JP2953403B2 - Settling determination method using mode separation and head positioning control device using the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a settling determination method using mode separation and a head positioning device using the same, and more particularly, to a magnetic disk drive, which quickly and accurately performs settling determination to reduce seek time. The present invention relates to a head positioning control device that is suitable for

[0002]

2. Description of the Related Art Positioning control of a magnetic disk drive includes an access section in which a magnetic head is moved to a position near a target position at a high speed, a follow section in which a magnetic head is positioned at a target position to read and write data, and an access section and a follow section. And a settling (settling) section which is a transition section.

The voice coil motor (VC)
M) There is a limit to the increase in the current-to-power conversion gain, and there is a limit to the current that can flow through the VCM due to the upper limit of the power supply capacity. Has limitations. That is, it is necessary to shorten the settling time in order to shorten the entire seek time. If the settling judgment is insufficient, there is a risk that the magnetic head will off-track from the target position after judging that the settling is completed, and erroneously read / write data on an adjacent track.

Conventionally, as a criterion for judging the completion of settling, a fixed settling in which settling is completed after a predetermined sample time has elapsed after a position error signal between a magnetic head and a target position first becomes smaller than a certain threshold value. Is widely used.

However, this method is a method in which the worst value of the completion of settling is set in advance as the settling time, which is very conservative and inefficient. Further, since only the time is managed, there is a risk that the settling range is deviated after it is determined that the settling is completed, and that data is read / written to an adjacent track.

As an improvement measure of this method, variable settling (for example, JP-A-2-123576) in which settling is completed only when the position error has become smaller than a certain threshold value more than a set number of times from the start of settling, and There is a method of evaluating the position error and shortening the settling time (for example, Japanese Patent Application Laid-Open No. 5-282814).

However, these methods are more efficient than fixed settling in which the settling time is set in advance, but since only the position error signal is observed, the position error signal changes the settling range after the settling is completed. It may come off.

In recent years, the sector servo system has become mainstream. However, in order to increase the data storage capacity of a disk, the number of servo sectors in which servo information is written cannot be increased in proportion to the increase in track density. As a result, the sampling time has reached a plateau despite the increase in track density. Therefore, in order to shorten the seek time, it is further required to reduce the number of samples from the start of settling to the completion of settling.

[0009]

In the above-described conventional fixed settling, the settling is stopped at a predetermined time in advance. In the conventional variable settling, only the position error is observed and evaluated. After completion, the settling range may be lost, and reading and writing may be performed on an adjacent track.

Also, in the conventional fixed settling, settling is not completed until a certain time elapses.
In addition, in the conventional variable settling, the settling is not completed unless a set number of times elapses after the position error falls within the settling completion range. However, there is a problem that the seek time is extended more than necessary and seek time is extended.

An object of the present invention is to improve the inconvenience of the conventional example, and in particular, to utilize the fact that the behavior of the head positioning mechanism to be controlled can be expressed by superimposition of vibration modes, and to determine the position error from the components of each vibration mode. A settling determination method that calculates a time required for a signal to accurately enter a settling range, performs necessary and sufficient settling, and thereby effectively reduces the time required for positioning a head and the like, and head positioning control using the same It is to provide a device.

[0012]

According to the settling judgment method using mode separation of the present invention, the control target is secondary (for example, the position and speed of a magnetic head in a head positioning device) and the state quantity thereof is observed. Targeted positioning servo mechanism is possible. In such a servo mechanism, the components of the separated vibration modes orthogonal to each other are calculated from the state quantities observed at the start of settling using the mode separation method. Then, the time required to enter the settling completion range is calculated from the components of each vibration mode, and settling is completed with the shortest sample time exceeding this time.

Thus, the components of the vibration modes that are similar to each other and separated by the mode separation method can be calculated from the state quantities of the control system at the start of settling, and the calculated components of each vibration mode can be calculated beforehand at the start of settling. The time to complete settling can be calculated.

In the present invention, when this settling determination method is used in a magnetic head positioning device, the position and speed of the magnetic head are detected, and at the switching point where the position of the magnetic head reaches an arbitrary settling start range, From the position and speed of the magnetic head and the internal state quantity of the servo compensator
Calculate the state quantities of mutually orthogonal vibration modes separated by a matrix orthonormalization method ( mode separation method ) ,
The position of the magnetic head was determined in advance from the results
It is characterized in that a settling time required for stabilization within an arbitrary settling completion range is obtained at the start of settling.

Further, a magnetic head for reading and writing data on one or a plurality of rotating magnetic disks, a drive amplifier for controlling an actuator for positioning and driving the head, and a magnetic head for controlling the position of the magnetic head relative to the data surface A head position detecting means for detecting a position, a head speed detecting means for detecting a seek speed of the magnetic head with respect to the data surface, a target trajectory of the position and speed of the magnetic head being generated and following the target trajectory. Target trajectory generation means for generating a target acceleration trajectory; feedforward compensation means for generating an input signal to a drive amplifier for driving the actuator in advance according to the target trajectory using the target acceleration trajectory; Servo compensation for controlling the drive amplifier so that the position matches the target trajectory A step, mechanical vibration removing means for removing a mechanical vibration component from an output of the servo compensating means, and a position and speed of the magnetic head and a servo compensator at a switching point at which the position of the magnetic head reaches an arbitrary settling start range. Calculate the state quantities of vibration modes orthogonal to each other separated using the mode separation technique from the internal state quantities of
Settling determining means for determining a settling time at the start of settling until the position of the magnetic head is stabilized within the settling completion range from the result.

By using the determination method of the present invention, it is possible to avoid that the position error signal deviates from the settling completion range after the completion of the settling and erroneous information is read or written to the adjacent track. Further, since the time until the completion of the settling can be calculated in advance, it is possible to prevent the settling time from being unnecessarily long, and the time from the start of the seek to the reading and writing of data (seek time) can be reduced.

[0017]

Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. Referring to FIG. 1, a magnetic disk medium (hereinafter, referred to as a magnetic disk) 2 is fixed to a rotating spindle 1, and a read / write operation is performed on a data surface 3 of the magnetic disk 2 via a composite magnetic head 4. The composite magnetic head (hereinafter, referred to as a magnetic head) 4 includes a thin-film inductive head for reading and writing data from and on a data surface on the thin-film magnetic disk 2 and an MR (magnetoresistive effect).
The head and the head are integrated.

Each side of the data surface 3 is divided into sectors for recording data, and each sector is divided into a data sector 5 for reading and writing data and a servo sector 6 for recording servo information. Further, a servo sector signal 9 including a track number and a position signal is recorded in the servo sector 6, a coarse track position in the radial direction is determined by the track number, and an off-track amount from the track center is detected by the position signal. Can be.

The head position detecting means 10 uses the track number and the position signal included in the servo sector signal 9 to determine the position of the magnetic head 4 with respect to the data surface 3 irrespective of the traveling direction of the composite head 4. The head position signal 11 is generated by continuously detecting from the innermost circumference to the outermost circumference.

The head speed detecting means 12 detects the seek speed of the magnetic head 4 with respect to the data surface 3 from the head position signal 11 and the input signal 22 to the drive amplifier 8, and generates a head speed signal 13.

The target trajectory generating means 14 is one of a low-pass filter and a band elimination filter.
And a mathematical approximation model of a VCM actuator (hereinafter referred to as an actuator) 7. The output of the mathematical approximation model of the actuator 7 is output to the target position of the magnetic head 4 in consideration of the effect of the back electromotive force of the actuator 7. The trajectory 15, the target trajectory 16 of the target speed, and the target trajectory 17 of the target acceleration are generated.

The feedforward compensating means 18 uses the target trajectory 17 of the target acceleration and generates an input signal A19 to the drive amplifier 8 for driving the actuator 7 in advance according to the target trajectory. The feed-forward compensating means 18 has means for estimating the loop gain of the actuator 7, and adjusts the internal gain in accordance with the individual difference of the magnetic disk drive.

The servo compensating means 20 is a trajectory-following type input to the drive amplifier 8 for causing the head position signal 11 and the head speed signal 13 to follow the target trajectory 15 at the target position and the target trajectory 16 at the target speed. The signal B21 is generated, and the magnetic head 4 is positioned. The servo compensating means 20 has means for estimating the loop gain of the actuator 7, and adjusts the internal gain according to the individual difference of the magnetic disk drive.

The mechanical resonance removing means 23 receives the input signal A19.
The input signal 22 is obtained by adding the input signal B21 and the input signal B21.
4 is generated.

The settling determining means 25 uses the head position signal 11, the head speed signal 13, the input signal B21, and the internal state quantity of the servo compensating means 20 to set the magnetic head 4.
Of the data plane 3 is set, and a settling determination signal 26 is output to the servo compensating means.

In this embodiment, a case will be described in which the closed loop system of the entire control system including the head position signal 11 and the head speed signal 13 as a part of the internal state quantity is an n-order digital system.

First, in the settling process, if the target trajectory 15 of the target position, the target trajectory 16 of the target speed, and the target trajectory 17 of the target acceleration are set to "0", the head position signal 1
The closed-loop state equation of the entire control system including 1, the head speed signal 13, and the servo compensating means 20 is represented by equation (1), and the output equation is represented by equation (2).

X (k + 1) = A × X (k) (1) Y (k) = C × X (k) (2) where A is a system matrix (n × n), and C is an output matrix (n × l). X indicates the internal state quantity of the closed loop, and Y indicates the output.

Next, a modal matrix is obtained in order to make the matrix A of the equation (1) orthonormal, and this is set to V.
State quantity conversion is performed as shown in equation (3).

X (k) = V · X _m (k) (3) where X _m is an internal state quantity after variable conversion.

Subsequently, equation (3) is substituted into equations (1) and (2), respectively, and both sides of equation (1) are multiplied by the transposed matrix of V to obtain equations (4) and (5). Also, the transposed matrix of V is i
nv (V).

X _m (k + 1) = inv (V) AV · X _m (k) (4) Y (k) = CV · X _m (k) (5) Further, the coefficient matrix is expressed by the following equations (6) and (7). ), The equations (4) and (5) become the equations (8) and (9), respectively.

A _m = inv (V) AV (6) C _m = CV (7) X _m (k + 1) = A _m × X _m (k) (8) Y (k) = C _m × X _m (k (9) Further, since _Am is orthonormalized, the internal state quantity is separated into each vibration mode. For example, i
The next (i <n) mode is indicated as follows.

[0035] _{x im (k + 1) =} a ii · x im (k) (10) However, x _im, a _ii is a state quantity of i-th order vibration mode respectively, the coefficient of i-th row i column of A _m matrix The components are shown.

Here, since the equation (10) is a simple recurrence equation separated into each mode, the settling start time is set to κ = 0, and the initial state quantity of the i-th mode is used. The internal state quantity after k samples from the start of settling can be calculated as in equation (11).

[0037] _{x im (k) = a ii} · x im (0) (11) Therefore, if the known A _m matrix, a sample of how much from the initial state amount at the time of settling the start, the respective vibration modes Whether the response converges can be obtained by calculation at the start of settling. Further, the position error signal is a superposition of these vibration modes, and is represented by Expression (12).

X ₁ (k) = b ₁ (θ) · x _1m (k) + b ₂ (θ) · x _2m (k) +... + B _n (θ) · x _nm (k) (12) b ₁ (θ), b ₂ (θ), ..., b _n (θ) are
Weight indicating transmission characteristics from each vibration mode to the position error signal, and can be calculated in advance. And equation (11),
From (12), x ₁ (k) = b ₁ (θ) · a _11k · x _1m (0) + b ₂ (θ) · a _22k · x _2m (0) +... + B _n (θ) · _annk · x _nm (0) (13), and the position error signal at time k is the initial value of the state amount of each vibration mode can be calculated out with a _i b _i (θ). Then, assuming that the settling completion range is ε ₂ , k that satisfies x ₁ (k) <ε ₂ (14) may be obtained, and this may be set as the settling time.
However, in reality, it is difficult to calculate Expression (13) including the phase due to the limitation of the processing time. Further, since the main vibration mode for extending the settling time is a vibration mode having a pole having the lowest order, the expression (13) is simply expressed as an example only in the vibration mode of the lowest order excluding the phase. , Which are evaluated as in equation (15).

X ₁ (k) = b ₁ · a _11k · x _1m (0) <ε ₂ (15) From the above, by obtaining the minimum k that satisfies the expression (15), an accurate and shortest settling time is obtained. Can be set. Further, since the a ₁₁ b ₁ matrix can be uniquely obtained if the control system is determined, it is calculated in advance and stored in the memory. Moreover, because it can calculate a modal matrix V also A _m matrix, which is also stored in the memory in advance calculated.

Next, the settling judgment operation of the settling judgment means 25 will be described with reference to the flowchart of FIG.

First, the settling determining means 25 captures the head position signal 11 during the head positioning operation (step 1). Then, to calculate whether the head position signal 11 is smaller than any settling start range epsilon ₁ (Step 2).

When the condition of step 2 is satisfied, the head speed signal 13 and the internal state quantity of the servo compensator 20 are fetched (step 3). Here, the internal state of the control system is X = {x ₁ , x ₂ , x ₃ ,..., X _n }. However, x ₁ is the head detection position, x ₂ represents the head detection speed.

Next, the mode-separated internal state quantity is obtained by equation (3) (step 4). Here, the state quantities separated by mode are X = ｛x _1m , x _2m , x _3m ,.
_xnm } and Xm = inv (V) X. Where V
Denotes a modal matrix, and inv (V) denotes a transposed matrix.

Subsequently, a sample time at which the position error signal becomes smaller than an arbitrary settling completion range ε ₂ is calculated from the state equation of the position error signal of the state quantity (step 5).
Here, the state equation of the position error signal is x ₁ = b ₁ a ₁₁ ^k x
_1m (0) +... + B _{n ann} ^k x _nm (0). Here, a _ii is the i-th component of the system matrix after the mode separation, and b _i is a weight coefficient indicating a transfer characteristic from each vibration mode to the position error signal.

Then, the obtained sample time is output to the servo compensating means 20 (step 6). Accordingly, the state equation of the position error signal can be appropriately selected depending on which vibration mode is evaluated, such as the lowest-order slower-pole vibration mode or the largest amplitude vibration mode.

Next, the effect of the settling judgment method will be described with reference to FIGS.

First, FIG. 3 shows a settling judgment method using a conventional variable settling. In the variable settling,
Settling is completed only when the position error from the start of settling becomes smaller than a certain threshold value by a set number of times or more. In this example, the set number is four, and it is necessary to wait until the settling of four samples is completed even though the position error signal falls within the threshold value.

Also, the set number must be determined by trial and error. If the set number is not enough, there is a risk that the position error signal will deviate from the settling range after the completion of the settling. If this happens, the settling time is extended.

Next, FIG. 4 shows a settling judgment method proposed in the present invention. In this determination method, since the time during which the position error signal falls within the settling range and does not deviate thereafter is known, if this time is regarded as the completion of settling, the settling time can be reduced.

[0050]

As described above, the settling judgment method using mode separation of the present invention separates each vibration mode and uses the components of the vibration mode directly in the judgment of the settling. A necessary and sufficient settling time at which the error signal does not deviate from the settling completion range can be set at the start of the settling.

This also has the effect of preventing data destruction of adjacent tracks and shortening the settling time.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a flowchart showing a flow of processing of a settling determination unit of FIG. 1;

FIG. 3 is an operation diagram for explaining a relationship between a position error signal at the time of settling and a settling determination flag according to a conventional technique.

FIG. 4 is an operation diagram for explaining a relationship between a position error signal during settling and a settling determination flag according to the present invention.

[Explanation of symbols]

 Reference Signs List 1 rotating spindle 2 magnetic disk (thin film magnetic disk) 3 data surface 4 magnetic head (composite magnetic head) 5 data sector 6 servo sector 7 actuator (CM actuator) 8 drive amplifier 9 servo sector signal 10 head position detecting means 11 head position signal 12 Head speed detecting means 13 Head speed signal 14 Target trajectory generating means 15 Target trajectory of target position 16 Target trajectory of target speed 17 Target trajectory of target acceleration 18 Feedforward compensating means 19 Input signal A 20 Servo compensating means 21 Input signal B 22 Input Signal 23 Mechanical resonance removal means 24 Drive signal 25 Settling determination means 26 Settling determination signal

Claims (6)

(57) [Claims] 1. A settling determination method for a positioning servo mechanism in which a control target is secondary and its state quantity is observable, wherein at a switching point when an output of the servo mechanism reaches an arbitrary settling start range, The state quantities of mutually orthogonal vibration modes separated by a matrix orthonormalization technique (mode separation technique) are calculated from the internal state quantities of the servo mechanism, and the output of the servo mechanism falls within the settling completion range from the result. A settling determination method using mode separation, wherein a settling time until stabilization is obtained. 2. A settling determination method for controlling positioning of a magnetic head at a target position, comprising detecting a position and a speed of the magnetic head, and setting a switching point at which the position of the magnetic head reaches an arbitrary settling start range. Then, from the position and velocity of the magnetic head and the internal state quantities of the servo compensator, the state quantities of mutually orthogonal vibration modes separated by a matrix orthonormalization technique ( mode separation technique ) are calculated. The position of the magnetic head is determined in advance.
A settling determination method comprising: determining a settling time required for stabilization within a given settling completion range at the start of settling. 3. A magnetic head for reading and writing data on one or a plurality of rotating magnetic disks, a drive amplifier for controlling an actuator for positioning and driving the head, and a magnetic head for the data surface of the magnetic head. A head position detecting means for detecting a position, a head speed detecting means for detecting a seek speed of the magnetic head with respect to the data surface, a target trajectory of the position and speed of the magnetic head being generated and following the target trajectory. Target trajectory generating means for generating a target acceleration trajectory; feedforward compensation means for generating an input signal to a drive amplifier for driving the actuator in advance according to the target trajectory using the target acceleration trajectory; Servo compensation means for controlling the drive amplifier so that the position matches the target trajectory; Mechanical vibration removing means for removing a mechanical vibration component from the output of the servo compensating means, and the position and speed of the magnetic head and the internal state of the servo compensator at a switching point where the position of the magnetic head reaches an arbitrary settling start range. Calculate the state quantities of vibration modes orthogonal to each other separated from each other by using the mode separation method from the quantity, and determine the settling time until the position of the magnetic head stabilizes within the settling completion range at the start of settling from the result. A head positioning control device comprising: a settling determination unit. 4. The target trajectory generating means comprises one of a low-pass filter and a band elimination filter and a mathematical approximation model of an actuator, and outputs an output of the mathematical approximation model of the actuator to a target position of the magnetic head. 4. The head positioning control device according to claim 3, wherein the target trajectory, the target trajectory of the target speed, and the target trajectory of the target acceleration are generated. 5. The head according to claim 3, wherein the feedforward compensator has a function of estimating a loop gain of the actuator, and adjusts an internal gain in accordance with an individual difference of the magnetic disk drive. Positioning control device. 6. The servo compensating means is of a trajectory following type, has a function of estimating a loop gain of the actuator, and adjusts an internal gain according to an individual difference of a magnetic disk drive. Item 4. A head positioning control device according to Item 3.