JP2000306348A - Magnetic disk drive and head position signal correction method - Google Patents

Abstract

(57) Abstract: In a magnetic disk drive, a method for generating a position signal having good linearity at all radial positions on a disk 11 is provided. A relational expression g (x) = dy / dx of a gain (g (x)) of a positioning control system with respect to a head position (x) is measured, and the following expression y = is obtained by integrating g with respect to x. ∫g (x) dx = f (x) is calculated, and information for calculating the inverse function f ⁻¹ of f is stored in the non-linear information storage unit 36, and the information is used every time sampling is performed. By accurately calculating the accurate head position (x) from the position signal (y), the head position is detected with high accuracy and the head is positioned.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a head position of a magnetic disk drive, and more particularly to a magnetic disk drive which generates a position signal having good linearity with respect to the head position and positions the head based on the position signal. It is about.

[0002]

2. Description of the Related Art In a magnetic disk drive, when reading or writing data information, it is necessary to perform an operation of positioning a head with respect to a track formed concentrically on a disk with high accuracy.

Usually, in a magnetic disk device, position information is recorded on a disk in advance by a device generally called a servo track writer so that the position of a head on a disk surface can be specified. FIG. 2 shows an example of the position information recorded on the disc. FIG. 2A is a schematic view of a disk surface of a sector servo system, and FIG.
It is the schematic of the position information recording area which expanded M part of FIG. 2 (A).

The position information is usually composed of a track number section 48, an A burst section 42, a B burst section 43, a C burst section 44, and a D burst section 45 as shown in FIG. The head position is determined by the track number reproduced when the head crosses the track number section 48, the N signal defined as the output difference between the A burst and the B burst when the head crosses each burst section, or the output of the C burst and the D burst. Determined by the Q signal defined as the difference.

For example, at the position of the reproducing head 41 in FIG. 2B, the track number = J, and the output waveform when the head crosses each burst portion is as shown in FIG. 2C.
Next, when each waveform is input to a peak hold (P / H) circuit, an output waveform as shown in FIG. 2D is obtained. Therefore, at the position of the reproducing head 41 in FIG. 2B, since the outputs (P / H values) of the A burst and the B burst are equal, the N signal = 0, and the head does not cross the D burst portion.
Since the burst output (P / H value) = 0, the Q signal =
This is the output (P / H value) of the C burst.

Next, the switching of the position signal and the relationship between the N signal, the Q signal and the head position will be described in detail. FIG.
(A) shows the track number and N signal for the head position,
FIG. 5 is a diagram showing a relationship between the Q signal and the Q signal. Reproduction head 41
2B is gradually moved downward from the position shown in FIG. 2B, for example, in FIG. 2B, the A burst output increases and the B burst output decreases.
The signal gradually increases with the movement of the head. When the reproducing head 41 is further moved, the width of the reproducing head 41 is smaller than the track pitch.
Is completely within the A burst width and is not affected by the B burst.

In a recent magnetic disk drive, an MR (Magn) used for the purpose of improving the reproduction output of the head is used.
An eto-Resistive (magnetoresistive) head has high reproduction sensitivity, but its sensitivity distribution is trapezoidal or triangular in the track direction (disc radial direction).
The reproduction output is high near the center of the width of the reproducing head 41 and becomes lower toward the end of the width of the reproducing head 41.

As described above, the N signal has a small signal change with respect to a change in the head position near the position where the center of the reproducing head 41 coincides with the track boundary line 47,
The sine wave output shown in FIG. FIG.
As shown in (B), C burst and D burst are
A burst and B burst are halved in the disk radial direction.
Since the Q signal is recorded at a position shifted by the track pitch, the Q signal has a shape obtained by translating the N signal in the head position direction by a half track pitch.

In general, in a magnetic disk drive, the head position and the position signal of the above-described N signal and Q signal are so arranged that the head position and the position signal correspond one-on-one continuously at all radial positions on the disk. A range in which the change of the position signal with respect to the change is large is selected and connected. This method will be described below with reference to FIGS.

First, the absolute value of the N signal and the absolute value of the Q signal are compared, and a signal having a small absolute value is selected. At this time, the sign is inverted as necessary so that the signal becomes a straight line rising to the right to create the Ni signal and the Qi signal (FIG. 3).
(B)). Here, the point at which the absolute value of the N signal is equal to the absolute value of the Q signal is the switching point of the position signal. If the absolute values of the N signal and the Q signal at the switching point are normalized so as to be 1/4, the position signal is normalized by the track pitch, and the following description is simplified. Shows a signal standardized as described above.

Next, as shown in FIG. 3, the track number increases within the same range so that the position signal increases in response to the increase in the head position.
A Qc signal 51 is created by connecting the position signals by offsetting the Qi signal of (B) by １ ／ or − ／. As a result, a signal as shown in FIG. 3C is created.

Further, by adding the track number reproduced from the track number portion 48 shown in FIG. 2B and the signal shown in FIG. 3C, a continuous position signal (not shown) is created.

However, since the N signal and the Q signal have a sine-wave shape, the position signal generated by the above-described method has a linear line with respect to the head position as shown in FIG. It will come off. That is, the relationship between the head position and the position signal has a non-linear property having a periodicity that repeats at a 1/2 track pitch.

Since the MR head employed in recent magnetic disk devices is a read-only element, a recording / reproducing separation head is constructed in combination with an inductive magnetic head for recording. Since the center position of the upper reproducing head is shifted from the center position of the recording head, it is necessary to offset the reproducing head to a different position during recording and during reproduction.

In the conventional magnetic disk drive, since the head position and the non-linear amount of the position signal are not so large, positioning control has been performed by regarding the relationship between the two as being linear. Accordingly, when the track pitch and the width of the reproducing head 41 are reduced, the nonuniformity of the reproduction sensitivity distribution in the width of the reproducing head 41 is relatively increased, and the non-linear amount of the head position and the position signal is increased. Further, the offset amount for positioning the reproducing head at different positions during recording and reproducing relatively increases. Therefore, the need to accurately position the head at an arbitrary position on the track increases, and in order to realize this,
It is necessary to ensure linearity between the head position and the position signal at all radial positions on the disk.

It is also important to ensure the linearity of the position signal from the viewpoint of suppressing the gain fluctuation of the head positioning control system.

For example, in order to determine the nonlinear amount of the position signal, the relationship between the accurate position of the head which can be detected from the position information normally recorded on the disk and the position signal at that position is measured. From this relationship, an approximate curve is set to estimate the off-track position. Alternatively, there is a method in which a measurement track on which a pattern capable of detecting the accurate position of the head at more points is recorded is provided on a disk, and a more accurate approximate curve is set to measure and correct the nonlinear amount. It is disclosed in JP-A-11-7738.

In the latter method, the pattern in which the accurate position of the head can be detected at more points is shown in FIG.
For example, instead of the A and B burst portions and the C and D burst portions recorded with a shift of 1/2 track pitch in the disk radial direction with respect to the A and B burst portions, for example, 1/4 in the disk radial direction. The C and D burst portions are recorded at positions shifted by the track pitch.

[0019]

In the former method of obtaining an approximate curve from an accurate position of a head which can be detected from a position signal recorded on a normal disk and a relationship between the position signal at that position, Within the range of 1/2 track pitch, only three points which are the boundaries of bursts are disclosed.
Or the exact position of the head can be determined only at R4, R1 or R3), so that the nonlinear amount could not be corrected much depending on the position of the head even if an approximate curve was obtained.

In the latter method, it is necessary to add an extra manufacturing step such as recording a special pattern for measuring the nonlinear amount of the position signal on the disk. Further, the measurement of the non-linear amount is limited to a measurement track position where a special pattern is recorded.

An object of the present invention is to measure a non-linear amount at many positions without making any special changes to the hardware, using only the functions of the magnetic disk drive, and to calculate a new position by correcting the non-linear amount. The purpose is to create a signal and accurately position the head at any radial position on the disk.

[0022]

In order to achieve the above object, the present invention measures the head position (x) by measuring the gain (g (x)) of the positioning control system with respect to the head position (x). The function of obtaining the relationship (y = f (x)) of the position signal (y) and converting the observed position signal (y) into an accurate head position (x) is added to the apparatus.

For example,

[0024]

The following equation obtained by integrating the relational expression of g (x) = dy / dx (Equation 1) with respect to x

[0025]

Y = ∫g (x) dx = f (x) (Equation 2) is obtained, and information for executing the operation of the inverse function f ⁻¹ of f is stored, and based on this information, In addition, a function for correcting the non-linearity of the position signal is added to the device by calculating the accurate head position (x) from the position signal (y) every time sampling is performed.

[0026]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing the basic configuration of a positioning control system of a sector servo type magnetic disk drive using two-phase servo signals according to the present invention.

In this magnetic disk drive, when a recording or reproduction command is sent from the central control unit 1 (CPU) to the disk controller 2, the disk controller 2 controls the microprocessor 9 to position the head 9 at the target position 37. Issue a command to 3. The position information of the head 9 is created by reproducing a signal from the position information recording area 12 by the head 9. The position information is shown in FIG.
The signal obtained when the head 9 crosses the track number section 48 is supplied to a read circuit 17 via a read / write (R / W) IC 15 and an automatic gain control (AGC) amplifier 16.
After being converted into a track number by a track number reproducing circuit 35, the track number is taken into the microprocessor 3. The pattern that is position information for positioning is 2
Although an example of the phase is shown, it is needless to say that two or more phases may be used.

The A burst section 42 and the B burst section 4
3, signals obtained when the signal crosses the C burst section 44 and the D burst section 45 are read / write (R / W) IC1
5, peak hold (P / H) circuit (or integration circuit) 19 via AGC amplifier 16 and full-wave rectifier circuit 18
, And are reproduced (demodulated) as an A signal 23, a B signal 24, a C signal 25, and a D signal 26, respectively. Therefore, the (R / W) IC 15, the AGC amplifier 16, the full-wave rectifier circuit 18, and the P / H circuits (or integration circuits) 19 to 22 are collectively referred to as demodulation means.

A signal 23, B signal 24, C signal 2
The 5 and D signals 26 are sent to operational amplifiers 33 and 34, where two-phase position signal N signal 27 and Q signal 28 are reproduced. The N signal 27 and the Q signal 28 are output from the A / D converter 2
9 and is taken into the microprocessor 3.
The position information is recorded in advance on the disk 11 at regular intervals as shown in FIG. 2A, and the track number, the N signal 27 and the Q signal 28 are sampled at regular intervals.

In this embodiment, the arithmetic circuit and the A / D converter are used on the downstream side of the full-wave rectifier circuit of the demodulation means. However, the output of the AGC amplifier 16 is directly sampled.
An LSI for creating the N signal 27 and the Q signal 28 by digital operation has also been put to practical use, and such an LSI may be used.

Next, in the microprocessor 3, the position signal synthesizing unit 30 combines the N signal, the Q signal and the track number so that the head position and the position signal correspond one-to-one at all radial positions on the disk. FIG.
A signal as shown in (C) is created, and the nonlinear correction execution unit 31
Then, a correction calculation is performed so that the position signal becomes linear with respect to the head position, and the head position is specified. This correction calculation is
The information is executed every time sampling is performed, but information necessary for correction calculation is created at a certain timing (for example, one manufacturing process before shipment from a factory) by a method described later and stored in the nonlinear correction information storage unit 36. .

In order for the head 9 to follow the target position 37 commanded by the disk controller 2, the digital filter 32 is used to determine the difference between the current head position and the target position 37 from the error position so that the difference becomes zero. A so-called feedback control for generating an operation amount up to the target position is performed. The operation amount 38 is converted into an analog signal by the D / A converter 4, and then the VCM that drives the voice coil motor 6.
The head 9 sent to the drive circuit 5 and supported by the head support system 8 is positioned. When the head 9 is positioned, the write circuit 14 or the read circuit 17
Thus, the head 9 records or reproduces data on the track of the disk 11.

The outline of the disk drive according to the present invention has been described above. Next, the non-linear correction method which is the center of the present invention will be described in detail with reference to FIG.

FIG. 4 shows the D / A converter 4 in FIG.
To A / D converter 29 and track number reproducing circuit 3
5 are arranged as a control object 61, the microprocessor 3 which is a main part of the positioning control is arranged at the center, and the configuration of the nonlinear amount measuring unit 64 is described.
The nonlinear amount measuring unit 64 includes a positioning control system open loop gain measuring unit 62 and a position signal nonlinear amount parameter determining unit 63 that determines a parameter of Expression 1 described later. Each block in the microprocessor 3 is
This is a function of a program to be executed by the microprocessor 3, and there is no need to add special hardware.

As described above, since the position signal in the magnetic disk device has a periodicity that repeats at a 1/2 track pitch, it can be Fourier series expanded. That is,
The relational expression between the head position (x) and the position signal (y) is x
And y are normalized by the track pitch,

[0036]

Y = x + Asin (4πx + α) + Bsin (8πx + β) +... = F (x) (Expression 3)

Next, an equation obtained by differentiating Equation (3) with respect to x,

[0038]

(Equation 4) think of. Dy / dx in (Equation 4) can be measured as a gain variation of the open loop characteristic of the head positioning control system with respect to the head position. Specifically, the head is positioned at a measurement position within the track pitch by offsetting the target position 37 by a fixed value, and at this position, the position control system open loop gain measurement unit 62 shown in FIG. A vibration signal 65 is applied to vibrate the head. The measurement is also performed by measuring the signals before and after the point where the excitation signal is applied, and calculating the amplitude ratio of the excitation frequency components of the measurement signal 66 and the measurement signal 67. Next, parameters A, α, B, β,... Are determined in the position signal nonlinearity parameter determination unit 63 by subjecting the relationship of gain fluctuation to Fourier series expansion.

By the above processing, a function f of the head position (x) and the position signal (y) is obtained, and therefore, an inverse function f ^{−1 of} f is obtained.
Is used, the detected position signal (y) can be converted to an accurate head position (x). Specifically, the parameters A, α, B, β,... Are determined at a certain timing (for example, one manufacturing process before shipment from the factory), and the non-linear correction information storage unit 36 (for example, a predetermined area of the disk or By storing the information in a non-volatile memory such as a flash memory, the nonlinear correction execution unit 31 can calculate f ⁻¹ , that is, correct the head position signal using this information.

In order to reduce the processing load on the microprocessor 3, an approximate expression of the inverse function f ^-1 and a table for executing a linear interpolation operation are prepared and stored in the nonlinear correction information storage section 36. The correction calculation based on the information may be executed by the non-linear correction execution unit 31.

As described above, since the amount of nonlinearity of the position signal at any and many positions can be measured by using the gain fluctuation, a more accurate approximation formula is applied to the case where the nonlinearity is corrected using the approximation formula. can do.
In addition, since accurate non-linear correction can be performed at each position, more accurate head positioning control can be performed.

Since the nonlinearity of the position signal depends on the characteristics of the MR head, if the magnetic disk drive has a plurality of heads, nonlinear correction information is created and stored for each head.

Further, since the non-linearity of the position signal is affected by the angle (yaw angle) between the head and the track, the non-linear correction information is created and stored in, for example, the inner, middle and outer tracks of the disk. The calculation for interpolating the information is executed by the nonlinear correction execution unit 31.

When measuring the gain variation with respect to the head position as described above, the target position 37 is offset by a constant value. However, due to the non-linearity of the position signal, the moving width of the head is different from the constant value. However, sufficient correction accuracy may not be obtained by only the above-described method. As a countermeasure against this, it is effective to add a function of repeatedly estimating the parameters in Equation 1 to the nonlinear quantity measuring unit 64. Specifically, head positioning control is performed using the inverse function f _i-1 ^-1 of the function f _i-1 of the head position (x) and the position signal (y) determined by the repetition up to the ( _i-1 ) _th time. while the gain variation is measured, this relationship of the gain variation parameter obtained by Fourier series expansion _{a i, α i, B i} , β i,
, And the relational expression between the head position (x) and the position signal (y) when the gain variation is repeatedly measured I times is

[0045]

F ₀ (x) = x (Equation 5)

[0046]

(Equation 6)

[0047]

(Equation 7)

[0048]

## EQU8 ## This is a function where y = f (x) = f _I (x) (Expression 8).

As another method for improving the correction accuracy, a function of finely adjusting each parameter centering on the first measurement value and searching for a combination that minimizes the gain variation is provided by a function of the non-linear amount measuring section 64. Is also effective. Specifically, the parameters A, α,
B, β,... Are A + ΔA, α + Δα, B + ΔB, β +
Δβ,..., The gain variation when the gain variation was measured, and the parameter A + Δ when the measured gain variation was minimized
A, α + Δα, B + ΔB, β + Δβ are replaced by A, α,
.., B, β,.

FIG. 5A shows the method presented in the present invention.
(B) shows an experimental result when the gain fluctuation is measured by the method according to the related art. In both figures, the horizontal axis represents the head position normalized by the track pitch, and the vertical axis represents the gain variation, which is obtained by normalizing the gain measurement value by the average value of the measurement values, and shows the gain fluctuation. According to the method, the gain variation is small over the entire area within the track pitch,
That is, it is understood that a linear position signal is obtained with respect to the head position.

Since the above-described nonlinear correction method is mainly a function of a program executed by a microprocessor, it may be sufficient to simply rewrite a program of a microprocessor of an existing product.

[0052]

According to this embodiment, the linearity of the position signal is improved at all radial positions on the disk, and the head can be accurately positioned at any position in the disk radial direction. .

[Brief description of the drawings]

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

FIG. 2A is a conceptual diagram showing a position information recording area on a disk surface. (B) is a conceptual diagram showing a burst pattern recorded on a disc. (C) A conceptual diagram of a burst pattern reproduction waveform. (D) is a conceptual diagram of a burst pattern peak hold waveform.

FIG. 3A is a conceptual diagram showing a relationship between a head position and an N signal and a Q signal. FIG. 3B is a conceptual diagram illustrating the selection and inversion operations of the N signal and the Q signal. (C) It is a key map showing the joining method of a position signal.

FIG. 4 is a schematic diagram of a magnetic head positioning control system according to an embodiment of the present invention.

FIG. 5 (A) is a measurement example of a head positioning control system gain variation of the apparatus of the present invention. (B) is a measurement example of a head positioning control system gain variation of a conventional device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... CPU, 2 ... Disk controller, 3 ... Microprocessor, 4 ... DA converter, 5 ... Voice coil motor drive circuit, 6 ... Voice coil motor, 7 ... Carriage rotation axis, 8 ... Head support system, 9 ... Head, 10 ... Spindle motor drive shaft, 11 ... Disk, 12 ... Position information storage area, 13 ... R / W (read / write) IC, 14 ...
Write circuit, 15: Read / write IC, 16: Automatic gain adjustment (auto gain control) amplifier, 17:
Read circuit, 18 ... full-wave rectifier circuit, 19, 20, 2
1, 22 ... peak hold circuit, 23 ... A signal, 24 ...
B signal, 25 ... C signal, 26 ... D signal, 27 ... N signal,
28 Q signal, 29 AD converter, 30 Position signal synthesizing unit, 31 Non-linear correction execution unit, 32 Digital filter, 33, 34 Operational amplifier, 35 Track number reproducing circuit, 36 Non-linear correction information storage Unit, 37 target position, 38 operation amount, 41 reproducing head, 42 A burst unit, 43 B burst unit, 44 C burst unit, 4
5 D burst portion, 46 track center, 47 track boundary line, 48 track number portion, 51 Qc signal,
Reference numeral 61: control target, 62: positioning control system open loop gain measurement unit, 63: position signal nonlinear amount parameter determination unit, 64: nonlinear amount measurement unit, 65: excitation signal, 66: measurement signal, 67: measurement signal.

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Hirosuke Seo 2880 Kozu, Kozuhara, Kanagawa Prefecture Inside the Hitachi, Ltd.Storage Systems Division (72) Inventor Kazuhisa Shishida 2880 Kozu, Kozu, Hitachi, Ltd.Storage Corporation Hitachi, Ltd. F-term in the System Division (Reference)

Claims (4)

[Claims] 1. A magnetic head, a magnetic disk on which a pattern for positioning the magnetic head is recorded, means for demodulating position information from the pattern, and means for calculating the position information to generate a position signal A magnetic disk drive having positioning control means for controlling the positioning of the magnetic head based on the position signal; a measuring unit for measuring a gain fluctuation of the magnetic head positioning control system with respect to the magnetic head position; A magnetic disk device comprising: a calculation unit for calculating a parameter for correcting a nonlinearity of a position signal based on a value; and a correction unit for correcting a position signal of the magnetic head based on the parameter. 2. The magnetic disk drive according to claim 1, wherein the calculation of the parameter is executed based on a result of repeatedly executing the gain fluctuation measurement by the measurement unit. 3. A magnetic head, a magnetic disk on which a pattern for positioning the magnetic head is recorded, means for demodulating position information from the pattern, and means for calculating the position information to generate a position signal. A magnetic disk drive having positioning control means for controlling the positioning of the magnetic head based on the position signal; a measuring unit for measuring a gain fluctuation of the magnetic head positioning control system with respect to the magnetic head position; A magnetic disk drive that creates a relational expression between the head position of the magnetic head and the position signal based on the value and corrects the position signal of the magnetic head. 4. A head is positioned by offsetting a target position, a vibration signal is applied at that position to vibrate the head, and a vibration frequency component of a signal before and after a point at which the vibration signal is applied. Calculate the amplitude ratio, measure the gain variation of the head positioning control system with respect to the head position from this calculation,
A head which obtains a parameter by Fourier series expansion of the relationship between the gain fluctuation and the head position, determines a function between the head position and the position signal from the parameter, and corrects the position signal using an inverse function of the function. How to correct the position signal.