JPS6040104B2 - Recording inspection method for magnetic recording devices - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a recording inspection method in a magnetic recording device such as a 6250 RPI magnetic tape device.

GCR of 6250RP1 (Row Per Inch)
(Group Coded Recording) type magnetic recording medium has an
The recording density is 3014R on the 6th track near the position where the 0T (MoginingofTape) marker is attached.
In addition to recording the 6250RPI recording density mark IDBmst of the PI, an ARA signal (Au mati
cReadAmplificationBmst).

The entire track 904 of the ARA signal is composed of a portion ARABu where RPI signals are recorded and an ARA identification character portion ARAID that follows this. This ARA
Following the signal, a data block DATABLK is recorded. The ARA signal can eliminate deviations in output level due to differences in recording media. 625 like this
In order to cause the 0RPI to write to or read from the magnetic tape, as shown in FIG. 2, when the central processing unit CPU gives an input/output command to the channel CH, the channel CH receives the command from the main memory MM. and transfers it to the magnetic tape controller MTC. The control MTC has multiple magnetic tape devices MTU #0 to #7.
It has the function of controlling and operating in parallel. As shown in FIG. 3, following the command, the data to be written is transferred from the channel CH to the buffer BU of the control MTC.
When transferred to F, the encoder IECD changes the 4 bits to 5.
At the same time as it is converted into bits, an ECC creation circuit generates 1 byte of ECC (Error Corr) for each 7 byte data unit.
rectjngC()de) is added so that even errors that occur simultaneously across two tracks can be corrected. The 5-bit data sub-group is NRZI modulated by the modulation circuit MOD, transferred to the magnetic tape unit MTU, and written onto the magnetic tape by the write head WTHD.

On the other hand, data read from the magnetic tape by the read head RDHD is demodulated by the demodulation circuit DEM via the amplifier AMP, and the data bit string for each track is sent to the deskewing buffer DSKW along with the phase error.
The data is stored in the BUF and the skew is corrected.

Next, in the decoder DEC, conversion from 5 bits to 4 bits is performed.

On the other hand, the FF track pointer PNT receives the phase FF from the drop out deskewing buffer DSKW BUF from the demodulation circuit DEM, or the error code from the decoder DEC.
Error detection circuit ERRDE
Send to T.

The error correction circuit ERRCO calculates the syndrome from the decoder DEC.
R compares the data with the data converted by the decoder DEC, corrects any errors, and then transfers it to the buffer BUF. In this way, in the 6250 RPI GCR system, during normal playback (i.e., read) and write (i.e.,
Conventionally, when an error occurs on two tracks at the same time during playback, it is corrected and read, and when writing, errors such as dropouts are checked. Even if an error occurs, if it occurs on only one track, it will not be detected as an error, and at the same time, it will increase the efficiency of media usage.
The overall throughput of information during writing is improved.

However, when the device has already failed and an error occurs in one track due to the failure, it is not preferable to continue writing without treating it as an error due to maintenance and data quality issues.

The purpose of the present invention is to solve such problems by naturally detecting an error in only one track during writing due to a device failure, while detecting an error in only one track during writing due to a defect in the medium. so that it is not detected,
The objective is to improve media usage efficiency and reliability.

In the present invention, at the time of writing, after detecting the BOT marker, the write command for each data block is used to read each data block. error
If there is, it is treated as a write error, and after detecting that there are no errors across all tracks, even if a drop out error occurs in one track when reading a data block with a subsequent write command, This achieves the above purpose by not causing an error.

4 and 5 are block diagrams of a recording inspection circuit showing an embodiment of the present invention. In Figure 4, the phase error PHSERRI from the deskewing buffer DSKWBUF and the drop error from the demodulation circuit DEM are shown.
output DRPOUTI and error code ERRCDI from the decoder DEC, respectively, to the timing signal TM.
AND gates A1, A2, A3 with A, TMB, TMC
through the OR gate ORI to set the flip-flop FF to indicate that the track TRKI is an error track.

Then, the same processing is performed on each track TRKI to 9 to detect whether or not it is an error track ERRTRK.
It passes through R2 and is input to the NAND gate NA in synchronization with the write command WTCMD.

If there is an error in even one track, NAND gate N
The output of A becomes "0", but if there is no error across all tracks, the output of NAND gate NA becomes "1".

In this case, as shown in FIG.
Corresponding to U#0-#7, memories (here, flip-flops FFO-7) are provided for storing select signals MTUO-7 and 80T detection signals DBOT of the magnetic tape device.

This memory, namely flip-flop FFO~7, is
Command end CMDE output at the end of the command
It is set by MD and the output of the NAND gate NA in FIG. 4, that is, the signal NOERRTRK that has no errors over all tracks, and the set output is sent to the AND gates AO-7. In the present invention, at the beginning of the magnetic tape,
If there is an error on a track, there is a possibility of a hardware failure, so strict checks are performed when writing. However, if it is detected that there are no errors on all tracks from the beginning or during writing, then Even if an error is detected, at least it is assumed that there is no hardware failure, and from that point on, the checking is relaxed and writing is performed.

For example, when machine number 1 is selected and the BOT marker is detected, flip-flop FFI is reset by the output of AND gate A12.

At this time, even if the write command WTCMD is issued,
AND gate AII will not open and flip-flop FFI will not be set unless the all-track no-error signal NOERRTRK is asserted. Therefore, unless the flip-flop FFI is set, even a one-track error due to drop-out or the like is treated as a write error.

When the all-track no-error signal NOERRTRK is given at the command end CMDEMD, the flip-flop FFI is set, and after the set output and the machine number 1 select signal MTUI are ANDed, the OR gate is A signal WTIT that passes through OR and allows even one track error during writing to not be treated as a write error.
RKNOERR is output. Permission signal WTITRKN
When OERR is given, the check conditions of the error detection circuit ERRDET in FIG. 3 are changed by the microprogram. Similar processing is performed during retry.

That is, backspace unevenness P is performed due to a write error, the write data is erased (ERS), and then rewriting is performed, but unless the all-track no-error signal NOERRTRK is given, even one track will have an error. If there is, it will be retried a specified number of times, then it will be determined that there is a hardware failure and the device will be disconnected. Command end CMDEMD and all track no error signal NO
After ERRTRK is given, it is determined that the error is caused by the medium rather than the hardware, and a one-track error is not treated as a write error. A magnetic tape device is a very basic file and records data for storage, so it is necessary to write data of considerably high quality.

According to the invention, tef.

If there is an error in even one track when starting writing from the start end, it will be considered a write error, and once all the tracks have been passed without error, even if there is a single track error, it will be a permanent hardware failure. Since it is determined that the error is not caused by a write error, the throughput of the entire system may decrease somewhat, but it is possible to write high-quality data. Brief Description of the Drawings Fig. 1 is an explanatory diagram of the recording format of the 6250RPI magnetic tape medium, Fig. 2 is a block diagram of the control system of the magnetic tape device, Fig. 3 is a block diagram of the magnetic tape controller, and Fig. 4 is an illustration of the recording format of the 6250RPI magnetic tape medium. FIG. 5 is a block diagram of an all-track no-error signal generating circuit showing an embodiment of the present invention. FIG. 5 is a block diagram of a recording inspection permission signal generating circuit showing an embodiment of the present invention.

PHSERR: PHSERR. F-, DRPOUT: Drotsuf.

・Out, ERR CD: Efu. Code, RST:
Reset signal, FF, FFO~7: flip-flop, ORI~3: OR gate, NA: NAND gate,
AI~3, AO I~72, A'0~7: AND gate, NOERRTRK: All track no error signal, WT・ITRKNOERR: Enable to prevent one track error from becoming a write error during writing. signal, WTC
MD: Write command, MTUO~7: Magnetic tape device select signal. Figure 1 Figure 2 Figure 3 Figure 4 Figure 5

Claims (1)

[Claims] 1 Simultaneous reading and modification of at least two tracks is possible62
In a 50 RPI magnetic recording device, when a BOT marker is detected, predetermined information is stored, and when it is detected that there is no error across all tracks by a write command after the BOT marker is detected, the predetermined information is stored. A memory means to be erased is provided, and when the predetermined information is stored in the memory means, even one track error is treated as a write error, and when the predetermined information in the memory means is erased, one track error is considered as a write error. - A recording inspection method for magnetic recording devices that is characterized by not converting errors into write errors.