JPH11273285A - Disk device and its control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correct fluctuation in a position signal detection gain due to minute shift amounts of a head at a reproducing time and a recording time in a combined head and to reduce a difference between loop gains of a control object at the reproducing time and recording time by fluctuating the head with amplitude of large/small two stages, obtaining a ratio between respective gain predictive values and obtaining a change in the position signal detection gain. SOLUTION: A detective position signal 17 is obtained from position reference signals from respective heads 3 with a position detective means 6, an A/D converter 7. When the gain is predicted, a gain identification means 10 obtains respectively the gain predictive values from plural target signals with different amplitude generated by a target position generation means 13 with a control signal 16 and the detective position signal 17, and obtains the position signal detection gain with the ratio between the gain predictive values to store it in a memory 12 through a switch 15. At a positioning control time, the memory 12 selects the position signal detection gain corresponding to target number, off-track position and head number to set it in a detective position signal correction means 18 through the switch 14.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a magnetic disk drive, and more particularly, to a magnetic disk drive suitable for high-density recording.

[0002]

2. Description of the Related Art In a magnetic disk drive, tolerances in parts due to manufacturing tolerances, variations in assembly, changes in the force constant of a voice coil motor due to track movement of a magnetic head, mass, core width and gap of a magnetic head. The loop gain of the head positioning control system fluctuates depending on conditions such as the variation of the position signal detection gain due to the individual difference in width.

Since the position control system deviates from the optimum design due to the fluctuation of the loop gain of the head positioning control system,
When the head of the position control system follows the track, the positioning accuracy of the head is deteriorated.

In order to solve the above problem, a position control system in a magnetic disk drive uses a loop gain between tracks including a position signal detection gain and a loop gain for each head as disclosed in Japanese Patent Application Laid-Open No. 05-26619. Is obtained by the gain identification means, thereby suppressing the variation in the loop gain of the control target.

[0005]

In the conventional head positioning method, a position error signal indicating a positional deviation amount from a track center (hereinafter referred to as an off-track amount) is obtained by using a servo signal read from an MR head. I have. The MR head has the highest sensitivity near the center of the head, and the sensitivity decreases toward both ends. Therefore, the position error signal calculated from a head having sensitivity characteristics such as an MR head does not become a straight line with respect to the off-track amount of the head, and the position signal detection gain changes depending on the off-track amount.

For this reason, the position error signal at the time of off-track includes a possibility of erroneous detection due to a change in the head position signal detection gain. Also, due to the fluctuation of the loop gain of the head positioning control system due to the position signal detection gain,
Since the position control system deviates from the optimum design, there is a problem that the positioning accuracy of the head is lowered when the head of the position control system follows the track.

The position signal detection gain is affected by the flying height of the head and the angle between the head and the track direction, in addition to the off-track amount described above, and these vary depending on the product. Therefore, considering the measurement of the position signal detection gain, the off-track amount obtained from the reproduction signal of the head and the actual off-track amount in a state where the head is incorporated in the magnetic disk drive are required. Is very difficult.

In a composite head such as an MR head in which a recording element and a reproduction element are separated from each other, a head position during reproduction and a head position during recording are determined according to the position of an actuator supporting the composite head. Is shifted. Although the position signal detection gain fluctuates due to this shift, the conventional method only considers the fluctuation of the position signal detection gain due to individual differences such as the core width of the head, and the head position at the time of recording as seen in the composite type head. And the fluctuation of the position signal detection gain due to the shift amount of the head position during reproduction. The change in the loop gain at the time of recording and reproducing the information is caused by the change of the position signal detection gain.

SUMMARY OF THE INVENTION It is an object of the present invention to correct a variation in a position signal detection gain due to a slight displacement of a head between a reproducing time and a recording time in the composite type head, and to reduce a difference between a loop gain to be controlled between a reproducing time and a recording time. An object of the present invention is to provide a disk device for reducing the size and a correction means therefor.

[0010]

In order to achieve the above object, according to the present invention, when the head is fluctuated with two levels of amplitude, the ratio of the respective gain estimation values is obtained to obtain the ratio in the composite type head. In addition, the configuration is such that the change in the position signal detection gain is obtained in consideration of the minute displacement of the head during reproduction and during recording.

[0011]

FIG. 1 is a block diagram showing a head positioning control device of a magnetic disk drive according to one embodiment of the present invention.

In FIG. 1, a plurality of disks 2 are mounted on a shaft driven by a spindle motor 1, and a head 3 is provided corresponding to the disks 2.
The head 3 is supported by a head support system 4, and the head support system 4 is fixed to a voice coil motor 5. With the movement of the voice coil motor 5, the head 3 moves from the outer periphery of the disk 2 to the inner periphery or in the opposite direction, and reads and writes information arranged on tracks of the disk 2.

The magnetic disk drive of this embodiment employs a sector servo system in which position signals discretely arranged on a disk are read by a head 3 and positioning is performed.

The execution procedure of the embodiment of FIG. 1 is divided into a positioning control system and a gain estimating system, and the selection is made by a command generated from a command generator (not shown). The command is issued together with information such as a target head number and a target track number. by this,
The target head number and the target track number are determined.

The position reference signals obtained from the individual heads are converted into position signals by a head 3 and position detecting means 6 having an operation unit for processing the signals from the head 3. The position signal generated by the head position detecting means is converted by an AD (analog-digital) converter 7 to become a detected position signal 17.

When the gain estimation is selected by a command, the switch 14 is closed and the switch 15 is opened. At this time, the gain identification means 10 outputs the control signal 16 and the detected position signal 1
7 is used to measure the loop gain of the control target when the target head is positioned at the target off-track position of the target track. Here, the loop gain to be controlled is a gain from the DA (digital / analog) converter 8 to the power amplifier 9, the voice coil motor 5, the head support system 4, the head 3, the position detecting means 6, and the AD converter 7. And the sampling function. Also, gain identification means 1
A value of 0 determines a gain estimation value from a plurality of target signals having different amplitudes generated by the target position generation means 13, and calculates a position signal detection gain by a ratio of these gain estimation values. The memory 12 stores the track number at which the command has occurred, the off-track position, the head number, and the gain of the detected position signal correcting means 18 based on the position signal detection gain. The amplitude generated by the target position generating means 13 is an amplitude larger than the core width of the reproducing element and an amplitude smaller than the core width of the reproducing element.

When the positioning control is selected by a command, the switch 14 is opened and the switch 15 is closed. At this time, the memory 12 selects the position signal detection gain corresponding to the target track number, the off-track position, and the head number specified by the command.
Set to 8.

The detected position signal correcting means 18 corrects the detected position signal 17 based on the position signal detection gain. The control means 11 generates a control signal 16 so that the deviation between the position signal corrected by the detected position signal correction means 18 and the target position signal generated by the target position generator 13 becomes zero.

The control signal 16 is converted into an analog signal by the DA converter 8, amplified by the power amplifier 9, and a current corresponding to the control signal is input to the voice coil motor 5 to move the voice coil motor 5.

FIG. 3 is a flowchart for explaining the execution procedure of an embodiment of the present invention.

The command generator includes a target position generator 13.
The amplitude of the target position signal at the time of gain estimation is set to a value larger than the core width of the reproducing element (step 300).
The head 3 is varied at the set amplitude, and the gain
At 0, the gain estimation value θh1 is obtained using the gain estimation function (step 301). Subsequently, the command generator sets the amplitude of the target position signal at the time of gain estimation in the target position generating means 13 to a value equal to or less than the core width of the reproducing element (step 302), and changes the head 3 at the set amplitude. Let
A gain estimation value θh2 is obtained by the gain identification means 10 using a gain estimation function (step 303). Gain estimate θ
The ratio θhKp between h1 and the estimated gain value θh2 is obtained and stored in the recording device (memory 12) as the position signal detection gain (step 304).

FIG. 2 shows the relationship between the off-track amount and the head position error.

As shown in FIG. 2, the gain estimation value θh1 at the time of large amplitude does not change depending on the off-track position, and can be approximated to a substantially constant value. On the other hand, the gain estimation value θh2 at the time of the small amplitude greatly fluctuates depending on the off-track position, and this is considered to be a change in the position signal detection gain.
For this reason, it is possible to measure a change in the position signal detection gain by normalizing the gain estimation value θh2 at the time of small amplitude with the gain estimation value θh1 at the time of large amplitude.

The gain estimation values θh1, θh to be measured
2 includes the gain of a mechanism such as a VCM to be controlled, but it can be considered that these gains hardly change in the same head and the same cylinder.
Therefore, the change in the ratio θhKp of the gain estimation values can be considered as it is as a change in the position signal detection gain.

The position signal detection gain θhKp is measured at each of the head positions during recording and reproduction, and the detected position signal correction means 18 that corrects the detected position signal 17 using the values is used for recording and reproduction. Regardless of the position of each head, the position signal detection gain can be made equal. As a result, it is possible to equalize the loop gain of the control target at the time of recording and at the time of reproduction.

The gain identification means 10 has a model of a control object shown in the block diagram of FIG. 1, and performs gain estimation using this model. The procedure of modeling the controlled object shown in the block diagram of FIG. 1 is described below. At the time of modeling, the control target is assumed to be a linear secondary system. This assumption holds for many disk devices by setting the frequency of the target position signal generated at the time of gain estimation to be near or below the zero cross frequency.

In the case of FIG. 1, the modeled discrete transfer function from the control signal u (k) to the detected position signal y (k) is represented by the following equation, where T is the sampling time. Here, z ~ ¹ is a delay operator, and k is the number of steps.

[0028]

(Equation 1)

However, the loop gain θ to be controlled is given by:

[0030]

(Equation 2)

Here, the physical meanings and units of the symbols used in Equation 2 are as follows.

Y (k): sample position signal u (k): sample control signal M: equivalent mass of head and head support system kDA: gain of DA converter kA: gain of power amplifier kF: power constant of voice coil motor kP: gain of the position detecting means kAD: gain of the AD converter Equation 1 can be expressed in the following identification expression.

[0033]

(Equation 3)

However,

[0035]

(Equation 4)

[0036]

(Equation 5)

Next, a method for estimating the loop gain of the control object will be described. The purpose is to estimate θh (k) which is a good approximation (estimated value) of θ from y (k), ζ (k) and η (k). Here, θ is estimated by the least square method. Since the control signal u (k) and the detected position signal y (k) include noise components, the least squares method is effective for advanced estimation of θ. For this purpose, an order identification model is constructed for Equation 3.

[0038]

(Equation 6)

Here, the identifier error ε (k) is determined as follows.

[0040]

(Equation 7)

At this time, the adaptive adjustment rule updates the parameter estimation value θh (k) using the following algorithm.

[0042]

(Equation 8)

[0043]

(Equation 9)

[0044]

(Equation 10)

FIG. 4 is a flowchart for explaining the procedure for calculating the gain estimation by the gain identification function in step 301 or 303 shown in FIG. First, the number of repetitions is set to 0 (step 400), and the number k of steps is set to 0 (step 401). The initial value of the loop gain estimation parameter to be controlled and the updated value of the adjustment rule are determined (step 402).

The detection position signal y (k) is fetched from the AD converter 7, and the control means creates a control signal u (k) (step 403). Here, the target position generator generates a small amplitude target position signal centered on the track designated by the command. The frequency of the target position signal requires that the control target be in a linear frequency band. Normally set to a frequency near or below the zero cross frequency. Here is 200H
Set to z.

Based on the amplitude of the target position signal,
(k) and calculation of η (k) in Equation 5 (Step 404), y in Equation 6
The calculation of h (k) (Step 405) and the calculation of ε (k) in Equation 7 (Step 406) are performed in order. And θh in Equation 8
Calculate (k) to obtain the estimated value of the loop gain of the controlled object,
After updating γ (k) in Equation 9 (Step 407), the step is incremented by one (Step 408).
The calculation from step 3 to step 408 is repeatedly executed.

When the number of steps k is 100 or more, the number of repetitions is increased by one (step 41).
0), the calculations of steps 401 to 410 are repeatedly executed. The maximum value and the minimum value of the obtained five gain estimated values are removed, and the average value is calculated from the remaining three estimated values, and is set as a gain estimated value (step 413).

[0049]

According to the present invention, the correction value of the gain of the detection position signal calculated based on the gain identification function is used.
Even in the case of a composite head in which the recording element and the reproduction element are separated from each other, the gain of the constant detection position error signal can be made constant regardless of the time of recording or reproduction. The loop gain of the control system at the time of reproduction can be equalized, the variation in head positioning accuracy is suppressed,
Servo instability due to gain fluctuation can be prevented. Since the loop gain of the control system at the time of recording and at the time of reproduction can be made equal, the difference between the positional error amounts synchronized with the disk rotated at the time of recording and at the time of reproduction can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram of a control system showing one embodiment of the present invention.

FIG. 2 is a diagram illustrating a relationship between an off-track amount and a position error signal.

FIG. 3 is a flowchart illustrating an execution procedure at the time of gain estimation.

FIG. 4 is a flowchart illustrating a calculation procedure of an identification algorithm.

[Explanation of symbols]

1 ... Spindle motor, 2 ... Disk, 3 ... Head, 4
... head support system, 5 ... voice coil motor, 6 ... position detection means, 7 ... AD converter, 8 ... DA converter, 9 ... power amplifier, 10 ... gain identification means, 11 ... control means, 12 ...
Scale, 13: target position generating means, 14: switch, 1
5 switch, 16 control signal, 17 detection position signal,
18 ... Detection position signal correction means.

 ──────────────────────────────────────────────────の Continuing on the front page (72) Inventor Masato Kobayashi 502, Kandachicho, Tsuchiura-shi, Ibaraki Pref.

Claims (2)

[Claims] 1. A composite type head comprising a disk on which information is recorded, a reproducing element for reading information from and a recording element for writing on the disk, and a detection position for detecting the position of the head. In a disk device comprising: a head detection unit that outputs a signal; and a control unit that generates a control signal for the head and controls the head, the gain from the control signal to the detection position signal when reading information is determined by: A disk device wherein the gain from the control signal to the detection position signal at the time of writing information on the same track and the same head as at the time of reading the information is 0.9 times or more and 1.1 times or less. 2. A step of detecting a position of a head for reproducing or recording information from / to a disk on which information is stored in a track and outputting a detected position signal; and a step of generating a control signal for the head. In the method of controlling a disk device, the step of generating the control signal corrects the detection position signal, and from the respective control signals at the time of reproducing and recording information on the same head and the same track, A method of controlling a disk drive, wherein a difference in gain up to the detection position signal is set to 10% or less of the gain at the time of reproducing information.