CA2008477A1

CA2008477A1 - Method to compensate for tape slope and head azimuth errors

Info

Publication number: CA2008477A1
Application number: CA002008477A
Authority: CA
Inventors: Anatoly Aranovsky; William A. Buchan
Original assignee: Archive Corp
Current assignee: Seagate Technology LLC
Priority date: 1989-09-27
Filing date: 1990-01-24
Publication date: 1991-03-27
Also published as: US5001580A; EP0420374A2; EP0420374A3

Abstract

METHOD TO COMPENSATE FOR TAPE SLOPE AND HEAD AZIMUTH ERRORS
Abstract of the Disclosure A method is provided to compensate for tape slope (A) and read/write head block (102) azimuth (B) errors in a tape drive system (100). The method writes a data pattern (504) on a magnetic medium (104) at a known slope (C). The portion of the magnetic medium (104) that is encoded with the data pattern (504) is then moved across the read/write head block (102), so that first one read head (154) and then the other read head (1523 detects the recorded data pattern (504). The time difference (.DELTA.TON) between the event of each head (154, 152) first sensing the data pattern (504), and the time difference (.DELTA.TOFF) between the points where each head (154, 152) no longer detects the data pattern (504), are both recorded. The recorded information is used to analyze the angular offset (A-B) between the centerline (192) of the tape (104) and the centerline (194) of the read/write head system (102). The angular offset (A-B) is used to determine a lateral displacement (D) between the two read heads (154, 152) in a direction perpendicular to the centerline (192) of the tape (104). The lateral displacement (D) is used to adjust the stepping distance between data tracks recorded in a first (i.e., forward) direction and a second (i.e., reverse) direction on the tape (104) so that the respective read head (152, 154) is positioned substantially on the center of the respective recorded track.

Description

2~

~RCHA.33A PATENT
~T~OD TO ~O~P~NSATE F~R TAP~ SL~P~ AND ~D AZINUT~ EaR~RS
Field of the Tnvention This invention pertains to streaming cartridge tape drives. In particular, the pres~nt invention is directed to a method for correcting write and read errors cau~ed by head block and tape misalignment.
Backqround of the_Invention It is well knQWII that information can be encoded onto a magnetic medium, such as a magn~tic disc or magnetic tape. Such encoding is usually accomplished by generating magnetic flux changes in close proximity to the magnetic medium. The magnetic flux cha~g~s can be gen~rated by a magnetic readJwrite head system. In such a system, electronic signals are convertecl into magnetic flux by induction at a write gap in a write head. Furthermore~
the signals ~rom the magnetic medium are decoded by sensing magnetic flux changes at a read gap in a read h~ad, the magnetic flux changes being produced by the movement of the ~agnetic medium past the reacl gap. The magnetic ~lux changes sensed at the read gap are converted into electronic signals via induction at the read gap.
~ onventional magnetic head systems for tape drives comprise at least one read head and one write head. Often, such systems utilize multiple read and/or wri~e heads mounted upon one head block. In an exe~plary read/write system, there are three ~aynetic heads aligned on a he~d block. This type of head block comprises a ~orward read head, a write head, and a reverse read head. Normally, the write head is positioned between the two read heads3 In the ideal case, the head centers are aligned in a straight line. Ideally, this straight line is al80 the center line along which the magnetic tape moves, so that there i~ no error in the alignment of the head block with respect to the center of the tape.

U~ually, ~ome error i8 introduced in the manufacturing process for both the tape cartridge and th~
head block. This generally unavoidable manufacturing error results in a misalignment of the tape with respect to the edge of the cartridge a~ well as a mi~alignment of the individual heads on the head block. In addition to thi~, the entire head block may be a~k~w by some angle with respect to the base plate of the tape player~recorder.
When reading data in co~Yentional Gtea~ing cartrldge tape drives, it is necessary to alternate between a forward read head and a reverse read head ~ach ti~e the head block i6 moved to an adjacent data track. For this rea~on, it i8 advantayeous to know the displacement error between the forward read head and the rever~e read head, so ~hat ~he read heads may be stepped accordingly as the head block switches to an adjacent data track.
It can be safely assumed for the purpo~eæ of the following discussion that th~ adgle of the tape cartridge i~
aligned with the base plate of the tape plzyer/recorder 80 that the two form an identical reference line. From th~s re~erence line, referred to hereinafter as the B-plane, the angle o~ displacement of the head block and the angle of displacement of the tape c~nter line can be measured.
~elative to the B-plane, the tape itself may have some characteristic slope which can be ~pecified as a ~um~er of minutes of arc, or a number of ~illimetars per inch.
Additionally, the horizontal ~lignment of the magnetic head~ may also have som~ ~lope relative to the B-plane.
~he net effect is that, i~ one head is eentered on a track of the tape, another head ~hich is horizontally displaced from the first head may be ~o~f track" by a ~ignificant distance.
Previously, this problem of head-tape ~isalignment has not been considered a serious deficiency. This i~ because the track~ on conventional 60-100 Megabyte tape~ h~ve b~en wide enough to allow for such error~O However, with the t~7 advent of high density magnetic tapes having s~orage capacities on the order of 300 Megabytes~ track widths have decreased to a point where a small misalign~ent betw~en the head and the tape can lead to int~rferenc~ with information written on adjacent tracks. For ~his reason, problems with head-tape misali~nmen~ mu~t be addre~sed.
Prior methods used to corr~ct for head-tape misalignment have been sketchy, approximate technigue~.
For example, the h~ad block ~ay be ~oved lat~rally acro~
the width of the track via a stepper ~otor until the amplitude of the ~ignal 6tored on th~ tape wa~ at a maximum. The di~tance stepped i8 then recorded.
Therea~ter, each time that head was used on that particular track, the head i stepped by the recorded distance. This method is de~icient for a number of reasons~ First of all, be~ause the information may not be written at the ~ame scaling amplitude over the entire tape, the amplitude may appear to be a maximum at a certain point when it i8 actually just an inconsistenc:y in the recording.
Secondly, since this type of te~st occurs over a small interval of timel many ~uch t~sts would have. to be performed in order to obtain ~ome average stepping increment. Finally, this method is inexact and it may ~e difficult to determins where th~ amplitude of tha sig~al i~
at its peak.
In anothar invention, disclosed in European Patent Application No. 87118762.1 (Publication No. 276,451), the edge o~ the magnetic medium i8 ascert~ined. Following this, the write head is moved laterally acro~s the wid~h of the tape ~ome known distance Yia the stepper motor. The write head then records a data line along the tape~ When the data ~ream is recorded, the Qdge o~ the tap~ i~ again determined, and the read head i~ moYed up fro~ the edge of ~he tape, fir~t to the lower edge o~ the data line, then to the upper edge o~ the data lin~. The upp~r and lower pvsition values of the read head are stored and ~veraged, ~ f~

so that the read head's theoretical position over the center of the data line is detenminedO The differenca between the read head's a~raged center position, and the write head recording position i8 ~tored as the error o~
alignment.
This method also has some inherent disad~antages.
First of all, there is the added complexity of recognizing the edge o~ the tape. 'rape edge recognition is su~ceptible to probleme with t~e consi~tency of the magnetic ~edi~
since, in the manufacturing proce~s, coating the edge of a tape with a magnetic ~aterial in a uniform manner i~ guite dif~icult. Also, it is ofken difficult ts determine khe exact location of the tape edge due to an uncertai~ty in the width of the track recorded by the write h~ad.
Secondly, since the head block mu~t be moved across the tape as measurements are taken, ~echanical complexity i8 introduced in the measuring process. Flnally, in the me~hod disclosed in European Pat~nt Application No.
87118762Al (Publication No. 276,45'1), the read head must be positioned so that the read head i6 neax the written data line~ otherwise the process could use a great deal of tape, as the head ~oves towards ~he t:rack one increment at a timeO
; Summary of the InvPntion A method is provided to compensate for tape 510pe and read/write head block azi~uth errors in a tape dri~e system. The method writes a ~tream of data on a ~agnetic medium at a known ~lope. ~he portion of the magnetic medium that is encoded with the data stream is the~ ~oved across the read/write head blQck, so that first one read head and then the other read head det~cts the recorded data stream. The time dif~erence between when each read head first ~enses the data stre~, and the time difference be~ween when each read head ~o longer detect~ the dat~
stream, are both recorded. This information is used to analyze the angular o~set between the ~enterline o~ the tape and the centerline of the read/write head sy~tem~ Theangular off~t is used to determine a lateral displ~ce~ent between the two read head~ in a direction perpendiculax to the centerline of the tape. The lateral di~placement i~
used to adjust the ~tepping distance between data tracks recorded in a first (i.e., ~orward) direction and a se~ond (i.~., reverse) direction on the tape ~o that the r~æpectiv~ read head is positioned sub~tantially on the ~enter of the respective recorded track.
The ~ethod of the pre~ent inYentlon calibrates a read/write head ~y~tem of a tape drive that ~rites dat~
onto and reads data ~rom a ~agnetic tap~ in a tape cartridge w:ithin the tape drive. The read/write head system comprises a write head that writes data onto the tape in the form of magnetic flux tran~itions. First and second read heads are provided to sense the ~agnetic flux transitions and generate electrical signal output~
responsiv~ thereto. At least one read circuit ~
responsive to the electrical sigmal outputs o~ the f~rst and second read heads to generate electronic data ~ignals.
The first and second read head~ are aligned with respect to each other along a read head cent~rline th~rebetween. The tape ~oves within the cartridge in a dire~tion defined ~y a tape centerline. The calibration method determinea a displacement of the first read head with respect to the seco~d read head in a dir~ction perpendicular to the tape ¢enterline. The method co~prisQ~ ~he step~ o~ writing ~
predetermined data pattern on a portion of ~aid tape ~uch that said data pattern is obligue to ~id tape cent~rl~ne and then positioning the read/write h2ad ~ystem 80 that the ~ir~t and second read heads are positioned ah~d o~
the predet~r~ined data pattern. The ~ethod further comprises the step of ~oving the tape pa~t the ~irst and second read heads whil~ monitoring the electrical signal outputs from the read headc to deter~ine a ti~e dl~fQrenc~
between when the first re~d head ~enses the predetermined 2~

data pattern and whPn the second read head ~en~es the predetermin~d data patternO Ther~after, the method calculates the displacement of the first read head with re~pect to the second read head using the time dif~rence.
Brief Descri~tiQn of the ~rawings ~ igure 1 is a simplifi~d bloc~ diagra~ of an exe~plary e~bodi~ent of a processor controlled tape drive including a ~chematic representation of the tape heads.
, Fi~ure 2 pictorially illu-ctrates an ex~plary tape c~rtridge showing a se~tion o~ tape between the t~o reel6 o~ the cartridge, the tape being sho~n with an exaggerated slope with respect to the ba~e o~ the cartridge.
Figure 3 pictorially illustrates the tape cartridge and tape of Figure 2, and ~urther illustrates a tape h~ad system superi~posed on the tape, the tape head system being shown at an exaggerated angle wit:h respect to the base of the cartridge.
Figure 4 pictorially illustrates a data track written onto a section of tape by the write head, and further 20 illustrates the offset of the read heads from the ~ata track caused by the angular misalignme~t of the read/write head system with re~pect to the tape.
Figure 5 illustrates an exemplary flow chart for the overall ~thod of the present invention.
Figure 6 illustrates a data pattern written onto a tape at a slope such that tha centerline o~ the data pattern is at an angle C wi~h re~pect to ~he c~nterline of the tape.
Figures 7A and 7B depict a data pattern moving past tAe head block as the tape move~.
Figure 8 illustrates an exemplary ~low chart for the write data pattern ~ubroutine of the method of the pre8ent invention.

;~Q~7 Figure 9 illustrate6 an exemplary flow chart of the read data pattern s~broutine of the method of the pr~sent invention showing the alternative paths for controlling the ON TIMER and th~ OFF TIMER~
Figure 10 illustrate~ a ~a~ple data pattern written onto the tape in accordance with the pre~ent invention.
Figures llA and llB illustrate diagrams that show the time relationship for read head pattern detection in two s~parate cases.
Figures 12A, 12B and 12C illustrate the angular and scaler relationships betweQn the read/write head system and the tape, further illustrating the 810pe of the data track written in accordance with the present invention.
Figures 13A and 13B illustrate the e~fect o~ the present invention in adjusting thl~ number of steps between adjacent data tracks when readtng forward and reading reverse in accordance with the calculated lateral displace~ent between the read ~orward head and the read reverse head.
Detailed_Descri~tion ~f the ~e~r~e~ F~odi~en~
Fi~ure 1 illustrates a block diagram of an ex~mpla~y tape drive systam 100 into whirh the prPsent inYentiOn is inco ~ orated~ The tape drive sy tem includes a read/write head system 102 which writes data onto and retrieves data 25 from a magnetic tap~ 104. The tape drive ~ystem 100 ~s connectable to a computer syste~ 110 (~hown in phantom), or the like. For example, the computer system 110 m~y be ~n IB~ PC, AT, PS/2, or ~ny of a number o~ oth~r commercially available co~puter syste~s. The co~put~r ystem 110.g~nerates com~ands to ~he tape drive sy~te~ 100 to initiate operations. For exa~ple, such operatio~s ~ay include ~elect, position, write data, write file mark, read data, read file ~ark, and read ~tatu~ command~. The tape drive system 100 responds to the co~mands by performing the 2~ 7 selected func~ion (e.g~, writing data onto the tape 104 or reading data ~rom the tape 104~ Commands, data and status are communicated ~etween the computer system 110 and the tape drive system 100 via a bus 112. The bus 112 is constructed to conform to on~ of a nu~ber of industry standards (e.g., the SCSI ~Small Computer Sy~tems Interface) standard)~
The tape drive system 100 includes a microproce~sor 120. The microprocessor 120 includes a~sociat~d interface lp circuits (not ~hown~ which allow it to communicate ~ith the comput~r system 110 via the hu~ 112 and further allow it to send data and commands to and rec~ive data and status from other circuîts within the tape drive sy~tem 100.
The tape drive system 100 further includes a ~otor 1~0, a capstan 132, a power ampli~ier 134 and a tachome~er 138. The ~icroprocessor 120 outputs control signals to ~he power amplifier 134 which in turn provides power to the ~otor 130 to drive the cap~tan 132 in a conventional manner. ~he rotation of the capstan 132 causes the tape 104 to move longitudinally across the read/wr~te head system 102. The tachom~ter 138 senses ~he rotational velocity of the capstan 132 and p:rovides a ~eedback signal to ~he microprocessor 120. The ~icroproce~sor ~20 responds ~o the ~eedback signal and Yaries the control signals applied to the power amplifier 134 so that the capstan 132 i~ dxiven at a constant predetermined rotational velocity.
The motor 130 can be operated in either of two rotational directions to move the t~pe in a ~orward direction ~right to l~ft in Figure 1~ or a rev~rse direction (le~t to right in Figure l)o The tape drive ~y6te~ 100 further includes an ampli~ier 142 which drive~ a stepper ~otor 140. The microprocessor 120 output~ signals to the ampl~fier 142 to drive the stepper ~otor 140 in one of two rotational direct~ons. The ~tepp~r motor 140 i~ mechan$cally connected to the read/write head ~y6tem 102 ~ia a l~ad screw 144. When the stepper ~otor 140 is stepped, the r~ad/write head 6ystem 102 is ~oved laterally across the tape 104 ~i~e., in a direction substantially perpendicular to the direction of ~otion of the tape 104). In the tape drive system described herein, the stepper ~otor 140 can be incrementally stepped ~o that the readfwrite head sy~tem 102 move~ laterally across the width of the tape 104 ~n steps of 0.000125 inches.
The read/write head sy~tem 102 comprises an era e bar 1~ 150, a read forward head 152, a read reverse head 154, ~nd a write head 156. As illustrated, the write head 156 is positioned between the read forward head 152 and the read reverse head 154.
The tape drive system 100 operates in a conventional manner to write data onto and reacl data from the tape 104.
A write circuit 160 receives data from the computer system 110 via the microprocessor 120 and its associated circuits, and produce~ drive signals that generate magnetic flux cha~ges on the tape 104. A read circuit 162 sense~ the magnetic flux changes on the ta~pe 104 a~d convert~ the sensed flux changes to data signals that are co~municated to the computer sy6tem 110 via the microprocessor 120 and its associated circuits.
~In the embodiment illustrated in Figure 1, the forward read head 152 and the reverse read head 154 are electrically connec~ed to the read circuit 162 via a multiplexer 170. The multiplexer 170 ~elects which of the read heads 152, 154 is electrically connected to the read circuit 162 at any particular time. Generally, when the tape 104 is drive~ in the forward direction, th~
multlplexer 170 conneots the forward read head 152 to the read circuit 162, and when the tape 104 i~ driven in the reverse direction, the ~ultiplexer 170 connect6 the reverse read head 154 to the read circuit 162. As will be di~cussed below, the two read head~ 152, 154 can be sampled by alternately switching the multipl~xer 170 at a rapid _g_ ~ t7 rate to connect first one and then the other read head to the read circuit 162. As used herein, the forward direction i~ the direction in which the tape 104 i~ moving such that a portion of the tape 104 traverses the write head 156 immediately prior to traversing the forward read head 152. Thus, the data (i.e., flux changes) written onto the tape 10~ can be i~ediately sensed by the ~o~ward read head 152 and converted to data ~y the read circuit 162.
The sensed data can be compared to the written data to determine whether the data was properly written onto th0 tape 104. Conversely, the rever~e direction i~ the direction in which the tape 104 is moving such that a portion of the tape 104 traverses the write head 156 prior to traversing the reverse read head 154 ~o that the data 15 written in the reverse direction can be ~en~ed by the reverse read head 154. As set f orth above, the tapa 104 moves from right to left in the forward direction and from le~t to right in the reverse direction.
In an alternative em~odiment (not ~hown~, separate read circuits may be provided for each of the forward read head 152 and the reverse read head 154 so that it is not necessary to ins::lude the multipl~xer 170. The output~ of the two read circuits can be provided as inputs to the microproce~;~or 12 0 and its associated circuits . Th~ data `fromfone of the read circuits is selectively transmitted to the computer sy~tem 110 via ~he ~us 112.
Control signals ar~ ~ent to the write circuit 160, to the read circuit 162 and to ~he ~ultiplexer 170 from the microproces~or 120 ~o seleatively activate the appropri~te circuits when writing data and reading data.
Figure 2 pictorially illu~trate~ the t~pe 104 within ~
tape cartridge 180. As i~ well known in the art, a conventional tape cartridge 180 for a ~tre~ming tape drive, comprises a base plate 182 onto which a pair of reels 184 35 and 186 are mounted. Th~ tap~ 104 traveræe8 b~twe~n the two reel5 184, lB6. ~hen the tape cartridge 180 i~

'10--7~

inserted into the tape drive ~00, the read/write head system 102 is positioned proximate to the tape 104 between the two reel~ 184, 186~ a~ illustrated pic~orially by the j~xtaposition of the r~ad/write head sy tem 102 o~er the tape 104, in Figure 3. The t~pe 104 is also in contact with the capstan 132 ~not hown in figure 3) which cau~e~
th~ tape 104 to move when th~ ~otor 130 iR activated.
Ideally, th~ tape 104 ~oves longitudinally between ~he two reels 184, 186 in a ~traight line parallel tn ~he ba~e plate 182. The base plate 182 def~ne~ a plane which ~ 8 referred to herein as the B plane 190. A~ illustra~ed in Figures 2 and 3, the portion of the tape 104 ha~ a tape centerline 192~ The tape centerline 192 may be akewed Prom the B-plane 190 by an angle NA~ as depicted. It Rhould be noted that the an~le "A" has been exaggerated in Figures 2, 3, and 4 in order to clearly depict the offset of the tape centerline 192 rom the caxtridge B-plane 190.
As illustrated in Figure 3, th.e xead/write head sy~t~
102 has a centerline 194. The centerline 194 of the read/write head syskem 102 may be mi~aligned by an angle WBW (also exaggerated) from the B-plane 190. The alignment dif~erence be~ween the readJwrite head ~yste~ centerline 194 and the tape centerline 192 i~ shown a~ the angle ~A-B".
If the ~isalignment of ~he tape centerlinP 192 and the misalignment o~ the readJwrite head ~y~te~ 19~ are additive (e.g., NA" positive and ~B0 negativs, a~ ~hown, or ~A" negative and "B" positive~, then the net displace~ent can be ~igni~icant. Figure 4 illustrate. the e~fect of such a net displace~ent. A~ illustrated, a data track 200 i~ written onto the t~pe 104 in parallel wi~h ~he tape centerline -192n Since the read/write h~ad system 102 ls oriented at an angle with reGpect to the tape centerline 192, the read forward head 152 and the read reverse head 35 154 are displaced from the center of the da~a tracX 200.
The net displacement of the read forward head 152 ~rom the '7~7 center of the data track 200 depends upon the siz~ of the angle '~A-B" and the ~pacing b~kw~en the read forward head 152 and the write head 156. (Again, th~ angle i'A-B" and the displacement are highly exagg~rated in the drawing figures. In actual e~bodiment~, ~he angle "A-B" is small, and the two heads will be laterally displaced with respect to each other by less than one track widtho ) Similarly, the net di placement of the read rever~e head 154 ~rom the center of the data track 200 d~pends upon the 8ize of the angle "A-B" and the spacing between the read reverse head 154 and the write head 156. ~he net displacements may be sufficiently great that after the data track 200 is written, the read forward head 152 and the read reverse head 154 will not adeguately sense the flux changes in ~he data track 200. Thus, the read/write head ~y~tem 102 needs to be moved laterally (i.e., ~ tantially perpendic~lar with respect to the tape centerline 192~ in order to ~osition the respective read forward head 152 or read reverse h~ad 15~ proximat~ to t:he data track 200 when reading the data trac~ 200 in the forward or reverse direction.
The present invention is a ~nethod that automatically detects the skew (i.e., angle ~A-B~) between the tape centerline 192 and the read/write haad sy~tem centerline 194. The ~ethod will be described in detail with reference to the flowcharts of Figures 5, 8, and 9 and the pictorial illustrations of Figures 6 and 7.
Figur~ 5 is an exemplary flow chart of the overall ~ethod of a preferred embodiment for calculating the align~ent error n~ ln accordance with the invention.
Initially, the beginning of ~he tap~ 104 i~ found in a process block 500. The detsction o~ ~he beginning o~ the tape 104 is accomplished in a conv~ntional mannar. After finding the beginning of the tape 104, the method enters a write pattern ~ubroutine block 502 wherein a predetermined data pattern is written onto the tape 104 at an angle ~C"

relative to the edge of the tape 104. The data pattern ~ust be written oblique (i.e., neither parallel nor pexpendicular) to the centerline of the tap2 104. The data pattern i~ repre~ented in F~gure 6 by a line 504 on a section of the tape 104. The written data pattern 504 i8 readily detectable by the read circuit 162. For exa~ple, the data pattern 504 may advantageou~ly compri~e a conventional gap signal (alternating one~ and zeros at the higheæt writing fre~uency). The write subroutine 502 will lp be described in greater detail below ~n connection wi~h ~igure 8.
Returning to Figure 5, following the write ubrout~ne502, the ~ethod enter~ a process block 510 whsrein the tape 104 is rewound ~o that the read/write head ~ystem 102 i~
positioned ahead of the written pattern 504. After the tape 104 has been rewound, the method enters a read pattern ~ubroutine 512 wherein the writt:en data pattern 504 i~
detected by the read forward he~d 152 a~d the read r~verse head 1~4. This is illustrated p:ictorially in Figures 7A
and 7B. Fiqure 7A shows that a~ t:he tape 104 mov~s in the forward direction (i.e., from right to left), the writ~en data pattern 504 i5 fir~t detected by the read reverse head 154. Thereafter, a~ the tape 104 move~ further to the left, as illustrated in Figure 7B, the written data pattern 504 is detected by the read forward head 152. Thera will be a difference in the times when the two read head~ 152, 154 detcct the written data pattern 504. The ti~
difference depends upon the distance between the two head~
~shown a~ ~2dN, where l'dN is the distance Pro~ the ~enter of the write head 156 to the center o~ ei~her of the two re~d head~ lS2, 154), the velocity o~ the tape ~shown ~s ~V~ in Figures 7A and 7B), the angle ~C~, ~nd the angle ~A~B~o Th~ times at which the read heads detect the data pattern are recorded. The read ~ubxoutine 512 i~ describ~d more fully in connection with Figure 9.

~g~

A~ discussed above in connection with Figure 1, the multiplexer 170 allows both read h~ad heads 152, 154 to be sampled at substantially the same time. The sampling rate iæ selected to be sufficiently high te.g., every 25 to 50 ~icroseconds) so that the sampling error is insigni~icant compared to the mea~ured detection times. Alternatively, as di cussed above, a separate read circuit can ~e provided ~or each head so that the read head~ 152, 154 are monitor~d continuously.
lp Following the read subroutine 512, the method enters a proce~ block 520. In the proce8s block 520, a di~tance d~
iG calculated using the time ~easure~ents obtained in the read subroutine 512. As illu~trated in Figures 7A and 7~, the distance dL is the distance that the tape moves during the time between the detection of the written data pattern 504 by the reverse read h~ad 154 and the detection of the written data pattern 504 by the forward read head 152.
After calculating the di~tance dL in the process block 520, the method enters a proces6 block 530 wherein the angle 'IA-BI' between the read/write haad sy~tem centerline 194 and th~ tape c2nterline 192 i~ calculated. In addition, the lateral displacement ~etween the read forward head 152 and the write head 156 is calculated. As discu~sed a~ove, the latexal displacement is the distance from the center of the written data track 504 to the center of r~ad forward head 152. Similarly, the lateral di~placement of the read reverse head 154 from write head ~6 is also calcul~ted. Therea~ter, in a process block 532, the calculat~d displacements are ~tored. A will be discu sed ~elow, ~he cal~ulated displa¢ements are later referenced in order to correct ~he azimuth ~rror between the tape centerline 192 and the read~write he~d ~y~t~
centerline 194.
Figure 8 illustrates the details of the write 35 ~ubroutine 502 of Figur~ 5, wh~rein a mei:hod i~ outlin~d for writing a data pattern at an angle "C" relati~e to the '7 edge of the tape 104. In a first process block 802, the read/write head system 102 is positioned approximately in the center o the tape 104 by applying appropriate signals to the stepper motor 140. Next, in a prores~ block 804, a step increment, ~5~ and a total count are set. The step increment ~3 ~1 iS usually the smallest increment size available. The total count determine~ the number of ~tep~
to be taken across the tape, and iB stored in an internal step counter in the microprocessor 120 (e.g., a register or ~ ~emory location). For axa~ple, when th~ pre~ent invention i5 implemerl~ed in an exemplary ~;tre2mirlg tape drive to implement the QIC-1350 standard, the ~tep increment "q" is equal to 0.000125 inche~.
A~ter the step increment is set, the writlng of the data pattern begins. Since the lateral portion o~ the read/write system 102 is controlled by the stepper motor 140, a straight line at the angle ~C" cannot be generated.
Instead, an approximation of a straight line i8 generated by writing the pattern in a cycle wherein the methGd alternates between writing data, and stepping the read/write head system 102 up by one incre~ent. The ~etho~
enters a process block 806 wherein a data pattern 1~
written in a straight line relative to the edge of the tape 104. Thi~ data lin~ is written onto the tape 104 for ~X~
milllseconds (this wrîte ti~e ~ay be on the order of 1 ~illisecond). The length of the written data line ~or each step can be calculated as "X~ (the write time) multiplied by "V~', the tape velocity. Following ~he writ~ng of the data lin~, the method ent r~ a proceæs block 808 wher~in the read/write head system lQ2 i~ advance~ ona ~tep increment by the tepper ~otor 140. In a proces block 810, the step counter is decrem~nted. As ~hown in a decision block 812, if the ~tep counter i~ not equal to zero, then the method returns to the pxocess block 806 and 3S a straight data line is written again ~or X milli~econd~ at the new lateral location on the tape 104. Thereafter, the f~7 method again enters the process block 808 and the read/write head system 102 is advanced one st~p ~p.
Following the advancement of the read/writ~ head block system 102, the cycle enters the process block 810 and the step counter is again decr2mented. Thereafter the method enters decision block 812 and th~ test for ero in the step counter is repeated. This cycle continue~ until the step counter raad~ zero.
The total distance laterally ~tepped acro~s the width of the tap~ 104 during ~he write cycle i~ equ~l to the total number of counts, mult~plied by the step increment "s~. This distance should be greater than the lateral displacement (a~ measured perpendicular to the tape centerline 192) between the reverse read h~ad 154 and the ~orward read head 152. In this way, when the read/write head syste~ 102 is moved to the nominal position ~approximately in the center of the tape 104) both read heads may pass below the upper portion of the written data pattern 504, and above the lower portion o~ the written data pattern 504.
The re~ulting written data pattern has an appearance similar to the sample data pattern illustrated in Fi~ure lO. Ag illustrated, the data pattern comprises a plurality of small steps. Th~ step increment, ~s", is selected to b~
sufficiently small so that the overall data pattern appears to be a continuous line with a slope "C" relative to ~he edge of the tape 104. Thi~ angl~ "C~ can be calaulated fro~ known values a~:
C = Arctan~ xY~] (l) 3a where "s" is the stepping increment of the stepper motor 140, "Xl~ is the write time for each line of data, and is the tape 104 velocity.
once the step counter reads zero, the ~ethod branches from the decision block 812 to a proces~ block 814 wherein the head is moved down v~rtically one-half the total number of counts. This is done in order to position the read 7~

heads lS2, ~54 in the nominal posi~ion approximately in the middle o~ the data track so that th~ read heads 152, 154 do not miss the written data pattern 504 by passing either above or below the written data.
Following the write pattern subroutine 502, the tape 104 is rewound in a conventional ~anner. The read heads 152, 154 are positioned on a blank portion of the tape 104 ahead of the written data pattern~ 80 that as the tape 104 begins to strea~ (i.e., from right to left as depicted in 10 Figures 7A and 7B) the written data pattern 504 will traverse acros~ the read/write head sy~tem 102. In ord~r to obtain ~he best error measurament result~, the incline of "C" should be shallow (on the order of 20 minutes of arc). This will allow tha r~ad heads 152, 154 to ~ove across the written data pattern 504 gradually during the read ~ubroutine 512.
Figure 9 shows the details of the read pattern subroutine 512 of Figura 5, After the read pattern subroutine 512 is initiated, the method enters a proce~
20 block 900 wherein a signal i5 sent to begin ~he tape 104 ~ovement. The read~write head system 102 i8 initially po~itioned in the write ~ubroutine 502 cO that both read heads 152, 1S4 are assured of pas6ing through the written data p~ttern 504 as the tape 104 moves. T~e method next ~ente~s a declsion block 902, where a test is performed to determine if the written data pattern 504 is detected on ~he rever e read head 154. once the written data pattern 504 is detected on the rever~e read head 154, the method branche~ from the deci~ion block 902 to a process block 904, wherein a timer (ON TI~ER) i8 ~tart~d. ~he ON TI~ER
and an OFF TIMER to be de~cribed below are pre~er~bly implemented as internal counter~ ~e.g., regi~ter~ or memory locations) of the microprocessor 120.
After the ON TIMER is started in the process block 35 904, th~ method enter~ a dual deci~ion loop 907. This dual decision loop 907 comprise~ a decision block 906 and a decision block 908. Th~ function o~ the dual decisio~ loop907 is to determine which method to implement ~or a given sequence of data pattern det~ction. The written data pattern 504 may be detected in two different seguences, illustrated in Figures llA and llB. Figure llA depicts a case where the written pattern 504 i sensed at the ~orward read head 152 while the written pattern 504 continues to be sensed at the reverse read head 154. Figure llB depict~ a case wher~ the written data patten 504 ceases to be lQ detected at the reverse read head ~5~ before it begin~ to be ~ensed at the forward read head 152. Becau~e ~he written data patt~rn 504 may be detected in either o~ the two sequences, the ~ethod of reading the written data pattern 504 ie divided into two branches described below.
Initially, the method enters the decision block 906.
If the data pattern is detected on the forward read head 152 ~irst~ then the method branche~ a proce~s block 910, wherein the ON TIMER is stopped. If the data pattern i5 not detected on the ~orward read head 152 îirst, then the method continues to the decision block 908. In the decision block 908, a te~t 1~ performed to see if the written data pattern 504 ha~ pa~ed (e.g., i~ no longer detected at) the reverse read head 154. If not, then the method branches back to the decision block 906. This cycle `repea~ts itself until one of the test conditions i~
satisfied.
For the sake of illu~trating the flow chart in an orderly manner, it shall be a~u~ed ~hat the writt~n data pattern 504 i~ detected at the forward read head 152 before the written data pattern 504 ha~ passed the reverse r~ad head 154 (a~ depicted in Figure llA). The other ca~e where the written data pattern 504 ha~ pas~ed the reveree read head 154 before the written data pattern 504 îs detected at the ~orward read head 152 will be di cussed later in connection With Figure llB~

As illustrated in Figur~ 9, if the written data pattern 504 is detect~d at the forward read head 152 while continuing to be detected at the reverse read head 154, the method exits the decision block 906 and branche~ to the process block 910 wherein th~ ON TIMER is stopped. The ti~e corresponding to th~ count in the ON TIMER i6 ~tored a6 ~ToN. The method nex* enters ~ deci~ion block 912 where ik waits until the written data pattern 504 has passed the reverse read head 154~ After the pattern i~ no longer 0 sensed at the reverse read head 154, the method enters a proce~ block 920 wherein the OFF TIN2R is started. The m~thod now enters a deci~ion block 922 where a test is performed to determine if the written data pattern 504 i~
still detected at the forward read head 152. When the written data pattern 504 moves past the forward read head 152, the decision block 922 is exited to a proces~ block 924, where the OFF TIMER is ~topped. The new time corresponding to the csunt in thl3 OFF TIMER is stored a~
~TOFF -Returning to the second case, as illustrated in Figure llB a situation could arise where the re~erse read head 154 is ~off" (the written data pattern 504 is not detected at the reverse read head 154) before the forward read head 152 goes "on" (the written data p~ttern 504 is d~tect~d ~t ~5 `~he forwaxd read head 152). ~his situation may occur, Por example, if the written data pattern 504 i8 writt~n at a steep sl~pe (e.g., C is large relative to A or B).
I~ the written data pattern 504 iB ~off" at the rever~e read head 154 before the wri~ten data pattexn 50~
3 0 ie "on" at the forward read head 152, then the method proceeds to a process block 934 wherein a ~econd timer, the OFF TIMER, is initiated. After thi8, a te8t occurs in a decision loop 936 to determine if the written data pattern 504 is detected at the forward read head 152. When the written data pattern 504 is detected at the forward read head 152~ the ~ethod e~ters a process block 938 wherein the ~19 -t~

ON TIMER is stopped, and the ON TIMER reading is stored a~
~ToN. Thereafter, the method enters the decision klock 922 where i~ waits until the written data pattern 504 iæ no longer sensed at the foxward read head 152, at which ~ime the OF~ TIMER is stopped in the process block 924. The ti~e corresponding to the count in the OFF TI~ER i stored a~ ~TOFF-Although described above in connection with theparticular em~odim nt of the write pattern ~ubroutine 502 and read pattern subroutine 512, ~t s~ould be understood that other subroutines can be used to ~easure the propagation time between the detection of a data pattern by the read reverse head 154 and the raad forwarcl head 152.
With the measured values of ToN and ~OFF~ and the known values of the tape ~elocity "V" and slope l'C" of the written data pattern 504, the error of alignment between the read/write head block centerline 194 and the tape centerline 192 can be calculated.
The values of ~ToN and ~OFF can be used to calculate the distance dL. As illustrated in Figures 7A and 7B, the distance dL is the distance the tape 104 ~oves between the reverse read head 154 detection t:ime and the forward read head 152 detection time. As illustrated in Figure 7A, the center of the written data pattern 504 crosses the rever~e `read Jhead lS4 at a point "P". At a later time, a~
illustrated in Figure 7B, the center of the written data pattern 504 crosses the forward read head 152. If ~he read/write head system 102 i~ misaligned wlth re~pect to th~ tape centerline 192, the forward read head 152 will cross the center of the written d~ta pattern 504 at 80~e point other than point "P"O For ~hi8 rea~on, t~e point wpw will have traveled a distance d~ which may not be egual to the distance 2d between the read head~ 1~2, 154. The angles in Figures 7A and 7B have been exaggexated in order to more clearly depict the difference between the distance dL and the distance 2d.

In Figures llA and llB more clearly illustrate the significance of ~ToN and ~ToFF. In order to accurately calculate the distance d~, the center crossing points for ~ach head are determined in the patterned embodiment of the S present invention. Since the written data pattern 504 has a certain width, there is a variance introduced when determining the points where each read head "crosses" the written data pattern 504. For example, if the croasing point is determined to b~ at the left ~dge of the writt~n data pattern 504 ~or th forward read head 152, and at th~
right edge of the written data pattarn 504 ~or the re~erse read head 154, the distance dL, would be greater than a case where the crossing points were both determined to be at the left edge o~ the written data pattern 504 for aach read head 152, 154. In order to avoid this error, a time averaging is performed to determin,e the point in time where each read gap passes across ~he cxnter of the written data pattern 504.
In Figures llA ~nd llB, two time bars are depicted, one above the other. The top time bar in each figure represents the time period that the reverse read head 154 passe~ through the data patternO TRoN is the ti~e at which a s~lected threshold amplitude is exceeded at the output of the reverse read head 154 and the written data pattern 504 is deitected, and TRoF is the time at which the amplitude of the written data pattern 504 is no longer det~cted above the threshold level at the reverse read head 154. The botto~ time bar in each ~iqure r~presents the time period over which the ~orward read head 152 passes through the written data pattern 504 ~ON and TFoFF
time the written data pattern 504 is detected at the forw2rd read head 152, and the ti~e the writt~n data pattern 504 is no longer detected at the forward read head 152 respectively. Each time bar depicted has a center mark which represents the ti~e, halfway between the ToN and ToFF
ti~es, that the center of the respective read head 152, 154 7~

passes acro~s the ~::enter of th~ written data pattern 50~.
Figure 11~ depicts the ~irst ca e where the time bars overlap, while Figure llB depicts the second case where the time bars do not overlap.
A~; depicted in Figure llA, the time difference dT
between the two c::enter points of the read heads 152, 154 c:an be calculated as:

FON FoFF)/23 ~ [ (TRoN ~ TRO )/2~ t2) Equation (2) can be rearranged algebraic~lly to read:
d~ = [ (TF - T~oN)/2] + [ (TFoFF RoFF
It can be seen that TF ~T}~O is equal to ~ToN, and TFoFF TROFF is equal to ~ToFF . Therefore Equation ( 3 ) reduces to:
dT = AToN/2 + ~TOFF/2 (4) 15 In other words, the time differenc~ between the points where the read heads 152, 154 cro~s the cent~r of the written data pattern 504, i8 calculat:ed as the average of ON and ~ToFF which were measured as described above.
In Figure llB, the ~ame equations apply, however, the 20 different ~equ~nce of events necessi1:ate~; a modifi~d route through the read subroutine 512 flowchart (Figure 9~. In the evellt that the reverse r~ad h~ad 154 passes completely across the written data pattern 504 before the forward read head 152 initially detects the written data pattern 504, 2 5 the method mus~ assure that the OFF TIMER i~ itiated promptly, rather than waiting until the ON TIMER stops.
Now that the time difIerence dT ha~ been deter~ined, the distance dL can be comput~3d a~ dTxV, where "V" is the tape 104 velocity. Using the calculated distance dL and 30 the known distance betwe~n the read heads 2d, a phy~;ical diagram can be drawn to determine the amgle of displacement A-B, between the tape centerline 192 and the read/write head centerline 194.

Figures l~A and 12~ diagram the geometry învolved in the calculation of the angle A-Bo In order to ~acilitate calculation oP A-B, some assu~ptions should be made. Since the angles under consideration are small, it can be assu~ed that the tangents of each of the angles A, B, and C are equal to the ~ines of their respective angles. The sines in turn are equal in magnitude t~ their respective angle~
(e.g~, tanA = ~inA = A; tanB = ~inB - B; and tanC - sinC = C).
Figure 12A illustra~es the differ~n~ angle~ involYed in the calculation of A-B. In Fiyure~ 12A and 12B, the blocks R~ and RR represent the ~orward and reverse read heads 152, 154, separated by a distance 2d. The read/write head centerline 194 passes through RF and RR, and i~ Qhown to have the angle "B'~ relative to the B-plane. The tape centerline 192, labelled as "TAPE SLOPE", is offset by the angle "A" relative to the B-plan~3. Finally, the written data pattern 504 is represented in Figure 12A by a line of~set from the tape centerline 192 by the angle "C"~
The same angles can be repos.itioned aæ illu~trated in Figure 12B to construct ~ triangl~3 having a first ~ide 950 of length dL and a ~econd side 952 of length 2d. As diagramed, the ~irst side 950 lies along the tape centerline 192, and the second ~ide 952 lies along the 2~ readfwrite head centerline 194. The angle between the ~irst side 950 and the ~econd side 952 i8 the angle ~A-B", as shownO A third side 954, opposite the angle "A-B~, corresponds to tbe written data pattern 504 and is ori~nted ~t the angle ~C~ with respect to the tape centerline 192 and thus with respect to the fir~t side 950.
The triangle depicted in Figure 12B can be analyzed to deter~ine the angle of ~isalignment betwe~n the ~enterl~ne o~ the tape 104, a~d the centerline of the read/write head block sy~tem 102.
U~ing the ~n~wn value~ for ~he angle ~C" and ~he sides dL and 2d, the law of sines can be used to calculate the unknown angle 7~A-B". In accordance with the trigonometric law of sine~, the relationship between the two known sides and the oppo~ite angles i8:
2d/sin(C) = dL~sin(180-~A-B~-C) = dL/sin((A-B)+C3 (53 Equation (53 can be rearranged as.
~in((A-B3~C)/d~ c ~in~C)/2d (6, As set ~orth above, for the small angles o~ offset e~pected herein, the sine o~ an angle i~ ~ubstantially ~qual to ~he . angle itself~ Thus, Equation (6) can be r~duced to:
~A-B+C)/dh = C/2d (7) Finally, by rearranging Equation (7), the angle between the tape c~nterline 1~2 and the read/write head centerline 194 is calculated to be:
A-B = CtdL-2d)/2d (8) A~ter the angle o~ displacement, A-B, is calculated ~n E~uation (8), the distances ~o be stepped for the different cases of reading and writing can ~hen be obtained. As illustrated in Figure 12C, a lat~ral distance 'IDI' between the two read heads 152, 154 in a direction perpendicular to the tape centerline 192 is equal to 2d times the sine of the angle "A-B". Since the sine of the angle "A-B'~ for the smal~ angle is su~stantially equal to the angle, the lateral di~tance "Dl' can be found by ~ultiplying A-B by the di~tance 2d between the two read head~ 152, 154. On~
will appreciate that for ~he s~all ~ngle~ involved, the distance "D" is substantially e~ual to the di~tance required to ~tep the re~d~write head sy~tem 102 fro~ a location where the read ~orward head 152 i8 centered on a data track on the tape 104 to a location where th~ r¢ad reverse head 154 i~ centered on ~hat ~ame data track. For example, in Figure 12C, a centerline 192' i8 the center of a data track on which the read forward head i8 originally centered. When the read/write head ~y~te~ 102 is ~tepped ~y the distance ~D", the read reverse head 154 will be -2~-sub~tantially centered on the same track (i.e., approximately on the cent~rline 192').
In a conventional streaming cartridge tape drive, it is advantageous to know the lateral displacement between the forward read head 152 and the rRverse read h~ad 154.
once the later~l displace~ent between the ~orward rPad head 152 and the rever~e r~ad head 154 is known, the readJwrite head sy~tem 102 can be ~tepped to the next data trac~ by the appropriate value.
~n example of th~ ~tepping proces~ i~ illu~trated in Fi~ures 13A and 13B. A6 depicted in Fiqure 13A, ~he distance be~ween adjacent written data tracks is t~enty steps since the read/write head system 102 is stepped by ~qual amounts between adjacent tracks. ~he lateral displacement between the forward read head 152 and the reverse read head 154 is two step& (a8 determ~ned by dividing the calculated distance "D" by the known stepping distance N5~ ) . Figure 13B depict~ the read/write head system 102 as it reads data alternately in ~orward and reverse mode on adjacent track~. ~8 further illustrated in Figure 13B, the total incrament E;tepped from reading with the ~orward read head 152 to reading with the reverse read head 154 is 22 steps, while t~e total increment stepped from reading with the reverse read head 154 to reading with the ~orward read head 152 is 18 8tep8. In this w~y, the rea~Jwrite head sy6tem 102 can be po~itioned accurately onto each data ~rack when data i~ being read. ~he in~ormation calculated with respect to the lateral displacement o the two read head~ 152, 154 can be saved ~or a particular tape drive (~or exa~ple, in non-volatile read/write RAM). Th~re~fter, whenever a tape i8 read, the tape drive can b~ adjusted to provide ~axi~um playback amplitude for ~he read forward head 152~ for example.
Therea~ter, the read/write head system 102 can be stepped from track to track making the stepping ad~ustment des ribed above when stepping from a forward written data track to a reverse written data track. I~ is not necessary to make separate adjustments ~or the forward and rever~e data tracks.
The abov~ described embodiment has be~n illustrated with a read/write head sy~te~ 102 wherein the write head 156 is po~itioned midway between the forward read head 152 and the reverse read head 154. This is particularly u~eful for reading d~ta immediately after writing the data~ It æhould b~ understood that the pra~ent invention can b2 used lp ~or other r~ad~write head ~yste~s wherein the write head i~
not positioned on the 6a~ centerline be~ween th~ two read heads. Ho~ever, in the e~bodiment herein where the write head 156 lies midway between and in alignment with th~ two read heads 152, 154, the lateral offset between the ~rite 15 h~ad 156 and o~e o~ ~he read heads 152 or 154 ~an be calculated as being one-half the d.istance D.
The present invention can al~;o be used in combination with a tape drive which has a read,/write head ~y~tem having two heads, at least one o~ which i~ ut~lized for both 23 writing data onto a tap~ as well as reading data fro~ the tape. It can be readily understo~d that once the data patterD is written onto the tape by a head op~rating a~ a write head, the lateral di~placement between the two heads can ~e found ~y operating both heads as read heads in accordance with the abo~e-describ~d method.

Claims

1. A method for calibrating a read/write head system of a tape drive that writes data onto and reads data from a magnetic tape in a tape cartridge within said tape drive, wherein said read/write head system comprises a write head that writes data onto said tape in the form of magnetic flux transitions, first and second read heads that sense said magnetic flux transitions and generate electrical signal outputs responsive thereto, and at least one read circuit responsive to said electrical signal outputs of said first and second read heads to generate electronic data signals, said first and second read heads aligned with respect to each other along a read head centerline therebetween, and said tape moving within said cartridge in a direction defined by a tape centerline, said method comprising the steps of:
writing a predetermined data pattern on a portion of said tape such that said data pattern is oblique to said tape centerline;
positioning said read/write head system so that said first and second read heads are positioned ahead of said predetermined data pattern;
moving said tape longitudinally past said first and second read heads while monitoring the electrical signal outputs from said read heads;
determining a time difference between when said first read head senses said predetermined data pattern and when said second read head senses said predetermined data pattern: and calculating a displacement of said first read head with respect to said second read head in a direction substantially perpendicular to said tape centerline using said time difference.

2. The method as defined in Claim 1, wherein said step of writing said predetermined data pattern comprises the steps of:
writing data in a straight line with respect to said tape centerline for a predetermined period of time;
moving said read/write head system by a known distance in a first direction substantially perpendicular to said tape centerline; and repeating said writing step alternately with said moving step for a predetermined number of steps, such that said predetermined data pattern is written at an effective angle with respect to said tape centerline.

3. The method as defined in Claim 1, wherein said step of determining said time difference comprises the steps of:
measuring a first elapsed time between when first read head first senses said predetermined data pattern and when said second read head first senses said predetermined data pattern;
measuring a second elapsed time between when said first read head first ceases sensing said predetermined data pattern and when said second read head first ceases sensing said predetermined data pattern; and averaging said first elapsed time and said second elapsed time to determine said time difference.

4. The method as defined in Claim 3, wherein said step of determining said time difference comprises the steps of:
sensing a first edge of said predetermined data pattern at said first read head at which time a first timer is initiated;
sensing a second edge of said predetermined data pattern at said first read head at which time a second timer is initiated;

sensing said first edge of said predetermined data pattern at said second read head at which time said first timer is stopped with a first value stored therein representing the first elapsed time;
sensing said second edge of said predetermined data pattern at said second read head at which time said second timer is stopped with a second value stored therein representing the second elapsed time;
and calculating the average value of said first and second values to determine said time difference.

5. The method as defined in Claim 2, wherein said positioning of said read/write head system ahead of said predetermined data pattern includes the step of moving said read/write head system in a second direction opposite said first direction by approximately half said predetermined number of steps.

6. A method of determining the lateral displacement of first and second read heads in a tape drive wherein a tape is moved longitudinally past said first and second read heads, wherein said first and second read heads have a known longitudinal displacement therebetween, and wherein said lateral displacement is in a direction substantially perpendicular to the longitudinal movement of said tape, said method comprising the steps of:
writing a predetermined data pattern on said tape at a selected oblique angle with respect to said longitudinal movement of said tape;
positioning said tape so that said predetermined data pattern is ahead of said first and second read heads;
moving aid tape longitudinally past said first and second read heads until both read heads have sensed said predetermined data pattern;
measuring a time difference between when said first read head senses said predetermined data pattern and when said second read head senses said predetermined data pattern; and calculating the lateral displacement of said first read head with respect to said second read head using said measured time difference, said known longitudinal displacement, and said selected angle of said predetermined data pattern.