JPH0834043B2 - Magnetic recording method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a magnetic recording method using a magnetic tape,
In particular, the present invention relates to a magnetic recording method for realizing a large-capacity and highly reliable magnetic recording device suitable as a peripheral device for a computer.

[Prior Art] With recent advances in magnetic recording technology, as seen in VTRs and DATs, a helical scan type recording / reproducing technology using a rotary (rotary) head has been put into practical use, and the recording density is improved, and In combination with digital signal processing technology, the development of computer devices using magnetic tapes for VTRs and DATs has become popular.

In these devices, voice and the like are converted into digital signals and recorded on tape. Therefore, the recorded data can be used by the computer instead of the voice.

Examples of such devices include Japanese Patent Publication No. 60-25286 and Japanese Patent Laid-Open No. 60-171678.

In addition to the track recorded by the rotary head, a fixed head is used for searching in order to have a searching function for selectively recording and reproducing at any recording location of the magnetic tape required as a peripheral device of a computer. Then, control data is recorded above and below the track, and this portion is used as search control data.

In this method, since the control data is recorded in a place independent of the data, it can be easily searched.

However, the location where the control data is recorded is near the edge of the magnetic tape, and there is a problem that it is susceptible to dirt and damage.If this control data becomes unreadable, the search function required as a computer device cannot be realized. , Reduce the reliability of the entire system.

Further, in a magnetic tape device, access time is generally slow. In the control using the fixed head, when the tape is run at high speed, the control information cannot be read,
High speed access is not possible.

Also, with a system that uses a method that changes the azimuth on adjacent tracks, such as a DAT recorder, especially in a device that slightly overlaps adjacent tracks with a wide head, it is possible to re-record on the previously recorded track. When recording, the data is also recorded on the adjacent track, and the data cannot be read correctly from this track. Therefore, even if the control data as in the conventional example is used, re-recording cannot be performed for each track. is there.

These have not been considered in the prior art.

[Problems to be Solved by the Invention] The above-mentioned prior art does not take into consideration uncontrollability due to dirt or damage on the magnetic tape, or inability to record on a track-by-track basis in an azimuth recording system. There is a problem in that it cannot be used as a computer device.

Also, the track pitch is made as small as possible in order to increase the storage capacity of the magnetic tape. Therefore, there is a problem that the area for recording control data provided above and below each track becomes small and effective data cannot be recorded.

An object of the present invention is to provide a magnetic recording method using a helical scan type magnetic tape, and particularly to a method most suitable for using the magnetic recording method using the azimuth recording method in a data recording device of a computer. is there.

[Means for Solving Problems] Among the above-mentioned objects, the problem that rewriting cannot be performed for each track by the azimuth recording method is that a plurality of tracks are set as a read / write unit (block), and at least the last one track of the plurality of tracks, Alternatively, the tracks before and after the block are not used for data but as gaps. When rewriting that block, record from the first track in order,
Gap absorbs the unreadableness of the adjacent track due to the scraping of the adjacent track, which enables rewriting in block units.

On the other hand, in order to reduce reliability due to dirt or damage on the magnetic tape, an area for recording control data is provided in each track, and data for detecting an error is added and recorded. Alternatively, this is used in a device such as DAT in which a region for recording control data is secured in advance in the standard. Then, as described above, a plurality of tracks are set as a block, and the number of this block is recorded on each track in the block as one of the control data. Therefore, the area where the control data is recorded is near the center of the magnetic tape, and is the area where damage is minimal, and since the block numbers are recorded over multiple tracks, even if there are some errors, the correct location It is possible to search.

In addition, high-speed access using the DAT search function is also possible.

[Operation] Each block is composed of consecutive frames (tracks), and the frames before and after it operate not as valid data but as a gap for absorbing unreadable data due to re-recording. Control data for access to the tape is recorded in a region near the center of the tape which is less likely to be damaged, or is recorded in a region secured in advance to record control data other than the data, and the same in the track in the block. By repeatedly recording data, even if a part of the block causes an error and cannot be read, it can be accessed from the control data of another track.

In addition, since the data has an inspection word for error detection and correction, it operates to correct the error even if part of the data cannot be read properly due to dirt or damage on the tape. If the value exceeds the range, the flag is output as an uncorrectable flag, so that no malfunction occurs.

Furthermore, if the data cannot be read correctly or cannot be recorded even after error correction due to the damage of the tape due to repeated use of the magnetic tape and the occurrence of data error due to unevenness of the medium, this block is referred to as a bad block. Since this block operates so as not to be used, it does not malfunction.

[Examples] Examples of the present invention will be described below.

This embodiment is based on the format of the magnetic tape used in the DAT recorder.

[Block Configuration] FIG. 1 shows a block configuration diagram on the magnetic tape of the present embodiment.

In the DAT recorder, a head that rotates together with a cylinder attached to the circumference of the cylinder like a VTR is used, and a track 2 as shown in FIG. 1 is recorded on the magnetic tape obliquely with respect to the running direction of the magnetic tape.

Further, the DAT recorder uses the azimuth recording method, and the adjacent tracks are recorded with different azimuth angles.

In order to use this DAT recorder as a storage device of a computer, 32 frames (for DAT, 2 tracks are 1
A block 1 is referred to as a frame), and read / write is executed in units of this block 2.

The first 3 frames and the last 1 frame of the block 1 are not used for data but are set as the gap area 3 between the blocks. Therefore, the data area 4 for recording data in 28 frames between the gap area 3
And

Figure 2 and Table 1 show the DAT track format standardized at the DAT conference. The track is divided into 16 areas, and each area is defined as shown in Table 1.

Area 9 is an area for recording PCM data. Areas 3 and 1
Subcode area 4 is defined as an area for recording information such as music number and time. Areas 6 and 11
Is called an ATF (Automatic Track Finding) area, and is an area for recording a signal for applying servo so that the rotary head can trace the track correctly.

The actual magnetic tape is formatted as shown in FIG.

In DAT, two heads, + azimuth and -azimuth, trace the tracks alternately. Each head is an ATF for a track recorded in azimuth that matches your azimuth.
Read the area, apply servo, and read the PCM area and subcode area.

In the case of recording, as shown in FIG. 4, recording is performed in the above-mentioned track format by the head A slightly wider than the track, and when recording the next track, the position previously recorded by the head B is recorded. And a newly recorded track is slightly overlaid and recorded. As a result, the previously recorded track is controlled to have the width defined by the standard.

In order to use it as a storage device of a computer, it is necessary that a search function in any place and recording in any place are possible.

However, as shown in FIG. 5, when a certain track is rewritten, recording is performed with a head having a width wider than the width of the track, and the adjacent track is deleted. Therefore,
The track next to the rewritten track becomes unreadable.

Therefore, even if gap tracks are provided before and after the block as shown in FIG. 1 and rewriting is performed in block units, only the gap tracks of the adjacent blocks become unreadable, and the data area is not affected. become.

"Subcode format" In DAT, the PCM area and subcode area in the track are divided into smaller blocks.

 FIG. 6 shows the structure of a small block.

The MSB (most significant bit) of the small block address is used to indicate whether this small block belongs to the PCM area or the subcode area.

As shown in the figure, when this bit is 0, it belongs to the PCM area, and when this bit is 1, it belongs to the subcode area.

The format of the subcode area of the present invention is shown in FIG.

 One subcode is composed of two small blocks.

 Table 2 shows the contents of W1 and W2.

The control ID shall be "0" for all 4 bits according to the DAT standard.

The data ID is set to "0001" to distinguish it from audio data ("0000").

Since seven packs are used as shown in FIG. 7, the format ID is set to "111" according to the DAT standard. One pack is represented by 64 bits, and defines 8 codes of 8 bits as shown in the figure. PC1, 2
Records the block 1 number of the basic unit of access. The block numbers that can be represented are up to 65536, and the upper part is recorded in PC1 and the lower part is recorded in PC2. PC3 is
A frame number is recorded to indicate the number of frame in block 1 to which the track to which this small block belongs is recorded.

PC4 is an error detection parity defined by the following equation, and detects the errors of PC1 to PC3.

P1 = PC1 + PC2 + PC3 (1) However, + represents exclusive OR (mod2).

PCs 5 and 6 are areas for recording area marks, and represent codes for discriminating which area the block 1 belongs to.

PC7 is an error detection parity defined as follows, and detects the errors of PC1 to PC6.

P2 bit 0 = PC1 parity P2 bit 1 = PC2 parity P2 bit 2 = PC3 parity P2 bit 3 = PC4 parity P2 bit 4 = PC5 parity P2 bit 5 = PC6 parity P2 Bit 6 = Parity of bits 0 to 5 Bit 7 of P2 = 0 PC8 is the parity for detecting the error of PC1 to 7 defined by the following equation.

P3 = PC1 + PC2 + PC3 + PC4 + PC5 + PC6 + PC7 (2) There are seven packs, but try to record the same contents.

"PCM-ID format" Tables 3 and 4 show IDs of small blocks in the PCM area for recording data.

In W2, the numbers of 128 small blocks in the PCM area are recorded in the DAT standard. The small blocks are grouped into eight blocks, and eight IDs in W1 are defined as shown in Table 4 in the present invention.

ID-1 shall be "01" to indicate usage other than audio.

ID-2 indicates that "00" configures the block 1 with 32 frames as in the present embodiment.

Data ID: 0001 Format ID: 111 Control ID: 0000 "01" to "11" are reserved for future expansion.

ID-3 represents the use of error correction, format. "0
“0” indicates a layered ECC format of this embodiment described later. “01” and “10” indicate spares for expansion, and “11” indicates a mode in which layered ECC is not used.

ID-4 indicates an access format as a peripheral device of a computer. "00" indicates the same sequential access as a magnetic tape device such as a streamer, and "01" indicates a magnetic tape that can be directly accessed such as a floppy disk or a hard disk. "10" and "11" are reserved.

ID-5 indicates that this block 1 is a block that can be correctly read / written by "00", and "11" indicates that it is a BAD block that cannot be correctly read / written due to damage of the tape or the like.

"01" and "10" are reserved.

ID-6 defines the track pitch, "00" is the standard 13.6
μm, “01” was recorded at a track pitch of 20 μm,
Alternatively, it indicates a magnetic tape to be recorded.

ID-7 indicates permission or prohibition of software-like copying, and "0
"0" indicates permission and "10" indicates prohibition. "01" and "11" are reserved.

 ID-8 is always 0.

"Tape Format" FIG. 8 shows an explanatory diagram of the tape format of the present invention.

The magnetic tape is shown divided into a PCM area and a subcode area.

On the magnetic tape, as shown in the figure, data used by the user as data in the PCM area and error detection / correction parity of this data, which will be described later, are recorded.

Although many data as described above are recorded in the PCM-ID, the BAD block information which is particularly important in the tape format is shown.

In the subcode area, data that plays an important role in access is shown.

As the program number in the subcode ID, 11 bits higher than bit 3 of the block number in the subcode pack data are assigned. Therefore, the program number is incremented by 1 every 8 blocks. This program number can be used for rough searches.

The area mark is set to "0" to indicate that it is a normal data block. The block number is continuously recorded in each block.

 Details of the blocks are shown in the lower part of the figure.

All PCM data in the gap area shall be "0". As for frame numbers, the first frame of block 1 is set to "0" and sequentially numbered to "31".

Figure 9 shows the lead-in area and EOI (End Of Informatio).
n) An explanatory diagram of the area is shown. Magnetic tapes for DAT have transparent tapes as leader tapes at the beginning and end of the tape. The recording device uses a BOT (Begin
Of Tape) and EOT (End Of Tape) are detected.

However, since the mechanism for pulling out the magnetic tape from the cartridge and loading it in the cylinder varies from device to device, the position of the first block varies.
In order to absorb this, a lead-in area is provided that is not treated as a valid data area. In DAT, the length of this area is 100 mm or more.

In this embodiment, 16 blocks (about 130 mm) are used, and "0" is recorded for all data. As a subcode, record the program number as "-2" or "-1" as shown in the figure. For the area mark, "BB" is recorded with an ASCII code. In addition, increase the block number in order from "-16"
Recording is performed so that the lead-in area ends at -1 ".

On the other hand, as an area showing the end of data and indicating that no valid data is recorded after this,
Establish EOI area. This area has a length of one block, all data is recorded with "0", and the program number, the block number, and the frame number are recorded in the same manner as a normal block. In order to distinguish it from the lead-in area and the data area, "EE" is recorded with an ASCII code as an area mark.

This makes it possible to know where the data starts and where the data is recorded.

 FIG. 10 shows an explanatory diagram of the BAD block format.

After recording the data in block 1, when reading out whether it was recorded correctly and comparing it with the written data,
If they do not match, the block is regarded as a BAD block.

In the figure, it is assumed that data is recorded while formatting a block number m + 9 after this block. If this block becomes a BAD block, put the tape back a little and format the block again as "0" while recording the code indicating the BAD block in ID-5 of the PCM-ID in this block. .

When formatting the next block, leave the block number as m + 9 and re-record the data that should have been recorded in the previous block. Therefore, blocks having the same block number are arranged.

On the other hand, when new data is recorded in a block where data has already been recorded by random access and it becomes a BAD block due to damage to the tape, etc., the BAD block is added to ID-5 of the PCM-ID as described above. Record the code indicating "0" and "0" data. At this time, the content of the subcode is the same as the content previously recorded.

However, in this case, since it is not possible to determine in the device where to replace the block, the host instructing the writing is notified that there is an error, and the operation thereafter is
Follow the instructions from the host.

The “Layered ECC” DAT performs error correction using a duplicated Reed-Solomon code that is completed within track 2. As a result, when the symbol error on the magnetic tape is 10 ^-3 , the error rate after error correction is 10 ^-20 or less.

However, it has been confirmed by experiments that the symbol error on the magnetic tape deteriorates to 10 ^-2 or more due to the dirt and damage of the magnetic tape and the life of the tape due to repeated use.

The error rate after error correction at this time is 10 ⁻¹² , which is a low-reliability computer data recording device.

Therefore, even if the symbol error on the magnetic tape becomes worse than 10 ^-2 , the parity for error correction is added to the data so that the error rate after error correction becomes 10 ^-15 or less. This is called layered ECC.

FIG. 11 shows the layered ECC format of this embodiment.

The format is completed with 28 frames for data in one block. Therefore, the data used by the user and the parity for error detection and correction are recorded in 28 frames of frame numbers 3 to 30 in the block 1.

This error correction also uses Reed-Solomon code,
Codes of code length 56, information word 50, and check word 6 are used. As shown in the figure, the data of 2880 bytes in the track is arranged vertically, the data of each track is arranged in order horizontally, and one code word is selected from each track to select 50 information words and 6 check words. Configure.

Since the distance is 7, it is possible to correct a maximum of 3 errors, but we do not know where the error is occurring. ”
In the "error correction", it takes a great amount of time to correct three errors, so in this embodiment, up to two "error corrections" are performed. As a result, a flag indicating that the data read from each track was error-corrected but could not be corrected is added, so the location of the error can be specified by this flag. When "erasure correction" is performed, a maximum of 6 errors can be corrected. This correction algorithm is shown below.

N (E) = 0 NO error N (E) = 1 1 error correction N (E) = 22 2 error correction N (E)> 2 N (F) = 0 F = 1 N (F) = 1 1 Double erasure +2 error correction N (F) = 2 double erasure +2 error correction N (F) = 3 triple erasure +1 error correction N (F) = 4 quadruple erasure +1 error correction N (F) = 5 quintuple erasure correction N (F) = 6 6 erasure correction N (F)> 6 F = 1 where N (F): number of uncorrected symbols by double Reed-Solomon code N (E): number of error symbols by layered ECC F: Error Flag With this algorithm, even if a burst error of up to 6 tracks occurs, it can be corrected and a very reliable device can be realized.

With this format, the data capacity that can be used by the user per block is approximately 140.6 Kbytes. However, since data that is usually handled by a computer is an integer multiple of 64 Kbytes, 128 Kbytes is the data capacity that can be used by a user per block.

"Operation" A functional diagram of this embodiment is shown in FIG.

A cylinder 10 equipped with a rotary head, a cylinder motor 11 for rotating the cylinder, a capstan motor 12 for constant running of a tape, a reel motor 13 for fast forward and rewind, a servo circuit 14 for controlling these motors, a rotary head Read / write circuit 15 for writing and reading signals using, DAT signal processing circuit 16 for performing signal processing according to DAT format, subcode control circuit 17, device mode instruction, mechanical drive instruction Drive control circuit 18, a mechanical drive circuit 19 that receives this signal and drives a mechanism, a DAT code / composite circuit 20 that encodes and decodes an error correction code according to the DAT format, a layered ECC circuit 21, data Memory 22, for performing layered ECC coding / compositing, and for reading and storing data from this memory 22 Memory control circuit 23, system control circuit 24 that controls all of these, interface SCSI that connects to the host
(Small Computer System Interface) 25.

FIG. 13 shows the operation sequence of the system in this embodiment.

First, it is checked whether a magnetic tape is loaded in the system. If the magnetic tape is not loaded, wait until it is loaded.

When the system control circuit 24 detects that a magnetic tape has been inserted into the system by a detection circuit (not shown), it drives a loading motor (not shown) to load the magnetic tape, pull out the tape from the cartridge,
Wrap around 10.

Once the magnetic tape is loaded or loaded, the system control circuit 24 sends a signal to the drive control circuit 18 to "rewind" the tape.

Upon receiving this signal, the drive control circuit 18 sets the servo circuit 14 in the "rewind" mode.

The servo circuit 14 drives the reel motor 13 to rewind the tape. The drive control circuit 18 detects the transparent tape portion of the magnetic tape and knows that the tape has been rewound.
Therefore, the servo circuit 14 is set to the "stop" mode, and the system control circuit 24 is notified that the "rewind" is completed.

The system control circuit 24 uses the drive control circuit 18 in order to know how much data is recorded on the magnetic tape.
Instruct "EOI search". The drive control circuit 18 sets the servo circuit 14 in the “high-speed search (described later)” mode to drive the reel motor 13 and the cylinder motor 10. Cylinder 10
The R / W circuit 15 captures the output from the rotary head attached to
To enter. The signal processing circuit 16 reads a subcode from this signal and sends it to the subcode control circuit 17. The sub-code control circuit 17 checks whether the reproduced sub-code is correct, and if it is correct, the sub-code is controlled by the drive control circuit.
The data is sent to 18, and the drive control circuit 18 checks the area mark in the subcode described above to check whether it is the EOI area.

When the area code indicating the EOI is detected, the drive control circuit 18 puts the servo circuit 14 in the stop mode once, rewinds the overrun tape a little in the rewind mode, and reads the EOI block in the "read (described later)" mode. .
The subcode control circuit 17 sends the correctly read subcode to the drive control circuit 18. The drive control circuit 18
The block number in the subcode is sent to the system control circuit 24.

The system control circuit 24 registers the block number received from the drive control circuit 18 as the final block number.

At the same time, the signal processing circuit 16 transfers the PCM-ID of the data area of the EOI block to the drive control circuit 18, and sets the above-mentioned various parameters.

If the EOI area is not found even after searching the end of the tape, or if nothing is recorded on the tape, register it as a new tape.

Next, the system control circuit 24 is in a state of waiting for a command sent from a host (not shown) via the interface 25.

The system control circuit 24 receives a command from the interface 25, analyzes the command, and executes a read, write, format, etc. command to drive the drive control circuit.
18. Drive the memory control circuit 23.

When the host issues an eject command for the cartridge or an instruction from an eject switch (not shown) attached to the apparatus, the EOI area is recorded at the end of the data area recorded on the magnetic tape, and the tape is rewound.

Since the system control circuit 24 stores the last block number of the current data area, in order to "search" for this block number as described above, the drive control circuit 18
Further, the drive control circuit 18 drives the servo circuit 14 and the subcode control circuit 17 to find the final block.
After this, after the last block, E in the EOI format
Record the OI area. At this time, the system control circuit 24 sends EOI to the subcode control circuit 17 via the drive control circuit 18.
Send the subcode data of the block. Further, data to be recorded in the data area in the block is prepared in the memory 22, and the code for error detection and correction is added to the memory control circuit 23 by the layered ECC circuit 21, and this is added to a predetermined location of the memory 22. Store to.

Encoding / decoding circuit in the order in which this data is written to the track
The data is sent to 20, where it is encoded into a data format conforming to the DAT standard.

After searching the last block, drive control circuit 18
Sets the servo circuit 14 to the "play" mode and reads the data. When this block is completed, the signal processing circuit 16 is switched to the "write" mode while checking the frame number of the subcode read by the subcode control circuit 17.
The signal processing circuit 16 instructs the encoding / decoding circuit 20 to read data. In response to this, the encoding / decoding circuit 20 outputs the recorded data to the signal processing circuit 16. The signal processing circuit 16 combines this data and the subcode from the subcode control circuit 17 with the data according to the data format and records the data via the R / W circuit 15.

After recording the EOI block, drive control circuit 1
In No. 8, after rewinding the tape in the same manner as the above-mentioned rewinding, the cartridge is ejected by a loading motor (not shown).

"Read" FIGS. 14 and 15 show read timing diagrams of the blocks of this embodiment.

The drive control circuit 18 sets the servo circuit 14 in the "play" mode and the signal processing circuit 16 in the read mode.

The servo circuit 14 controls the reel motor 13 to drive the magnetic tape at a constant speed of 8.15 mm / s, which is specified in the standard.
The servo circuit 14 takes in and controls the rotation speed of the cylinder 10 so that the 10 rotates at a constant speed of 2000 rpm. The switching signal of the two rotary heads mounted facing the cylinder 10 by 180 degrees is a signal for switching the operating head every 15 ms.

In the figure, a signal is read by the head A of + azimuth from time T ₁ to T ₂ , and a signal is read by the head B of −azimuth from time T ₂ to T ₃ .

Each head reads the ATF recorded on the tape, and the R / W circuit
The signal processing circuit 16 inputs a servo signal to the servo circuit 14 via 15. The servo circuit 14 controls the rotation speed of the cylinder 10 so that the track can be traced correctly.

In this state, the output from the head through the R / W circuit 15 becomes the signal shown in the figure.

This signal contains a signal for one track. The signal processing circuit 16 receives this signal and outputs the signal in the subcode area to the subcode control circuit 17 at time T ₁₁ .
Upon receiving this signal, the subcode control circuit 17 reproduces the subcode from this signal and performs error detection using the parity indicated by the subcode ID format. Regarding pack data, error detection is performed using parity in the small block format of DAT, and at the same time, error detection is performed by comparing a plurality of pack data and determining whether they match.

The results at the time T _12, and sends to the drive control circuit 18.

On the other hand, of the signals input to the signal processing circuit 16, signals in the PCM area are temporarily stored in the signal processing circuit 16.

At time T ₂ , switch the head to-Azimuth head B,
Read the data as well.

After reading the data for one frame, the data of PCM-ID is transferred to the drive control circuit 18 at time T ₃ .

The data for one frame read by both heads is transferred to the DAT code / decoding circuit 20 by the signal processing circuit 16 from time T ₄ .

Encoding / decoding circuit 20 that has received data for one frame
Performs error detection and correction according to the DAT format, and requests the memory control circuit 23 to transfer data.

If the frame number in the subcode read from time T ₁ to T ₃ is 2, this frame is a gap area as shown in the block format and is not treated as valid data, so the memory control circuit twenty three
Does not store the data transferred from the encoding / decoding circuit 20 in the memory 22, but stores the data of the next frame number 3 in the memory 22.

The memory 22 shown in FIG. 12 is divided into three areas, which are referred to as memories A, B, and C, respectively.

The block data signal of FIG. 15 represents the timing at which the memory control circuit 23 stores the data of frame numbers 3 to 30 of the nth block in the memory A22. At this time, as a result of the error correction for the DAT output from the encoding / decoding circuit 20, the uncorrected flag is stored in the memory A22 together with the data. Memory A for 1 block of data
After storing in 22, the memory control circuit 23 at time T ₂₁ ,
This data is read and transferred to the layered ECC circuit 21, and error correction is performed according to the error correction algorithm of the layered ECC. At this time, up to six errors are corrected using the previously stored uncorrected flag.

 The corrected data is stored in the memory A22.

However, if ECC prohibit mode is specified in the PCM-ID data, the above error correction is not performed,
The data from the encoding / decoding circuit 20 is stored in the memory A22 as it is.

Thereafter, the system control circuit 24 at time T _22, the memory A22
The decrypted data is transferred to the host via the interface 25.

At time T ₂₃ , since the data signal of the next block is output from the encoding / decoding circuit 20, this data is stored in the memory B,
Store in C22. Thus, when the data transfer to the host is completed, the data from the encoding / decoding circuit 20 is stored in this memory area, so that each processing can be executed without waiting time. This is because three transfers need to be executed at the same time when the host's data transfer preparation is delayed and the timing indicated by the broken line in the figure is reached.

“Write” FIGS. 16 and 17 show timing charts for recording data.

At time T ₃₁ , the host requests to record data, and the system control circuit 24 causes the memory control circuit 23 to store the data.
Instruct to transfer data to A22. Memory control circuit 23 that receives a data transfer request via interface 25
Receives the data and stores it in the memory A22.

After this, at time T ₃₂ , the system control circuit 24 instructs the memory control circuit 23 to transfer data with the layered ECC circuit 21, and performs encoding in the layered ECC format. However, layered EC is instructed by the host
If the mode that prohibits C is specified, this encoding is not performed.

 This data is stored in the memory A22.

When there is a subsequent data transfer request from the host, the system control circuit 24 sets the memory control circuit 23 to store this data in the memory B. If there is a further data transfer request, it is stored in the memory C.

In the part b, the data in the memory A22 is encoded, and the data from the interface 25 is stored in the memory B22. Similarly, in the c section, data is stored in the memory C22. Now that the memory 22 is all in use, the host is informed that it is in operation and requested to suspend the data transfer.

Thereafter, the encoding of the memory A22 is ended at time T _34,
The system control circuit 24 instructs the memory control circuit 23 to encode the memory B22 (section d). After this, time T ₃₅
Then, the memory control circuit 23 transfers the encoded data in the memory A22 to the DAT encoding / decoding circuit 20. Also,
When the encoding of the memory B22 is completed, the system control circuit 24
Instructs from time T ₃₆ the encoded memory C22 to the memory control circuit 23 (e portion).

When a block of data from the memory A22 to code / decode circuit 20 transfer is completed at time T _37, the memory A22 becomes idle.

Therefore, the system control circuit 24 notifies the host that the interruption of the data transfer is released, and stores the data transferred from the host in the memory A22. In section f in the figure, since all the memories are in use, the system control circuit 24 again requests the host to interrupt the transfer.

At the same time, the encoded data is transferred from the memory B22 to the encoding / decoding circuit 20.

When the data transfer from the host ends at time T ₃₈ , the system control circuit 24 instructs the memory control circuit 23 to encode the data in the memory A 22 (g portion).
This is repeated thereafter.

 FIG. 17 shows a frame recording timing chart.

First, data is read in the read mode. Assuming that the block number to be recorded is N, the drive control circuit 18 inspects the block number in the subcode data and detects the block number immediately before the target block number, here N-1. Furthermore, when the frame number is inspected and the frame number 30 is detected, it is transferred to the encoding / decoding circuit 20 in advance at time T _41, and the encoded data according to the DAT standard is transferred to the signal processing circuit 16. . Here, the signal processing circuit 16 creates a small block format in the track, receives the subcode data from the subcode control circuit 17, and when the frame number 31 has been read, switches the read mode to the write mode and waits for the time T. ₄₂ with R / W circuit
Through 15, the block number N is recorded in the order of + azimuth track and −azimuth track.

At the time of this mode switching, when the signal from the ATF area, which was the servo reference in the read mode, is switched to the write mode, it is switched to the internal reference signal, causing track disturbance. This is absorbed by setting the first three frames of the block as a gap.

[Post-recording] When rewriting the data of the already recorded block, it is not necessary to rewrite the subcode area and ATF area in the track format described above. Therefore, in this case, only the PCM area is rewritten. (This is called after-recording: dubbing.) The recording method is almost the same as the method described above.

The difference is that in the above recording method, the signal processing circuit 16 to the R / W circuit 1
All the data for one track output via 5 was recorded, but in dubbing, the subcode area is read as shown in Fig. 18, the ATF area is read, the servo is applied, and then the write mode is set. To record.

"Search" In this embodiment, high-speed search is possible as in the case of DAT.

 An explanatory view is shown in FIG.

When searching, the tape is run at high speed with the tape wound on the cylinder 10. At this time, the recorded data can be read by controlling the relative speed between the tape and the rotary head so as to be the same as during normal read / write.

In order to perform the search at 200 times the normal speed, assuming that the tape traveling speed is 200 times the normal speed, the rotation speed of the cylinder 10 (normally 2000 rpm) becomes 3000 rpm and 1000 rpm in the forward search and the reverse search, respectively.

Therefore, the number of heads that cross a track is about 10 each.
It is decided to be 0 and 400.

Therefore, the reel motor 13 is controlled so that the number of crossing this track becomes a predetermined number.

In the case of forward search, the servo circuit 14 first controls the reel motor 13 to run the tape 200 times.

In this case, the locus of the head that crosses the tracks on the tape is as shown by the broken line in the figure. The output from the head at this time is as shown in FIG. A large output is obtained when a head of + azimuth passes through a track recorded with the same azimuth, and a small output is obtained when the head of a different azimuth is crossed. Therefore, the servo circuit 14 controls the number of revolutions of the cylinder 10 so that this large output (envelope) is counted and becomes about 50 which is half the predetermined number.

In this way, the data read intermittently in the signal processing circuit 16, the sub-code control circuit 17, the drive control circuit 18 inspects the data, by reading the sub-code data and the contents of PCM-ID, You can search for the target block.

Similarly, in the case of reverse search, the servo circuit 14 runs the tape in the reverse direction by 200 times. At this time, the locus of the head that crosses the track is as shown by the solid line in the figure. Cylinder 1 so that the number of envelopes is about 200
Controls 0 rpm.

As described above, according to the present embodiment, 32 frames are set as one block, and by accessing in block units, it becomes possible to rewrite data in the middle of recording on the tape, and as a data storage device of a computer. It has the effect of realizing the function. In addition, by recording the block number to which the frame belongs in the subcode data of each frame, the search function of an arbitrary place that is indispensable as a data storage device of a computer is realized, and a normal read operation for searching a target block is realized. / It can be performed at 200 times the writing speed, which has the effect of enabling high-speed searches that were previously impossible.

Also, by adding layered ECC to the data area, there is an effect that sufficient data reliability can be obtained even when the tape is dirty or damaged.

Furthermore, by setting the first three frames of the block as a gap, it is possible to absorb the track disturbance that occurs when switching from the read mode to the write mode when recording a new block after the last recorded block. It is possible to prevent the data from becoming unreadable at the seam between the blocks and improve the reliability of the data.

According to this embodiment, a storage capacity of 1 Gbyte can be recorded on a 120-minute tape for DAT, and this data can be accessed in 20 seconds on average.

In the present embodiment, one block is composed of 32 frames, but it is clear that this is not necessarily the case in order to obtain the same effect. Further, although the Reed-Solomon code is used as the layered ECC and the information word is 50 and the check word is 6, the same effect can be obtained by using a check word length other than this or an error correction method.
it is obvious.

In this embodiment, an apparatus using two rotary heads has been described, but an apparatus having four heads for reading out this data immediately after writing and comparing it with the written data can also achieve the same effect.

[Effects of the Invention] According to the present invention, since the read / write in block units is realized, the search function and the rewrite function of an arbitrary place necessary for the storage device of the computer can be realized.
There is an effect that it can be used as a data storage device.

Further, by adding an error correction code to the data area to prevent the deterioration of the data reliability due to the dirt and damage of the data used by the user and the tape, the data reliability can be improved.

Furthermore, since the block number recorded in the subcode area of the frame in the block can be searched,
In a magnetic tape device for recording data on a computer,
This has the effect of realizing unprecedented high-speed access.

[Brief description of drawings]

FIG. 1 is a format diagram showing a block configuration of the present invention,
2 to 5 are explanatory views of an embodiment of the present invention, FIG. 6 is a format diagram showing a small block structure of DAT, FIG. 7 is a subcode format of the embodiment, and FIGS. 8 to 10 are tapes of the embodiment. Format explanation diagram, Fig. 11 is the layered EC of the embodiment
C format diagram, FIG. 12 is a functional block diagram of the embodiment,
FIG. 13 is an operation sequence diagram of the embodiment, FIGS. 14 to 15 are read timing diagrams, FIGS. 16 to 18 are write timing diagrams, and FIG.
19 Fig. Is an explanatory diagram of the search. Explanation of symbols 1 ... Block, 2 ... Track, 3 ... Gap area, 4 ... Data area 10 ... Cylinder, 11 ... Cylinder motor, 12 ... Capstan motor, 13 ... … Reel motor, 14 …… Servo circuit, 15 …… R / W circuit, 16 …… Signal processing circuit, 17 …… Subcode control circuit, 18 …… Drive control circuit, 19 …… Mechanical drive circuit, 20 …… Encoding / decoding circuit, 21 ... Layered EC
C circuit, 22 ... memory, 23 ... memory control circuit, 24 ... system control circuit, 25 ... interface

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kazutoshi Konno, 292 Yoshida-cho, Totsuka-ku, Yokohama, Kanagawa Prefecture Home Appliances Research Laboratory, Hitachi, Ltd. (72) Inventor Toru Mihei 292 Yoshida-cho, Totsuka-ku, Yokohama, Kanagawa Home Appliance Research Laboratory, Hitachi, Ltd. (56) References JP-A-63-251971 (JP, A)

Claims (8)

[Claims] 1. A magnetic recording method for recording / reproducing digital data diagonally on a magnetic tape having a rotary head on the circumference and wound around a cylinder rotating at a slight angle with respect to the running surface of the magnetic tape. A plurality of tracks on the magnetic tape are used as basic units (blocks) for recording and reproduction, and blocks (gap frames) in which valid data is not recorded are provided on at least one track at both ends of the block or on one side of the block. This is a magnetic recording method that rewrites the track of the
And an inspection word for detecting an error that occurs, an area (main code area) is arranged in the central portion of the track, and user data such as a data recording position in the main code area is managed. An area (subcode area) composed of data is arranged outside the main code area in a track. When rewriting a track in a block, only the main code is rewritten and the subcode A magnetic recording method characterized in that the area is not rewritten. 2. The magnetic recording method according to claim 1, wherein the block has a configuration in which a plurality of frames each including two tracks are assembled, and the subcode area region is formed in the subcode area region. The magnetic recording method, wherein the number of a frame to which the track belongs, the number of a block to which the frame belongs, and an area code indicating the type of the block are recorded. 3. The magnetic recording method according to claim 1, wherein the inspection word is added to the user data when the user data is recorded and recorded, and an error that occurs when the user data is reproduced is recorded. A magnetic recording method in which user data is reproduced by detecting / correcting with an inspection word, and error detection / correction is completed within each block of the magnetic tape. 4. A magnetic recording method according to claim 3, wherein a Reed-Solomon code is used as an error detection / correction code, and a code is selected so that a plurality of code words are not selected from the same track in a data track. The word is extracted, the error correction result by the double Reed-Solomon code completed in the track used in DAT (Dejital Audio Tape) is stored, and the error correction result is added to the data in the block by the Reed-Solomon code. A magnetic recording method characterized by being used as location information for erasure correction during error correction. 5. The magnetic recording method according to claim 1, wherein when the data cannot be correctly read due to damage or dirt on the tape, or when an error occurs during reading immediately after recording, the data is recorded. Data indicating that the containing block is unusable is recorded in the main code area in at least one or more tracks in the block, and the data indicating that the block is unusable is read during reproduction. The magnetic recording method is characterized in that the block containing the data is invalidated when the data is erased. 6. The magnetic recording method according to claim 2, wherein user data is not recorded on tracks in blocks included in the first at least 100 mm or more of the magnetic tape. A magnetic recording method characterized in that a code different from other blocks is recorded in the area code area in the sub-code area of a track. 7. The magnetic recording method according to claim 2, wherein EOI (End Of) indicating that user data recorded on the magnetic tape is not present in a block behind the user data.
Information) A magnetic recording method characterized by recording a code different from other blocks in the area code area in the sub-code area of the track in the block. 8. The magnetic recording method according to claim 2, wherein 32 frames are set as one block, the first 3 frames and the last 1 frame of the block are set as the gap frame, and the user data is set as a block. A magnetic recording method characterized by being configured as 128 Kbytes each.