JPS63173274A - Data recording system - Google Patents

Abstract

PURPOSE:To surely detect a start part of a unit data area of a recording medium with high accuracy by constituting a mark data of a code having run-length which can not exist in a prescribed run-length limited code. CONSTITUTION:A mark data consists of the first data part and its following second data part, the first data part is constituted of a repetition of a code having run-length which does not exist in a prescribed run-length limited code, and the second data part is the same bit pattern as a xynchronizing data which follows thereafter. The first data part is constituted of a repetition of a code having run-length which does not exist in the prescribed run-length limited code, and the second data part records a mark data being the same bit pattern as the synchronizing data which follows thereafter, in a start part of a unit data area of a recording medium. In such a way, the start part of the unit data area like a sector of the recording medium can be detected surely with high accuracy.

Description

[Detailed description of the invention] The present invention will be explained in the following order.

A. Field of industrial application B. Overview of the invention C. Prior art D. Problems to be solved by the invention E. Means for solving the problems (Figures 1 and 2) F. Effect G. Example G1 Recording formant (Fig. 1) Format of G2 address mark (Fig. 2) Detection of G3 address mark (Fig. 3) Rewriting of G41D field (Fig. 4) Effects of the invention A Industrial application field The present invention is directed to magnetic disks, The present invention relates to a data recording method suitable for use in data recording on magnetic tapes and the like.

B. Invention) Ready-made The present invention sequentially records synchronization data, control data, and main data, all of which are composed of a predetermined run-length limited code, in a unit data area of a recording medium, and also records the synchronization data, control data, and main data, which are all composed of a predetermined run-length limited code, in a unit data area of a recording medium, and In a data recording method that records mark data, the mark data is composed of a code with a run length that cannot exist in a predetermined run length limited code, so that the start part of the unit data area of the recording medium is raised. This allows reliable detection with accuracy.

C. Prior Art First, with reference to FIG. 6, 5, called 5T-506,
The recording format of a 2-inch hard disk will be explained. This explanation is based on the May 1984 issue of the magazine "Interface J."

In FIG. 6, the average track capacity is 10,416 bytes. The number in parentheses represents the byte length. The part in m represents a hexadecimal bit pattern. When bit 7 of the head byte in the ID field, which will be described later, is 1, it indicates a bad sector. However, cylinder O is error free. Total data byte length per track is 8
It is 192 (=256×32) bytes.

Gap 1 (Index gap) Only one gap is provided for each rotation of the hard disk. This gap is provided to absorb the deviation of the index signal, and has a length of 16 bytes and has "4E" written therein.

Synchronization (VFO synchronization field) This is used to synchronize the controller's VFO synchronization field prior to address search.
Used to lock up (PLO). The data is all "0", that is, only the clock.

In MFM recording, all "1"s (FF) also have the same signal pattern, so the controller recognizes that this part is all clocks and subsequently separates the clocks and data.

ID Field In this field, following address marks "A1" and "FE", cylinder, head, sector, and check code for this area are written.

AM (address mark) A1'' written in this part is called a missing clock (dropped clock) and is written in a pattern that is impossible according to MFM rules.The controller detects this special pattern using hard logic. It is possible to identify whether the following data is an ID field or a data field depending on whether it is "FE" or "F8".

The cylinder, head, and sector following the address are 1 byte, but the actual cylinder has 256 or more, and is 1 byte (8 bits).
It cannot be expressed. The most significant bit of the head address is
This indicates that this sector is a defective sector.

Gap 2 This gap is called a write space gap and is provided for 2 or 3 bytes. When writing to the data area, the read state is switched to the write state, and this part is provided to provide time for this switching.

Synchronization This synchronization has the same function as the synchronization field at the beginning of the ID field, but its feature is that it is rewritten at the same time as the next data field using the same write clock.

Data field Data is written to this part. A 2-byte CRC or 3-byte ECC' is added following the data.

Most recent controllers employ ECC with an error correction function. Next, there is a gap called a write turn-off gap of 2 or 3 bytes. This write turn-off gap is provided to ensure the switching time of the write operation.

Gap 3 This gap is an inter-record gap (Inter
-Record Gap: I RG), there may be fluctuations in the motor rotation while writing to the previous sector, which may destroy the next sector. Or fill it with 15 Hyde "4E". Although the IRG is written to the above-mentioned length at the time of formatting, there is a possibility that it becomes shorter if data writing operations are repeated.

Gap 4 This gap is called the speed tolerance gap. It continues after the last sector in one track and is written until the index signal arrives.

This gap accommodates spindle motor speed fluctuations. The basic format is 352 bytes and 4E'', but the actual length depends on spindle speed variations during formatting.

Next, the ESD I hard sector format will be described with reference to FIG. Area (2) is merely an example and may be configured according to the needs of individual users. The number of check bytes in area (■) is determined by the user. Area■
PLO synter fields and ISOs are broadcast in response to request configuration commands. The total number of bytes shown in area 3 is the minimum except for the address field length. The format speed tolerance gap in area ■ is
It is necessary if the reference clock does not depend on the rotation speed, and the validity of this gap depends on the configuration data.

l5O (Intersector Gap) The minimum length of this ISO is determined by configuration data. This part is a dividing part between each sector.

Next, address areas will be explained.

PLO Synchronization Field This byte is used for reading data from the disk drive.
The LL oscillator generates data bits recorded on the disk and
Required by disk drives to be synchronized in frequency and phase.

The controller sends OO during that time.

Byte Synchronization Pattern This byte forms a byte synchronization and tells the controller:
Indicates the start of address field information.

Address Field This byte is determined by the user, and in one example consists of 5 bytes, 2 bytes for the cylinder address, 1 byte for the head address, 1 byte for the sector address, and 1 byte for the flag status.

ADR Check Byte (Address Field Check Code) An appropriate error detection mechanism is generated by the controller and passed to the address for correct file. This code is written to disk during formatting. Data accuracy is maintained by the controller recalculating the address field check code to ensure it is correct when the address field is read. This ADR check byte is determined by the user.

ADR Pad (Address Field Bad) This address field pad must be written by the controller and requires that the last bit recording and recovery of the address field check code be ensured by the data drive. This byte becomes 00.

Next, the data area will be explained.

Write Splice This byte area is required by the data drive to allow the time required for the write driver to transition and reach a sufficient recorded value to ensure data recovery.

PLO Sync Byte This byte is required by the disk drive so that the disk drive's preamble data PLL oscillator is synchronized in frequency and phase with the data bits recorded on the disk.

The controller sends OO during that time.

Byte Sync Pattern This byte forms a byte sync and tells the controller,
Indicates the start of address field information.

Data Field The data field is equivalent to the host system's data file.

Data check bytes Data check bytes, part error checking codes, are generated from the controller and written to disk along with the data fields. Disk accuracy is maintained by the controller recalculating the data field check code to ensure it is correct, i.e., applying an error correction algorithm, if appropriate, when the address field is read. Ru. This data check field is determined by the user.

Data Pad The data field pad must be generated by the controller, the data drive must be written by the controller, and the data drive is required to ensure the last bit recording and recovery of the address field check code. Ru. This byte is 00
becomes.

Format Speed Tolerance Gap This gap is required if bit 14 of the configuration response flag is set to 1. If this gap is necessary,
between each sector.

Next, the format of the ESDI soft sector system will be explained with reference to FIG. Area (2) is merely an example and may be configured according to the needs of individual users. The number of check bytes in area (■) is determined by the user. The PLO synter field of area ■ is larger than 11 bytes. The total number of bytes shown in area ■ is the minimum. Area■
is part of the PLO synter field so that the read gate activation delay controller has to treat it as an additional byte in the PLO synter field. The format speed tolerance gap in area (2) is necessary if the reference clock does not depend on the rotation speed, and the validity of this gap depends on the placement data.

The format of this ESD I soft sector method is
Most of the format is the same as the ESDI hard sector format shown in Figure 7 above, but the only difference is that the E
I in the hard sector formant of SD I
A 3-byte address mark (AM) and a 1-byte pad are inserted between the SO and the PLO sink.

D Problems to be Solved by the Invention Now, if there is a defect in a certain sector of a disk and the recording of main data is prohibited in the data field of that sector, a control for prohibiting the recording is installed in that sector. Data must be recorded. for that purpose,
The control data area (ID field or address area) of that sector must be rewritten (rerecorded).

In this way, in order to rewrite the control data area of a certain sector, write E in the beginning of that sector.
Like the soft sector method formant of SDI,
Mark data (address area) must be recorded, and if there is no mark data at the beginning of the control data area, such as the 5T-506 format shown in Figure 6, the control data area cannot be rewritten. It is possible.

In addition, in the case of the ESDI hard sector format shown in Figure 7, it is necessary to generate a mark pulse in hardware at the beginning of each sector, which increases the price of the disk drive and also causes the mark pulse to be generated at the beginning of each sector. Since the pulse generation timing is fixed, there is a drawback that the data length per sector is fixed.

By the way, there is no particular regulation for the pattern of the 3-byte mark data (address mark) of the ESD I soft sector format in Figure 8 above, so SDM
If one pattern, that is, 3 bytes, is no signal (all 0) as in the method (refer to the DP 8463 ENDEC manual), some of the bits may become 1 due to an error.
If the mark data becomes 0, or some bit data is missing, that is, the bit 1 becomes 0, the mark data will not be detected.

In view of this, the present invention proposes a data recording method that can reliably detect the start of a unit data area such as a sector of a recording medium with high accuracy.

E Means for Solving the Problems The present invention sequentially records synchronization data, control data, and main data, all of which are composed of a predetermined run-length limited code, in a unit data area of a recording medium, and In a data recording method in which mark data is recorded at the start part, the mark data consists of a first data part and a second data part following it, and the first data part is a predetermined run length limited code. It consists of a repetition of a code with a run length that cannot exist, and the second data part is characterized by having the same bit pattern as the synchronization data that follows.

F Effect According to the present invention, the first data part consists of repetitions of a code having a run length that cannot exist in a predetermined run length limit code, and the second data part consists of a repetition of a code that has a run length that cannot exist in a predetermined run length limit code. Mark data having the same bit pattern as the data is recorded at the beginning of the unit data area of the recording medium.

G. Embodiment Hereinafter, an embodiment in which the present invention is applied to data recording on a hard disk will be described in detail with reference to the drawings.

C. Recording Format First, with reference to FIG. 1, the recording format of a hard disk to which the present invention is applied will be explained. still,
In Figure 1, () indicates the number of bytes of data length, and “
indicates a hexadecimal number.

Gap 4 There is only one gap per track, in the last sector (
It is provided between the gear 7 of the Nth sector (Nth sector) and the gap in front of the first sector (first sector), which will be described later, to absorb uneven rotation of the spindle motor. The average data length is 412 bytes.

Gear 1 This gear is placed one (place) per track, and is provided between gap 4 and the first synchronization of the first sector (first sector), and is for head switching margin. Its data length is 40 bytes.

AM (address mark) This is for detecting the start of a sector and is provided at the beginning of each sector. The data length is 3 bytes.

This is the margin time provided between the address mark of each sector of the band and synchronization, which will be described later. Data is “FF” (hexadecimal)
It is.

Synchronization This is provided before the ID field and data field of each sector, and a clock synchronization signal for reading each data of the ID field and data field is written here. This synchronization signal is rewritten every time the ID field and data field are rewritten. The data length is 13 bytes. Data is “FF”
(hexadecimal).

Next, the ID field for each sector, in which address information is mainly written, will be explained.

Byte Sync This immediately follows the sync and indicates the byte deadline for converting run-length limited code to NRZ code.

The data length is 1 byte. The data is a special value “7E”
(hexadecimal).

Cylinder H and Cylinder L These indicate cylinder numbers, and the total data length is 2 bytes.

Head This indicates the head number (disk surface number), and the data length is 1 byte. Note that of this 1 byte, ie, 8 bits, the lower 4 bits are used as the head number. Also, the MSB 1 bit of the 8 bits is a pad (ba
d) Used as a sector identification flag. When this flag is 1, it indicates that there is a defect in the data field of that sector and the sector is unusable. Indicates that there are no defects in the data field and the sector is usable.

sector# It shows the sector number, and the data length is 1 byte.

CR is an error check code that is added to the data in the ID field when writing, and when reading, this CRC is checked to check for errors in the data. The data length is 2 bytes.

Gap 2 This gap is the margin provided between the ID field and subsequent synchronization, and the data length is 3 bytes. The data is “FF” (16 forward).

Next, the data field into which data from the host is written will be explained.

Byte Sync This immediately follows the sync and indicates the separation of bytes when converting run-length limited code to NRZ code.

The data length is 1 byte. The data is a special value “7E”
(hexadecimal).

Data This is the part that records data, that is, the main data to be written on the disk, and the amount of data can be switched to 256.512.1024 bytes upon request from the host side.

EC is an error check code, which adds 7 bytes of ECC to the entire data field when writing, and checks this ECC when reading to check for data errors.

It has the ability to correct continuous errors within 19 bits in length.

Write splice This is a margin for finishing writing when rewriting a data field. The data length is 3 bytes.

Therefore, writing must be completed within these 3 bytes.

Gap 3 This gap is the margin between each sector, and includes the rotational fluctuation of the spindle motor when rewriting each sector, and the margin for switching between rewriting data in the previous sector and reading the beginning of the next sector. is the sum of The data length varies between 44 and 59 bytes depending on the data length of one line.

Each part except the address mark mentioned above is, for example, 2-
Consists of 7 run length limited (RLL) codes,
Only address marks consist of repeating codes with run lengths that do not follow the 2-7 run length limited code. An example of such a run length limit code conversion table is shown in FIG. In addition, in this Figure 5, × is “0”
"," or "1" may be used. The left column shows message data, and the right column shows 2-7 run length limited code data corresponding to the message data.

G2 Address Mark Format Next, referring to Figure 2, the address mark (AM
) Describe the format of the signal.

This AM signal is 3 bytes in NRZ data conversion as mentioned above, and 6 bytes in RLL low (RAW), for a total of 48
Consists of bits. In FIG. 2, T indicates a 2-bit period. This AM signal consists of an initial 2T period, followed by three 6T periods, and two subsequent 2T periods as shown. The first 2T period is all 110+1.

All three 6T periods are codes with run lengths that do not follow the 2-7 run length limited code described above;
Here, it is "100000000000". The last two 2T periods are in the same format as the pads that follow (see Figure 1), for example "10001000".

G3 Address Mark Detection Next, address mark (AM) detection will be explained with reference to FIG. Here, after the first '6T period is detected, one 6T pattern and two 2T patterns are detected within T1=2μsec (when a surface is detected,
It is assumed that an address mark has been detected. Figure 3A is
The AM area of the formant in FIG. 1 is shown. When a hard disk recorded with data according to the format shown in FIG. 1 is played back, if no noise pulse occurs in any of the three 6T periods as shown in FIG. 3B, then as shown in FIG. An AM detection signal is generated at the end of the AM signal.

As shown in the third diagram, within the first 6T period, the noise pulse
1'', an AM detection signal is generated at the end of the AM signal as shown in FIG. 3E.

When a noise pulse "1" occurs within the second 6T period as shown in FIG. 3F, an AM detection signal is generated at the end of the AM signal as shown in FIG. 3G.

As shown in Figure 3H, within the last 6T period the noise pulse''
1'' occurs, an AM detection signal is generated at the end of the AM signal as shown in FIG. 3I.

If a noise pulse "l" occurs within <2T period following the last 6T period as shown in FIG. 3J, then AM detection is detected as shown in FIG. A signal is generated.

If the noise pulse “1” occurs in the first and second 6T periods as shown in Figure 3, then <A> as shown in Figure 3M.
No M detection signal is generated. In this case, the first 6
6T within T1=2μsec after T period is detected.
1 pattern, 21 patterns of 2T [! If detected, it means that the conditions for detecting an address mark are not satisfied.

Rewriting G41D Field Next, referring to FIG. 4, the operation of rewriting the data in the ID field based on the detection of the above-mentioned AM signal will be described. Consider the case where the Nth sector is a bad sector and its ID field is to be rewritten.

FIG. 4A shows the address mark (AM), bad, synchronization, and ID fields of the (N-1)th sector and the Nth sector, respectively. First, when the ID field of the (N-1)th sector is detected, a detection pulse is obtained as shown in FIG. 4B. When this detection pulse is obtained, as shown in FIG. 4C (period T (1 sector)
The address mark (AM) of the Nth sector is searched for a period of 240 to 880 μsec). When the Nth address mark (AM) is detected, a detection pulse is obtained as shown in the fourth diagram. When this detection pulse is obtained, the synchronization and ID of the Nth sector are determined as shown in Figure 4.
Rewrite the field. In this case, in particular, rewrite the MSB value of the head from "O" to "l",
The values of other parts are simply rewritten as they are.

In the above-described embodiment, a hard disk as a magnetic disk was used as the recording medium for recording data, but of course a floppy disk or a magnetic tape may also be used.

In the case of magnetic tape, data is recorded to form slanted tracks, and one slanted track is attached to one side of the magnetic disk.
It can correspond to a sector or multiple sectors.

H Effects of the Invention According to the invention described above, a data recording method is provided in which the start of a unit data area of a recording medium can be reliably detected with high precision, thereby making it possible to easily rewrite control data. Obtainable.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing the recording format of a hard disk to which an embodiment of the present invention is applied, Fig. 2 is a diagram showing the format of its address mark, Fig. 3 is an explanatory diagram of address mark detection, and Fig. 4 is a diagram showing the format of the address mark. An explanatory diagram of ID field rewriting, FIG. 5 is a diagram showing a run-length limited code conversion table, and FIGS. 6, 7, and 8 are diagrams each showing the recording format of a conventional hard disk. . Figure 3 explaining how to find the dress mark of Hidemori Matsukuma 7

Claims (1)

[Claims] Synchronous data, control data, and main data, all of which are composed of a predetermined run-length limited code, are sequentially recorded in a unit data area of a recording medium, and mark data is written at the beginning of the unit data area. In the data recording method in which the mark data is recorded in the first data section and the second data section following it,
The first data part consists of a repetition of a code having a run length that cannot exist in the predetermined run-length limited code, and the second data part consists of the following data part. A data recording method characterized by having the same bit pattern as synchronous data.