WO1996007180A1

WO1996007180A1 - Method and apparatus for virtual edge detection of magnetic tape

Info

Publication number: WO1996007180A1
Application number: PCT/US1995/008058
Authority: WO
Inventors: Mark R. Ayres
Original assignee: Exabyte Corporation
Priority date: 1994-08-30
Filing date: 1995-06-23
Publication date: 1996-03-07
Also published as: AU2910695A

Abstract

A virtual edge of magnetic tape (22) is located by successively moving a head unit (28) toward a physical edge (22EB) of tape and, at each of a plurality of transverse locations, recording a pattern P of digital information. By either successive read-back or rewind followed by reading, sensing of a read-back or acquisition signal is attempted at each location. In connection with the reading operation, a pattern is acquired from the acquisition signal using a method of amplitude invariant data recovery such as peak detection, the pattern acquired being invariant regardless of any overall amplitude envelope variations. A virtual edge of tape is determined when an acquired pattern derived from the acquisition signal does not match a predetermined pattern of information which was recorded as a corresponding pattern of flux directional transitions.

Description

METHOD AND APPARATUS FOR VIRTUAL EDGE DETECTION OF MAGNETIC TAPE

BACKGROUND

1. Field of Invention

This invention pertains to magnetic tape drives and methods of operation thereof, and particular to method and apparatus for detecting a virtual edge of magnetic tape.

2. Related Art and Other Considerations

Some magnetic tape recorders (also known as tape drives) record information on magnetic tape along a plurality of tracks which are parallel to the direction of travel (e.g., longitudinal direction) of the tape. Such tape drives have sometimes been called "streaming" or "serpentine" tape drives. Such tape drives typically include at least one pair of information transducing elements. In a record mode of operation, a first transducer (write element) of the pair records information on a selected track of the tape, which information is read back (for checking purposes) by a second transducer (read element) of the pair. In some embodiments, the second transducer of the pair is horizontally spaced slightly apart from the first transducer in such a manner to follow the first transducer as the transducer pair together traverse the track. This allows the read back to be essentially immediate. In other embodiments the read back occurs considerably later (e.g., after rewind of the tape) .

Generally, the transducer pair of such tape drives is mounted on a head unit. The head unit comprises a carriage capable of selective travel in a direction which is transverse to the longitudinal direction of the tape. By controllably moving the head unit in the transverse direction, the transducers are transversely positioned for recording/reading with respect to selected one(s) of the plurality of tracks horizontally extending along the tape. Movement of the head unit in the transverse direction is typically accomplished using a stepper motor. See, for example, United States Patent 4,422,111 to Moeller et al. , incorporated herein by reference.

Controlling the head unit, and consequentially controlling the transverse positioning of the transducers relative to the tracks, subsumes a known format regarding such dimensional considerations as the transverse width of the track and the (transverse) location of the plurality of tracks with respect to a horizontal edge of the tape. Accordingly, when a tape is to loaded into such a tape drive for formatting, there must be a mechanism or procedure for detecting a horizontal edge of the tape.

One approach for detecting a horizonal edge of the tape is a read signal amplitude method such as that disclosed in United States Patent 4,422,111 to Moeller et al . and United States Patent 4,476,503 to Solhjell. In such read signal amplitude method, a tape edge is located when the amplitude of a signal acquired from the read transducer either exceeds a predetermined reference or has changed by predetermined amount.

United States Patent 4,786,984 to Seeman detects the presence of media by counting the number of flux transitions in a given interval, but does not recognize specific patterns of flux transitions for tape edge detection.

SUMMARY A virtual edge of magnetic tape is located by successively moving a head unit to a physical edge of tape and, at each of a plurality of transverse locations, recording and a pattern of digital information. By either successive read- back or rewind followed by reading, sensing of a read-back or acquisition signal is attempted at each location. In connection with the reading operation, a pattern is acquired from the acquisition signal using a method of amplitude invariant data recovery such as peak detection, the amplitude being invariant regardless of the transverse positioning of the head unit. A virtual edge of tape is determined when an acquired pattern derived from the acquisition signal does not match a predetermined pattern of information which was recorded as a corresponding pattern of flux directional transitions. As used herein, the term "pattern" means more than the existence of a predetermined number of bits or flux directional transitions per unit tape, but instead requires a predetermined time ordering of bits or flux directional transitions relative to one another. In one embodiment, the predetermined pattern of information is a preamble. A formatter detects the preamble by scanning the digital acquisition data stream for the predetermined pattern corresponding to the preamble. The acquisition data stream (acquisition pattern) is generated by detecting changes in the direction of the magnetic flux in a sensed acquisition signal and synchronizing a clock to the detected direction changes.

In one embodiment, changes in the magnetic flux are detected using a qualified peak detection technique. The qualified peak detection technique of the invention generates information for inclusion in the acquisition pattern when the derivative of the (gain) adjusted acquisition signal exceeds a predetermined fraction of a local averaged amplitude. In another embodiment, changes in the magnetic flux are detected using a partial- response, maximum likelihood (PRML) technique.

In a preferred embodiment, the predetermined pattern of digital information consists sixteen sets of a digital ONE followed by a digital ZERO (i.e., 10) , for a total of thirty two bits.

The method and apparatus of the present invention establishes a virtual edge of tape, which is distinct from a physical edge of tape. As used herein, a "virtual edge" of tape is a transverse location on tape whereat, in contrast to a neighboring preceding transverse location, a pattern of data recorded has transitioned from being readable to being corrupted. Thus, the virtual edge of tape is determined with reference to a location whereat data can no longer be reliably recovered, not with reference to locations at which a signal amplitude exceeds a predetermined level .

Moreover, virtual edge detection is accomplished utilizing hardware that is already present for the primary purpose of the tape drive (i.e., data recovery) . Advantageously, the present invention does not require additional components such as amplitude detection filters or flux transition counters. Furthermore, the invention effectively prequalifies the edge of the medium for recording and reading in exactly the manner in which it will be utilized. The invention does not merely test an indirect quantity (such as amplitude or a temporally-unqualified pulse count) and assume that it will be correlated with the capacity of the medium to retain data.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other objects, features, and advantages of the invention will be apparent from the following more particular description of preferred embodiments as illustrated in the accompanying drawings in which reference characters refer to the same parts throughout the various views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

Fig. 1 is a schematic block diagram of a tape drive according to an embodiment of the invention. Fig. 1A is a schematic block diagram of a pattern recovery circuit for the tape drive of Fig. 1 which uses a qualified peak detection technique. Fig. IB is a schematic block diagram of a pattern recovery circuit for the tape drive of Fig. 1 which uses a partial-response, maximum likelihood (PRML) technique.

Fig. 2 is a schematic view of a head unit relative to a magnetic tape.

Fig. 3 is a flowchart showing steps executed in connection with a tape edge detection procedure according to an embodiment of the invention.

Fig. 4 is a schematic view showing successive positions of a head unit during the tape edge detection procedure depicted by Fig. 3.

Fig. 5A - 5C are graphs illustrating signal waveforms during the tape edge detection procedure depicted by Fig. 3.

DETAILED DESCRIPTION OF THE DRAWINGS

Fig. 1 shows a magnetic tape drive system 20 for recording information on (and reading recorded information from) an information storage medium which, in the illustrated embodiment, is magnetic tape 22. As shown in Fig. 2, in drive system 20 tape 22 is transported either in a forward direction (denoted by arrow 24) or a reverse direction (denoted by arrow 26) past head unit 28. In a recording mode, depending on direction of movement of tape 22, one of two write elements 30 _F, 30W_R (mounted on head unit 28 in proximity to tape 22) causes magnetic signals to be recorded in a horizontal track on tape 22. Tape 22 has a top horizontal physical edge 22E_T and a bottom horizontal physical edge 22E_B.

As explained hereinafter, head unit 28 is vertically adjustable (in the direction depicted by arrow 34) . In a recording mode, repeated vertical adjustment of head unit 28 yields a plurality of parallel, horizontal tracks (unillustrated) . Given such a pattern of track recording, the drive system 20 is known as a streaming or serpentine magnetic tape drive system.

As shown in Fig. 1, tape drive system 20 (also known as a tape "deck" system) includes a deck assembly section 70. Deck assembly 70 in turn includes head unit 28. Head unit 28 has mounted thereon the two write elements 30 _F, 30W_R and two read elements 30R_F, 30R_R shown in Fig. 2. As used herein, the term "element" includes any structure suitable for transducing information, including gaps and magnetoresiεtive material.

Deck assembly 70 also has means for positioning head unit 28 so that head unit 28 has a component of travel in the vertical direction depicted by arrow 34. In particular, head unit 28 is mounted on a head positioner or "element positioner" 74 (see Fig. 2) which includes a positioner motor 76. In the illustrated embodiment, positioner motor 76 is a stepper motor which incrementally repositions head unit 28 relative to a width or transverse dimension of tape 22 in conventional manner. As used herein, the width dimension of tape 22 refers to the vertical direction depicted by arrow 34, and thus is substantially orthogonal to the horizontal direction depicted both by forward direction arrow 24 and reverse direction arrow 26. In conventional manner, deck assembly 70 also includes a capstan motor 78 which serves to rotate an unillustrated capstan. The capstan is in contact with an unillustrated cartridge drive mechanism which transports tape 22 in the horizontal direction (either forward or reverse, depending on the direction of drive applied to capstan motor 78) during recording and reading operations . As also shown in Fig. 1, tape drive system

20 includes a basic drive control section 80. Drive control section 80 includes a drive or transport controller 82 which is connected via various circuits to hardware installed in deck assembly 70. For example, transport controller 82 is connected through motor control driver 84 to capstan motor 78 and through positioner circuitry 86 to positioner motor 76.

Tape drive system 20 also includes a signal processing circuit 90. An output or record side of signal processing circuit 90 in turn comprises a write driver 92. Write driver circuit 92 is connected inter alia to apply recording signals to the two write elements 30W_F, 30W_R during recording operations.

An input or read side of signal processing circuit 90 receives on line 97 an acquisition signal sensed from a read element mounted on head unit 28. In addition, the input side of signal processing circuit 90 comprises a pattern recovery circuit 100.

As further shown in Fig. 1, tape drive system 20 includes a microprocessor or system controller 102 which communicates with other constituent members of drive system 20 in conventional manner using a microprocessor bus 104. In particular, microprocessor 102 communicates over bus 104 with a formatter/encoder 110, a preamble generator 106, and a buffer memory 112.

Microprocessor 102 also communicates with the drive control processor 82 using a serial bus 113 and is ultimately connected to an unillustrated host (e.g., utilization) device (in a manner understood by the man skilled in the art, for example by reference to United States Patent Application Serial Number 08/181,317, filed January 13, 1994 by Chuck Klomp et al . [incorporated herein by reference] ) . Fig. 1A illustrates a qualified peak detection circuit (QPDC) 100A which serves as a pattern recovery circuit for one embodiment of the invention. In the Fig. 1A embodiment, circuit 100A comprises an automatic gain control (AGO circuit 118A which controls an AGC amplifier 119A. AGC amplifier 119A receives the acquisition signal on line 97 and generates a gain-adjusted acquisition signal which is applied both to a differentiator 120 and to automatic gain circuit (AGC) 118A. AGC 118A controls the gain to AGC amplifier 119A. Differentiator 120 has both an inverting output port and a non-inverting output port. The inverting output port of differentiator 120 is connected to a first input port of a comparator 122, while the non-inverting output port of differentiator 120 is connected to a second input port of a comparator 122. Comparator 122 generates an output signal transition or qualified peak when the output from both ports of differentiator 120 is equal to zero. An output signal from comparator 122 is applied to a pulse generator 124A.

Connected between formatter/encoder 110 and pulse generator 124A of circuit 100A is synchronizer 125A. Since the operation of such a synchronizer is known to the man skilled in the art, it suffices here to state that synchronizer 125A basically fits time windows to a stream of pulses outputted from pulse generator 124A.

In the illustrated embodiment, qualified peak detection circuit 100A is of a type commercially known as a National Semiconductor 8464 chip. Qualified peak detection circuit 100A maintains the gain of the acquired signal at an averaged local amplitude, with the localness depending upon such external setting as time constants, etc., in accordance with conventional practice. Moreover, the shape of pulses produced by qualified peak detection circuit 100A is controllable using external filtering and by settings of capacitances and resistances as is known to the man skilled in the art .

In another embodiment of the invention illustrated in Fig. IB, a pattern recovery circuit 100B is provided with an automatic gain control

(AGC) circuit 118B which controls an AGC amplifier 119B. As in the Fig. 1A embodiment, AGC amplifier 119B receives the acquisition signal on line 97 and generates a gain-adjusted acquisition signal which is output which is applied to a partial-response, maximum likelihood (PRML) circuit 126; to synchronizer 125B; and to automatic gain circuit (AGC) 118A. AGC 118A controls the gain to AGC amplifier 119A. Synchronizer 125B signals PRML circuit 126 when to take a sample. An output port of the PRML circuit is applied to formatter/encoder 110.

As is understood by the man skilled in the art, a partial-response, maximum likelihood (PRML) technique such as that implemented by circuit 100B has two unrelated but complementary components: partial response equalization and maximum likelihood detection. Partial response equalization refers to a specific method in which a data stream is filtered and sampled. In order to maximize the rate at which symbols may be transmitted, neighboring symbols are allowed to overlap in a controller manner. In a magnetic recording system, for example, an isolated positive flux transition may produce a pulse which is sampled to yield a first sequence 0, 0, +1, +2, +1, 0, 0, 0. If this first sequence were allowed to overlap a negative pulse offset by 1 (e.g., a second sequence -0,0,0,-1,-2,-1,0,0) , the resulting sequence would be 0, 0, +1, +1, -1, -1, 0, 0.

Maximum likelihood (ML) detection refers to the rule used to detect the symbols in the data stream. Under this rule, symbols are selected which make the actual data stream the most probable. Typically this is done in a manner which makes the most likely symbol sequence that which is geometrically closest to the observed data stream. In this regard, a detector in the PRML circuit 100B compares a data stream sequence to possible sequences formed by a combination of perfect, noise free alternating positive and negative pulses, preferably using a fast technique known as the Viterbi Algorithm. Using the PRML technique, symbols are detected within the data stream without regard to the overall amplitude envelope. Thus, PRML circuit 100B provides an alternate technique for recovery of an acquisition pattern from an acquisition signal in an amplitude invariant manner.

OPERATION Fig. 3 basically depicts steps executed by drive system 20 in connection with an operation for seeking a horizontal virtual edge of tape 22, and particularly a virtual edge in the direction of bottom horizontal edge 22E_B. Fig. 4 is referenced in the discussion of Fig. 3 for the purpose of showing relative positions of head unit 28 relative to tape edge 22E_B. Fig. 5A, 5B and 5C are also referenced in the discussion of Fig. 3 for depicting signal waveforms at various points as hereinafter described. It should be understood, however, that the waveform "pulse detected" is not applicable to the above-described PRML embodiment. Many of the steps of the virtual edge seek operation of Fig. 3 are conducted in connection with execution by processor 102 of programmed instructions.

As used herein, a "virtual edge" of tape is a transverse location on tape whereat, in contrast to a neighboring preceding transverse location, a pattern of data recorded has transitioned from being readable to being corrupted. As used herein, the term "pattern" means more than the existence of a predetermined number of bits or flux directional transitions per unit tape, but instead requires a predetermined time ordering of bits or flux directional transitions relative to one another. Execution of the virtual edge seek operation of Fig. 3 begins with a starting or initializing step 300, at which inter alia various counters and registers are initialized.

At step 302, capstan motor 78 is actuated so that the capstan moves tape 22 in the forward direction (e.g., the direction indicated by arrow 24) . Capstan motor 78 is actuated by signals applied via motor control driver 84 under control of the drive control processor 82, which (in turn) is ultimately under the control of the microprocessor 102.

As the virtual edge seek operation will typically occur upon installation of a new tape (e.g., tape cartridge) into the tape drive system 20 for formatting, after tape installation the head unit 28 must be positioned (see step 304) so that the elements 30 thereon can transduce information with respect to tape 22. Initially, the head unit is positioned in a nominal or default position, such as the position depicted by head unit 28 (a) in Fig. 4. Inasmuch as Fig. 4 is utilized to show the same head unit at differing (e.g., successive) positions, the positions are shown using broken lines and reference is made to the head unit has head unit 28 (a) in a first position, head unit 28(b) in a second position, and so forth.

At step 306 microprocessor 102 initiates recordation on tape 22 of a write burst consisting of one or more series "preambles" P. In particular, microprocessor 102 requests that a predetermined pattern of flux directional transitions, corresponding to a predetermined pattern of digital information, be recorded (on tape 22, when tape 22 is present) by write element 30W_F. A first such preamble P_a is shown recorded at head unit position 28 (a) .

Although the predetermined pattern of digital information (to which the predetermined pattern of flux directional transitions corresponds) can be any digital pattern, a conventional preamble is preferably employed. A preamble is typically a recognizable pattern which precedes data blocks in order to facilitate electronic synchronization during playback. For the particular embodiment herein discussed, a pattern consisting of sixteen repetitions of the digital set "10" is preferred. However, other patterns can be utilized, or even optionally selected from a plurality of possible patterns.

The preamble P is generated by preamble generator 106. In the illustrated embodiment, preamble generator 106 is a flip-flop circuit which is clocked to provide sixteen repetitions of the digital set 10 (a total of 32 bits) . In another embodiment, preamble generator 106 can be a memory having the preamble pattern stored therein.

Formatter/encoder 110 receives the pattern of digital information for preamble P from preamble generator 106 and applies the pattern to write driver 92, which in turn drives write element 30W_F to generate the pattern of flux directional transitions depicted as pattern P_a. Preamble P is recorded (one or more times) at step 306 only for a predetermined time window. After expiration of the time window (step 308) , at step 310 write element 30W_F is turned off and then (in a readback embodiment) read element 30R_F is turned on. Turning off write element 30W_F helps prevent cross-talk. Read element 30R_F is turned on (step 310) so as to sense magnetic flux encountered thereby (which will be at the same transverse location as that at which the preamble was recorded at step 306) .

After read element 30R_F is turned on at step 310, read element 30R_F outputs an acquired signal (on line 97 in Fig. 1) to signal processing circuit 90. In head unit position 28(a) shown in Fig. 4, upon reading preamble P_a the acquired signal will be as shown in the top graph of Fig. 5A. For such a situation, when the read element 30R_F is entirely over the tape 22 and preamble P_a fully recorded thereon, the gain output of AGC circuit 119A or 119B (depending on the embodiment employed) is relatively low, as shown in the second graph of Fig. 5A. The analog acquisition signal from AGC amplifier 119A or 119B is processed by pattern recovery circuit 100 in order to find flux directional transitions.

In the embodiment of Fig. 1A, the gain- adjusted acquisition signal is differentiated by differentiator 120. A flux directional transition causes both output ports of differentiator 120 to be zero. However, since the derivative of the acquired signal may be zero between peaks, the output of the peak detection circuit 100 must be "qualified" . For this reason, the amplitude of the undifferentiated acquisition signal must exceed a predetermined fraction of the local average amplitude in order for the zero output of differentiator 120 to qualify as a peak (i.e., a flux directional transition) for use by pulse generator 124A in generating pulses.

As explained previously the local average used to qualify peaks is determined by the AGC amplifier 119A. AGC amplifier 119A is used to normalize the acquisition signal to produce a signal with a constant amplitude envelope. Thus, at the output of the AGC amplifier 119A, the local amplitude of the signal has been normalized to a constant level, so a constant threshold voltage can be used for peak qualification. The size of the local area considered is determined by the time constants used by the AGC amplifier 119A to adjust the gain applied to the playback signal.

Thus, as peaks are detected by circuit 100A, a series of pulses (shown in the fourth line of Fig. 5A) is applied to synchronizer 125A.

Synchronizer 116 clocks the pulses to develop an acquisition pattern. Formatter/encoder 110 then determines whether the acquired pattern matches the predetermined pattern of digital information of the preamble. When formatter/encoder 110 recognizes that the acquisition pattern (e.g., the series of acquired pulses) matches the predetermined pattern of digital information of the preamble, it is determined (at step 312) that the preamble has been detected. When a preamble is thusly determined (as occurs in head position 28(a) of Fig. 4) , steps 314 and 316 are executed prior to looping back to step 306. Otherwise, after the time window has expired (step 318) , no preamble is detected, and execution continues at step 320.

Assuming that a preamble has been detected for head unit position 28(a) for Fig. 4, at step 314 (in a readback embodiment) the read element 30R_F is turned off. Next, at step 316, head unit

28 is moved in a direction toward tape edge 22Ξ_B to a next transverse position. Movement of head unit 28 is effected by positioner motor (e.g, stepper motor) 76, under control of positioner circuitry 84 and drive control processor 82, and ultimately controlled by microprocessor 102. Such head unit movement would, in the example of Fig. 4, move the head unit to head unit position 28(b) .

Considering head unit position 28(b) of Fig. 4, steps 306, 308, 310, and 312 of Fig. 3 are executed with a similar result of preamble detection and looping back to step 306 after a further positioning of head unit 28. Such repositioning and looping are repeated until head unit Dosition 28(n+l) is accruired. When head unit position 28(n+l) of Fig. 4 is acquired, the acquired signal from preamble P_n+1 and subsequently processed signals (including AGC level) are as shown in Fig. 5B. As explained below, at head unit position 28(n+l) , the read element 30R_F is positioned at a virtual edge of tape 22. In such case, the acquired signal sensed by read element 30R_F is as shown in the top line of Fig. 5B. Upon receipt of such an acquisition signal, the gain level of AGC amplifier 119A must be increased (as shown by the second line of Fig. 5B in order to maintain the constant gain envelope) . Line 3 of Fig. 5B, and particularly the pulse waveform of line 4 of Fig. 5B, show the corrupted acquisition signal as read back by read element 30R_F and detected by signal processing circuit 90. The corrupted signal results when the head unit 28 is at the virtual edge of tape. Line 4 of Fig. 5B illustrates such corruption occurring by shifted bits, missing bits, and extra bits.

Upon its clocked comparison of the acquired pattern (e.g., line 4 of Fig. 5B) with the predetermined pattern constituting the preamble, at step 312 (and after expiry of the time window [step 318] ) formatter/encoder 110 determines that the preamble was not detected.

When head unit 28 reaches the virtual edge of tape 22 in the manner aforedescribed with respect to head unit position 28(n+l), step 320 is executed. At step 320, the current position of the positioner motor 76 is stored as a candidate for the (bottom) virtual edge of tape 22.

After storage of the candidate edge position according to step 320, at step 322 a determination is made whether a predetermined number (e.g, ten) candidates have been ascertained. If ten candidates are not yet made, then execution loops back (after various initializations) to repeat the entire operation (e.g. even numbered steps 304 - 322) .

After obtaining ten candidates, at step 324 tape 22 is stopped. Then, at step 326, the ten candidates are averaged to obtain a final determination of the (bottom) virtual edge of tape 22. Such final determination of the virtual edge of tape can then be utilized as a reference point for locating the tracks and further operations of the tape drive system 20. For example, controller 102 may determine a bottom-most track position with reference to the virtual edge (e.g., by spacing the bottom-most track from the virtual edge by a tolerance distance) .

Fig. 5C shows signal waveforms obtained in a situation in which head unit 28 is substantially off-tape. In an off-tape situation, the acquisition pattern (see the fourth line of Fig. 5C) markedly differs from the predetermined pattern constituting the preamble. Accordingly, should read element 30R_F be off-tape when formatter/encoder 110 is trying to detect a preamble, a preamble will not be detected and step 320 (and ensuing steps) will be executed.

The foregoing discussion has assumed that head unit 28 travels from an on-tape position to an off-tape position. It should also be understood from the foregoing, particularly with reference to Fig. 5C, that detection of the virtual edge can occur with head unit 28 first being in an off-tape position and moving toward an on-tape position. For simplification the foregoing discussion of Fig. 3 has been confined to an operation for seeking only the virtual edge proximate bottom physical edge 22E_B. However, in another embodiment, using analogous steps but working instead in an opposite direction, the location of a virtual edge proximate the top physical edge 22E_T is also determined. Such can be done by transporting tape 22 in the reverse direction (depicted by arrow 26), and by using elements 30W_H and 30R_R. Although elements 30W_R and 30R_R are not shown in Fig. 4, their positions and operations are amply understood with reference to the preceding discussion and Fig. 2 in particular. Moreover, principles of the invention can be utilized in an embodiment which does not provide for immediate readback. In such non-readback embodiment, preambles can be recorded (but without immediate readback) in a direction approaching a tape edge as in the foregoing description. Thereafter, the tape is rewound for a sequential reading of each preamble and associated pattern detection in the same manner as described above.

It should also be realized that the pattern recognition function accorded formatter/decoder 110 of the illustrated embodiments can be assigned or delegated to other circuitry, including controller 102, for example. Furthermore, although operation has been described above primarily in the context of the Fig. 1A embodiment, from the foregoing it will be understood by the man skilled in the art how operation is conducted using the PRML technique of the Fig. IB embodiment.

It is important to note that the invention does not detect a virtual edge of tape by resort to a single value (e.g., amplitude value) read-back from the tape, but instead by determining a point of transition from recovery to non-recovery of an expected predetermined pattern of information. In contrast to an amplitude-dependent detection scheme, for accuracy the present invention requires an amplitude invariant pattern recognition scheme (e.g., maintenance of a constant amplitude envelope signal applied to a peak detector in one embodiment) , so that (regardless of the transverse position of the head unit) an invariant pattern is recovered despite any amplitude envelope variations . While the invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood by those skilled in the art that various alterations in form and detail may be made therein without departing from the spirit and scope of the invention.

Claims

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A method of operating a magnetic storage tape drive, the method comprising:

(a) driving a write element, when the write element is positioned at a first transverse position relative to the tape, to generate a predetermined pattern of flux directional transitions, the predetermined pattern of flux directional transitions corresponding to a predetermined pattern of digital information; (b) detecting an acquisition signal at the first transverse location;

(c) deriving an acquisition pattern from the acquisition signal obtained in step (b) ,- and

(d) comparing the acquisition pattern of step (c) to the predetermined pattern of digital information for locating the first transverse position relative to a virtual edge of the tape.

2. The method of claim 1, further comprising adjusting the gain of the acquisition signal whereby the acquisition signal has a constant amplitude.

3. The method of claim 1, wherein the write element is moved to a second transverse position and driven to generate the predetermined pattern of flux directional transitions thereat; the acquisition signal is detected at the second transverse location; and steps (c) and (d) are repeated using in step (c) the acquisition signal detected at the second transverse location.

4. The method of claim 1, wherein the acquisition signal is differentiated to produce a differentiation output and, when the acquisition signal exceeds a local average amplitude, the differentiation output is used to derive the acquisition pattern.

5. The method of claim 1, wherein the acquisition pattern is derived from the acquisition signal using a PRML technique.

6. The method of claim 1, the write element is moved to a second transverse position and driven to generate the predetermined pattern of flux directional transitions thereat; the acquisition signal is detected at the second transverse location; and steps (c) and (d) are repeated using in step (c) the acquisition signal detected at the second transverse location.

7. A method of operating a magnetic storage tape drive, the method comprising:

(a) positioning a write element at a first transverse position relative to the tape; (b) driving the write element to generate a predetermined pattern of flux directional transitions at the transverse position at which the write element is positioned, the predetermined pattern of flux directional transitions corresponding to a predetermined pattern of digital information;

(c) detecting an acquisition signal at the transverse location at which the predetermined pattern was generated in step (b) ; (d) deriving an acquisition pattern from the acquisition signal obtained in step (c) ; and (e) comparing the acquisition pattern of step (d) to the predetermined pattern of digital information; (f) using the comparison of step (e) to locate a virtual edge of the tape if the acquisition pattern of step (d) does not match the predetermined pattern of digital information; otherwise (g) positioning the write element at a second transverse position relative to the tape and repeating at least steps (b) - (e) .

8. The method of claim 7, wherein step (f) locates a candidate transverse position for a virtual edge of the tape, and wherein the candidate of step (f) is averaged with at least one other candidate transverse position.

9. The method of claim 7, wherein the predetermined pattern of digital information consists of sixteen sets of a digital ONE followed by a digital ZERO.

10. The method of claim 7, wherein, while repeating steps (b) - (e) , the amplitude of the gain of the acquisition signal detected at the second transverse location is adjusted to the amplitude of acquisition signal detected at the first transverse location, whereby the acquisition signal has a constant amplitude envelope.

11. The method of claim 10, wherein the adjusted acquisition signal is differentiated and, when the adjusted acquisition signal exceeds a predetermined fraction of a local average amplitude, the result of the differentiation is used to derive the acquisition pattern.

12. The method of claim 7, wherein the acquisition pattern is derived from the acquisition signal using a PRML technique.

13. A method of operating a magnetic storage tape drive, the method comprising: selecting a predetermined pattern of digital information; successively positioning a write element at a plurality of transverse positions relative to the tape and, at each such transverse position:

(a) moving the tape in a direction parallel to the longitudinal direction of the tape and generating magnetic flux having flux directional transitions corresponding to the predetermined pattern of digital information and thereby recording the pattern on the tape when the tape is present; and then (b) activating a read element to sense magnetic flux and to generate an acquisition signal in accordance therewith, the read element being activated and the tape having been moved such that any flux directional transitions generated in step (a) which were recorded on the tape are sensed by the read element;

(c) deriving an acquisition pattern from the acquisition signal obtained in step (c) ; and (d) comparing the acquisition pattern of step (c) to the predetermined pattern of digital information;

(e) using the comparison of step (d) to locate a virtual edge of the tape if the -SO- acquisition pattern of step (d) does not match the predetermined pattern of digital information.

14. The method of claim 13, wherein step (e) locates a candidate transverse position for a virtual edge of the tape, and wherein the candidate of step (e) is averaged with at least one other candidate transverse position.

15. The method of claim 13, wherein the predetermined pattern of digital information consists of sixteen sets of a digital ONE followed by a digital ZERO.

16. The method of claim 13, further comprising: selecting a predetermined constant amplitude envelope; and adjusting the gain of the acquisition signal of step (b) if necessary to maintain the constant amplitude envelope.

17. The method of claim 16, wherein the adjusted acquisition signal is differentiated and, when the adjusted acquisition signal exceeds a predetermined fraction of a local average amplitude, the result of the differentiation is used to derive the acquisition pattern.

18. The method of claim 13, wherein the acquisition pattern is derived from the acquisition signal using a PRML technique.

19. A method of operating a magnetic storage tape drive, the method comprising:

(a) selecting a predetermined pattern of digital information; (b) positioning a write element at a plurality of transverse positions relative to the tape and, at each such transverse position, moving the tape in a direction parallel to the longitudinal direction of the tape to at least attempt to record magnetic flux having flux directional transitions corresponding to the predetermined pattern of digital information; and subsequently

(c) activating, at each of the plurality of transverse positions, a read element to sense magnetic flux and to generate an acquisition signal in accordance therewith, the read element being activated and the tape having been moved such that any flux directional transitions which were recorded on the tape are sensed by the read element;

(d) deriving an acquisition pattern from the acquisition signal obtained in step (c) ; and (e) comparing the acquisition pattern of step (d) to the predetermined pattern of digital information;

(f) using the comparison of step (e) to locate a virtual edge of the tape if the acquisition pattern of step (d) does not match the predetermined pattern of digital information.

20. The method of claim 19, further comprising: selecting a predetermined constant amplitude envelope; and adjusting the gain of the acquisition signal of step (c) if necessary to maintain the constant amplitude envelope.

21. The method of claim 19, wherein the acquisition pattern is derived from the acquisition signal using a PRML technique.

22. A magnetic tape drive comprising: a write element which generates a predetermined pattern of flux directional transitions, the predetermined pattern of flux directional transitions corresponding to a predetermined pattern of digital information; a read element for sensing magnetic flux proximate the read element and for generating an acquisition signal in accordance therewith; a head unit upon which the write element and the read element are mounted, the head unit being movable in a transverse direction which is orthogonal to a longitudinal direction of the tape; a tape transport for moving the tape in the longitudinal direction of the tape; a circuit which compares the acquisition pattern and the predetermined pattern of digital information to determine information relative to the location of a virtual edge of the tape.

23. The apparatus of claim 22, further comprising a signal processing circuit which adjusts the acquisition signal to maintain a constant amplitude waveform therefor regardless of the transverse positioning of the head unit and which derives therefrom an acquisition pattern.

24. The apparatus of claim 23, wherein the signal processing circuit comprises: an automatic gain controlled-amplifier which receives the acquisition signal and adjusts the gain applied to the acquisition signal to yield a constant amplitude waveform therefor; and a peak detection circuit which generates information for inclusion in the acquisition pattern when a qualified peak is detected in the adjusted acquisition signal, the qualified peak being at least a predetermined fraction of a local averaged amplitude.

25. The apparatus of claim 23, wherein the signal processing circuit comprises a PRML circuit.

26. The apparatus of claim 22, wherein the predetermined pattern of digital information consists of sixteen sets of a digital ONE followed by a digital ZERO.