble after the change. In the overwriting step, the preamble is changed short,
2. The data reading method according to claim 1, wherein in the reading step, after reading the sync mark, reading of the data is started after an interval corresponding to the amount of shortening of the preamble. 3. The data reading method according to claim 2, wherein in the reading step, the data reading is repeatedly executed while changing the length of the interval stepwise. In the overwriting step, the preamble is changed longer,
In the reading step, dummy data having a length corresponding to the decompression amount is prepared in advance, and the data is read immediately after detecting the sync mark, and the dummy data and the read data are obtained. 2. The data reading method according to claim 1, wherein error correction processing is performed to obtain true data. 5. The data reading method according to claim 4, wherein, in the reading step, the error correction process is repeatedly executed while changing the size of the dummy data stepwise. An output unit for outputting at least one of a preamble, a sync mark, and data to a preamplifier connected to a head for recording data on the storage medium in order to record data on the storage medium;
The output mode of the output unit is a first mode in which the preamble, the sync mark, and the data are output in this order, and the length of the preamble is changed from the length in the first mode, and the preamble after the change A read channel comprising: a switching unit that switches between a second mode that sequentially outputs sync marks. A reading unit for reading the data;
The read channel according to claim 6, wherein the read unit reads the data in consideration of a length of the preamble. In the second mode, when the length of the preamble is changed short,
8. The read channel according to claim 7, wherein after the sync mark is detected, the read unit starts reading the data after an interval corresponding to the amount of shortening of the preamble. In the second mode, when the length of the preamble is changed long,
The read channel according to claim 7, wherein the read unit prepares dummy data having a length corresponding to the expansion amount in combination with the reading of the data. A preamplifier connected to a head for executing data recording / reproducing with respect to a storage medium;
A storage device comprising: the read channel according to claim 6 connected to the preamplifier.

Images