JP2000040202A - Magnetic disk device and method of detecting thermal asperity in the device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make surely detectable the place of TA(thermal asperity) generation, which used to be undetected due to the in-phase noise removing characteristics of a DSMR head in spite of generation of a TA. SOLUTION: Using a gain-variable head amplifier 22-1, 22-2 for a pair of amplifiers that amplifies signals reproduced by the MR elements 11-1, 11-2 of a DSMR head 1-i (i=0, 1), the detection of TA by CPU 6 is performed using a TA detection circuit 4 with the gain α1, α2 switched to different value through a D/A converter 23-1, 23-2 and set, in addition to using a TA detection circuit 4 with the gain α1, α2 set as a standard value α. The sector in which RA generation is thus detected is registered as a defect sector in a defect sector control table 62. In this case, an alternate sector is assigned to such sector.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk drive using a double stripe magnetoresistive head having a pair of MR elements for reproducing data from a disk, and a method for detecting thermal asperity in the magnetic disk drive.

[0002]

2. Description of the Related Art In recent years, in order to realize a high recording density (high capacity) of a magnetic disk drive, a magnetic resistance (Magneto-Resistant) has been developed.
Magnetic disk devices employing read-only (playback) heads called sistive (MR) heads are increasing. However, in a magnetic disk device using a magnetoresistive head, when a minute protrusion is present on the surface of a magnetic disk as a recording medium, a reproduction waveform (read) is generated by heat generated by the head coming into contact (collision) with the protrusion. (A signal waveform) is disturbed, that is, the reproduction waveform changes in a DC manner (a change in time of about 1 to 2 μs, the frequency is low), and the base line (the center line of the amplitude of the reproduction waveform) shifts. Due to a phenomenon called asperity (Thermal Asperity; TA), there is a problem that causes a read error or the like.

[0003] Japanese Patent Application Laid-Open No. 9-231505 discloses a dual stripe magnetoresistive (hereinafter referred to as DSMR) head having a pair of magnetoresistive elements (MR elements). D
A technique for removing a low-frequency fluctuation signal (noise) due to the occurrence of TA included in a signal read by the SMR head by a differential amplifier and a technique for detecting the occurrence of TA are described.

According to the above publication, in a magnetic disk drive using a DSMR head, a fluctuation signal included in a signal read by a pair of MR elements constituting the DSMR head appears as an in-phase noise component. The component can be canceled (attenuated) by the differential output (differential amplifier for amplifying the) of the pair of MR elements. Since the fluctuation signal (TA component) caused by the occurrence of the TA also appears as an in-phase signal, the fluctuation signal is canceled to some extent by the same principle. That is, in the magnetic disk device using the DSMR head, it is possible to suppress the baseline shift due to the TA due to the protrusion on the disk surface by the characteristic of common mode removal.

[0005]

As described above, the DS
In a magnetic disk drive using an MR head, TA
The common-mode noise component (fluctuation signal) resulting from the generation is canceled to some extent by the differential output of a pair (two layers) of MR elements (MR films). Compared to the case using the MR head of
The effect of the occurrence of A (that is, the amount of baseline shift caused by the occurrence of TA) can be reduced.

However, a pair of M of the DSMR head
If the magnetoresistance characteristics (temperature rise characteristics, heat radiation characteristics) of the R element change differently over time, the common-mode noise removal characteristics cannot be sufficiently exhibited. Then, if there are minute projections on the disk surface, those that were not detected as TAs in the initial state can be seen without canceling the baseline shift caused by the TAs due to the projections. In other words, in the prior art described in the above-mentioned publication, TA that was not visible (not detected) in the initial state is D
Lagging due to aging of the magnetoresistive characteristics of the SMR head,
There is a risk of causing a read error or the like.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object that despite the occurrence of thermal asperity (TA), it has not been detected by the prior art due to the common mode noise removal characteristics of the DSMR head. Thermal asperity (T
It is an object of the present invention to provide a magnetic disk drive capable of reliably detecting the location where A) occurs and a thermal asperity detection method in the drive.

[0008]

SUMMARY OF THE INVENTION The present invention relates to a magnetic disk drive using a DSMR head (double stripe magnetoresistive head) having a pair of MR elements for reproducing data from a disk. A pair of variable gain amplifiers for independently amplifying and setting gains for amplifying signals reproduced by the MR elements, and a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers A TA channel detecting the occurrence of a thermal asperity (TA) based on the signal differentially amplified by the differential amplifier;
With the gains of the pair of variable gain amplifiers set to the same value, the presence / absence of a TA is determined for each sector on the disk in accordance with the result of detection by the TA detection circuit. The first TA occurrence sector identification (first TA identification) in which an alternative sector is assigned to the relevant sector is performed.
When the gains of the paired variable gain amplifiers are set to different values, the result of the TA detection circuit is used to determine whether or not a TA has occurred in at least each sector except for the sector detected by the first TA washout on the disk. Judge according to
TA detection means for performing a second TA detection for allocating a substitute sector to a sector in which the occurrence of a TA is detected as a defective sector.

In such a configuration, a sector in which the occurrence of a TA is detected is registered as a defective sector (TA-occurring defective sector) by the first TA washout, and an alternative sector is assigned to the sector. Therefore, when an access request to the defective sector is issued thereafter, the access to the substitute sector is performed and the access can be performed correctly.

However, in the group of sectors that are not detected as defective sectors (TA-occurring defect sectors) in the first TA washout, even though TA is actually generated, that is, a pair of MR elements of the DSMR head is used. Although the in-phase fluctuation signal (noise component) caused by the occurrence of TA is included in the reproduced output signal of the differential amplifier, the fluctuation signal is attenuated by the common-mode noise component removal function of the differential amplifier, and as a result, TA is generated by the TA detector May be included in some sectors. In such a sector,
If the magnetoresistive characteristics (temperature rise characteristics, heat radiation characteristics) of a pair of MR elements of the DSMR head deviate due to aging, TA which is not visible in the initial state may be detected.

Therefore, in the above configuration, after the first TA washing out, the gains of the pair of variable gain amplifiers are switched and set to different values, so that the magnetoresistive characteristics of the pair of MR elements of the DSMR head are changed. Is virtually realized, and in this state, at least the first
The presence / absence of a TA is determined for each sector other than the sector detected by the TA detection according to the result of detection by the TA detection circuit.

In this manner, the common mode noise component elimination function of the differential amplifier reduces the TA due to a projection on the disk which has not been detected even though it has actually occurred.
Detection is performed in a gain setting state in which a change over time in the magnetoresistance characteristics of a pair of MR elements of the MR head is expected.
Therefore, the sector in which this TA is detected, that is, D
When the magnetoresistance characteristic of the SMR head changes over time, TA
By registering, as a defective sector, a sector in which the influence of the occurrence appears on the output of the differential amplifier and may cause a read error, even if the magnetoresistance characteristic of the DSMR head changes over time, TA occurs later. Can be prevented.

Here, the above TA detection is not limited to the case of TA identification for each sector on the disk, but is also performed at the time of a read error in a normal disk read or at the time of a write fault in a desk write. Good.

For example, in the case of a read error, the above 1
TA detection is performed by setting the gain of the pair of variable gain amplifiers to the same value as in the normal state. If TA is not detected, the above-described operation is performed to virtually realize the secular change of the magnetoresistance characteristic of the DSMR head. It is preferable to switch and set the gains of the pair of variable gain amplifiers to different values, and perform the TA detection again. Then, if the read retry is successful and the occurrence of TA is detected, the corresponding sector may be registered as a defective sector. Here, by writing data that can be read from the corresponding sector in the read retry to the replacement sector, the data can be stored in the replacement sector. Even if the magnetoresistive characteristics of the DSMR head change thereafter, the TA Can be prevented from occurring later.

Also, at the time of a write fault, the gain of the pair of variable gain amplifiers is set to the same value as in the normal state, TA detection is performed. If TA is not detected, the magnetoresistive characteristic of the DSMR head is detected. In order to virtually realize the change over time, the gain of the pair of variable gain amplifiers may be switched and set to different values, and TA detection may be performed again. If the occurrence of a TA is detected by the TA detection at the time of this write fault, the corresponding sector may be registered as a defective sector. Here, a substitute sector is assigned to the sector, and write data is written to the substitute sector. Therefore, even if the magnetoresistive characteristics of the DSMR head subsequently change, TA can be prevented from occurring later.

[0016]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a magnetic disk drive according to one embodiment of the present invention.

In FIG. 1, DSMR heads 1-0, 1-1
Is a read-only head, each of which is a pair of MR elements 11
-1, 11-2. Here, a magnetic disk device having a single disk (magnetic recording medium) (not shown) is assumed. The DSMR head 1-0 is provided corresponding to one surface of the disk, and the DSMR head 1- 1 is provided corresponding to the other surface of the disk.

As is well known, the DSMR head 1-i (i = 0, 1) utilizes the fact that the resistance value changes due to an external magnetic field change, and is used to read data magnetically recorded on a disk. Pair of MR elements 11-1 and 11-2
Are connected so as to output differentially. That is, DSM
One end of each of the MR elements 11-1 and 11-2 of the R head 1-i forms a signal terminal for differential output, and the other end is commonly connected to form a common terminal COM. DSMR head 1-i
Of the magnetic recording data by the MR elements 11-1 and 11-1
-2 is supplied by supplying a constant sense current (bias current) to detect data recorded on the disk as a voltage change across the MR elements 11-1 and 11-2 due to a change in the magnetic field. . In FIG. 1, the recording head provided as a pair with the DSMR head 1-i is omitted. The pair of the DSMR head 1-i (reproducing head) and the recording head is called a composite separation type head.

Each of the two signal terminals of the DSMR heads 1-0 and 1-1 is connected to an input of a head selection circuit 21 provided in the head channel 2. Head selection circuit 21
Selects and outputs one of a set of voltage signals (differential output) appearing at two signal terminals of each of the DSMR heads 1-0 and 1-1 according to a head selection signal from the CPU 6 described later.

The two outputs of the head selection circuit 21 are single-ended amplifiers (head amplifiers) 22-1 and 22-1, respectively.
22-2. This head amplifier 22
-1 and 22-2 can be switched and set to arbitrary gains (amplification gains) α1 and α2 in a certain range in an analog manner.

The gain α of the head amplifiers 22-1 and 22-2
1 and α2 are switched and set via D / A (digital / analog) D / A converters 23-1 and 23-2 under the control (head channel control) from the CPU 6. D / A
The converters 23-1 and 23-2 convert the gain switching control data D1 and D2 of a plurality of bits for designating the gain of the head amplifiers 22-1 and 22-2 provided from the CPU 6 into analog data, and The switching setting signal is output to the head amplifiers 22-1 and 22-2.

Outputs of the head amplifiers 22-1 and 22-2 are connected to respective inputs (positive input and negative input) of a differential amplifier (differential amplifier) 24. The head channel 2 has
The above head selection circuit 21, head amplifiers 22-1 and 22
-2, D / A converters 23-1 and 23-2, and differential amplifier 24
In addition, a circuit necessary for outputting a write signal (write current) to the recording head in accordance with write data sent from a write channel (not shown) is provided, but is not directly related to the present invention. Is omitted. In the present embodiment, the head channel 2 is an integrated circuit (I
C).

The read channel 3 is connected to the head channel 2. The read channel 3 receives an analog output (read signal) subjected to differential processing by the differential amplifier 24 in the head channel 2 and performs, for example, a decoding function of performing signal processing necessary for an operation of reproducing data into an NRZ code, and a head. It has a signal processing function for extracting data (burst data) necessary for servo processing such as positioning control. Note that a write channel having an encoding function for performing signal processing necessary for recording data on a disk is omitted.

The TA detection circuit 4 monitors the amplitude of the read signal from the head channel 2 input from the read channel 3 to generate a TA signal containing a fluctuation signal (noise component) of a specified level or more. And outputs a TA detection signal 41 indicating that to the CPU 6. As the TA detection circuit 4, a thermal asperity detector as described in JP-A-9-231505 can be used.

In the present embodiment, the read channel 3,
The write channel and the TA detection circuit 4 constitute a read / write circuit (read / write IC) formed as an IC.

The CPU 6 forms a main control unit for controlling each unit in the magnetic disk drive. The CPU 6 has a flash ROM (Read Only) as a rewritable nonvolatile memory.
Memory) 60 is built-in. A flash ROM (hereinafter, referred to as FROM) 60 stores a control program 61 required for controlling the CPU 6. Also FRO
M60 stores a table (defect management table) 62 in which a list of defective sectors such as a sector in which a TA is detected is registered.

The CPU 6 performs processing for detecting a sector in which a TA occurs in all sectors on the disk (hereinafter referred to as a "sector").
When performing the TA washing process, TA washing with the gains α1 and α2 of the head amplifiers 22-1 and 22-2 matched and TA washing with the gains α1 and α2 different are performed. Is executed twice.

When a read error occurs, the CPU 6 sets T
A TA detection process using the A detection circuit 4 is performed. The feature of this TA detection processing is that TA detection for a sector in which a read error has occurred is performed up to the head amplifiers 22-1 and 22-2.
TA detection with the gains α1 and α2 of
This is to execute TA detection twice in a state where the gains α1 and α2 are different.

When a read error occurs, the CPU 6 retries the read operation (hereinafter, referred to as a first retry). If the first retry fails, the CPU 6 changes the retry condition to execute the read operation. A retry (hereinafter, referred to as a second retry) is performed. Further, when the first retry is successful and the TA is detected in the TA detection processing, the CPU 6 sets the corresponding sector as a defective sector.
An alternative process (reassignment process) for assigning an alternative sector to the sector and writing data read to the alternative sector is performed.

Further, the CPU 6 also performs
The same TA detection processing as that at the time of the read error is performed.
Further, when a TA is detected in the TA detection processing, the CPU 6 performs an alternative processing of assigning an alternative sector to the relevant sector with the corresponding sector as a defective sector.
Is not detected, the write operation is retried (write retry).

An HDC (disk controller) 8 is connected to the CPU 6. The HDC 8 controls communication of commands and data with a host device (not shown), and controls data communication with the read channel 3 (at the time of reading) or the write channel (at the time of writing). The HDC 8 has an ECC function of performing an ECC (error detection and correction) process on the read data decoded by the read channel 3. The HDC 8 stores read / write data in a cache system, for example, a buffer memory (a buffer R
AM) 9 is connected.

Next, a description will be given, with reference to the flowchart of FIG. 2, of the TA washing process of the configuration of FIG. First, it is assumed that the TA washing process is performed under the control of the CPU 6 in response to an instruction from the host device (test device) in the manufacturing stage of the device.

The CPU 6 outputs the same gain switching control data D1 and D2 to the D / A converters 23-1 and 23-2, so that the head amplifiers 22-1 and 22-1 in the head channel 2 are output.
22-2 Gain α1, α2 values at normal use (standard value)
(Step S1). In this case, the gain α1,
The value of α2 is the same value α (α1 = α2 = α) in order to sufficiently exhibit the in-phase noise removal characteristic of the DSMR head 1-i.

Next, the CPU 6 repeats the following control operation for all sectors on the recording surface of the disk. That is, CPU
6 writes test data to the target sector through the DSMR head 1-i and controls the written data to be read out by the DSMR head 1-i (step S2), and normally performs the control without generating a TA. It is determined whether or not the sector is a TA occurrence defect sector based on whether or not the sector has been read (step S3). Specifically, the TA detection circuit 4 detects the occurrence of TA, outputs a TA detection signal 41 indicating the detection, and outputs
When a read error is notified by 8, the CPU 6 determines that a read error in which a TA has occurred in the corresponding sector, that is, the sector is a TA occurrence defect sector. In this case, the CPU 6 assigns one of the alternative sectors secured on the disk to the defective sector, and registers a set of information on the defective sector and information on the alternative sector in the defect management table 62 (step S4).

As described in the section of [Prior Art], the cause of TA is that when a minute projection is present in a sector to be read on a disk, the head (here, a DSMR head) is attached to the projection. The contact (collision) of 1-i) generates heat and disturbs the reproduced waveform of the head.

Now, the MR element 11- of the DSMR head 1-i
The reproduction output signal (read signal) by 1, 11-2 is converted to vMR.
1, vMR2. The reproduction output signals vMR1 and vMR2 from the MR elements 11-1 and 11-2 are supplied to the head selection circuit 2
1 is input to the head amplifiers 22-1 and 22-2. The head amplifiers 22-1 and 22-2 are connected to the MR element 11
-1 and 11-2 are amplified with gains α1 and α2 (here α1 = α2 = α).

In this case, the outputs VMR1 and VMR2 of the head amplifiers 22-1 and 22-2 are as follows: VMR1 = α1 × vMR1 = α × vMR1 VMR2 = α2 × vMR2 = α × vMR2

Output VMR of head amplifiers 22-1 and 22-2
1 and VMR2 are input to the differential amplifier 24. The differential amplifier 24 takes and amplifies the differential output of the outputs VMR1 and VMR2. Here, a minute protrusion exists in the corresponding sector,
Assuming that TA has occurred, vMR1, vMR2
The fluctuation signal (TA component) caused by the occurrence of the TA included in the signal appears as an in-phase signal. Therefore, its vMR1, v
Amplified output VMR1 of MR2 by head amplifiers 22-1 and 22-2 = α1 × vMR1 (= α × vMR2), V
By differentially amplifying MR2 = α2 × vMR2 = (α × vMR2) by the differential amplifier 24, a variation signal (baseline shift) caused by the occurrence of TA due to the common-mode noise component removal function of the differential amplifier 24. Can be canceled (attenuated).

As described above, when TA is generated, that is, when the fluctuating signal due to the occurrence of TA is included in vMR1 and vMR2 and the fluctuating signal is canceled by the differential amplifier 24, TA detection is performed. In circuit 4, T
Not detected as A. That is, the corresponding sector is a sector in which a TA occurs, but is not treated as a TA occurrence defect sector and is not registered in the defect management table 62.

If the apparatus of FIG. 1 is shipped in this state,
Thereafter, if the magnetoresistance characteristics (temperature rise characteristics, heat radiation characteristics) of the MR elements 11-1 and 11-2 change due to aging of the DSMR head 1-i, the common-mode noise removal characteristics cannot be sufficiently exhibited. . In such a case, in the initial state of the apparatus, the TA detection circuit 4 indicates that TA has occurred in a sector that has not been detected as a TA occurrence defect sector.
To be detected. In other words, the TA that is not detected in the initial state occurs later due to the secular change of the magnetoresistive characteristic of the DSMR head 1-i, and causes a read error and the like.

Therefore, in the present embodiment, the head amplifiers 22-1 and 22-2 described above are used so that the sector in which a TA has occurred can be reliably detected in the initial state (TA washout processing).
Following the TA washing (first TA washing) in a state where the gains α1 and α2 of 2-2 are set to the same value (standard value) α, a state where the gains α1 and α2 are shifted (α1 ≠ α)
The TA washing (second TA washing) in 2) is performed.

That is, the CPU 6 controls the head amplifiers 22-1 and 22-2.
With the gains α1 and α2 of 2-2 set to the same value α (step S1), the above steps S2 to S5 are repeated to obtain the first sector for all sectors on the recording surface of the disk.
Of the head amps 22-1 and 22-2
Are set to different predetermined values (α1 ≠ α2) via the D / A converters 23-1 and 23-2 (step S6).

The subsequent operation is the same as the above-described steps S2 to S5. That is, the CPU 6 determines whether or not all sectors on the disk (excluding the sectors detected as the TA-occurring defect sectors in the first TA washout) can be normally read without generating a TA. It is determined whether or not the sector is a TA occurrence defect sector, and the operation of registering the TA occurrence defect sector in the defect management table 62 (steps S7 to S7).
S9) is repeatedly executed (step S10).

Here, the values of the gains α1 and α2 of the head amplifiers 22-1 and 22-2 are set to different values (α1 ≠ α2). If the values of α1 and α2 are α1 = α, α2 = k
If α (k ≠ 1), the outputs VMR1 and VMR2 of the head amplifiers 22-1 and 22-2 are as follows: VMR1 = α1 × vMR1 = α × vMR1 VMR2 = α2 × vMR2 = kα × vMR2 The value of VMR2 in this case is apparently D
This is equivalent to the fact that the magnetoresistive characteristics (temperature rise characteristics, heat radiation characteristics) of the MR element 11-2 of the SMR head 1-i have changed by k times from the initial state. That is, the head amplifiers 22-1 and 22-2
By shifting the values of the gains α1 and α2 of -2, the aging of the MR elements 11-1 and 11-2 of the DSMR head 1-i having different magnetoresistive characteristics (temperature rise characteristics and heat radiation characteristics) is virtually realized. It will be done.

Therefore, by performing TA detection in this state, the gain α of the head amplifiers 22-1 and 22-2 can be obtained even though the sector actually generates TA.
In the first TA washout (in a state where 1, α2 is set to the same value α), it is possible to detect the TA canceled by the common-mode noise component removal characteristic of the differential amplifier 24 and additionally register it in the defect management table 62. Becomes That is, a sector in which a TA may occur later due to aging of the MR elements 11-1 and 11-2 of the DSMR head 1-i can be registered in the defect management table 62 in advance as a TA occurrence defect sector. Becomes

In the above description, test data is written and read in units of one sector.
After writing test data continuously to all sectors, 1
The write data may be read in sector units to determine whether or not the data was read normally without the occurrence of a TA. Further, for simplicity of description, whether or not the sector is a TA-occurring defect sector is determined by writing / reading test data to each sector once, but this is called reliability of the magnetic disk device. In this respect, it is preferable to determine that a sector is a defective TA sector by repeatedly writing / reading test data in each sector a plurality of times and, for example, if a read error accompanied by a TA has occurred even once.

The storage destination of the defect management table 62 is not limited to the FROM 60, but may be a rewritable nonvolatile storage device such as a special area (generally called a system area) on the disk which is not visible to the user. I just need.

Next, the read processing of the configuration of FIG. 1 during normal operation (for example, at the time of user use) will be described with reference to the flowchart of FIG. First, it is assumed that a read command is given from the host device to the magnetic disk device of FIG. 1 and a disk read specified by the read command has been performed (step S11). Here, head amp 2
The gains α1 and α2 of 2-1 and 22-2 are the same value (standard value)
It is set to α. If you can't read normally, HD
When a read error is notified from the C8 to the CPU 6 (step S12), the CPU 6
TA detection processing using the TA detection circuit 4 when the gains α1 and α2 of the head amplifiers 22-1 and 22-2 have the same value α and different values (α1 = α, α2 = kα) (step S13) )I do. However, if the gains α1 and α2 have the same value α, that is, if TA is detected under normal gain conditions, the operation of changing the gains α1 and α2 becomes unnecessary.

The CPU 6 determines that the gains α1 and α2 have the same value α
Or with the gains α1 and α2 changed
When TA is detected by the TA detection circuit 4 (step S14), a TA flag (not shown) indicating this is turned on (set) (step S15), and the process proceeds to step S16. On the other hand, when the TA is not detected by the TA detection circuit 4 even when the gains α1 and α2 are changed (step S14), the CPU 6 proceeds to step S16.

In step S16, the CPU 6 retries (first retries) the read operation for the sector in which the read error occurred first. If the first retry fails (step S17), the CPU 6
Performs a second retry under conditions relaxed from the first retry (determination conditions of retry success) (step S
18). One of the determination conditions for this retry success is HDC
8, the upper limit correction bit length when the read data is corrected in the ECC (error detection and correction) processing is used. In other words, even if a correct value can be corrected by the ECC process, if a correction exceeding the upper limit correction bit length is performed, it is determined that the retry has failed. It is to be noted that this bit length is generally set to a shorter value in the second retry than in the first retry.

If the second retry succeeds, that is, if the first retry fails, but the second retry succeeds under a looser condition, the CPU 6
Assigns a substitute sector on the disk to the corresponding sector and registers it in the defect management table 62, and performs a substitute process (reassign process) for writing data that could be read by the second retry to the substitute sector to perform a series of read processes. To end. If the second retry also fails, the CPU 6 performs a predetermined error process and ends a series of read processes.

On the other hand, if the first retry succeeds, the state of the TA flag is checked to determine whether a TA has occurred in the corresponding sector (step S19). If TA has occurred in the corresponding sector, CP
U6 determines that a TA is fixedly generated in the sector due to the presence of a minute projection or the like, that is, the sector is determined to be a TA generation defective sector, and an alternate sector on the disk is assigned to the sector. An alternative process of writing data that could be read by the first retry to the alternative sector is performed (step S20), and a series of read processes ends. On the other hand, if no TA has occurred in the corresponding sector, the CPU 6 determines that the read error has occurred in the sector due to factors other than the TA, and there is a possibility that the sector can be read normally next time. The substitute process (step S20) is skipped, and the series of read processes ends.

Next, write processing during normal operation of the configuration of FIG. 1 will be described with reference to the flowchart of FIG. First, it is assumed that a write command is given from the host device to the magnetic disk device of FIG. 1 and a disk write specified by the write command has been performed (step S2).
1). If a write fault that cannot perform the write operation occurs for some reason (step S22), C
PU6 is a state in which the DSMR head 1-i (and the composite separation type head including the recording head) is located in the target sector.
TA detection processing using the TA detection circuit 4 (step S2
3) is changed to the gains α1, α of the head amplifiers 22-1, 22-2.
2 is the same value α and different values (α1 = α, α2 = k
It is performed sequentially in the case of α). However, if the gains α1 and α2 have the same value α, that is, if TA is detected under normal gain conditions, the operation of changing the gains α1 and α2 becomes unnecessary.

The CPU 6 determines that the gains α1 and α2 have the same value α.
Or with the gains α1 and α2 changed
When the TA is not detected by the TA detection circuit 4 (step S24), a retry of the write operation is performed (step S25), and a series of write processing ends.

On the other hand, if the TA is detected by the TA detection circuit 4, that is, if the occurrence of TA in the sector to be written is detected (step S24), the CPU 6
Assigns an alternative sector on the disk to the sector and registers it in the defect management table 62, performs an alternative process of writing the designated write data to the alternative sector (step S26), and ends a series of write processes.

In the embodiment described above, a case has been described in which the gains α1 and α2 of the head amplifiers 22-1 and 22-2 are set to different values, but only one of the gains is changed from the standard value α. However, both the gains α1 and α2 may be changed from the standard value α. In consideration of the case where the magnetoresistance characteristics of the MR elements 11-1 and 11-2 of the DSMR head 1-i have already changed, the gains α1 and α2 are
First, a value that satisfies α1> α2 is set, TA detection is performed, and if TA occurrence is not detected, then α1 <
If switching to a value that satisfies α2 is performed and TA detection is performed again, the occurrence of TA can be detected more reliably.

[0057]

As described in detail above, according to the present invention, D
A variable gain amplifier is used as a pair of amplifiers (head amplifiers) for amplifying signals respectively reproduced by a pair of MR elements of the SMR head, and TA is extracted in a state where the gain of the pair of amplifiers is set to a standard value. A state in which the gain of the pair of amplifiers is switched to a different value,
In other words, TA washing is performed in a state where the secular change of the magnetoresistive characteristics of the pair of MR elements is virtually realized, so that the difference amplifier actually removes the TA due to the common-mode noise component removing function. It is possible to reliably detect a TA due to a protrusion or the like on the disk that has not been detected. Therefore, by registering the sector in which the TA is detected as a defective sector, even if the magnetoresistive characteristics of the DSMR head subsequently change over time, it is possible to prevent the TA from occurring later and causing a read error or the like.

According to the present invention, even when a read error occurs in a disk reread or a write fault occurs in a disk write, the corresponding sector is set in a state where the gains of a pair of amplifiers are switched to different values as necessary. Is detected, when a TA actually occurs in the sector, the occurrence of the TA can be detected and registered as a defective sector.
Thereafter, even if the magnetoresistive characteristics of the DSMR head change over time, it is possible to prevent TA from occurring late and causing a read error or the like.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of a magnetic disk drive according to an embodiment of the present invention.

FIG. 2 is an exemplary flowchart for explaining a TA washing process according to the embodiment;

FIG. 3 is a flowchart for explaining an operation at the time of reading in the embodiment.

FIG. 4 is a flowchart for explaining an operation at the time of writing in the embodiment.

[Explanation of symbols]

Reference numerals 1-0, 1-1: DSMR head (double stripe magnetoresistive head) 2: Head channel 3: Read channel 4: TA detection circuit (thermal asperity detection circuit) 6: CPU (thermal asperity generation sector identification means,
Thermal asperity generation sector determination means, read retry means, sector replacement means) 8 HDC (disk controller) 21 head selection circuit 22-1, 22-2 head amplifiers (variable gain amplifiers) 23-1, 23-2 D / A converter 24: differential amplifier (differential amplifier) 60: FROM (flash ROM) 62: defect management table

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G11B 20/18 576 G11B 20/18 576Z

Claims (8)

[Claims] 1. A pair of MRs for reproducing data from a disk
In a magnetic disk drive using a double-stripe magnetoresistive head having an element, gain can be independently switched and set to amplify signals reproduced by the pair of MR elements of the double-stripe magnetoresistive head. A head channel having a pair of variable gain amplifiers, a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers, and a heat channel based on the signals differentially amplified by the differential amplifier. A thermal asperity detection circuit for detecting occurrence of asperity; and means for detecting presence / absence of thermal asperity on the disk by the thermal asperity detection circuit in a state where the gains of the pair of variable gain amplifiers are set to different values. A magnetic disk drive comprising: 2. A pair of MRs for reproducing data from a disk.
In a magnetic disk drive using a double-stripe magnetoresistive head having an element, gain can be independently switched and set to amplify signals reproduced by the pair of MR elements of the double-stripe magnetoresistive head. A head channel having a pair of variable gain amplifiers, a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers, and a heat channel based on the signals differentially amplified by the differential amplifier. A thermal asperity detection circuit for detecting occurrence of asperity; and a thermal asperity detection circuit for determining whether or not thermal asperity is generated for each sector on the disk with the gains of the pair of variable gain amplifiers set to the same value. Judgment is made according to the detection result, and the sector in which the occurrence of thermal asperity is detected is regarded as a defective sector. The first thermal asperity-occurring sector on the disk while the gain of the pair of variable gain amplifiers is set to a different value. The presence or absence of occurrence of thermal asperity is determined for each sector except the detected sector according to the detection result of the thermal asperity detection circuit, and the sector in which thermal asperity occurrence is detected is set as a defective sector, and a second alternative sector is assigned to the sector. A magnetic disk drive comprising: a heat asperity-occurring sector identification means for identifying a heat asperity-occurring sector. 3. A pair of MRs for reproducing data from a disk.
In a magnetic disk drive using a double-stripe magnetoresistive head having an element, gain can be independently switched and set to amplify signals reproduced by the pair of MR elements of the double-stripe magnetoresistive head. A head channel having a pair of variable gain amplifiers, a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers, and a heat channel based on the signals differentially amplified by the differential amplifier. A thermal asperity detection circuit for detecting occurrence of asperity; a read channel for decoding data from a signal differentially amplified by the differential amplifier; and a read error when the read data is not decoded to correct data by the read channel. With the gain of the variable gain amplifier set to the same value, the heat in the sector where the read error has occurred A thermal asperity occurrence sector determining means for determining whether or not asperity has occurred according to a detection result by the thermal asperity detecting circuit; In a state where the gain of the variable gain amplifier is set to be changed to a different value, a thermal asperity occurrence sector determining means for determining whether or not thermal asperity is generated in the sector according to the detection result by the thermal asperity detection circuit; A read retry unit for performing a read operation retry for a sector in which an error has occurred; a read retry by the read retry unit succeeded; and a thermal asperity is generated in a corresponding sector by the thermal asperity occurrence sector determination unit. If it is determined that Magnetic disk apparatus characterized by comprising a sector alternatives for allocating an alternate sector to the sector of the sector as a defective sector. 4. A pair of MRs for reproducing data from a disk
In a magnetic disk drive using a double-stripe magnetoresistive head having an element, gain can be independently switched and set to amplify signals reproduced by the pair of MR elements of the double-stripe magnetoresistive head. A head channel having a pair of variable gain amplifiers, a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers, and a heat channel based on the signals differentially amplified by the differential amplifier. A thermal asperity detection circuit for detecting occurrence of asperity, and a write fault in which the write operation to the disk fails, the gain of the pair of variable gain amplifiers is set to the same value, Thermal asperity which determines whether or not thermal asperity has occurred in accordance with the detection result of the thermal asperity detection circuit. When it is not possible to determine that thermal asperity has occurred in the sector, the sector generating means determines that the sector of the pair of variable gain amplifiers is switched to a different value. Thermal asperity generating sector determining means for determining whether or not thermal asperity has occurred in the sector according to the detection result by the thermal asperity detecting circuit; and if the thermal asperity has occurred in the write faulted sector, the thermal asperity generating sector determining means And a sector replacement means for assigning a replacement sector to the sector as a defective sector when the determination is made by (1). 5. A pair of MRs for reproducing data from a disk.
In a magnetic disk drive using a double-stripe magnetoresistive head having an element, gain can be independently switched and set to amplify signals reproduced by the pair of MR elements of the double-stripe magnetoresistive head. A head channel having a pair of variable gain amplifiers, a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers, and a heat channel based on the signals differentially amplified by the differential amplifier. A thermal asperity detection circuit for detecting occurrence of asperity; a read channel for decoding data from a signal differentially amplified by the differential amplifier; a read error when the data is not decoded to correct data by the read channel; In the case of a write fault in which the write operation of the pair has failed, the gains of the pair of variable gain amplifiers are set to the same value. In a state set to a single value, a thermal asperity occurrence sector determining means for determining whether or not thermal asperity has occurred in the sector in which the read error or write fault has occurred according to a detection result by the thermal asperity detection circuit. If it is not possible to determine that thermal asperity has occurred, the thermal asperity detection is performed by determining whether or not thermal asperity has occurred in the sector with the gain of the pair of variable gain amplifiers switched to a different value. Thermal asperity-occurring sector determining means for determining in accordance with a detection result by a circuit; read retry means for retrying a read operation for a sector in which a read error occurred at the time of the read error; and read retry means for the read error. Successful retry When it is determined by the thermal asperity occurrence sector determining means that thermal asperity has occurred in the corresponding sector, at the time of the write fault, the thermal asperity generating sector determining means determines that thermal asperity has occurred in the corresponding sector. A magnetic disk drive comprising: a sector replacement unit that assigns a replacement sector to the sector when the determination is made that the sector is a defect sector. 6. A method for reproducing data from a disk.
A double-stripe magnetoresistive head having a pair of MR elements; and the pair of Ms of the double-stripe magnetoresistive head.
A pair of variable gain amplifiers for independently amplifying and setting the gain for amplifying the signals reproduced by the R elements, and a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers. A thermal asperity detection method for a magnetic disk drive, comprising: a head channel having a head channel; and a thermal asperity detection circuit for detecting thermal asperity generation based on a signal differentially amplified by the differential amplifier. With the gain of the variable gain amplifier set to the same value, the presence / absence of thermal asperity is determined for each sector on the disk according to the result of detection by the thermal asperity detection circuit, and the sector in which thermal asperity occurrence is detected is determined. A first thermal asperity-occurring sector wash that assigns an alternate sector to the sector as a defective sector After that, in a state where the gains of the pair of variable gain amplifiers are set to different values, at least each of the sectors on the disk except for the sector detected by the first thermal asperity-occurring sector identification is targeted. The presence or absence of occurrence of thermal asperity is determined in accordance with the result of detection by the thermal asperity detection circuit, and a sector in which thermal asperity occurrence has been detected is determined as a defective sector, and a second sector in which an alternative sector is assigned to the sector is identified. A method for detecting thermal asperity, characterized in that: 7. A method for reproducing data from a disk.
A double-stripe magnetoresistive head having a pair of MR elements; and the pair of Ms of the double-stripe magnetoresistive head.
A pair of variable gain amplifiers for independently amplifying and setting the gain for amplifying the signals reproduced by the R elements, and a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers. A head channel, a thermal asperity detecting circuit for detecting the occurrence of thermal asperity based on the signal differentially amplified by the differential amplifier, and a read channel for decoding data from the signal differentially amplified by the differential amplifier A thermal asperity detection method in a magnetic disk device comprising: when a read error has not been decoded to correct data by the read channel, while performing a retry of a read operation for the sector in which the read error occurred, When the gains of the pair of variable gain amplifiers are set to the same value, the read error may occur. The presence / absence of thermal asperity in the sector is determined in accordance with the detection result of the thermal asperity detection circuit. If it is not determined that thermal asperity is occurring in the sector, the asperity of the pair of variable gain amplifiers is further determined. With the gain switched to a different value, the presence or absence of thermal asperity in the sector is determined according to the result of detection by the thermal asperity detection circuit, the read retry succeeds, and thermal asperity occurs in the corresponding sector. A thermal asperity detection method, wherein when it is determined that the sector is defective, the sector is set as a defective sector and an alternative sector is assigned to the sector. 8. A method for reproducing data from a disk.
A double-stripe magnetoresistive head having a pair of MR elements; and the pair of Ms of the double-stripe magnetoresistive head.
A pair of variable gain amplifiers for independently amplifying and setting the gain for amplifying the signals reproduced by the R elements, and a differential amplifier for differentially amplifying both output signals of the pair of variable gain amplifiers. A thermal asperity detection method for a magnetic disk drive, comprising: a head channel having a thermal asperity detection circuit for detecting generation of thermal asperity based on a signal differentially amplified by the differential amplifier. At the time of a write fault in which the write operation has failed, whether or not thermal asperity has occurred in the sector where the write fault has occurred is detected by the thermal asperity detection circuit with the gains of the pair of gain variable amplifiers set to the same value. Judgment according to the result, if it is not determined that thermal asperity has occurred in the sector, Further, in a state where the gains of the pair of gain variable amplifiers are switched and set to different values, the presence or absence of occurrence of thermal asperity in the sector is determined according to the detection result by the thermal asperity detection circuit. A thermal asperity detection method, wherein when it is determined that thermal asperity has occurred, the sector is set as a defective sector and an alternative sector is assigned to the sector.