CA1079846A - Record track following and seeking - Google Patents

Abstract

RECORD TRACK FOLLOWING AND SEEKING
Abstract An improved servo block pattern bands a plurality of parallel record tracks into a track seek and follow band.
Servo block positions in the record tracks, plus the lon-gitudinal duration of the signal bursts enable simultaneous track seeking and following within a band of tracks. Servo apparatus operable with such patterns adapt to amplitude variations of the servo readback signal to reduce the effect of amplitude variations on servo performance. The servo blocks may have differing frequencies or correlation patterns for enhancing track seek and follow functions.

Description

Related Patent U.S. Patent 3,185,982 to Sippel shows interspersed data and positioning information signals.
Zimmerman et al, U.S. Patent No. 4,087,843, issued May 2, 1978 and entitled "Positioning Device for the Access Arm of the Magnetic Head of a Magnetic Disk Storage".
Baca et al, U.S. Patent No. 4,052,741, issued October 4, 1977 and entitled "Track Seeking and Following".

r~

.. ,.~ :

. . . - . ., ~ . .. . .

- , .: .... ~ .:. . .::: : : -, . . .. . . .. . . . .

, . ~ . : . : .
. . . . . . -- . : :. ., . ~ : . :
- - . . - - : ::
- . - : . :
- - . . :
... . . .

1 The above-identified, commonly assigned,

2 U.S. patents relate to track following and

3 location using patterns relatable to the present

4 invention.
Background of the Invention 6 The present invention relates to record 7 track following and seeking, particularly to improved 8 patterns plus apparatus for utilizing such patterns , `
9 for both track following and seeking.
In magnetic data recorders, particularly -11 of the rotating disk type, high track densities, 12 i.e., densities greater than 500 tracks per inch 13 (tpi), require electronic servo controls for precisely 14 positioning recording and sensing transducers with respect to a given record track. This is achieved 16 by two modes of operation. First, the track to 17 be accessed must be first located. This actio~
18 is termed "track seeking" and is achieved by rela-19 tively moving the transducer with respect to the record member in a direction transverse to the 21 Length of the tracks until the appropriate track , ~ .
22 is located. Once the track is located, a second 23 mode of operation, called "track following", keeps ;

24 the transducer centered as close as possible on the track being accessed.

26 In a multidisk recorder, one of the disk 27 surfaces is reserved for containing such servo 28 patterns whereby a so-called "comb" head assembly can B~975016 -2-, .
.. . . . . . . .- , , . .' , 1 move radially of the record disk for accessing record 2 or data tracks on the other record surfaces. In 3 other disk recorders, where only a single disk surface 4 is accessed at a given time and employing a transducer assembly designed to operate with but a single disk 6 surface, the luxury of a separate servo signal disk 7 surface is not available. In those situations, so-8 called "sector servoing" is employed, such as taught g by Sippel, supra. In these latter systems, sector patterns, i.e., positioning information signals, are 11 interleaved between data signals such that as the 12 record member relatively passes the sensing transducer, 13 positioning and data signals are alternately sensed 14 by the transducer. This arrangement enables the transducer to be automatically^and precisely positioned 16 with respect to the record track being accessed.
17 While the above-stated actions on their 18 face appear simple and straightforward, as track density ~;
19 increases in order to create a greater storage capacity per disk surface, locating a record track from a plura-2~ lity of record tracks on a random basis becomes an 22 exceedingly difficult proposition. This is particularly 23 true if one desires to servo position the recording 24 transducer directly to the track to be accessed without overshoot plus a successful track access on each and 26 every try. An additional problem arises in that, as 27 the track widths are decreased and the track pitch 28 (center-to-center spacing of adjacent txacks) is --. ~ ,. - . . , . :
- . . .
.

107~846 1 decreased, the position indicating or servo signals 2 have correspondingly less amplitude. This means that 3 the amplitude of the readback signal becomes smaller 4 and more unpredictable. That is, the servo signals become more susceptible to dropouts, transducer-to-6 record-surface spacing variations, and intermodu]ation 7 in the transducer of signals detected from two adja-8 cent record tracks. The above-identified problems g are magnified when flexible record media are employed.
Further, the substrates of flexible record disks 11 exhibit such dimensional instability that position 12 indicating servo signals and the information signals 13 must be on the same disk, at least at higher record-14 ing densities (track density in particular).
Accordingly, it is desired to provide rapid 16 track acquisition and precise following. All of the 17 above should be achieved at minimum cost because of 18 the competitive nature of record storage apparatus.
19 Another important factor is the reliable detection and evaluation of the positioning or servo 21 signals received by the servo apparatus. Such servo 22 apparatus should exhibit a maximal tolerance to ampli-23 tude variations, intermodulation perturbations, record 24 track eccentricities, and yet be fully responsive to signals indicating a position error signal with 26 respect to positioning a transducer with respect to 27 a record track.

, 1 Summary of the Invention 2 It is an object of the present invention 3 to provide improved servo patterns for use in posi-4 tioning two members relative to each other, and for providing enhanced and simplified apparatus for inter-6 preting the signal patterns.
7 A servo signal pattern in accordance with 8 certain aspects of the invention comprises first and 9 second groups of servo signals aligned transversely to track locating lines on a record member. The ll servo signal groups are spaced longitudinally along 12 the lines. A servo signal in each of the groups has 13 an extent transverse to the lines of about the spacing 14 of adjacent ones of the locating lines and having different lengths or durations along the length of 16 the lines. It is preferred that servo signals having 17 shorter lengths are longitudinally adjacent signals 18 of longer lengths in the respective groups. Predeter~
l9 mined ones of the servo signals abut at least one of said locating lines, while other predetermined 21 ones of the servo signals are centered on other ones 22 of said locating lines.
23 In a preferred method of operation, the 24 servo signals are arranged in radial bands such that 25 coarse positioning can be directed toward first ~ ;
26 locating a band of servo signals. The servo signals 27 forming a block pattern within a band are not neces-28 sarily continuous. ~pparatus of the invention then ~.

~079846 1 identifies and locates a given one of the lines 2 within the band, 3 Transverse spacing between servo blocks or 4 patterns in at least one of the groups is approxi-mately one-half the line pitch, i.e., one-half the 6 spacing between adjacent ones of said locating lines.
7 Servo blocks in the two groups include pairs of 8 position indicating signals, including a pair having 9 a signal in one group centered on a locating line and another signal of the pair in the second group 11 being disposed on one side of such locating line.
12 Other servo signal pairs have signals from one group 13 abutting a given locating line on one side and a signal 14 from a second group abuts the same line from the opposite side.
16 Apparatus employable with the above-described 17 patterns perform track seek and follow functions for 18 two relatively moving members including first means 19 for measuring and indicating the duration of a plurality of servo signals~ Control means are responsive to 21 the indication to supply a set of control signals 22 in accordance with such indications. Integrator means 23 receive the servo signals and are responsive to the 24 control signals for integrating the xeceived servo signals at a given rate in accordance with the control 26 signals to supply an indication usable for track follow-27 ing or ultrafying positioning. Finally, compare means 28 are responsive to both indications to supply a position : . ' , 1 error signal for effecting relativ~ motion of the 2 members toward a given ox desired relative position.
3 In a variation of the basic fundamentals 4 of the present invention, the various servo signals have differing frequencies or other correlation type 6 characteristics for further enhancing the discrimination 7 power of the servo apparatus when sensing the servo 8 positioning signals. In a preferred form of the inven-9 tion, the servo patterns are combined with the appara-tus of the invention to provide a so-called "sector 11 servoing" wherein a given servo sector contains posi-12 tioning information which is interleaved among data 13 sectors of a record member. In a most preferred form, 14 the record member is a rotating magnetically coated record disk. A transducer moves radially of the disk 16 for accessing any of the plurality of tracks on the 17 disk for scanning same to supply sensed signals to 18 a servomechanism and to data signal handling apparatus.
19 A tachometer indicates precise angular position for 20 gating the servo and data signals to the appropriate ;
21 destination.
22 Another aspect of the invention is the adap-23 tion of apparatus to amplitude variations of the read-24 back signals, particularly when employing amplitudes as indications of position error. Each locating line 26 may signify location of one or more record tracks.
27 The foregoing and other objects, features, 28 and advantages of the invention will become apparent - . ~, . . .
- . :. . - . . : ~ . .. .
. . . - . : : : ~
. .

~079846 1 from the following more particular description of 2 the preferred embodiment, as illustrated in the ac-3 companying drawing.
4 The Dr~wing FIGURE 1 is a diagrammatic simplified plan 6 showing of a record storage disk which may employ ~
7 the present invention. ;
8 FIGURE 2 is a diagrammatic showing of a 9 prior sector servo pattern with which the apparatus of the present invention may be employed and with 11 which the present invention provides certain improve-12 ments thereover.
13 FIGURE 2A shows a simplified set of signal 14 waveforms illustrating the operation of servo functions with respect to the FIGURE 2 illustrated sector servo 16 patterns.
17 FIGURE 3 is a diagrammatic showing of sector 18 servo patterns of a constructed embodiment of the 19 present invention.
FIGURE 3A shows a set of idealized readback 21 signal waveforms usable to describe the operation 22 of sector servoing when using the FIGURE 3 illustrated 23 sector servo patterns.
24 FIGURE 4 is a simplified logic flow diagram of apparatus which incorporates the present invention.
26 FIGURE S is a combined logic flow and sche-27 matic diagram illustrating the operation of the present 28 invention and shows apparatus usable with the EIGURE 4 29 illustrated apparatus.

~0975016 -8-. ~ . . .

~079846 1 FIGURES 6-9 of which FIGURE 9 is shown on the sheet of drawings bearing FIGURE 2A show simplified signal and integration waveforms illustrating the operation of the FIGURE 5 illustrated apparatus for various signal and position conditions.
FIGURE 10 on the sheet of drawings bearing FIGURE 1 is a simplified logic flow diagram of a modification of the FIGURE 5 illustrated ap-paratus for incorporating a two-frequency discriminator function in the servo patterns.
FIGURE 11 diagrammatically illustrates an application of a multi-frequency discriminator function into the FIGURE 2 illustrated servosignal pattern.
FIGURE 12 is a diagrammatic showing of apparatus employing the FIGURE 11 illustrated discriminator function.
Detailed Description Referring now more particularly to the drawing, like numerals -indicate like parts and structural features in the various diagrams.
Preferred Environment FIGURE 1 diagrammatically illustrates a record storage disk 10 rotatable about axis 11 relative to radially positionable transducer 12 electrically connected to data and servo circuits 13. The recording area of disk 10 is divided into a plurality of separate recording areas, such as outer area 14 and inner area 15, each area having a given number of concentric record tracks scannable by transducer 12. In each of the record areas, the data recordings in ... -.. . . .- . ........ , ........ . .... ..... . . .... - :
. : . ~ - . ... : .

1079~46 1 such areas are interspersed by a plurality of inter-2 leaved positioning or se~tor servo signals indicated 3 by the hatched areas 16 ~nd 17, respectively, in 4 areas 14 and 15. The circumferential displacement at the radially inwardmost portion of the two record-6 ing areas between adjacent sector servo signal areas 7 16 and 17 are made identical in the two record areas.
8 In this manner, the time required for adjacent ones 9 of the sector servo areas 16 and 17 to pass under transducer 12 is at a predetermined minimal time.
11 For angular addressing purposes, an index line 20 12 extends radially through both record areas 14 and 13 15. Index line 20 resets all angular position indi-14 cating circuitry, as is well known in the art. The angular position of record disk 10, with respect to 16 transducer 12, is additionally indi~ated by a set 17 of tachometer marks 21 disposed at the outer circum-18 ference. While record disk 10 is a preferred form 19 of practicing the invention, the invention can be practiced with equal facility on rectilinearly trans-21 ported record members such as 1/2" tape; and the broad 22 pri~cipIes of the invention can also be applied to 23 relatively positioning two members, whether one is 24 a record member or not.
A Pr_ r System 26 FIGURE 2 illustrates a prior sector servo 27 pattern taken from record area 15 and including a 28 diagrammatic showing of two sector servo areas 17A

;, .

.
: . . . , , . . . ~ , ~791~46 :~

l and 17B. The horizontal lines represent so-called 2 ~track locating" lines. In the most preferred form, 3 such track locating lines identify the center of the 4 record track as illustrated at track C by track locating line 22A. Track C has a width, indicated by dash 6- lines 23, scannable by a sensing/recording transducer 7 gap diagrammatically illustrated by small rectangle 8 24. The tracks are banded into identifiable groups 9 of three. To address tracks B, the three tracks are respectively identified by the alphabetic characters 11 A, B, and C in bands 25, 26, and 27. The FIGURE 2 12 illustrated servo pattern assumes that a positioner 13 (not shown) can position a transducer 12 having a 14 gap 24 radially within a so-called "positioning window"
having an extent not greater than the radial extent 16 of a band of tracks. When this is done, the center 17 track B within a given band can be located by 18 examining the readback signal from the transducer 19 12 sensing the servo signals in the three servo groups 30, 31, and 32 in either of the servo sectors 21 17A and 17B. Each and every track to be addressed for 22 positioning purposes is a center track of a band of 23 three tracks, as wi11 become more apparent.
24 Referring to FIGURE 2A, the readback signals from the servo signals of FIGURE 2 are shown for a 26 track centered transducer, such as transducer gap 27 24 being centered over track C about center locating 28 line 22A. In this case, the track C is a center track.

., ~
"

~079846 AS gap 24 scans the sign~l from group 30, a first half 2 amplitude readback signa~ 30C is followed by full 3 amplitude readback signa~ 31B, which in turn is fol-4 lowed by a half amplitude signal 32C. This sequence of amplitudes is unique to track C thereby identifying 6 same. The full amplitude signal is termed a registra-7 tion servo signal or pulse. In a similar manner, 8 track A is identified by the sequence of half amplitude 9 signals 30A, followed by signal 31A, which in turn is followed by full amplitude (registration) signal 11 32A. Similarly, a track B is identified by a full 12 amplitude (registration) signal 30B, followed by two 13 half amplitude signals 31B and 32B. This pattern 14 yields a location window of + one track for each track to be located.
16 Inventive Di-Block Patterns . . ~
17 The present invention provides certain improve-18 ments over the FIGURE 2 and 2A track seeking and fol-19 lowing patterns, as will become apparent. In one aspect 20 of the invention, rather than having three tracks 21 in the band of tracks, a constructed embodiment 22 employing the present invention has a band of six 23 tracks A'-F', as best seen in FIGURE 3, in bands 35 24 and 36. Again, for each track to be located, there is 25 a set of six tracks constituting an initial positioning 26 "window". Each track A has a six-track window D to C, 27 tracks B from E to D, etc. The FIGURE 3 illustration 28 is also based upon a showing of record area 15 having .. . . . ... : .
.

107~8~6 :

1 two adjacent servo sectors 17A and 17B being scanned 2 by gap 24 of transducer 12. As shown in FIGURE 3, 3 a track location window of + 2-1/2 tracks is achieved.
4 The FIGURE 3 illustrated pattern has two groups of servo signals, a first group and a second group respec-6 tively transversely aligned with respect to the track 7 location lines 22. Gap 24 is shown as being centered 8 over track D' of band 35. Idealized readback signals 9 from the FIGURE 3 illustrated sector servo signals ... ... . .
are shown in FIGURE 3A. Track type A' is identified 11 by a half amplitude signal lA followed by a full ampli-12 tude signal 2A. Also note that the durations of the 13 signals are relatively short and long, respectively.
14 Track B' is indicated by a relatively short full ampli-tude signal lB separated by a long interval from a 16 half amplitude, long duration signal 2B. A C' type 17 track is identified by two successive half amplitude 18 long duration signals lC and 2C. Track D' is indi-19 cated by a long duration, one-half amplitude signal lD separated from a short duration, full amplitude 21 signal 2D. Track type E' is indicated by long dura-22 tion, full amplitude signal lE, closely followed by 23 a short duration, half amplitude signal 2E. The track 24 type F' is identified by two successive half amplitude, short duration, widely spaced signals lF and 2F. For 26 ease of reference later on, the signals are identified 27 as occurring in seven time periods 1-7.

"~ .

~79846 1 Turning to FIGURE 3, the signals for track 2 type D' are described. ~n the first group, gap 24 3 scans one-half the width of the first group signals 4 yielding a half amplitude, long duration signal lD
while scanning the entire width of the second group 6 signal yielding full amplitude, short duration signal 7 2D. As will become apparent, track position errors 8 are similarly indicated enabling efficient track fol- , 9 lowing operations. Each of the tracks and associated signals in FIGURE 3A can be analyzed to show the rela-11 tionship of the FIGURE 3A signals to the transverse 12 locations of the servo signal shown in FIGURE 3.
13 Some comments about the geometry of the 14 FIGURE 3 illustrated pattern are first in order. The transverse locations of the signals in the respective 16 first and second groups alternate between being cen-17 tered on a track locating center line 22 and being 18 disposed intermediate two adjacent ones of said lines.
19 While the two groups of servo signals are generally longitudinally aligned, the servo signals, each of 21 which have a width approximating the pitch of the 22 tracks represented by the transverse spacing of the 23 lines 22, are each offset from another providing the 24 amplitude changes in the FIGURE 3A illustrated signals, the exception being a second type of track indication 26 for tracks C' and F'. Further, the duration of the 27 signals in the first group is relatively short when 28 substantially longitudinally aligned with longer signals 1 in the second group, and vice versa. Track C' is 2 identified by two identiaal length signal bursts, 3 each of which abut the track center line, as is track 4 F'. By intermixing such signals in the two groups, a minimal longitudinal extent is required for pro-6 viding positioning signals. The hatched areas in 7 FIGURE 3 represent constant frequency recorded sig-8 nals. In one embodiment, five cycles were recorded 9 for each servo time period (1-7). In other embodiments of the invention, as will be explained in more detail 11 later, adjacent tracks may be distinguished by bursts 12 of differing frequencies or other correlating type 13 of signal patterns.
14 While FIGURE 3 illustrates a two-group pattern for uniquely identifying six record tracks in a group 16 of tracks in a reliable manner, other patterns having 17 a greater number of groups of servo signals are within 18 the contemplation of the present invention, which 19 results in a greater number of tracks in each band.
For example, other patterns (not disclosed herein) 21 resulted in identifying 25 tracks in each band of 22 tracks.
23 Having described the patterns for sector 24 servoing or other servoing positioning operations, factors involved in track seeking and following using 26 the concepts of the present invention, are described.
27 In a track seeking mode within a given band, all nearby 28 or adjacent tracks must be uniquely identifiable, such ~079846 1 as identifiable by the FIGURE 2A and 3A indicated 2 readback signals. For a relatively narrow band of 3 tracks, i.e., a small nu~ber of tracks in the band, 4 care has to be exercised to avoid overshooting the transducer outside the band. This means the narrower 6 the band of tracks, the slower the seek speed, hence, 7 possibly increased access time to a given track. Accor-8 dingly, by extending the number of identifiable tracks g in the band to six, the present invention enhances access time by allowing a more rapid seek speed.
11 Similarly, for coarse positioning transducer 22 with 12 respect to a band of tracks, a larger band permits 13 a more rapid access to a given band, as well as permitting 14 greater tolerances in the positioning accuracy of such a coarse positioner. Such a coarse positioner 16 could use a linear scale on the frame of the apparatus 17 mounting the record member, such as is employed in 18 present direct access storage devices of the disk 19 type. All of these factors affect the allowable maximum track density or minimum track pitch and, hence, has a 21 material effect on the storage capacity of a record 22 member.
23 In accordance with a preferred mode of ope-24 rating with the inventive servo patterns of this inven-tion, track seeking within a band of tracks is achieve~
26 by merely measuring the duration of the sector servo 27 signals within each sector. Examination of FIGURE
28 3A shows that each track is identified by two signals, 10798~6 1 each of which have a unique combination of signal 2 durations. Once ~ track to be accessed has been lo-3 cated by measuring the d~rations and comparing same 4 with the indicated track type, track following operations proceed. This is preferably achieved using integration 6 techniques. The readback signals, as represented 7 in FIGU~E 3A, are first rectified, then integrated.
8 The integration time constant of the integrators is 9 adjusted in accordance with the signal durations;
that is, in integrating signal lA, the integrator 11 is assigned a shorter time constant than integrating 12 signal 2A. An effect of these differing integration 13 times is to neutralize the effects of the burst dura-14 tion times, as well as normalize readback signal ampli-tudes for adapting the servo apparatus to signal per-16 turbations. Integration for the first signal is pre-17 ferably the inverse of the integration of the second 18 group signal such that on a track centered position, 19 the resultant net integrated signal is zero, and the sign of the any net integration signal indicating 21 direction and magnitude of position error. During 22 track following operations, the track seeking portion 23 of the servomechanism is preferably kept in the active 24 state for yielding transducer radial position correc-tions as may be necessary by large disturbances in 26 the servomechanism, which may cause the transducer to 27 move radially off-track. The active track seeking 28 portion can quickly recover from such an error for . .

10798~6 1 maximizing data throughp~t during signal exchanging 2 operations wit~ t~e record member.
3 A Preferred Implementatiqn 4 One implementation of the invention is shown in FIGURE 4. The description assumes that head posi- -6 tioner 60 has moved head support arm 61 into transducer 7 access such that transducer head 62 (corresponding ~ -8 to head 12 of FIGURE 1) is scanning a desired track 9 to be accessed. Such track seeking operations will be detailed later with respect to FIGURE 5. The descrip-11 tion also assumes a separate tachometer disk 10A having 12 a fixed angular position to disk 10 such that tachometer 13 fiducial mark 20A precedes radial index line 20 by 14 a small angle. The description starts with photosensing unit 63 sensing the tachometer index mark 20A supplying 16 a tachometer index signal over line 64 to reset the 17 angular positioning sensing and controlling circuits 18 which include resetting angular position counter 65, 19 angular address register 66, and buffer register 67. -Additionally, pre-index flip-flop 68 was set to the 21 active condition signifying that transducer 62 is 22 scanning an erased portion prior to the disk radial 23 index line 20 (FIGURE 1). Pre-index flip-flop 68 24 conditions AND circuit 69 to respond to a signal signi-fying detection of the disk fiducial line 20 signal 26 to supply a disk index signal over line 74.
27 The disk index signal and all other control 28 signals are produced initially by transducer 62 sensing . :
. :, ....... .... . . .. .
: - . , , .

~079846 1 the signals represented by the FIGURE 1 illustrated 2 format and supplying same over line 70 to servo and 3 data signal separator 71 and to envelope detector 4 (amplitude sensor) 72. 9ervo and data signal sepa-rator 71 supplies the servo-separated signals to head 6 positioner 60 for enabling transducer 62 to faithfully 7 scan the track being accessed. Additionally, separator 8 71 supplies the separated data signals to data handling 9 circuits 73 for processing in a known manner. Tech-niques of data handling circuits 73 bear no significance 11 on practicing the present invention and, hence, are 12 omitted for purposes of brevity. Envelope detector 13 72, upon receiving the signal generated by transducer 14 62 corresponding to disk fiducial line 17, supplies an active signal to AND circuit 69. AND circuit 69 16 was conditioned by pre-index flip-flop 68 to pass 17 the envelope detector 72 signal to line 74 as disk 18 index signal 75. This action corresponds in time 19 with transducer 62 scanning radial index line 20.
In preparation for normalizing the content 21 of angular address register 66 to the record areas 14 22 and 15 in which 62 is scanning a given track, the disk 23 index signal conditions AND circuits 80 to pass the 24 signal content of angular position counter 65 to adder 81. The angular location of servo sectors 16 differs 26 from servo sectors 17; hence, the angular position 27 indicator circuits are automatically adjusted or nor-28 malized to either area 14, 15. Photosensing unit 82 :
' :.
- BO975016 ~19-. .:

, .. . ~ . ' , . . .............................. .

1079~46 1 senses the tachometer marks 84 and supplies tachometer 2 signals to increment the content of angular position 3 counter 65. At the time disk index signal 75-occurs, 4 the signal contents of a~gular position counter 65 is No~ the known angular displacement between tachometer 6 index 20Aand radial index line 20. These signal con~
7 tents are supplied through adder 81 to buffer register 8 67 in preparation for insertion into angular address 9 register 66.
Insertion of the angular position counter 11 signal contents into angular address register 66 is 12 now described. The disk index pulse on line 74 also 13 resets servo sector trigger type flip-flop 82. As a 14 result, an activating signal travels over complement output line 83 which, in turn, causes the servo sector 16 trigger to be set; i.e., it acts as a monostable multi-17 vibrator. As such, a negative pulse is supplied over 18 line 84 to read-only memory 85 input address register 19 86. A portion of address register 86 contains track 2b addresses which are supplied to head positioner 60 21 over cable 100 to be used as later described. One -22 position of register 86 indicates whether head 62 23 is scanning servo signals or data signals. This indi-24 cating signal is supplied over line 87 to ROM 85 for addressing, as will be later described, to servo 26 data separator 71 and to data handling circuit 73.
27 Additionally, this negative pulse strobes signals ;
28 into buffer register 67 and angular address register 66.

; . .

.. , . . . - . , - . .
: . . ~: . : . . : : . .
, . .,, : : ~
. .
, ~079846 1 Hence, the No signals go to buffer register 67 and 2 angular address register 66. This action is so fast 3 that angular position co~nter 65 has not yet counted 4 past the No tachometer c~unts. Hence, digital compare circuit 88 indicates a compare successful signal on 6 line 89 to trigger servo sector trigger 82 to the 7 reset state indicating "not-servo" time.
8 Additionally, the disk index signal on line 9 74 sets the post index flip-flop 90 to the active condition. This action forces a binary 1 into address 11 register 86. Then, as set forth in the table below, ;
12 the addresses of register 86 access one of the five 13 registers in ROM 85. The addresses are set forth 14 in the lefthand column, wherein the X's indicate "don't cares", and the content of the registers is in the 16 righthand column. In the ROM address, the middle 17 symbol indicates the area (14=A, 15=B), the lefthand 18 symbol indicates the activity of the post index flip-19 flop, and the righthand symbol indicates the activity of the servo sector trigger 82. Each time trigger 21 82 is triggered by a signal on line 89, it acts as -22 a monostable multivibrator for the duration of the 23 servo time such that the appropriate numerical content 24 of the registers is supplied through adder 81 to buffer register 67 and angular address register 66.

:. .
~- BO975016 -21-.... .

.
- .
~ , . . . . ..
~ ' " ' '", . ",~'-''~' , . ' , ~079846 2 Address Content 3 0XX No 4 lA0 N
~a lAl Mja 6 lB0 jb 7 lBl Mjb 8 The terms Nja and Njb represent the angular 9 extents of servo sectors, while Mja and Mjb represent the angular extents of data sectors respectively in 11 areas 14 and 15.
12 Each time servo sector trigger 82 is trig-13 gered by compare 88, the signal contents of address 14 register 86 of ROM 85 are transferred to adder 81 and added to the contents of angular address register 16 66 for updating the angular address for the next data 17 sector or the next servo sector, as the case may be.
18 The above action is repeated for the duration of the 19 track. A recurrence of the tachometer index signal on line 64 restarts and recalibrates the angular address-21 ing of the track. Hence, even with error conditions, ~;
22 there is automatic recovery because of this automatic 23 recalibration. Construction of the ROM can be using 24 known read-only memory techniques, such as capacitive memories, conductive memories, mechanical pinboards, 26 electrically settable flip-flops, and the like.
27 Referring to FIGURE 5, the constructional 2A features of the head positioner 60 are next described.

.. , - ~ .
.
.

10'79846 1 Head positioner 60 per~orms two functions. The first 2 function ascertains the ~rack number within any group 3 of six tracks, wnether it be A-F. The second function 4 is following an addressed track. The arrangement is such that both functions can be performed simul-6 taneously such that a perturbation in track following 7 automatically invokes track seeking for rapid automatic 8 track positioning error recovery.
9 The track seeking function is implemented by burst length comparator circuit 101, which supplies 11 its output signals over cable 102 to pulse width modu-12 lator circuit 103. Pulse width modulator 103 combines 13 the outputs from track seeking circuit 101 and the 14 track following circuit 104 to supply a servo controlling position error signal (PES) over line 105 to servo-16 mechanism circuits 106 which, in turn, position head 17 carrier 61 with respect to record storage disk 10.
18 It is to be understood that head 62 has been positioned 19 within the group of tracks by a coarse positioning circuit including a linear tachometer 110 affixed 21 to head support 61 and providing positioning signals 22 over cable 111 to compare circuits 112. The track 23 address and the actual linear tachometer indicated 24 address are compared by circuits 112 to supply a coarse position error signal over line 113 to servo circuits 26 106 to initially coarse position transducer 62 radially 27 with respect to disk 10. Servo circuits 106 may include 28 dual mode circuits which are selectively responsive 1 to either the 113 line signal indicating a coarse 2 error or the line 105 si~nal indicating a fine posi- -3 tion error. It is to be understood that the 113 coarse 4 position error will override the fine position error signal on line 105. Rnown multimode servomechanism 6 techniques may be used for such selection.
7 Returning now to the description of circuits 8 101, the envelope indicating signal received from 9 envelope detector 72 (FIGURE 4) over line 72A is dif-ferentiated by negative transition differentiator 11 115. Each time the envelope falls in amplitude (as 12 at the end of the burst of signals as indicated by 13 lines 116 and 117 of FIGURE 3), circuit 115 supplies 14 an actuating pulse to AND circuit 120. AND circuit lS 120 supplies these negative end-of-burst indicating 16 signals only during the sector servo time. To achieve 17 this result, the servo sector pulse on line 84 sets 18 servo sector latch 122 to the active condition and 19 simultaneously sets burst select latch 125 to the actlve condition. Burst select latch 125 being set 21 indicates that the signal being received is in the 22 first-appearing burst in the first group of servo sig-23 nal bursts. The line 84 signal leading edge also 24 resets divide-by-two circuit 126 priming it for reset-ting the servo sector latch to signify the end of 26 the servo sector. For each servo sector, as shown 27 in FIGURE 3, circuit 115 supplies two negative pulses 28 over line 127 to divide-by-two circuit 126 corresponding, 10~9846 1 respectively, to burst e~ds 116 and 117. The divide-2 by-two circuit, upon receiving the second negative 3 pulse corresponding to b~rst end 117, resets servo 4 sector latch 122 (data time is indicated). Accor-dingly, burst select latch 125, when in the set state, 6 signifies the first group burst is being read; while 7 in the reset or second state, signifies a second group 8 of servo signals could be read.
9 It is remembered that the duration of the respective bursts signifies which track is being sensed.
11 A pair of AND circuits 130 and 131 respond respec-12 tively to the burst select latch 125 being set and 13 reset to supply pulses from timing oscillator 132 14 to actuate counters X and Y for metering servo signal duration. Initially, the leading edge of the line 16 84 pulse resets both counters to a reference state.
17 Accordingly, as long as the line 72A envelope indica~
18 ting signal is active, either AND circuit 130 or 131 19 will supply pulses to the two counters for metering the durations of the first or second group signal 21 bursts. Counters X and Y act as digital integrators 22 with memory for metering the elapsed time. Each of 23 the counters supply their respective output signals 24 over cables 132 and 133 to read only memory (ROM) decoder. The signals on cables 132 and 133 corres-26 pond to unique addresses within ROM at which there 27 are stored the track numbers A-F for a band of six 28 tracks. The inner band track address is determined ,...-, . ~ .

1 in combination with the burst select latch 125 output 2 signals indicating that the second burst has been 3 successfully read. When burst select latch 125 is 4 set to the initial condition, a leading edge pulse traveling over line 35 causes an output register (not 6 shown~ in ROM to be reset. Similarly, a negative-7 going pulse traveling over line 136 signifying burst 8 select latch 125 being reset causes the signal con-9 tents of the addressed register in ROM to be inserted into the output register for then being supplied over 11 cable 102 to pulse width modulator 103.
12 Pulse width modulator 103 includes a track 13 position detector having a track decoder 140 which 14 receives track addresses from register 86 (FIGURE
4) and converts same into addresses A-F for within 16 band track addressing. An arithmetic comparator 141 17 responds to the actual address on cable 102 and to 18 the decoded track address signals from decode 140 19 to supply track number error signals over cable 142 to compare circuit 143. As will be later described, 21 compare circuit 143 responds to the track number error 22 signals on cable 142, plus the off-track position 23 error signals from a properly addressed track, i.e., 24 when the signal on cable 142 indicates correct track, to supply a pulse width position error signal PES
26 over line 105.
27 Once transducer 62 has been properly posi-28 tioned over a track, track following circuits 104 .
. . . . .

1 cause the transducer to ~aithfully follow the track 2 center or location lines in accordance with the relative 3 amplitude measurements generated by a gap 24 scanning 4 the respective bursts, as shown in FIGURE 3 and as will be described in detail with respect to FIGURES
6 5-9. Track following circuit 104 receives the servo 7 sector signals over line 71A from separator 71 (FIGURE
8 4). Such signals may be equalized, but are not other-9 wise processed in order to maintain precise amplitude indications sensed by transducer 62. Rectifier 150 11 full-wave rectifies received servo signals and supplies 12 positive and negative output signals, respectively, 13 over lines 151 and 152. Field effect transistor (FET) 14 153 transfers the positive signal on line 151 during the first servo signal burst (first group), as 16 indicated by burst select latch 125 to summing node 17 155; while FET 154 supplies a negative signal from 18 line 152 during the second group signal burst, also 19 as indicated by burst select latch 125. Both signals pass node 155 to bipolar integrator 156. Integrator 21 156 includes differential amplifier 157, integrator 22 capacitor 158, squelch control transistor element 23 159, and time constant controlling transistor-type 24 switches 160-162. The ROM supplies track address signals over cable 102 to control the time constant 26 switches 160-162 for altering the integration rate 27 to bala~nce the integrations in accordance with the 28 desired amplitudes. That is, as shown for track D' 1 in FIGURE 3, the first b~rst in group 17A ending at 2 116 ideally has a ~ne-half amplitude signal, while 3 the second group signal ending at 117 has a full amp-4 litude signal. For prop~rly indicating track centering of gap 24 with respect to track D', the integrated 6 amplitude versus time for both successive slgnal bursts 7 should be made equal. This allows the servoing to 8 null at the track location line and facilitates servo 9 positioning control. This action is achieved by varying the integration rate, as will be described.
11 The above balancing is better understood 12 by referring to FIGURE 6. The track D' centered idealized 13 servo readback signals are shown in the top view wherein 14 the burst signal from the first group servo signal has one-half amplitude, and the burst signal from the 16 second group servo signal has a full amplitude readback.
17 To equalize the time-amplitude integral, the time ~ -18 constant imposed on integrator 156, when FET 153 is 19 supplying its signal, is three-fourths that used for the second group servo signal. For example, the first 21 group burst signal ending at 116 has 12 units of 22 amplitude-time integration, while the second group 23 signal ending at 117 has 16 amplitude-time units of 24 integration. To balance the resultant integration values (the area under the curves) based upon common 26 constant reference value VR, the first group signal 27 ending at 116 is integrated at three-fourths the rate 28 of the second group signal. Calculation shows that the 1 two integrations are the~ equal; and if subtracted a one from the other, the PES should be zero. As shown 3 in FIGURE 6, when gap 24 is disposed more toward track 4 location line C' than the track D' center line, a posi- -tive PES signal results; whereas, if gap 24 is disposed 6 toward track location line B', a negative PES signal 7 results.
8 The switch settings for the various track 9 addresses within a six-track band are given in the table below for the three time constant determining 11 switches of integrator 156. Switch settings are for 12 integrating the first group signals and the second 13 group signals. A binary 0 indicates a switch is open 14 (no current flows) and a binary 1 indicates the switch is closed (current flows). For a greater number of 16 tracks within a band, a greater number of switches are 17 provided.

19 First Group Settings Second Group Settings Time 160 161 162 Time 160 161 162 Track Constant R R/2 ~ Constant R R/2 R/6 21 A' 2 0 0 0 8 0 0 22 B' 4 0 1 0 3 1 0 0 23 C' 3 1 0 0 3 1 0 0 24 D' 3 1 0 0 4 0 1 0 E' 8 0 0 1 2 0 0 0 26 F' 2 0 0 0 2 0 0 0 27 The integrator 156 accumulates the amplitude-28 time integral as above described and supplies same to BO97501~ -29-.. ..
,, . , . : : ' .: . ~
.
,: , ~ . - .

1(~79846 1 analog-to-digital conver~er 165 in pulse width modu-2 lator 103. Converter 16S supplies a digitized version 3 of the position error si~nal PES to be used as later 4 described. Converter 165 also responds to a normal-izing signal received over line 166 from amplitude 6 normalizing circuit 167, which provides automatic 7 gain control for adjusting the position error signal 8 to be normalized irrespective of amplitude perturbations 9 of the readback signal. That is, during a first read operation of the servo sector, a first amplitude of, 11 for example, 2.75 millivolts, may be read. After 12 an elapsed time of several hours, the temperature 13 of the apparatus may changei for example, the tempera-14 ture may increase. As a result, the signal amplitude increases because of decreased transducer-to-medium 16 spacing, for example, up to 2.85 millivolts. Operation 17 of the positioning servo should be insensitive to 18 such perturbations. Accordingly, converter 165 responds 19 to the line 166 reference signal for adjusting the ,20 conversion, i.e., gain of the signal, in a known manner.
21 Normalization circuit 167 includes an inte-22 grator 168 identical to integrator 156.
23 The time constant switch settings for switches 24 170, 171, and 172 are shown in Table II below, with the first group settings indicating reading the full 26 amplitude first group servo signal; while the second 27 group settings indicate the time constant switches 28 for reading the full-amplitude signals occurring in 29 the second group of servo signals.

.,~

, . . .
- - - .: . . .
' -': .' . '' ''. ' ',,, ' ~. ', .. ..

1~'79846 TABLE~ II
2 First Group Setti~gs Second Group Settings 3 Time Time Track Constant R R/2 R/6 Constant R R/2 R/6 4 A' X X X X 8 0 0 B' 4 0 1 0 X X X X
6 C' 3 1 0 0 3 1 0 0 7 D' X X X X 4 0 1 0 8 E' 8 0 0 1 X X X X
9 F' 2 0 0 0 2 0 0 0 The X's signify that the integrator is not 11 used because the corresponding servo signal is a half-12 amplitude signal, while the other servo signal is a full-13 amplitude signal.
14 Normalization or reference circuit 167 also receives signals via a pair of FET ' s 174 and 175 16 respectively activated by the set and reset output 17 signals of burst select latch 125. Circuit 175 passes 18 the voltage reference VR during the reset phase of 19 burst select latch 125, while FET 174 supplles the negative signal from line 152 to integrator 168 during 21 the first portion of the servo signals. In this manner, 22 the full-amplitude signal received from rectifier 23 150 is compared in integrator 168 with VR. The net 24 difference of the two integrations is supplied over line 166 to converter 165 for normalizing the analog-26 to-digital conversion, as previously mentioned. Integra-27 tor 168 is squelched by transistor element 178 in 28 response to the signal from servo sector latch 122 ,. ..
.

1 going negative as determined by the NOT circuit 179 2 and delay 180.
3 Coaction of normalization circuit 167 and 4 track following circuit 104 is better understood by referring to FIGURES 7, 8, and 9. In FIGURE 7, track 6 A~ is shown having a first group half-amplitude signal 7 followed by a second group full-amplitude signal. I
8 The integrator 156 signal and the integrator 168 signal , -.
9 are shown. FIGURE 8 shows track type A' which is offset from track center with the signals from inte-11 grators 156 and 168 shown in the bottom portion of 12 the figure. Note that VR is the same amplitude and, 13 at PES, is adjusted upward.
14 FIGURE 9 shows the operation when two half-amplitude servo signals are provided for track type 16 C. Again, the integrated values for integrators 156 17 and 168 are shown.
18 Pulse-width modulator 103 responds to the 19 converter 165 output value and the digital value from arithmetic comparator 141 to generate variable pulse 21 widths on line 105 in accordance with the digital 22 difference between the signals on cable 142 and the 23 output ~f converter 165. To this end, counter 185 24 is activated by AND circuit 186 to count up to the ~ .
difference value of compare 143. When the value in .
26 counter 185 equals the values of signals supplied 27 by cables 142 and converter 165, an output pulse is .
28 terminated thereby yielding a pulse-width representa-', ~

. .

1 tion of digital values on cables 142 and converter 2 165. To this end, AND circuit 186 passes the oscil-3 lator 132 pulses for tallying time in counter 185.
4 The signal on line 187 from NOT circuit 179 activates AND 186 jointly with a positive output from compare 6 143 received over line 189. Therefore, as soon as 7 a compare is successful between counter 185 and the 1, 8 values on cable 142 and converter 165, the signal 9 on line 105 decreases. This decrease is transferred over line 189 disabLing AND circuit 186 thereby stop-11 ping counter 185. Stoppin~ counter 185 automatically 12 resets same in accordance with known techniques.
13 Multiple Frequenc~_T ack Discrimination 14 A further advantage can be provided by dis-crimination between sub-bands of tracks defined by 16 multiple frequency servo signals. For example, tracks 17 A' and B' may be respectively recorded at frequencies 18 Fl and F2. Such frequency difference can be used 19 to further distinguish and discriminate between the servo signals and thereby more reliably identify the 21 tracks. The servo signals received from separator 22 71 (FIGURE 4) over line 71A are given to filters 195 23 and 196 (FIGURE 10), respectively, for frequencies 24 Fl and F2. The outputs of the filters respectively -go to rectifiers 197 and 198 corresponding to the 26 full-wave rectifier 150 of FIGURE 5. In this instance, 27 the tracks A', C', and E' are rectified by rectifier 28 197; while the servo signals from tracks B', D', and - : , ,. . : . , :
.

1 F~ are rectified by rectifier 198. The outputs of 2 the rectifier~ ca~ be li~ited by limiters 200 and 3 201, respectively, and supplied to circuit 101 to 4 be pulse-width measured ~s previously described. Ad-ditionally, the rectifiers 197 and 198 supply their 6 output signals to differential amplifiers 203 and 7 204 to be compared with a voltage reference VRl. The 8 output of differential amplifiers 203 and 204 drives 9 the switching network and decoder in accordance with burst select signals from burst select latch 125 received, 11 respectively, over lines 205 and 206. Decode circuit 12 207 responds jointly to the signals from differential 13 amplifiers 203 and 204 and the signals on lines 205 14 and 206 to indicate the track address over cable 208.
The signals on cable 208 are then compared with the 16 decoded address from cable 140A from decoder 140 in 17 compare circuit 209 to supply control signals to the 18 pulse-width modulator 103 over cable 142A. Track 19 following is as previously described.
The principle of multiple frequency track 21 discrimination is equally applicable to the track 22 patterns illustrated in FIGURE 2. In this particular 23 situation, providing a pattern of differing frequencies 24 and combining same with the triplet amplitude coding (three-track identification or discrimination), the 26 band of identifiable tracks is expanded from three 27 to 24, an eight-fold increase. If two frequencies 28 Fl and F2 are recorded in the FIGURE 2 illustrated .

1~79846 1 patterns, the band shown in FIGURE 11 results wherein 2 t`he numerals 30, 31, and 32 correspond to the numerals 3 in FIGURE 2 for identifying the columns of servo bursts 4 extendi~g transverse to ~he length of the tracks.
Track numbers for a given band are shown in the lefthand 6 column. Resultant readback signals for each of the 7 tracks, when the transducing gap is centered over 8 the track location line, are shown in the righthand 9 portion with the numerals of the track, 1-24, also identifying the readback signals. The frequencies 11 of the differing bursts are shown underneath the enve-12 lope waveform. Examination of the signal bursts in 13 columns 30, 31, and 32 shows a binary progression.
14 Column 32 has the least significant binary digit repre-sentation, wherein every other burst is Fl, F2. Column 16 31 has the second digit position of a binary sequence, 17 wherein two successive bursts have the same frequency.
18 Finally, column 30 has the third digit position. It 19 is understood that other combinations of frequency assignments can also be employed with advantageous 21 results. For example, gray coding may be employed 22 or an arbitrary pattern may also be designed.
23 The illustrated frequency burst geometric 24 arrangement provides for hemibands, quadribands, and octabands. The table below illustrates the relationship 26 of the tracks to the frequency and amplitude identi-27 fiers.

, . ..

- ,, - ,~ . ' ~L079846 2 Track~requenqy Amplitude 27 Inspection of the above table shows that 28 the triplet amplitude coding corresponds to the fre-.
..

1 quencies in the rightmost column of the frequency 2 tabulation. That is, i~ tracks 1, 2, and 3, Fl is 3 in the rightmost column corresponding to the first 4 triplet; tracks 4, 5, a~d 6 have frequency F2 corres-ponding to a second triplet, etc. The center column 6 of frequencies represents quadribands (logical, not 7 physically contiguous), wherein each band of tracks 8 1-24 can be divided into four groups of tracks as 9 follows: 11 represents tracks 24, 1-5; 12 represents tracks 6-10, 23; 21 represents tracks 12-17; and 11 22 represents tracks 18-22, 11. Finally, the leftmost 12 column of the frequency tabulation defines the hemi-13 bands, wherein frequency Fl defines tracks 23, 24, 14 1-10; while frequency F2 represents tracks 11-22.
The octabands correspond to the triplet amplitude 16 coding groups as mentioned above.
17 Providing a two-frequency modulation of the 18 three group sector servo blocks increases the band 19 extent from three tracks to 24 tracks. While this change is an amazing increase, a more significant 21 result is the increase in off-track recovery capa-22 bilities. That is, once the apparatus is in a track 23 following mode, errors may cause the following trans-24 ducer to move completely away from the track being followed -- an off-track error. The positioning pat-26 tern ideally enables the apparatus to automatically 27 find the track being followed without going to a 2~ track seek mode or other time-consuming error recovery .. , , , ,- ~ . : :
.. . . .
.
: :

lV79846 1 procedures. Without the frequency additions, off-track 2 recovery capability is + one track. With the invention 3 frequency addition, off-track recovery increases to 4 + 11-1/2 tracks. Yet an~ther improvement is a reduc-tion in total track seek time.
6 FIGURE 12 diagrammatically illustrates an 7 implementation of the FIGURE 11 illustrated track 8 identification patterns. It is to be understood that ~ -9 coarse positioning places the transducer 62 within a band. Tertiary controls can select one of the sub-11 bands mentioned above. The coarse positioner is ac-12 tuated by receiving a desired address over cable 220 13 and decoding same in adder and stepper logic block 14 221. A control signal flows over line 222 to stepper motor 223 which operates based upon comparison of 16 the received signals with a tachometer system (not 17 shown) well known in the art. The stepper motor moves 18 the assembly 225 to a predetermined coarse positioning 19 location corresponding to a predetermined number of tracks, i.e., may be plus or minus ten tracks, for 21 example. A second carriage 226 mounted for movement 22 on assembly 225 supports head arm 61 for moving head 23 62 to the desired tracks within the constraints of 24 a single group of tracks, such as ten or 24. The latter operation is perormed by sensing signals 26 supplied by head 62 through preamp equalizer and 27 separator 71 with the output servo signals being 2~ supplied over line 227. Servo sector gating is achieved , . . .

1 as aforedescribed using tachometer 10A, which co-rotates 2 with disk 10 about spindle 11. Tachometer signals 3 flow over line 228 to co4nter 65A and to disk initiali- ~ -4 zation circuits 230. The disk initialization circuits are those described with respect to FIGURE 4 whPrein 6 counter 65A corresponds to angular index counter 65 7 of the FIGURE 4 illustration. The count or angular 8 position indicated by counter 65A is supplied to track 9 code logic and ROM 231 which operates in the same manner as described previously for FIGURE 4. The track location 11 and error are respectively supplied over lines 232 12 and 233 to adder and stepper logic 221 which, in 13 turn, supplies the error signals over 222 to motor 1~ 223.
Further, the servo signals on line 227 16 actuate track following circuits 235 as previously 17 described with the PES signal being supplied over 18 line 236 to carriage 226 which includes servomechanisms.
19 Further, the following circuits supply a track-centered indicating signal over line 238 to decode 231 which 21 can inhibit track seeking operations or indicate 22 that track seeking can be momentarily dispensed with.
23 Further, the multiple frequencies of servo signals 24 also travel to frequency discriminator and comparator 2S 240 which is constructed similar to the apparatus 26 shown in FIGURE 10. The detected frequencies Fl 27 and F2 respectively travel over lines 241 and 242 -28 to decode 231. Decode 231 decodes the Fl and F2 signals ,. :
'.

. . , ., - .- . . .
- . ~

~079846 1 in a time domain as desc~ibed and shown in Table ~:
2 III. The principles sho~n in FIGURE 10 are applicable 3 to the design of decode ~31. Further, disk initializa-4 tion circuits 230 supply gating signals over line 243 for controlling and timing the operations of circuits 6 235 and 240.
7 While the invention has been particularly 8 shown and described with reference to a preferred 9 embodiment thereof, it will be understood by those skilled in the art that various changes in form and 11 detail may be made therein without departing from :~
12 the spirit and scope of the invention.

. BO975016 -40-,, :

:

Claims (18)

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

1. A record member having a plurality of bands of parallel record tracks spanned at intervals by servo signal areas, the record member having groups of servo signals so recorded in each servo signal area that in each servo signal area each record track has a servo signal from each group recorded therein, the servo signals being located in a pattern in relation to the centre locating lines of the record tracks such that the relative amplitudes of the servo signals when read from any one record track by a transducer following the track serve to indicate the position of the transducer in the track, the servo signals being recorded to have a variable physical characteristic which alone or in combination with the said pattern is such that each record track can be distinguished from the others in the same band by reference to the servo signals recorded therein.

2. A record member according to claim 1 wherein the servo signals in each servo signal area are located in a pattern wherein some of the servo signals span a record track and some span the interval between adjacent record tracks.

3. A record member according to claim 1 wherein the said variable servo signal characteristic is that of the duration of the servo signals along the record track.

4. A record member according to claim 2 wherein the said variable servo signal characteristic is that of the duration of the servo signals along the record track.

5. A record member according to claim 3 or claim 4, wherein the servo signals are recorded as oscillations of constant frequency.

6. A record member according to claim 3 or claim 4 wherein the servo signals in alternate tracks are recorded as oscillations of a first frequency and the servo signals in the intervening tracks are re-corded as oscillations of a second frequency.

7. A record member according to claim 1 wherein the said variable servo signal characteristic is that of the frequency of the servo signals in the record track.

8. A record member according to claim 2 wherein the said variable servo signal characteristic is that of the frequency of the servo signals in the record track.

9. A record member according to claim 7 or claim 8 wherein there are three groups of servo signals in each servo signal area and the frequency of the servo signals is either a first frequency or a second frequency.

10. A record member according to any one of claims 2, 3 or 7 which takes the form of a magnetisable record disk having concentric record tracks.

11. Apparatus for reading a record member as claimed in Claim 1 the apparatus having a movable transducer means to support the record member for movement relative to the transducer, coarse positioning means to position the transducer to access the tracks within a selected band of the tracks on the record member, register means to register the selection of a track within the selected band, track seeking means to derive from the servo signals read by the transducer an indication of the position of the transducer within the selected band, the track seeking means being operative to distinguish between the tracks by reference to the said variable physical characteristic of the servo signals read by the transducer entries along or in combination with variation in the servo signals caused by the pattern of servo signals on the record member, means to respond to the indication from the track seeking means to cause the transducer to access the selected track, and track following means operative to derive an indication of the position of the transducer relative to the selected track by reference to variations in the servo signals caused by the pattern of servo signals on the record member.

12. Apparatus for reading a record member as claimed in Claim 2 the apparatus having a movable transducer means to support the record member for movement relative to the transducer, coarse positioning means to position the transducer to access the tracks within a selected band of the tracks on the record member, register means to register the selection of a track within the selected band, track seeking means to derive from the servo signals read by the transducer an indication of the position of the transducer within the selected band, the track seeking means being operative to distinguish between the tracks by reference to the said variable physical characteristic of the servo signals read by the transducer entries along or in combination with variation in the servo signals caused by the pattern of servo signals on the record member, means to respond to the indication from the track seeking means to cause the transducer to access the selected track, and track following means operative to derive an indication of the position of the transducer relative to the selected track by reference to variations in the servo signals caused by the pattern of servo signals on the record member.

13. Apparatus for reading a record member as claimed in Claim 3 the apparatus having a movable transducer means to support the record member for movement relative to the transducer, coarse positioning means to position the transducer to access the tracks within a selected band of the tracks on the record member, register means to register the selection of a track within the selected band, track seeking means to derive from the servo signals read by the transducer an indication of the position of the transducer within the selected band, the track seeking means being operative to distinguish between the tracks by reference to the said variable physical characteristic of the servo signals read by the transducer entries along or in combination with variation in the servo signals caused by the pattern of servo signals on the record member, means to respond to the indication from the track seeking means to cause the transducer to access the selected track, and track following means operative to derive an indication of the position of the transducer relative to the selected track by reference to variations in the xervo signals caused by the pattern of servo signals on the record member.

14. Apparatus for reading a record member as claimed in Claim 8 the apparatus having a movable transducer means to support the record member for movement relative to the transducer, coarse positioning means to position the transducer to access the tracks within a selected band of the tracks on the record member, register means to register the selection of a track within the selected band, track seeking means to derive from the servo signals read by the transducer an indication of the position of the transducer within the selected band, the track seeking means being operative to distinguish between the tracks by reference to the said variable physical characteristic of the servo signals read by the transducer entries along or in combination with variation in the servo signals caused by the pattern of servo signals on the record member, means to respond to the indication from the track seeking means to cause the transducer to access the selected track, and track following means operative to derive an indication of the position of the transducer relative to the selected track by reference to variations in the sero signals caused by the pattern of servo signals on the record member.

15. Apparatus according to claim 11, claim 12 or claim 13 for reading a record member wherein the track seeking means comprises a counter circuit to count the durations of the servo signals read by the trans-ducer.

16. Apparatus according to claim 14 for reading a record member wherein the track seeking means includes a frequency discriminator and comparator responsive to the frequencies of the servo signals read by the trans-ducer.

17. Apparatus according to claim 16 wherein the track following means include integrating means to integrate the servo signals read by the transducer thereby to provide an indication of the position of the transducer relative to the selected track.

18. Apparatus according to claim 17 wherein the integrating means is operative to integrate first and second servo signals from each servo signal area at rates dependent on the track selected by the track seeking means.