JPH0296985A - Magnetic recording and reproducing device - Google Patents

Abstract

PURPOSE:To accomplish read/write in block units and to make the title device appropriate for a storage device for a computer by dividing a series of tracks constituted on a tape every plural ones and performing the recording and reproducing of user data in block units in a helical scanning system magnetic recording and reproducing device. CONSTITUTION:The title device records and reproduces the data on a magnetic tape in an azimuth recording system and a guard bandless system by performing helical scanning by the use of a rotary head like a conventional digital audio tape recorder. However, it performs the write and readout of the data in block units by dividing the tracks formed on the tape every specified number and setting them as a block, which is different from the conventional tape recorder. Namely, for example, two blocks are set as one frame and 32 frames constitute one block. The head three frames and the last one frame out of the block constituted of 32 frames (64 tracks) are set as a gap area and 28 frames between the above-mentioned frames are used as a data area.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a helical scan type magnetic recording and reproducing device using a magnetic tape, and particularly to a magnetic recording and reproducing device suitable as a peripheral device for a computer.

[Conventional technology]

With recent advances in magnetic recording technology, a helical scan type magnetic recording device for audio, which converts audio signals into PCM and records them on magnetic tape using a rotating head, has come into practical use, i.e., DAT (Digital Audio Tape Recorder). has been made into

This is a system that samples two channels of audio signals at a sampler frequency of 48kHz, quantizes them into 16-bit binary codes, and records them on magnetic tape.It takes two hours of recording time on a small cassette tape. It is possible. Regarding such magnetic recording and reproducing devices, for example, Japanese Patent Application Laid-Open No. 58-188314 and Japanese Patent Application Laid-Open No. 59-215013
This is stated in the bulletin etc.

[Problem to be solved by the invention]

In the above DAT recorder, 2 channels are 48k
PC sampled at I(z and quantized to 16 bits)
Since it is possible to record and reproduce M data for 2 hours, this amount of data is approximately 1.4 gigabytes when converted into bytes, which is a very large storage capacity. For this purpose, one A
T-recorders are also attractive as external storage capacity for computers.

However, DAT recorders originally record audio signals 'B11.
Since it was developed for general use, there are problems as described below as a computer storage device, and sufficient reliability cannot be obtained.

In other words, DAT recorders perform complete error correction encoding for each track formed on the magnetic tape, and cannot sufficiently correct errors that occur randomly or relatively short burst errors. can.

However, if data for one track is completely lost due to head clogging or scratches on the magnetic tape in the head scanning direction, errors caused by this cannot be corrected. Therefore, in the DAT recorder, we focused on the fact that there is a correlation between adjacent samples in the PCM data of the audio signal, and recorded the PCM data with interleaving between two tracks in advance. By deinterleaving, the samples are returned to their original arrangement, and even if one of the two rotating heads becomes clogged, the error caused by this is dispersed.
It is possible to reproduce the original audio signal through operations such as data interpolation and error correction.

On the other hand, data used by computers has no correlation between adjacent data, and for this reason, when recording and reproducing data directly on a DAT recorder, even if the error correction method for audio signals described above is used, the head It is not possible to correct errors caused by clogging, etc.

In addition, DAT recorders use an azimuth recording method in which adjacent tracks are recorded with different azimuths, and a guard panning method in which recording is performed so that the azimuth is changed between adjacent tracks, and guard bands are eliminated between adjacent tracks. A dress method is used. By the way, data is constantly being rewritten in external storage devices for computers, and when a DAT recorder is used as the external storage device, when this data is rewritten, the track width of the rotating head becomes smaller than the track width on the magnetic tape. Because it is wider than the width, not only the track on which data is rewritten but also a part of the adjacent track is rewritten, making it impossible to read data from this adjacent track.

SUMMARY OF THE INVENTION An object of the present invention is to solve these problems and provide a highly reliable helical scan type magnetic recording/reproducing device that can be used as an external storage device for a computer.

[Means to solve the problem]

In order to achieve the above object, the present invention divides a series of tracks formed on a magnetic tape into a plurality of tracks, each division is made into a block, and user data is recorded and reproduced in units of blocks. In the block, a predetermined area at the beginning and end of the block is used as a data recording area, and the user data and its parity for error detection and correction are recorded in the data recording area.

Further, the present invention encodes the user data and the parity so that they are completed within the data recording area, and thereby reduces the number of error correction code words to one track or a plurality of tracks in the data recording area. Make it different from the number of words in one frame. The number of words of the same error correction code word recorded on the same track is made smaller than the number of redundant words in the error correction code word.

Further, in the present invention, the number of bytes of user data in the data recording block is a power of two.

Further, the present invention has a memory for storing at least one block of information words and redundant words in the recording/reproducing device or its control device, and when recording, the first information of the first code word is sent from the host computer to the memory. The data is transferred in the order of word, second information word of the first code word, third information word of the first code word, and once stored in the memory, and the data transferred from this host computer is transferred. The generated redundant words are stored in the memory, and the information words and redundant words in the memory are used as the first information word of the first code word, the second code word, and the second code word.
the first information word of the codeword.

A third code word, a third information word, and so on are read out and recorded on the magnetic tape.

[For production]

In each block, the portion excluding the data recording area becomes a gap area where no user data is recorded, and a gap is formed between blocks. For this,
When user data is recorded in block units, the recording state of user data recorded in adjacent blocks is not affected by the presence of the gap area. Furthermore, since recording is performed in blocks rather than tracks, the recording state of user data is good on all tracks even within a block.

In addition, parity is added to each block of data for error detection and correction, so even if part of the data cannot be read correctly due to a clogged head or damaged tape, the error can be corrected. If it works and is beyond the scope of correction.

It works to output a flag that says it cannot be corrected, so
There will be no malfunction. If the data cannot be read or recorded correctly even after error correction due to damage to the magnetic tape due to repeated use or data errors due to non-uniformity of the media,
Since this block is used as a bat block and the entire block is not used thereafter, there will be no malfunction.

In addition, the number of error correction code words in one block is configured so that it does not match the number of words in one track or one frame, and furthermore, interleaving is performed within one track or one frame during recording, and the interleaving is restored during playback. By performing deinterleaving, it becomes possible to disperse the information words and parity words constituting one code word on the magnetic tape. This makes it possible to prevent data errors from becoming uncorrectable when concentrated data errors occur due to dust or scratches. A Reed-Solomon code can be used as the intra-block error correction code,
This makes it possible to perform highly efficient encoding with high error correction ability, and it is also possible to make the circuits for error correction encoding and decoding operations smaller and faster.

Also, after performing intra-block error correction encoding, DA
Recording is performed by performing intra-track error correction encoding that is completed within a track such as T, and during playback, error correction is performed using the intra-track error correction code, and at the same time a flag indicating the correction status is added, and error correction within the block is performed. By using the flag as error position information when decoding a code, erasure correction becomes possible in intra-block error correction, and error correction capability is significantly improved.

A typical computer exchanges data with an external storage device in units of 512 or 1024 bytes. Therefore, by setting the number of bytes of the part used as user data among the information words in one block to a power of 2, the present invention becomes easy to use as a storage device for a computer.

Furthermore, by setting the data read order in the memory, the present invention can perform data transfer from the host computer and error correction encoding operation in parallel, increasing the data transfer speed from the host computer. It becomes possible to perform a recording operation at high speed without increasing the speed.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

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

For this purpose, a DAT recorder will first be briefly explained.

A DAT recorder uses two rotating heads, +azimuth and -azimuth, which alternately scan the magnetic tape to record data one track at a time diagonally on the magnetic tape. Play data. During playback, servo is applied so that each rotary head plays back tracks with matching azimuths. The track width of these rotating heads is set slightly wider than the width specified by the standards for tracks formed on magnetic tape, and when recording, the next track is overlapped with a part of the previously formed track. form a track. That is, as shown in FIG. 16, the other rotary head B forms the next track, overlapping a part of the trunk formed by one rotary head A, and the trunk formed by rotary head B. The rotary head A forms the next track so as to overlap part of the track. In this way, each track formed has a width defined by the standard.

By recording as described above, there is no guard band between adjacent tracks, increasing the recording density, and since the azimuth is different between adjacent tracks, crosstalk from adjacent tracks during reproduction can be prevented.

In a DAT recorder, tracks are formatted, and FIG. 17 shows the track format standardized by the DAT conference. According to this, the track is divided into 16 areas, and each area is defined as shown in Table 1 below.

Table 1> In Table 1 above, area 9 is the area where PCM data is recorded. Areas 3 and 14 are subcode areas, which are defined as areas for recording information such as song titles and performance times. Also, the area 6°10 is A T F (Auto
In this area, a signal for applying servo so that the rotary head can correctly scan a track during reproduction, that is, an ATF signal is recorded.

FIG. 18 shows a track pattern on a magnetic tape formatted in this manner.

In order to be able to use such a DAT recorder as an external storage device for a computer, it is necessary to be able to search for any location on the magnetic tape and rewrite data at this location.

However, as mentioned above, since the track width of the rotating head is wider than the track width on the magnetic tape, the first
As shown in FIG. 9, when an attempt is made to rewrite data on a certain track, the rotating head also hits the track adjacent to this track, and the adjacent track is scraped.

Therefore, the width of this adjacent track becomes narrower, making it impossible to obtain a reproduction output of a sufficient level, making it impossible to read data.

In the DAT recorder, the P shown in Figure 17 in the track is
The CM area 9 and subcode area 3.14 are further divided into smaller blocks.

Table 1 above also shows the number of small blocks in these areas. For areas other than areas 9, 3, and 14, the number of small blocks corresponding to each length is shown.

FIG. 20 shows the configuration of this small block. The small block includes an 8-bit synchronization signal area, an 8-bit identification code area W1.1, and an 8-bit small block address area W2.1 representing the address of the small block in the first track.
8-bit parity P for error detection in areas Wl and W2
It also consists of a 256-bit area for recording data and parity. Here, the parity P is P=W1+W2 However, the code sign represents exclusive OR (mod2).

It is expressed as In area W2, the MSB (most significant bit) is used to indicate whether the small block belongs to the PCM area or the subcode area; in the PCM area, MSB="0'''; in the subcode area, MSB="1". Regarding the remaining 7 bits in area W2,
These 7 bits become a small block address in a small block in the PCM area, and the lower 4 bits in a small block in the subcode area become a small block address.

The DAT recorder has been briefly described above, and next, an embodiment of the magnetic recording/reproducing apparatus according to the present invention will be described with reference to FIGS. 1 to 15.

"Block Configuration" It is similar to a DAT recorder that data is recorded and reproduced on a magnetic tape by helical scanning using a rotating head and by an azimuth recording method or a guard panless method. However, one difference from DAT is that the tracks formed on the magnetic tape are divided into a predetermined number of blocks, each division is defined as a block, and data is written and read in block units. Here, it is assumed that two tracks constitute one frame (in DAT, two tracks are also called one frame), and one block consists of 32 frames.

FIG. 2 is a diagram showing the block configuration on the magnetic tape 1. In a block consisting of 32 frames (64 tracks), the first three frames and the last frame are gap areas, and the 28 frames between these gap areas are It is used as a data recording area. Data is recorded in this data recording area. Gap areas are not used for data, but rather provide gaps between blocks.

Data writing and reading are performed in block units.

For this reason, data rewriting is also performed in block units. By rewriting this data, the last track of the preceding adjacent block and the 19th track of the following adjacent block are
As in the case shown in the figure, the respective widths are narrowed and signals cannot be reproduced from there, but these tracks belong to the gap area, and the data recording areas of these adjacent blocks are not affected in any way. Further, in the block where data has been written, since recording is performed on all tracks, the recording state on these tracks is good.

In this way, it is possible to prevent data in other parts from becoming unreadable due to the data rewriting described above, as in the case where the DAT is directly used as an external storage device of a computer.

"Subcode Format" The structure of the track 2 (FIG. 2) formed on the magnetic tape 1 is the same as that of the DAT recorder shown in FIG. 17.

FIG. 3 shows the format of the subcode area in this embodiment. In this subcode area, one subcode is made up of two consecutive small blocks, and FIG. 3(a) shows these two small blocks making up the subcode.

In FIG. 20(a), areas Wl and W2 are respectively areas W1 and W2 in FIG. ．． Corresponding to W2, 8 bits are allocated to each. These areas Wl.

The contents of W2 are shown in Table 2 below.

Table 2> Data ID: 0001 Format ID: 11] - Control ID: 0O00 As shown in Table 2, the area W1 of the preceding small block shown in the upper row of FIG. The bit control ID and data ID are recorded, and the area W
2 has the MSB of 11111 representing the sub-food area,
3-bit format ID.

Each 4-bit small block address is recorded.

The control ID is 4 pins 1 based on the DAT standard.
- Both are set to "On."The data ID is set to "O" to distinguish it from an audio magnetic tape ("OOOO").
OO1''. Also, as shown in FIG. 3(a), ]
Since 7 banks are used for one subcode, the format ID is "111" according to the DAT standard.
”.

In the next small block shown in the lower part of FIG. 33(a), as shown in Table 2 above, an 8-bit program number, which will be described later, is recorded in area W1, and M of "1" is recorded in area W2.
SB, a 3-bit program number and a 4-bit small block address are respectively recorded.

In the data+parity area in two small blocks shown in FIG. 3(a), seven packs 1 to 7 of 64 bits are provided, and the remaining 64 bit area is used as a parity C1 area. In each pack, as shown in FIG. 3(b), eight codes PCI to PC8, each having 8 pins I-, are defined.

PCI and PC2 record block numbers for access. The block number is represented by 16 bits, with the upper 8 bits being pci and the lower 8 bits being PC2. The block number represented by this is 0 to 6 in decimal notation.
It is 5536.

PC3 records the frame number within the block. This indicates the frame number within the block to which this small block belongs.

PC4 is parity P1 for error detection of PC1 to PC3 defined by the following equation.

P1=PC1+PC2+PC3 However, the code sign represents exclusive OR (mod2) (the same applies hereinafter).

PC5 and PC6 are codes, or areas, for determining which area the block belongs to (this does not refer to the gap area or data recording area shown in Figure 2; this area will become clear later). This is an area for recording marks.

PC7 is a parity P2 for error detection of PCI to PCB defined as follows. Here, this parity P
Starting from the least significant bit of 2, bit 0, bit 1, and so on. ...
..., bit 7 = bit 0 of P2 = parity P2 of PC1 bit 1 = bit 2 of parity P2 of PC2 = bit 3 of parity P2 of PC3 = bit 4 of parity P2 of PC4 = parity P2 of PC5 Bit 5 of PC6 = Bit 6 of parity P2 of PC6 = Bit 7 of parity P2 of bits O to 5 == LL Q 111) C8 is parity P3 for error detection of PC, l to PC7 defined by the following equation.

P3=PC, 1+PC2+PC3+PC4+PC5+P
C6+PC7 The above codes PC1 to PCB all have the same content, and the third
The information is recorded in each pack 1 to 7 in Figure (a).

(ID format in PC:M area) As in the DAT standard shown in Table 1 above, the PCM area in which data is recorded is divided into 128 small blocks. These small blocks basically have the same configuration as in FIG. 3, but the area W
2, a 7-bit small block address is recorded with the MSB set to "0", and the data+parity area is different from the subcode area in that data and parity are recorded.

In the PCM area, eight small blocks are grouped together, and each ID shown in Table 3 below is defined in these eight areas W1.

In Table 3, 2 bits of ID-, L are respectively stored in area W1.
.

ID-2 and a 4-bit frame address are recorded, and similarly, two 2-bit IDs and a 4-bit frame address are recorded in the third, fifth, and first small block areas W1, respectively. The areas W1 of the six second, fourth, sixth, and eighth small blocks to be recorded are areas for option codes. In addition, in the area W2 of these eight small blocks, as explained earlier, the MSB is
10" and 8-bit information is recorded with 7 bits as a small block address. Table 4 below shows the contents of each ID in Table 3.

Below, the margin, that is, the first small block of eight small blocks <Table 4
> In Table 4, ID-1 is set to rt O1pp to indicate that it is used for recording other than audio signals.

When ID-2 is 00'', it indicates that one block consists of 32 frames as in this embodiment.''01''
"10" and "11'" are reserved for future expansion.

ID-3 represents the use of error correction and its format. 100011 indicates the layered FCC format used in this embodiment, which will be described later.

1101 ++ and 11101'' are reserved for expansion, and "11" represents a mode in which layered FCC is not used.

ID-4 indicates the access format as a computer peripheral device. "00" is the same sequential access as a magnetic tape device such as a streamer, +101 ++
indicates a device that can be accessed directly, such as a floppy or hard disk, or a magnetic tape. "10",
Keep "11" as a reserve.

For ID-5, "00" indicates that this block 1 is a block that can be read/written correctly, and "11" indicates that it is a BAD block that cannot be read/written correctly due to damage to the tape.11011+
“10” is reserved.

ID-6 defines the track pitch, "OO" indicates a standard 13.6 .mu.m track pitch, and 1101+1 indicates a magnetic tape recorded or to be recorded at a 20 .mu.m track pitch.

ID-7 indicates permission or prohibition of software copying.
"o o '" indicates permission, and ill 10 ++ indicates prohibition. #OL II and H11++ are reserved.

ID-8 is always O.

"Tape Format" Figure 4 shows the tape format according to this embodiment. Figure 4 (a) shows the format divided into blocks, and Figure 4 (b) shows the format divided into blocks in detail. This is shown in .

Here, each block and each frame is shown divided into a PCM area and a subcode area, and each of these areas is shown divided into an ID recording area and a data recording area. However, as the ID of the PCM area, many IDs are recorded as shown in Table 4, but here, ID-5 (read/write is possible or not) is recorded, which is particularly important in terms of the tape format. Only the ID indicating the ID) is shown.

Good blocks are blocks that can be read/written.
Let the impossible block be a BΔD block.

Further, in the subcode area, only information necessary for access is shown.

In the data area of the T-' CM area, data used by the user and parity for error detection and correction, which will be described later, of this data are recorded.

In the subcode area, the program number is the 16-bit block number consisting of the codes PCI and PC2 in Figure 3(b), and the information from the 11th bit to the upper 11th bit, including the fourth bit from the lowest, is assigned. It will be done. In other words, the least significant bit of the block number is bit O, and bits 1, . . . Bit 2. ..., bit 15, bits 4 to 14 become the program number. In this 11-bit program number, as shown in Table 2 above, 8 bits are allocated to area W1 of the subcode area, and 3 bits are allocated to area W2. This program number has the same content in 8 consecutive block circles, and the value increases by 1 every 8 blocks. In FIG. 4(b), block m is the last block of eight blocks, and if the program number at this time is P, then the next block (m + 1)
The eight program numbers from then on are (P+1).

This program number can be roughly used for searching in units of eight blocks.

The area mark (area code in Figure 3(b)) has all pits '0'', indicating that the block is a normal data block.The block number is 1 for each block.
However, for each frame of the same block,
As shown in FIG. 4(b), the same block numbers are recorded. In addition, 0.1, 2.・
..., a frame number of 31 is assigned.

PC recorded in the gap area in each block
All M data are set to 0''.

The above is a description of the area (hereinafter referred to as the data area) consisting of blocks in which data actually used by the user is recorded. However, a DAT magnetic tape is provided with transparent tape as a leader tape at the beginning and end of the tape. With a DAT recorder, the beginning of the tape is
(Begin ofTape), end of tape E O
T (End of Tape) is being detected.

However, since the mechanism for pulling the magnetic tape out of the tape cartridge and loading it into the rotating cylinder differs from device to device, the position of the first block varies. In order to absorb this problem, as shown in Figure 5(a), the lead-in (Lead-in) area, which is not treated as a valid data area, is
in) area.

This lead-in area has a length of 100 mm or more in DAT, but in this example, it is composed of 16 blocks (approximately 130 mm), and the PC of each block is
All PCM data in the M area is set to "O". In the subcode area, the program number is as shown in the figure.
-2” for the first 8 blocks, rt- for the remaining 8 blocks
Let it be 1n.

The area mark shall be “BB” in ASCII code.

The block numbers are "-16"-15"...-1" in order from the first block in this lead-in area, and the block number is 0 in the first block of the data area.
Make it so that

Further, as shown in FIG. 5(b), following the last block of the data area, there is an E OI indicating the end of the data area.
(End of Information) area is provided. This EOI area consists of one block, and all PCM data in the PCM area is "OI+".

In addition, the program number, block number, and frame number following the data area are recorded in this block, but in order to distinguish it from the lead-in area and the data area, the area mark is set to "EE" in ASCII code.

These lead-in areas and EOI areas make it possible to determine the beginning and end of the data area in which PCM data used by the user is recorded.

Next, according to FIG. 6, BΔD which cannot be read/written
Explain blocks.

After data is recorded in a block, it is read out and checked to see if it has been recorded correctly, and if a defect is detected, the block is designated as a BAD block.

In the figure, it is assumed that data is being recorded while formatting block number m+9 backwards from this block. If this block becomes a BAD block, move the tape back a little and add ID-5 in the ID of the PCM area to this block.
(Table 32 Table 4 above) shows the code "1" indicating the BAD block.
While recording 1'', format this BAD block again with all PCM data as 0''. Then, format the first block, but at this time, leave the block number of this block as m + 9. Then, the PCM data that was supposed to be recorded in the previous block is recorded again.Therefore, blocks with the same block number are lined up.

On the other hand, when new data is recorded in a block where data has already been recorded due to random access, if the block becomes BAD due to damage to the tape, etc.
In the same way as above, the code 1'0" (7) PCM data indicating the BAD block is recorded in ID-5 of the PCM area ID. At this time, the contents of the subcode area are the same as those previously recorded. the same as the content.

However, in this case, since it is not possible within the device to decide where to place the block to be replaced, the host that instructs the writing is notified of the error, and subsequent operations follow instructions from the host.

"Layered ECCJ DAT recorders perform error correction using a duplicated Reed-Solomon code that is completed within a track. This allows symbol errors on magnetic tape to be reduced to one.
0-3, the error rate after error correction is 10-”
It is as follows.

However, it has been experimentally confirmed that the symbol error on the magnetic tape worsens to 10<-2> or more due to stains or damage to the magnetic tape, and due to the longevity of the magnetic tape due to repeated use. The error rate after error correction at this time is 10-
12, and as a computer data storage device,
It becomes unreliable. In addition to the normally occurring random errors and relatively short burst errors, data errors may occur due to head clogging, scratches on the tape in the head scanning direction, and the like.

Therefore, even in such a case, the error rate is 1o-1.
Parity is added to the data for error correction across a plurality of tracks so that the number of errors is 5 or less. This error correction code will be called layered F, CG.

Layered ECC uses format 1, which is completed with 28 frames for data in one block, in order to enable access in units of blocks.

Therefore, data in which PCM data used by the user and parity for error detection and correction are added is recorded in 28 frames belonging to data areas with frame numbers 3 to 330 within the block.

PCM data and parity are arranged as shown in FIG. The array shown in FIG. 1 indicates the format of layered ECC, and can also be seen as an array in memory for encoding into layered ECC and decoding thereof.

In this embodiment, one block of PCM data and parity is 38 bytes horizontally (in the X direction) and 4 bytes vertically (in the X direction).
They are arranged in 244 rectangles. Therefore, this PCM data and parity are represented by the xJI index, which takes an integer value from 0 to 37, and the
Let's use dimensional coordinates.

One codeword of the layered FCC consists of one row in the X direction. That is, one code word is composed of 38 bytes of data having the same X coordinate. Of this code word, 32 bytes with X coordinates from O to 31 are data, and 32 to 3
The 6 bytes up to 7 are used as the parity of this data.

A Reed-Solomon code with a code length of 38, number of information words of 32, and inter-code distance of 7 is used as the layered FCC code.

Here, when data is input from an interface and subjected to layered FCC encoding and recorded, the input data is stored in the memory in the order of the X direction for encoding. As a result, even if the transfer of all data for one block is not completed, if the data portion of one code word is transferred, it becomes possible to encode that code word and generate parity. . This reduces the access time during recording. Note that during playback, data is read from this memory in the same order in the X direction, and rQ raw data is output from the interface.

On the other hand, one block of coded data and parity are recorded on the tape in order in the X direction. That is,
5760 bytes from (0,0) to (1,1515) are recorded in frame number 3, and (1,151
6) to (2°3031) are recorded in frame number 4, and so on. However, the data of each frame is recorded after being subjected to intra-frame interleaving and intra-track error correction encoding used in DAT.

During playback, the FCC of the DAT is similarly decoded,
After the intra-frame interleaving is restored to its original state (deinterleaving), the data is stored in the memory in the order of the X direction, and layered FCC decoding is performed.

In the case of decoding, a flag (uncorrectable flag) is added to data that cannot be corrected by the FCC of the DAT.

Use this as a pointer to the location of the error.

Furthermore, as a result of FCC decoding of DAT, there is no uncorrectable error.
If it can be determined from the CI flag, uncorrectable flag, etc. that sufficient reliability is ensured, the layered F
It is also possible to omit the CC decoding process. Thereby, the access time during playback can be shortened.

The total data shown in Figure 1 is 38X4244=1612
Although it is 72 bytes, since the frame for data is 28 frames, 28×5760=161280 bytes, which leaves an 8-byte margin.

This is sufficiently small for the recording capacity of one group, so
In this embodiment, this extra 8-byte area is not used for recording data, but O is recorded therein.

Among the data of one block shown in FIG. 1, the X coordinates are from 0 to 31. The area in which the X coordinate ranges from 0 to 4095 is defined as a user data area where the user's PCM data is actually recorded. 32x4096=131072 in this part
Bytes of user data are recorded. The remaining X coordinate range from 0 to 31 and the remaining X coordinate range from 4096 to 4243 is the system data area. The system data area is 3
2x148 = 4736 bytes, and when one group is divided into logical records of a size such as 512 bytes, it records information for managing user data such as logical record delimiter information, its length, and address information. used for Since the size of the user data area is 131,072 bytes, which is exactly 2 to the 17th power, it is suitable for use as a conventional computer storage device or as a replacement for a conventional storage device.

Next, interleaving will be explained with reference to FIG.

FIG. 5A shows an example of a block format when no interleaving is performed. The y (vertical) direction is the longitudinal direction of the frame, and the x (horizontal) direction is the direction for each frame. The size in the X direction matches the data capacity of one frame,
According to FIG. 1, it is 5760 bytes. As explained earlier.

This is because 131,072 bytes of user data are recorded in the 28 frame data recording area.

131072/5760 = 23 frames are the data part, and the remaining 5 frames are the parity part.

At this time, the code word of layered FCC is as shown in Fig. 7(a).
It is composed of data arranged in the X direction, as shown by diagonal lines. In addition, since the data of each frame (data in the X direction) is recorded in an array in the longitudinal direction of the track on the magnetic tape, the data on the magnetic tape will end up as shown in the rows of . in FIG. 7(b). , error correction codewords are recorded in the longitudinal direction of the tape. However, in the figure, Nu indicates the running direction of the magnetic tape, and V< indicates the scanning direction of the rotary head.

By the way, there are tape guides that regulate the running of the magnetic tape and vintilonilla that drives the running of the magnetic tape.

The magnetic tape may be scratched in the longitudinal direction by capstans, etc., or dust may adhere to it.

In such a case, a data error occurs in the longitudinal direction of the magnetic tape, but if the layered FCC code word is recorded in the longitudinal direction of the magnetic tape as described above, this entire code word or An error occurs in a range that exceeds its correction ability, and cannot be corrected even with layered ECC.

To prevent this situation, it is preferable that the number of bytes in the X direction not match the number of bytes in one frame in the layered ECC format. The case of this example is shown in Fig. 7(c).
Shown below. Here, increase the X direction by 1 to 38,
Instead, the X direction is set to 4244 bytes. FIG. 1 shows the details.

In this block format, 1
When recorded on the magnetic tape as explained in the figure, the code words of each layered FCC are recorded in a distributed manner within the data recording area of the block on the magnetic tape, as shown in FIG. Generally, in magnetic recording, data errors tend to occur in a concentrated manner, and from this point of view as well, it is desirable that data be distributed.

However, the interleaving performed in FIG. 7(d) is simple and does not provide sufficient data distribution.

For this reason, interleaving is further performed in each frame. This embodiment further employs this intra-frame interleaving, and uses the same interleaving method as DAT. FIG. 7(e) shows the recording state.

Next, the relationship between layered FCC code words and interleaving will be explained.

According to the block format shown in FIG.

For example, there are parts where 2 bytes from the same frame are extracted into the same code word of the layered FCC, such as data at coordinates (0, O) and (0, 1) of frame 3, which is disadvantageous in terms of data distribution. It appears to be. But 1
As described below, this point is actually advantageous.

First, regarding intra-frame interleaving, the intra-frame interleaving used in DAT has an 8-byte cycle, and if you sequentially number the data of one frame from 0 to 5759, the remainder when you divide that number by 8 is 0.
1, 6.7 on a track with 10 azimuths, 2, 3,
4.5 is recorded on the -azimuth track. In addition, in this embodiment, the number of bytes in the X direction is 4,244, which is a value with a remainder of 4 when divided by 8, so that after performing intra-frame interleaving, the same code word will not be included in these same frames. data is distributed to tracks with different azimuths.

For example, considering a code word whose y coordinate is O, the 0th data and 4244th data of frame number 3 are included. Here, the 0th data is recorded on the +azimuth track because the remainder of 0 divided by 8 is ○,
The 4244th data is 4244 divided by 8 with a remainder of 4.
Therefore, it is recorded on one azimuth track. Similarly, all other parts will be recorded on tracks with different azimuths. This is effective not only for data distribution but also for data errors due to head clogging.

In this way, in the case of this embodiment, in the same frame of data, only one byte is included in the same code word in the same track, and data errors caused by scratches in the head scanning direction or head clogging can be avoided. However, the maximum correction ability can be demonstrated.

Also, as mentioned above, the size of the user data area is 1
Since it is 31,072 bytes, which is exactly 2 to the 17th power, it is suitable for use as a conventional computer storage device or as a replacement for a conventional storage device.

Furthermore, by setting the number of layered ECC code words in one block, that is, the number of bytes in the X direction (4244) to a value different from the number of bytes in one frame (5760), and using intra-frame interleaving, Since the two code words are distributed and recorded within the group, uncorrectable data errors are less likely to occur.

Further, by employing the Reed-Solomon code as the error correction code of the layered FCC, the efficiency and correction ability of the layered FCC are high, and the circuit for performing encoding/decoding processing can be made relatively small and high-speed.

In addition, when recording data, the data sent from the interface is stored in memory in the can be shortened.

A Reed-Solomon code is also used for this error correction, with a code length of 38 and an information word length of 32. A code with a check word length of 6 is used.

Since the distance is 7, it is possible to correct up to 3 errors, but in ``error correction'' where the location of the error is not known, it takes a lot of time to correct 3 errors. Therefore, in this embodiment, up to two "error corrections" are performed.

However, as a result of the error correction performed by the DAT recorder, if the error correction was performed on the data read from each track but could not be corrected, a flag indicating this is added, so this flag This allows you to pinpoint the location where the error occurred. Therefore, if "erasure correction" is performed based on this result, it is possible to correct up to six errors.

This correction algorithm is shown below.

N (E) =O N (E) =1 N (E) =2 N (E) >2 N (F) =O N (F) =1 1 error correction No error 2 error correction F = 1 1 double erasure Correction + 2 error correction Double erasure correction + 2 error correction N (F) = 2 N (F) = 3 Triple erasure correction + 1 error correction N (F) = 4 Quadruple erasure correction + 1 error correction N (F) = 5 5 Double erasure correction N (F
) =6 6-fold erasure correction N (F) >6
F=Madashi, N (F):
Number of uncorrected symbols by double Reed-Solomon code N (E) : Number of error symbols by layered FCC F : Error flag With this algorithm, even if burst errors occur in 6 or more tracks, they can be corrected and are extremely reliable. This makes it possible to realize a device with high performance.

"Structure and Operation" FIG. 8 is a block diagram showing an embodiment of the magnetic recording/reproducing apparatus according to the present invention. 10.
A cylinder motor 11 rotates this rotary cylinder, a capstan motor 12 performs constant running of a magnetic tape (not shown). Reel motor 1 for fast forwarding and rewinding
3. Servo circuit 14 that controls these motors
, a read/write circuit 15 that writes and reads signals using a rotating head, a DAT signal processing circuit 16 that processes signals according to the DAT format, a subcode control circuit 17, a device mode instruction, a mechanical A drive control circuit 18 that instructs driving, a mechanical drive circuit 19 that receives this signal and drives the mechanism, a DAT encoding/decoding circuit 20 that encodes and decodes an error correction code according to the DAT format, and a layered ECC circuit. 21. A memory 22 for storing data and performing layered ECC encoding/decoding, a memory control circuit 23 for controlling reading and storing data from this memory 22, a system control circuit 24 for controlling all of these, and a connection with the host. An interface S CS I (Small
Computer 5ystea+ Interfa
ce) Consists of 25.

The operation sequence of the system in this embodiment will be explained using FIG. 9.

First, check whether 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 is inserted into the system by a detection circuit (not shown), it drives a loading motor (not shown) to pull out the magnetic tape from the cartridge and load it into the rotating cylinder 10.
Wrap it around.

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

Drive control circuit 18 receives this signal and sets servo circuit 14 to a "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 "stop" mode,
It notifies the system control circuit 24 that the "rewind" has been completed.

The system control circuit 24 uses the drive control circuit 1 to know how far data is recorded on the magnetic tape.
Instruct 8 to ``Search''.

The drive control circuit 18 sets the servo circuit 14 to a "high speed search (described later)" mode and drives the reel motor 13 and cylinder motor 10. The R/W circuit 15 takes in the output from the rotary head attached to the cylinder 10, shapes the waveform into a digital signal, and then inputs it to the signal processing circuit 16.

The signal processing circuit 16 reads the subcode from this signal and sends it to the subcode control circuit 17. The subcode control circuit 17 checks whether the reproduced subcode is correct, and if it is correct, sends the subcode to the drive control circuit 18, and the drive control circuit 18 checks the area mark in the subcode described above and determines whether the EOI area is correct or not. Find out.

Upon detecting the area code indicating EOI, the drive control circuit 18 temporarily sets the servo circuit 14 to stop mode, and then
Rewind the overrun magnetic tape a little in rewind mode, and read the EO nib block in "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 sends the block number in the subcode to the system control circuit 24.

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

At the same time, the signal processing circuit 16 transfers the PCM region ID of the data area of the EOI block to the drive control circuit J8, and sets the various parameters described above.

Even if the search is performed to the end of the magnetic tape, the EO
If the I area is not found or nothing is recorded on the magnetic tape, register it as a new tape.

Next, the system control circuit 24 waits for a command sent from a host (not shown) via the interface 25.

The system control circuit 24 receives commands from the interface 25, analyzes the commands, and drives the drive control circuit 18 and memory control circuit 23 to execute commands such as read, write, and format.

Tape cartridge eject command from host,
Alternatively, if an instruction is given by an eject switch (not shown) attached to the device, an 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, it also sends the drive control circuit 18 to "search 1" for this block number as described above. , servo circuit 1
4. Drive the subcode control circuit 17 to find the final block. After this, an EOI area is recorded next to the last block in the format of the EOI. At this time, the system control circuit 24 sends the subcode data of the E○■ block to the subcode control circuit 17 via the drive control circuit 18. In addition, data to be recorded in the data area in the plotter is prepared in the memory 22, the code for error detection and correction is added to the memory control circuit 23 via the layered ECCM path 21, and the data is transferred to a predetermined location in the memory 22. Store.

Encoding/decoding circuit 2
0, and encoded into data formats 1 to 1 according to the DAT standard.

If the final block can be searched, drive control circuit 18
sets the servo circuit 14 to the 1'play' mode and reads the data.While checking the frame number of the subcode read by the subcode control circuit 17, when this block is completed, the signal processing circuit 16 is set to the 'write' mode. The signal processing circuit 16 instructs the encoding/decoding circuit 20 to read data.In response to this, the encoding/decoding circuit 2
0 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 into data according to the data funi mat and records it via the R/W circuit 15.

When the recording of the EOI block is completed, the drive control circuit 18 rewinds the magnetic tape in the same manner as described above, and then ejects the tape cartridge using a loading motor (not shown).

"Read" FIGS. 10 and 11 show timing diagrams for reading the blocks of this embodiment.

At the time of reading, the drive control circuit 18 controls the servo circuit 1
4 to Pabray" mode and the signal processing circuit 16 to read mode. As a result, the servo circuit 14 controls the reel motor 13 and moves the magnetic tape to the standard 8" mode.
, 15 mm/sec, and the rotational speed of the rotary cylinder 10 is taken in and controlled by the servo circuit 14 so that the rotary cylinder 10 rotates at a constant speed of 2000 rpm. The switching signals for the two rotating heads A and H attached to the rotating cylinder 10 180 degrees opposite each other are 15
This is a signal for switching between rotary heads A and B that operate every m5ec.

In FIG. 10, from time T1 to T2, the signal is read by the +azimuth rotating head A, and from time T2 to T3, the signal is read by the -azimuth rotating head B.

The rotary head reads the ATF signal recorded on the magnetic tape, supplies it to the signal processing circuit 16 via the R/W circuit 15, and inputs it to the servo circuit 14 as a servo signal. The servo circuit 14 controls the rotational speed and phase of the capstan motor 12 so that the rotary heads A, R can correctly trace the tracks on the magnetic tape.

In this state, the output of the signal from the rotary head through the R/W circuit 15 is as shown in the figure.

This signal includes 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□. The subcode control circuit 17 that receives this signal reproduces the subcode from this signal and performs error detection using the parity indicated in the format of the subcode ID. Regarding pack data, errors are detected using parity in the DAT small block format, and at the same time, errors are detected by comparing a plurality of back data to see if they match.

The result is sent to the drive control circuit 18 at time T1m.

On the other hand, among the signals input to the signal processing circuit 16, PCM
The signal of the area is stored in the signal processing circuit 16 once.

At time T2, from rotating head A to -azimuth rotating head B
and read the data in the same way. In this way, 1
After reading one frame's worth of data, the data of the ID in the PCM area is transferred to the drive control circuit 18 at time T.

One frame worth of data read by the rotary heads A and B is transferred to the DAT encoding/decoding circuit 20 by the signal processing circuit 16 from time T4. The encoding/decoding circuit 20 which has received one frame of data performs error detection and correction according to the format of f) AT, and requests the memory control circuit 23 to transfer the data.

If the frame number in the subcode read from time T1 to time T is 2, this frame is in the gear block area as shown in the block format and is not treated as valid data, so the memory control circuit 23 The data transferred from the encoding/decoding circuit 20 is not stored in the memory 22, but the data of the next frame number 3 is stored in the memory 22. Here, the memory 22 is divided into three areas, each of which is a memory area A.

Let them be B and C.

In FIG. 11, the block data signal represents the timing at which the memory control circuit 23 stores the data of frame numbers 3 to 30 of the n-th block into the memory area A of the memory 22.

At this time, as a result of error correction for the DAT output from the encoding/decoding circuit 20, the uncorrected flag is stored in the memory area A of the memory 22 together with the data, and one block of data is stored in the memory area A. 1 time after storage T2□
Then, the memory control circuit 23 reads out this data, transfers it to the layered ECC circuit 21, and performs error correction according to the layered FCC error correction algorithm. At this time, a maximum of six errors are corrected using the previously stored uncorrected flag.

The corrected data is stored in memory area A of the memory 22.

However, if the FCC prohibition mode is specified in the ID data of the PCM area, the above error correction is not performed and the data from the encoder/decoder circuit 20 is stored as is in the memory area A of the memory 22. .

Thereafter, at time T2□, the system control circuit 24 transfers the decoded data in the memory area A of the memory 22 to the host via the interface 25.

At time T23, the data signal of the primary block is encoded/
Since it is output from the decoding circuit 20, this data is stored in memory areas B and C of the memory 22. In this manner, after the data transfer to the host is completed, by storing the data from the encoding/decoding circuit 20 in this memory area, each process can be executed without waiting time. this is,
This is because when the host's preparation for data transfer is delayed and the timing indicated by the broken line in the figure occurs, three transfers need to be executed simultaneously.

"Write" Figures 12 and 13 show data recording timing diagrams.

When a data recording request is received from the host at time T□1, the system control circuit 24 instructs the memory control circuit 23 to transfer data to the memory area A of the memory 22. The memory control circuit 23 that receives the data transfer request via the interface 25 receives the data and stores it in the memory area A of the memory 22.

Thereafter, at time T32, the system control circuit 24 instructs the memory control circuit 23 to transfer data to and from the layered ECC circuit 2, and performs encoding in the layered ECC format. However, if a mode in which layered ECC is prohibited is specified by an instruction from the host, this encoding is not performed. This data is stored in the memory 22
It is stored in memory area A of .

If there are still data transfer requests from the host,
The system control circuit 24 sets the memory control circuit 23 to store this data in memory area B of the memory 22. If there is a further data transfer request, it is stored in memory area C of the memory 22. Time T 3
In part b between m t T 33, data in memory area A of memory 22 is encoded, and data from interface 25 is stored in memory area B. Similarly, data is stored in memory area C in section C from time T33. Now memory 22
are all in use, so they notify the host that they are in use and request that data transfer be temporarily suspended.

Thereafter, when the encoding of memory area A is completed at time T34, the encoding of memory area 3 is started by the system control circuit 2.
4 instructs the memory control circuit 23 (section d).

After this, at time "3S", the memory control circuit 23
is the encoded data in memory area A as l) A
'r transfer to the encoding/decoding circuit 20. Further, when the encoding of memory area B is completed, the system control circuit 24 instructs the memory control circuit 23 to encode the memory area C from time T36 (copy 0).

When the transfer of one block of data from memory area A to encoding/decoding circuit 20 is completed at time T37, memory area A becomes vacant. Therefore, the system control circuit 24
notifies the host that the suspension of data transfer will be canceled, and stores the data transferred from the host in memory area A. At part f in the figure, all the memories are in use, so the +11 and system control circuit 24 request the host to interrupt the transfer. At the same time, the encoded data from memory area B is transferred to the encoding/decoding circuit 20.

When the data transfer from the host ends at time T33, the system control circuit 24 instructs the memory control circuit 23 to encode the data in memory area A (g
Department). Repeat this from now on.

FIG. 13 shows a frame recording timing diagram.

First, data is read in read mode. Assuming that the block number to be recorded is N, the drive control circuit 18 checks the block number in the subcode data and detects the block number immediately preceding the target block number (N-1 in this case). Furthermore, when the frame number is inspected and frame number 30 is detected (time T41) -
Data that has been previously transferred to the encoding/decoding circuit 20 and encoded according to the DAT standard is transferred to the signal processing circuit 16. Here, the signal processing circuit 16 creates a small block format within the track.

After receiving the subcode data from the subcode control circuit 17 and reading frame number 31, the read mode is switched to the write mode at 9 time T48, and the R/W circuit 1
5, the block with block number N is recorded in the order of +azimuth track and -azimuth track.

During this mode switching, when the signal from the ATF area that was used as 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 making the first three frames of the block a gap area (Fig. 2).

"Dub Recording" When rewriting data in a block that has already been recorded, there is no need to rewrite the subcode area and ATF area in the track format described above. Therefore, in this case, only the PCM area is rewritten. Dubbing this (
After recording). 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
All the data for one track outputted through the W circuit 15 was recorded, but during dubbing, as shown in FIG. 14, after reading the subcode area and applying the servo that read the ATF area, Recording is performed in write mode. This makes it possible to rewrite only the PCM area.

[Search] In this embodiment, high-speed search is possible like DAT. This will be explained with reference to FIG.

When performing a search, the magnetic tape is run at high speed while being wound around the rotating cylinder 10. At this time, set the relative speed between the magnetic tape and the rotating heads A and B to -17548
If you control it so that it is the same as the time, you can read the recorded data.

In order to perform a search at 200 times the normal speed, if the running speed of the magnetic tape is 200 times the normal speed, the rotating cylinder 1
The rotation speed of 0 (normally 2000 rpm) is 3000 rpm and 11 in forward search and reverse search, respectively.
It becomes 00Orp. Therefore, the number of tracks traversed by the rotary heads A and B is approximately 100. It is determined to be approximately 400.

Therefore, the number of rotations of the rotary cylinder 10 is controlled so that the number of times the track is crossed becomes a predetermined number.

In the case of forward search, first, the servo circuit 14 controls the reel motor 13 to run the magnetic tape at 200 times. In this case, the trajectories of the rotary heads A and H across the tracks on the magnetic tape are as shown by the broken lines in FIG. 15(a). The outputs from the rotating heads A and B at this time are shown in the same figure (b).
become that way.

When the rotating head A of +azimuth passes through a track recorded in the same azimuth, a large playback output is obtained, and when it crosses a track of a different azimuth, a small playback output is obtained. Therefore, the servo circuit 14 controls the rotation speed of the rotary cylinder 10 so that this large reproduction output (envelope) is counted and the number becomes approximately 50, which is half of the predetermined number.

The data read intermittently in this way is taken out by the signal processing circuit 16, and the subcode control circuit 17 and drive control circuit 18 inspect the data, and the subcode data and PG
A target block can be searched by reading the contents of the M-ID.

In the case of reverse search, the servo circuit 14 similarly runs the magnetic tape in the reverse direction at a speed of 200 times. At this time, the trajectory of the rotating heads A and B crossing the track is the 15th
It looks like the solid line in Figure (a).

Then, so that the number of envelopes is about 200,
The rotation speed of the rotating cylinder 10 is controlled.

As described above, according to this embodiment, 32 frames are
By making the magnetic tape into blocks and accessing the magnetic tape in units of blocks, it becomes possible to rewrite data that is being recorded on the magnetic tape, which has the effect of realizing the function as a data storage device for a computer. In addition, by recording the block number to which each frame belongs in the subcode area of each frame, it is possible to realize a search function in any location, which is essential for a computer data storage device, and to search for a target block using a normal lead. /20 when writing
The search can be performed at a magnification of 0, which has the effect of enabling high-speed searches that were previously impossible.

Furthermore, by adding layered ECC to the data area, sufficient data reliability can be obtained even when the magnetic tape is dirty or damaged.

Furthermore, by using the first three frames of a block as a gap area, it absorbs the track disturbance that occurs when switching from read mode to write mode when recording a new block after the last recorded block. This prevents data from becoming unreadable at the joints between blocks, which has the effect of improving data reliability.

According to this embodiment, IG is added to the 120 minute tape for DAT.
Bytes of storage capacity can be recorded, and this data can be accessed in an average of 20 seconds.

In this 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.

In addition, as a layered FCC, a Reed-Solomon code was used, and the number of information words was 32 and the number of check words was 6.
It is clear that similar effects can be obtained using word lengths other than this.

In this embodiment, an apparatus using two rotary heads has been described, but exactly the same effect can be obtained with an apparatus equipped with four heads that read this data immediately after writing and compare it with the written data.

〔Effect of the invention〕

As explained above, according to the present invention, by realizing read/write in block units, it is possible to realize the search function and rewrite function of any location necessary for a computer storage device, so it can be used as a data storage device. This has the effect of making it easier to use.

Furthermore, the data reliability is improved by adding an error correction code to the data recording area to prevent a decrease in data reliability due to dirt or damage to the data used by the user and the tape.

Furthermore, since the block number recorded in the subcode area of the frame within a block can be searched, it is possible to realize a magnetic tape device for recording computer data that can provide unprecedented high-speed access. .

[Brief explanation of drawings]

1 to 15 show an embodiment of a magnetic recording/reproducing apparatus according to the present invention, and FIG. 1 shows a layered FC.
Figure 2 is a diagram showing the block configuration on the magnetic tape, Figure 3 is a diagram showing the subcode format, Figures 4 to 6 are diagrams each showing the tape format, and Figure 7 is a diagram showing the block configuration on the magnetic tape. An explanatory diagram of interleaving, FIG. 8 is a block diagram showing the configuration, FIG. 9 is a diagram showing the operation sequence,
10 and 11 are diagrams showing read timing,
Figures 12 to 14 are diagrams showing the write timing.
Figure 5 is an explanatory diagram of the search, and Figures 16 to 20 are D
FIG. 16 is an explanatory diagram of the AT recorder.
FIG. 17 is a diagram showing a recording method using a head, and a schematic diagram showing the structure of a track. FIG. 18 is a diagram showing a track pattern on a magnetic tape;
FIG. 19 is a diagram showing a rewriting state, and FIG. 20 is a diagram showing a small block configuration. 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, 1
8... Drive control circuit, 19... Mechanical drive circuit,
20... Encoding/decoding circuit, 21... Layered FC
C circuit, 22...Memory, 23...Memory control circuit, 24...System control circuit, 25...Interface. Atsushi drawing 7 frame ink-1)-bu' Compilation of drawings illustrations Sade read light
IZ Bird Figure 7 O 7 - Pure - Sensai no Heichi 7 Tori 118 ~ Sutake - 4 - To the prison 12 Doki wa kuzutsu + 7 Jimasu ・
A collection of Nodo's outputs.

Claims (1)

[Claims] 1. In a helical scan type magnetic recording and reproducing device that records and reproduces digital signals on a magnetic tape using a rotating head, a series of tracks formed on the magnetic tape is divided into a plurality of tracks, and each What is claimed is: 1. A magnetic recording and reproducing device characterized in that the divisions are made into blocks, and user data is recorded and reproduced in block units. 2. In claim 1, an area other than a predetermined area at a leading end and a trailing end of the block is defined as a data recording area,
A magnetic recording/reproducing device characterized in that the user data and parity for error detection and correction thereof are recorded in the data recording area. 3. The magnetic recording and reproducing apparatus according to claim 2, wherein the number of bytes of the user data recorded in the data recording area is a power of two. 4. According to claim 2 or 3, the user data and the parity are error-corrected encoded so as to be completed within the data recording area, and the data recording area is composed of a plurality of error-correcting code words, and the error-correcting code A magnetic recording/reproducing device characterized in that the number of words is different from the number of words in one track or one frame consisting of a predetermined number of tracks in the data recording area. 5. A magnetic recording and reproducing apparatus according to claim 4, wherein recording is performed by performing interleaving within one track or within one frame. 6. A magnetic recording and reproducing apparatus according to claim 4 or 5, characterized in that the number of words of the same error correction code word recorded on the same track is smaller than the number of redundant words in the error correction code word. 7. The magnetic recording and reproducing apparatus according to claim 4, 5 or 6, wherein the error correction code is a Reed-Solomon code. 8. In claim 7, 2 tracks are 1 frame, the recording capacity of 1 frame is 5760 bytes, the data recording area is 28 frames, and 13107 of the data
A magnetic recording/reproducing device characterized in that 2 bytes are used as the user data, and the remaining bytes are used as system data that is management information of the user data. 9. In claim 4, 5, 6, 7 or 8, after performing the error correction encoding, recording is performed by performing intra-track error correction encoding that completes within the track, and when playing back, the intra-track error correction encoding is performed. 1. A magnetic recording device characterized by performing error correction, adding a flag indicating the state of correction at the same time, and using the flag as error position information when decoding the error correction code. 10. In claim 9, the playback state is determined based on the flag, and the decoding of error correction within the block is performed only when good playback is performed and a state that cannot be corrected by intra-track error correction does not occur. , a magnetic recording/reproducing device characterized in that no correction processing is performed. 11. Claim 4, 5, 6, 7, 8, 9 or 10, further comprising a memory for storing at least information words and redundant words to be recorded in the data recording area, and when recording, the first the first information word of the code word, the second information word of the first code word, the third information word of the first code word, and so on, in the order of transferring and storing data in the memory; The error correction encoding is performed on the data transferred from the computer, the generated redundant word is stored in the memory, and the information word and redundant word in the memory are converted into the first information word of the first code word. , the second information word of the second code word, and the third information word of the third code word, and so on, and record them on the magnetic tape.