WO1983001859A1

WO1983001859A1 - Track identification code recording method

Info

Publication number: WO1983001859A1
Application number: PCT/US1982/000386
Authority: WO
Inventors: Systems Corporation Dma; William A. Pollock
Original assignee: Dma Systems Corp
Priority date: 1981-11-16
Filing date: 1982-04-01
Publication date: 1983-05-26
Also published as: BR8207978A; EP0093713A4; GB2121654A; EP0093713A1; JPS58501879A; GB8317981D0

Abstract

In a magnetic disc for a magnetic disc storage system, at least one of the surfaces of the disc being coated with a magnetic material, the disc being adapted to be mounted on a spindle for rotation relative to a magnetic transducer positioned for recording data on and retrieving data from the surface, a plurality of concentric annular tracks (2, 1, 0, 15) being defined on the surface of the disc, each of the tracks being divided into a plurality of sectors, each such sector having associated therewith prerecorded data for identification thereof, the prerecorded identification data including a track identification code, the track identification code being a unit distance code, there is disclosed a method for recording the unit distance track identification code wherein the code has a first uni-directional clock transition (61) at a fixed time in each bit cell and a second uni-directional transition (N-N) at a variable time in each bit cell, the time of the second transition relative to the first transition determining whether the data bit is a one or a zero.

Description

TRACK IDENTIFICATION CODE RECORDING METHOD BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to a method for recording a unit distance track identification code and, more particularly, to a method for recording a unit distance track identification code on a magnetic disc which preserves the unit distance characteristics of the code. ' 2. Description of the Prior Art

Magnetic disc storage systems are widely used to provide large volumes of relatively low-cost computer accessible memory or storage. A typical disc storage system includes a number of discs coated with a suitable magnetic material mounted for rotation on a common spindle and a set of transducer heads carried in pairs on elongated supports for insertion between adjacent discs, the heads of each pair facing in opposite directions to engage opposite faces of adjacent discs. The support structure is coupled to a positioner motor, the positioner motor typically including a coil mounted within a magnetic field for linear movement and oriented relative to the discs to move the heads radially over the disc surfaces to thereby enable the heads to be positioned over any annular track on the surfaces. In normal operation, the positioner motor, in response to control signals from the computer, positions the transducer heads radially for recording data signals on or retrieving data signals from a preselected one of a set of concentric recording tracks on the discs.

In such a system, it is desirable to record data on a disc to enable the transducer heads to locate the desired recording track. Accordingly, a number of track following systems for magnetic disc drives have been developed. Most commonly, a disc surface and a head have been dedicated to the recording of position information for use by the track following servo system. In these systems, position information is recorded continuously around the disc. Typical techniques for recording position information are disclosed in U.S. Patent No. 3,534,344 to Santana and U.S. Patent No. 3,691,543 to Mueller. In both of these patents, position information is derived from single pulse amplitudes which are time-gated from recorded clock pulses. In the continuous systems for which they were designed, these pulses are repeated continuously around the disc and position information is continuously derived at the output of a comparator.

In- such a continuous systems, each track crossing can be detected by the track following circuitry no matter how fast the head carriage might be moving. For this reason, track identifica ion information can be derived by simply decrementing a track difference counter until the difference is equal to zero, meaning that the transducer head has arrived at the desired trac .

For a variety of practical reasons, it is desirable to place track position information on the same surface as the data information and to eliminate the use of a dedicated surface and head for track position information. One reason is that misregistration of the disc center due to disc interchange or temperature variations can be accommodated since the head is moved directly to the track of interest. Another reason is that the physical alignment of the heads in a disc drive is not as critical as it is where there are multiple heads which must be aligned on multiple surfaces. As a result, no field adjustments are generally required. Another obvious reason is that an entire disc surface need not be dedicated to track position information.

As a result, most recently developed systems have employed embedded servo information (i.e., prerecorded identification information, on the same surface used for recording data, for use by the head tracking servo system) . In the most practical form of embedded servo system, each track is divided into a plurality of sectors and the track identification and fine position information is recorded at the beginning of each data sector. This information is then read by the same head that reads and writes data on the disc. Previous embedded servo systems are exemplified by U.S. Patent No. 4,208,679 to Hertrich, U.S. Patent No. 4,163,265 to Van Herk et al, U.S. Patent No. 4,149,201 to Card, U.S. Patent No. 3,812,533 to Ki ura, U.S. Patent No. 3,185,972 to Sipple and British Patent Application No. 2,017,364 to Droux.

In magnetic disc drives using embedded servo techniques, the requirement exists for knowledge of which track the head is currently positioned over. To accomodate this requirement, track boundary crossings are typically counted. In continuously recorded servo implementations, counting boundaries during high speed head moves is not a problem since information is available continuously. However, in an embedded servo system where track boundary information is available only on an intermittent basis, as here, it is possible to skip track boundaries. Hence, there arises a need for track identification. This identification could go to the extreme of recording an absolute track address for every track on the disc. A more practical implementation is to record repeated bands wherein the tracks are identified within the band, i.e., groups of sixteen tracks per band where the tracks within the band are number 0-15. This requires only four bits of data for track identification.

In recording this type of data, it is preferable to use a unit distance code, commonly referred to as a Gray code. These codes were developed so that only one bit of the code changes as a boundary between tracks is crossed. Use of a Gray code limits ambiguity to a head position uncertainty of ±1/2 track. The remaining problem relates to the combination of a unit distance code with magnetically recorded information. If conventional magnetic recording codes are used, such as NRZ or MFM, ambiguities arise on track boundaries, causing uncertainties of more than one bit, and destroying the required unit distance feature of the Gray code. This can be seen most clearly from FIGURES 1, 2, 3 and 4.

More specifically, FIGURE 1 shows the data that it is desired to record at the beginning of each sector in four consecutive tracks on the surface of a magnetic disc. For purposes of the present example, these tracks are identified by the numbers 2, 1, 0 and 15. FIGURE 1 also shows four consecutive bit cells and the "ones" and "zeros" which are to be recorded therein.

It is seen that the track identification code is a unit distance code, i.e., only one bit of the code changes as a boundary between tracks is crossed.

FIGURE 2 first shows the actual transitions that would be recorded on the tracks on the disc for NR2. encoding. It is seen that a "one" is indicated by a transition whereas a "zero" is indicated by no transition. Superimposed on tracks 2 and 1 are three possible positions of a transducer head, signified by A, B, and C. It is seen that when in position A, the transducer head is exactly aligned with track 2. When in position C, the transducer head is exactly aligned with track 1. When in position B, the transducer head is positioned midway between tracks 2 and 1. This convention will be used in each of the figures to follow.

FIGURE 2 also shows the signals 21, 22 and 23 which are derived from the head when in positions A, B and C, respectively. It is seen that when in position A, the magnetic transitions at 24 and 25 are detected at 26 and 27, respectively, and that when in position C, the magnetic transition at 28 is detected at 29. However, when the transducer head is in position B, the transition at 24 is still detected at 30, but the transitions at 25 and 28 cancel. Accordingly, the track positioning system will decode a position 0010 which obviously has no relevance to the actual track position.

FIGURE 3 shows the manner in which the Gray code binary data of FIGURE 1 would be recorded for FM encoding where there is a clock transition at each cell boundary and a transition in the middle of a cell to indicate the presence of a "one". Superimposed on tracks 2 and 1 are the head positions A, B and C. FIGURE 3 also shows the signals 31, 32 and 33 derived from the transducer head when it is in positions A, B and C, respectively. Again it is seen that ambiguity results in the cell where there is a data bit change. When in position A, the transitions at 34 and 35 are detected at 36 and 37, rspectively, and when in position C, the transitions at 38 and 39 are detected at 40 and 41, respectively. However, when in position B, the transitions at 34 and 38 cancel and the transitions at 35 and 39 cancel so that the clock transition at the cell boundary cancels as well as the mid-cell transition. This is an illegal FM signal which cannot be interpreted by the decoding circuitry.

FIGURE 4 shows the manner in which the binary data of FIGURE 1 would be recorded for MFM encoding where a transition at a mid-cell indicates a "one" and there is a clock transition at each cell boundary, except adjacent a "one". Superimposed on tracks 2 and 1 are the possible head positions A, B and C. Also shown are the signals 51, 52 and 53 which are derived when the head is in positions A, B and C, respectively. When in position A, the transition at 54 is detected at 55 and when in position C, the transition at 56 is detected at 57. However, when in position B, transitions are detected at both 54 and 56, as shown at 58 and 59, and this is an illegal signal which cannot be decoded by the decoding circuitry.

A prior art solution to this problem is shown in the beforementioned U.S. Patent to Kimura where the track identification is provided by a single pulse on each track indicating track identification by its time index so that on a track boundary, the addition of two time pulses would smear between the two adjacent time windows providing for an uncertainty of one track. However, Kimura must divide the different tracks into a multiplicity of windows requiring a significant amount of circuitry in order to be successful in decoding.

SUMMARY OF THE INVENTION

According to the present invention, there is provided a method for recording a unit distance track identification code on a magnetic disc which solves these problems in a manner unknown heretofore. The present recording technique completely eliminates ambiguities on track boundaries so that the decoding circuitry will know exactly between which tracks it is located. This is achieved simply and efficiently and in a manner which substantially simplifies the decoding circuitry required.

Briefly, in a disc having opposed surfaces, at least one of the surfaces being coated with a magnetic material, the disc being adapted to be mounted on a spindle for rotation relative to a magnetic transducer positioned for recording data on and retrieving data from the surface, a plurality of concentric annular tracks being defined on the surface of the disc, each of the tracks being divided into a plurality of sectors, each such sector having associated therewith prerecorded data for identification thereof, the prerecorded identification data including a track identification code, the track identification code being a unit distance code, there is disclosed an improvement wherein the code has a first uni-directional clock transition at a fixed time in each bit cell and a second uni-directional transition at a variable time in each cell, the time of the second transition relative to the first transition determining whether the data bit is a one or a zero.

It is therefore the object of the present invention to solve the problems associated with recording a unit distance code on a magnetic disc. It is a feature of the present invention to solve these problems by recording a track identification, unit distance code using a 1/3-2/3 interval code which preserves the required single track ambiguity by not introducing additional uncertainties as a result of the magnetic recording technique. An advantage to be derived is the preservation of the unit distance characteristics of the recorded code. Another advantage is the preservation of the required single track ambiguity. Still another advantage is the elimination of the introduction of additional uncertainties. Still another advantage is the elimination of illegal codes.

Still other objects, features and attendant advantages of the present invention will become apparent to those skilled in the art from a reading of the following detailed description of the preferred embodiment constructed in accordance therewith, taken in conjunction with the accompanying drawings wherein:

^C ' BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 shows diagrammatically the binary data to be recorded on a disc surface;

FIGURE 2 shows diagrammatically the manner in which the data of FIGURE 1 would be recorded on a disc utilizing NRZ encoding and the signals which will be derived from a transducer head positioned over such a disc in the positions A, B and C;

FIGURE 3 shows diagrammatically the manner in which the- data of FIGURE 1 would be recorded on a disc utilizing FM encoding and the signals which will be derived from a transducer head positioned over such a disc in the positions A, B and C;

FIGURE 4 shows diagrammatically the manner in which the data of FIGURE 1 would be recorded on a disc utilizing MFM encoding and the signals which will be derived from a transducer head positioned over such a disc in the positions A, B and C;

FIGURE 5 shows diagrammatically the manner in which the data of FIGURE 1 would be recorded on a disc utilizing the teachings of the present invention and the signals which will be derived from a transducer head positioned over such a disc in the positions A, B and C.

cv:-ι DESCRIPTION OF THE PREFERRED EMBODIMENT

Referring now to FIGURE 5, there is shown the preferred embodiment of encoding which will preserve the characteristics of a unit distance code. To preserve the value of a unit distance code, the recorded transitions must change only one transition or group of transitions from one track to the next. It is also desirable to have a transition per bit, minimum, in order to make decoding easy. These considerations lead to a code including a mandatory clock transition plus a mandatory data transition whose position determines whether the bit is a one or a zero. According to the preferred embodiment of the invention, the mandatory clock transition is a uni-directional clock transition at a fixed time in each bit cell and the data transition is also a uni-directional transition at a variable time in each bit cell, the time of the second transition relative to the first transition determining whether the data bit is a one or a zero. The implementation shown is a 1/3-2/3 code.

More specifically, and with reference to FIGURE 5, there is seen the manner in which the binary data shown in FIGURE 1 would be recorded on tracks 2, 1, 0 and 15. It is seen that each bit cell begins and ends with a transition and that such transition is always in the same direction. Thus, every track has a transition, as shown at 61, at a fixed time in each bit cell. Furthermore, each bit cell has a second uni-directional transition which is either 1/3 of a bit cell after a clock transition for a one or 2/3 of a bit cell after a clock transition for a zero. The significance of this can be seen in comparing the signals 62, 63 and 64 derived from a transducer head when in positions A, B and C, respectively. More specifically, in any of positions A, B or C, the head will receive the exact same signal except in the bit cell where there is possible ambiguity. In this cell, namely the third cell, when in position A, the transitions at 61A, 65 and 61B will be detected at 66, 67 and 68, respectively, whereas in position C, the transitions at 61C, 69 and 61D, will be detected at 70, 71 and 72, respectively. When in position B, the detected signals at 66 and 70 will add, as shown at 73, and the detected signals at 68 and 72 will add, as shown at 74. There will be some cancellation and therefore ambiguity between the signals detected at 67 and 71 so that the decoding circuitry may decode either a one or a zero. The decoding circuitry will, therefore, readily decode 00?1, which can be clearly interpreted to indicate that the transducer head is between tracks 2 and 1. Thus, not only is decoding permitted, but any ambiguity can in fact be utilized to indicate to the decoding circuitry exactly where the transducing head is located.

1/3-2/3 codes have been used previously in other recording environments, such as in low-cost digital cassette recorders. Accordingly, decoding circuitry for decoding signals 62, 63 and 64 are well known to those skilled in the art. However, the combination of this type of encoding with a unit distance code in a magnetic disc recording system provides advantages not to be obtained from other encoding techniques. While the invention has been described with respect to the preferred physical embodiment constructed in accordance therewith, it will be apparent to those skilled in the art that various modifications and improvements may be made without departing from the scope and spirit of the invention. Accordingly, it is to be understood that the invention is not to be limited by the specific illustrative embodiment, but only by the scope of the appended claims.

Claims

I CLAIM ;

1. In a disc having opposed surfaces, at least one of said surfaces being coated with a magnetic material, said disc being adapted to be mounted on a spindle for rotation relative to a magnetic transducer positioned for recording data on and retrieving data from said surface, a plurality of concentric annular tracks being defined on said surface of said disc, each of said tracks being divided into a plurality of sectors, each such sector having associated therewith prerecorded data for identification thereof, said prerecorded identification data including a track identification code, said track identification code being a unit distance code, the improvement wherein said code has a first uni-directional clock transition at a fixed time in each bit cell and a second uni-directional transition at a variable time in each bit cell, the time of said second transition relative to said first transition determining whether the data bit is a one or a zero.

2. In a disc according to claim 1, the improvement wherein said second transition is spaced from said first transition by a distance equal to one-third or two-thirds the width of each bit cell depending upon whether the data bit is a one or a zero.

3. In a disc according to claim 1 or 2, the improvement wherein said first clock transitions define the boundaries of each bit cell.