Note: Descriptions are shown in the official language in which they were submitted.
19 BACKGROUND OF THE INVENTION
Field of the Invention
21 This invention relates to apparatus using elements
22 which interact with one another, and more particularly to .
23 apparatus using lattice arrays of interactive elements where
24 the positions of the elements within the lattice are sub- :
stantially de*ermined by the interactive forces existing be- .
26 tween the elements.
YO972-063 -1-
'
.. .
.~
` q '
1~4q~3~;8
1 Descriytion of the Prior Art
2 In the past, various systems have been described
3 using elements which have the capability of interacting with
4 one another. For instance, cylindrical magnetic bubble domains
have stray ~agnetic fields which cause interactive forces to
6 exist between bubble domains which are sufficiently close that
7 the stray magnetic fields of each couple to one another.
8 U.S. Patent, 3,689,902 and 3,701,125 describe magnetic bubble
9 domain systems in which the functions of memory, storage, decoding, -~
writing, and reading are described. These prior art systems are
11 usually designed so that interactions between bubble domains are
12 - minimized. Since interactions can lead to adverse deflection
13 of the bubbles, such a design has always been considered
14 advantageous. Also, these prior systems used defined structures
for determining the path of all bubble domains in the system.
16 GeneraIly, these prior systems are characterized by information -~
17 storage in the form of the presence and absènce of magnetic
18 bubble domains.
19 ~ecentl~, there has been some work in magnetic -
bubble dom&in technology using information which is coded by
21 other than the presence and absence of màgnetic bubble domains.
22 For instance, U.S. application, Serial No. 319,130, filed Dec.~ 29/72, now
23 ~.S. patent 3,911,411 describes a magnetic bubblè domain apparatus
24 in whlch different size magnetic bubble domalns provide the ;
various information states. In this manner, all bit positions -
26 of the system can be filled, the size of the domain at each
27~ position determining the information state of that position.
~2~ In that copending application, functions such as writing,
29 storage, and reading of the information states are described.
Yos7a-o63 -2~
.
::: : .
.. .. . , ~, . .. . . i :. . . - - . . - .
3~
1 Another apparatus using magnetic bubble dom~ins which
2 have different properties is described in U.S. application, Serial No.
3 375,285, filed June 29/73 no~J U.S. patent 3,399,779. This copending application
4 utilizes difrerent vertical Bloch line configurations for
domain wall magnetization in order to code magnetic bubble
6 domains in accordance with their properties in a magnetic -
7 field tending to collapse the domains. It has been discovered ;
8 that the field at which a bubble domain collapses is a -
9 function of ~he number of vertical Bloch lines in its domain ~ .
10 wall; therefore, various logic states can be provided by domains : .
11 which have differing number~ of vertical Bloch lines in the
12 domain walls. Of course, this leads to levels of logic higher ~ .
.13 than merely binary levels.
14 Still another bubble domain apparatus using bubble
domains having different properties is shown in.U.S. application, Serial ~ :
16 No. 375,289, filed June 29, l973 now U.S. patent 3,890,605. This copending
17 application codes the magnetic bubble domains in terms of ~ :.
18 their properties of movement in a gradient magnetic field .
19 .normal to thç plane of the medium in which they exist. Depending :~
upon the~angl~ through which these domains are deflected,
21 various information levels can be provided. .
22 . A still further technique for coding magnetic :
23 bubble domains is descrlbed in the I~M Technical Disclosure~ ~ :
24 Bulletin, Vol. 13, No. 10, March 1971, at page 3021. In
this publication, G. R. Henry describes coding in terms of
26 the chirality.of wall magnetization of bubble domains. A
27 technique for reading different chiral states uses a reference
28 domain into which an unknown domain is forced, leading to a . .
YO972-063 -3-
' :,
~'' ' ', .
~ - . ,
... . .
3~
1 collision w~ich determines the chiral state of the unknown
2 domain. ~ -
3 ~hile the prior art has addressed various aspects
4 of magnetic bubble domain technology, and information storage
in general, very little emphasis has been placed on the provision
6 of systems which would have ultimate density configurations,
7 be very stable, and be as structureless as possible. Thus,
8 the prior art has attempted to o~tain high densities by,
9 for instance, using smaller and smaller magnetic bubble domains, .
and by reducing the line width of structures used to move
11 these domains (for instance, electron beam technology has been . :
12 used to make smaller T and I permalloy overlay bars). However,
13 the prior art has not attempted to divert from established
14 procedures in an attempt to find new approaches which may lead --
to significant improvements in system performance and in packing
16 density.
17 The present invention is directed to an entirely new
18 approach for providing apparatus which has a high degree of ;~
19 stability, significantly increased storage densities, and a
minimum of structural requiréments. In the present invention,
21 a lattice of interactive elements is utilized where the position~
22 Of the interactive elements with respect to one another are
23 largely determined by the force~ existing between the elements,
24 rather than by the locations of structures used to move the .
interactive elements. This leads to extremely high density,
26 which can be varied easily, and to structurele~s arrays of
27 elements having great internal stability. The interactive
.. . . . .
28 elements are any elements which can have positions determined .~
Yo972-063 -4- . . ~ :
, ,: .
:, : :
:..... ... :
, ~
L~3~i8
1 by forces existing between them, and are particularly exemplified
2 by magnetic elements such as cylindrical bubble domains. ~ -
3 Various means are used to manipulate the interactive elements
4 into and out of the lattice as well as within the lattice array.
If desired, information can be coded in the elements within
6 the array, thus providing an extremely high density, structure-
7 less storage having internal stability over a wide range of
8 operating conditionS.
9 The existence of arrays of interactive magnetic
bubble domains has been shown in the prior art. For instance,
11 the following technical publications describe some of the
12 physical chaxacteristics of bubble domain lattice arrays.
13 1. S. H. Charap et al, "Behavior of Circular
14 Domains in GdIG", IEEE Transactions on Magnetics,
Vol. Mag-5, No. 3, September 1969, page 566. ;~
16 2. J. A. Cape et al, "Magnetic Bubble Domain
17 Intsractions", Solid State Communications, Vol. 8,
18 page~ 1303-1306, 1970.
19 3. W. F. Druyvesteyn et al, "Calculations on
Some Periodic Magnetic Domain Structures; Consequences
21 for Bubble Devices", Philips Research Reports, Vol.
22 26, No. 1, pages 11-28, February 1971.
23 4. J. W. F. Dorleijn et al, "Repulsive Inter-
24 actions Between Magnetic Bubbles: Consequences for
Bubble Devices", IEEE Transactions on Magnetics,
26 Vol. Mag-7, No. 3, page 355, September 1971. -~
27 5. F. A. DeJonge et al, "Bubble Lattices",
28 American Institute of Physics Proceedings of 17th `~
"'' "
YOg72-063 -5-
,. ,:
... . ., , , ., . . ., , .. , ~ . . . . , , . . " .
43~i8
:
1 Annual Conference on Magnetism and Magnetic Materials,
2 Chicago, Illinois, 1971, Section 4, page 130.
3 Even though others have studied the various theories
and physical properties of lattices containing magneiic bubble
S domains, no one has heretofore thought of utilizing such lattices
6 in practical systems. Thus, these articles contain no suggestion
7 or statement directed to a usable system incorporating the
8 many features which can be present in lattice arrays of interactive
9 elements. Despite the known existence of lattices of various
elements, it remained for the present inventors to recognize
11 that many features can be obtained by the use of lattice
12 arrays to provide apparatus and systems having numerous
13 - advantages over thdse found in the prior art. Rather than utilize
14 known directions for providing storage and memory systems, the
present inventors have taken a fresh approach and have
16 obtained systems which are significantly improved over those ~`
17 of the prior art. ~, -
18 Accordingly, it is a primary object of the present ;~
19 invention to provide techniques for very high density storage
20 of information- ,
21 It is another object of this invention to provide
22 storage of information in accordance with natural interactive
.. . .. ..
23 phenomenon, rather than by imposed limitations due to auxiliary
24 strUcture-
25 It is still another object of this invention to ;
26 provide information handling apparatus using arrays of e1ements27 whose pos~tio~s aré substantially determined by interactions
28 existing therebetween.
. - . .. .
. .. . . .. ..
YO972-063 -6- ~ ~
, ' ~ : ` ' , .
.. :
: . :-
..:
.. :
3~
1 It is a further object of this invention to provide
2 an apparatus which requires only a minimum of structural
3 elements. ;~
4 It is a still further object of this invention to
provide a system for storing magnetic bubble domains with
6 extremely high density.
7 It is another object of this invention to provide
8 systems using lattice arrays of magnetic bubble domains.
9 It is still another object of this invention to provide
apparatus utilizing arrays of interactive magnetic elements.
11 It is a further object of this invention to provide
12 storage and memory with extremely high density using interaction
13 between magnetic elements for determination of storage positions.
14 It is a still further object of this invention to
provide techniques for storage of information at very high
16 densities with minimum cost.
17 It is another object of this invention to provide
18 techniques for high density confinement of interactive elements
19 with high degrees of inherent stability.
It is another object of this invention to provide
21 apparatus having high density storage where operating margins
22 are significantly enhanced.
23 It is still another object of this invention to
24 provide storage of information which achieves high density
without constraints due to auxiliary ~tructure.
26 It is a ~till further object of this invention -
27 to provide information storage using lattice arrays of coded
28 elements whose positions are substantially determined by
~0972-063 ~7~
-
..'
.:
36i3
l interactive forces existing between the elements.
2 It is still another object of this invention
3 to provide techniques for utilizing arrays of interactive
4 elements in new ways.
It is a further ob]ect of this invention to provide
6 apparatus for moving a plurality of interactive elements
7 whose positions are determined by interactions existing between
8 the elementS.
9 It is another object of this invention to provide
techniques for accessing interactive elements within lattice
ll arrays.
12 It is still another object of this invention to
13 provide techniques'for controllably moving interactive elements ` -
14 into and out of lattice arrays of such elements. ~-
It is a further object of this invention to utilize
16 multiple lattice arrays of interactive elements.
17 It is a further object of this invention to provide ~ ~
18 displays using lattice arrays of interactive elements. ~ ; --
19 It is another object of this invention to provide
interacting elements in confined arrays which have information
21 associated with them.
". ~, .. .
22 BRIEF SUMMARY OF THE INVENTION .,-
_ _ _
23 This invention~relates to techniques for utilizing ~;,,
24 elements which are held in an array by a con f ining means.
Interactive forces can exist among these elements if they are
26 sufficiently close to one another. Other than near the confining ~
, ,,;, ~ ~ ",
YO972-063 -8-
,"~'' ~"' " ," "'.
:,- ,.,. .;
., ,:' '.'
1~43~i~
1 means surrounding the array, the positions of the elements
2 within the array are substantially determined by interactions
3 existing among the elements. Input/output means are provided
4 for moving el~ments into the array and for removing elements
from the array. In an especially useful apparatus, the elements
6 within the array form a lattice arrangement which has the advan-
7 tage of high density packing. Depending on the parameters of
8 the system, the lattice can be of several types, such as a
g hexagonal lattice, or a square lattice.
Magnetic elements are particularly useful as the inter-
11 active elements, although other types of elements can also
12 be utilized. The magnetic elements can have information
13 associated with thçm, as for example when a high density infor-
14 mation storhge system is desired. Magnetic bubble domains
lS are especially suitable interactive elements.
16 The interactive elements are brought to a lattice
17 forming means which confines the elements within an area. The
18 positions of the interactive elements within this area are
19 largely determined by interactive forces existing between the
elements, rather than by external structure. -Depending on the
21 interaction forces between the elements, the elements can
22 arrange themselves in a lattice configuration which is a stable
23 arrangement. In fact, the stability of this arrangement
24 can be greater than that provided in systems ~here external
structure is used to determine the positions of storage
26 elements with respect to each other. ~ ;
: . .
YO972-063 ~9- ~
3~ :
1 In order to provide forces to hold elements which - -
2 are along the edges of the lattice area, a confinement means ~ -
3 is provided. This confinement means simulates the forces
4 which would be present on the elements in the outer periphery
5 of the lattice, if additional interactive elements were present
6 to interact with these peripheral elements. Thus, a lattice
7 of any size ~and any number of interactive elements) is provided
8 which, mathematically, has the appearance of an infinitely
g extending lattice. The lattice can extend in any of a
10 plurality of dimensions, as for instance a one dimensional or -~-
11 two dimensional lattice.
12 Input means is provided for moving interactive elements
13 into the lattice and output means is provided for taking
14 interactive e~ements out of the lattice. These input means and
output means provide forces to overcome the confinement forces
16 used to maintain the lattice. In certain systems, the lattice
17 is sufficiently elastic that the provision of input interactive
18 elements will cause elements within the lattice to move
19 out of the lattice at the other end, due to a wave propagation
of interactive forces between elements in the lattice.
21 Generally, the input means and the output means can be used to
22 remove or enter a plurality of interactive elements, or single -
23 interactive elements. Preferably, the lattice is maintained
:: ,.
24 although it is possible to alter it somewhat during the
input and output operations. ~ ~
26 A write means is provided for producing interactive ~ -
27 elements to be entered into the lattice. In one mode of
. .- . - ,
28 operation, information is associated with each of the interactive
.. . .
.. .
YO972-063 -10-
:, .:
,~
. , . -, . . .
: ..
~. ..- : : - : .:
.. ..
. , ' ~; ~'
,, ,"~,.. ,... ~ .
,-- .
1 elements as opposed to a system where information is stored
2 as the presence and absence of elements. For instance, if the
3 elements are magnetic bubble domains, coding can be in accor-
4 dance with a magnetic property of the domains. Therefore, the
write means provides interactive elements for the system and
6 preferably provides coded information where the coding is in
7 terms of the properties of the elements. As another example,
8 magnetic interactive elements can have information bearing
9 elements associated with them which appear optically different.
Also, the electrical or magnetic properties of the information
11 elements can be different in order to be able to detect differ- -
12 ent information states. Thus, the lattice can be ~illed
. : .
13 with elements having different properties or the interactive
14 elements can have information associated with them.
A reading means is provided to detect the interactive
16 elements used in the lattice. Preferably, this reading means -
"~ . . .
17 includes means for detecting different properties associated ; ~ -
18 with the elements. In this way, the coded information is read
19 and utilized.
20 Means is provided to manipulate interactive elements -~
21 outside the lattice area. Such means includes propagation
22 means for moving the elements, as well as means for performing
23 functions on the elements themselves, such as creating and
24 destroying elements, etc. -
25 A lattice display can be pro~ided by using a light ;--~
26 source to illuminate the lattice which contains interactive
27 elements having different properties. Thus, a viewer (or an
28 output detector3 is responsive to light passins through the
29 lattice or reflected from it in order to provide a representa~
'-'',' '. '
YO972-063
3~
l tion of the lattice itself.
2 A particularly suitable embodiment illustrating the
3 use of a lattice in a system utilizes magnetic bubble domains
4 in a magnetic medium. The magnetic bubble domains are free
to move within the magnetic medium and have stray magnetic
6 fields associ~ted with them. Due to the stray magnetic
7 fields, the magnetic bubble domains can interact with one ~;
8 another and, if no position determining structure is
9 provided, will seek positions determined by the interactive
forces existing between the bubble domains. Thus, if bubble
11 domains are brought into a confined area where they are free
12 to move, they will position themselves to minimize the -
13 total energy of the arrangement and can establish a lattice
14 of magnetic bubble domains.
When magnetic bubble domains are used as the magnetic
16 elements in a lattice, bias field means can be provided to
17 provide a magnetic bias field directed substantially along ~ ~ -
18 an easy direction of magnetization of the magnetic medium. ~-~
19 This bias magnetic field can have different values over
different regions of the magnetic medium. For instance,
21 it can be a substantially small (or zero bias) in the lattice
22 area while being a larger bias outslde this area.
, ;,,~,. ..
..
~,'". ' . .:
YO972-063 -12- -`
`'' ' '' "' " ''
" ;, "
- ~a~3~
1 Another alternative is that the bias field be uniform through-
2 out the magn~tic medium (i.e., within the lattice area and
3 outside the lattice area). Depending upon the range of bias
4 conditions chosen, the confinement force as well as the other
forces acting on the magnetic bubble domains is appropriately
6 adjusted. Additionally, the bias field means can include
7 means for providing a small modulating bias field which aids
8 in overcoming the coercivity of magnetic bubble domains in the ~
9 magnetic medium. ~-
10 Anotber suitable embodiment of a system using a ~-~
11 lattice of magnetic elements ¢omprises magnstic elements which ~
12 are supported by a medium in which they are free to movè. An -
13 example ls magnetic elements which are carried by a medium
14 such as a liquid. For instance, magnetic elements can be
located on styrofoam balls which are free to float on water.
16 These magnetic elements are essentially dipole elements having
17 stray magnetic fields which interact with one another. Thus,
18 if the elements are close enough to one another, the position ~-
19 of each of the styrofoam balls on the surface of the water ` ~ -
will be determined by interactive forces existing between each
21 of the magnetic elements. These interactive elements can be
22 coded by different physical properties, such as their color,
23 and when so coded can be used to represent information. When ;~ ;
24 they are confined within an area, the array of elements may ~ -
assume lattice positions.
26 Nean~ is provided to ve the s~yrofoam balls into the
- .
YO972-063 -13-
.-. .
,',
3f~8
1 confined area lattice and outside the lattice area. -
2 Additlonally, means is provided to detect different coded
3 properties of the styrofoam balls (for instance, their color)
4 in order to sense the information associated with the
magnetic elements.
6 A system using a lattice is particularly useful
7 for storage of information when the interactive elements
8 themselves have different properties. In such a case, a very ~-~
9 dense store is provided which uses a minimum of structure in the l~,
storage area for moving the interactive elaments. However,
11 the elements need not be coded in order to be useful in a
12 system. Such interactive element lattices can be useful as
13 sub-combinations of other useful apparatus. Thus, systems `~
14 having information associated with interacting elements are
presented herein which use an array (lattice) arrangement
16 of these elements as an integral part of the system. Whereas -~
17 periodic arrangements of elements have been physically observed,
,~ .
18 the following is the first such use of these arrangements in
19 useable systems. -
These and other objects, features, and advantages
21 will be more apparent from the following more particular ~ ~
22 description of the invention. - ;
23 BRIEF DESCRIPTION OF THE DRAWINGS ~ ~-
.. _ _ . __ . _ . .;,~,i.,.:
24 FIG. 1 shows a lattice arrangement of interactive '~
elements.
26 FI~. 2 is a block diagram of an information-
27 handling apparatus utilizing a lattice arrangement of inter-
YO972-063 -14-
- ;
3~
1 active elements.
2 FIC-. 3 is a block diagram of another information-
3 handling apparatus using a lattice array of interactive elements,
4 where the in~eractive elements can be returned to the lattice -~
array.
6 FIG. 4 is a block diagram of an information-handling
7 apparatus using two lattice arrays of interactive elements.
8 FIG. 5 is a block diagram of an information-handling
g apparatus using a lattice array of interactive elements, in
which a lisht source is used to llluminate the lattice array.
11 FIG. 6 is a block diagram of an information-handling
12 apparatus using a lattice array of interactive elements, where --
13 the interactive elements are magnetic bubble domains.
14 FIG. 7 is a plot of lattice constant aO and bubble
domain diameter d versus applied bias field, for a lattice
16 arrangement of magnetic bubble domains.
17 FIG. 8 illustrates the shape of magnetic bubble
18 domains in a lattice array when the applied bias field ~ is -
19 sufficiently negative. ~
20 ~ FIGS. 9A-9D represent various ~hapes of a lattice ~ . -
21 arrangement of interactive elements.
22 FIGS. lOA and 10B represent an input operation where `
23 interactive elements are to be put into a lattice array.
24 FIG. 11 is a schematic diagram illustrating
tolerances for the positions of peripheral interactive elements
26 in relation to the confinement means used to retain the lattice -~
27 array.
'~
YO972-063 -15-
. ' ,'~' ' ' ' , .'
: ','.. ' :
;`' ~ ' `'"' ,,,"'',' , '
: : ~
; . . .
1 FIG. 12 is a plot of the total bias field Hz versus
2 distance along a magnetic medium for a bubble domain lattice
3 array which has a different bias field outside the array than
4 within the ~attice array.
FIGS. 13A-13D illustrate various structures for
6 confining in~eractive elements within a lattice array.
7 FIGS. 14A-14D illustrate additional structures for -~-
8 confining interactive elements within a lattice array.
9 FIG. 14E shows a current-carrying conductor and the
magnetic field produced by this conductor as a function of
11 one of the dimen6ions of the conduetor.
12 FIG. 15 shows a eonfinement structure for magnetic
13 bubble domain interactive elements which utilizes a magnetic
14 discontinuity as an, aid in confining magnetic bubble domains -,
15 within the lattice. -
16 FIGS. 16A-16C show various ~truetures for providing
17 uniform applied bias fields across the entire magnetic sheet
". ...: . - : -
18 in which a lattice array exits.
19 FIGS. 17A-17B show structures for providing
~ .... ~ .. .. .
magnetic bias fields having different amplitudes within the
21 lattiee array and outside the lattice array.
22 FIG. 18 shows a structure for providing access means to
23 move interactive elements into and out of a lattice array.
24 FIGS. l9A-19E illustrate the operation of the
25 structure of FIG. 18 for movement of interaetive elements '
26 into the lattice array.
27 FIG. 20 shows another structure for movement of ~ ;
28 interaetive elements into a lattiee array.
YO972-063 -16-
,"','"''~.
''`", ' ' ~: '. .,.:
, . `:
-' ' ' ,' ,'. :
'lF3i~
1 FIG. 21 illustrates another structure used for the
2 movement of interactive elements into and out of a lattice ,
3 array.
4 FIr~'S. 22A-22G illustrate the operation of the
structure of FIG. 21 for various time sequences, showing the
6 operations of injecting interactive elements into the lattice
7 and ejecting these elements out of the lattice.
8 FI5. 23 shows a structure suitable for aiding
9 movement of magnetic bubble domains within a lattice array,
the structure providing forces which tend to overcome coercive
11 forces within the array.
12 FIG. 24 shows a structure for aiding movement of
13 interactive elements within a lattice array.
. . .
14 FIG. 25 illustrate~ schematically the movement of
interactive elements into and out of a lattice array using Q~
16 geometric fan-in and fan-out techniques.
17 FIG. 26 is a plot of the various bias fields which
18 can be used in the technique of FIG. 25, when the interactive
19 element~ are magnetic bubble domai~ns.
FIG~ 27 is a structure which accomplishes the fan-in
21 and fan-out of interactive elements into/from a lattice array
. .
22 according to the technique conceptually illustrated in ~`~
23 FIG. 25~
24 FIG. 28 is a block diagram of another structure
for moving interactive elements into and out of a lattice
26 arraY-
27 FI5. 29 is a detailed diagram of the structure shown
28 in block diagram form in FIG. 28. -- -
'.`~ .~ '
YO972-063 -17- ~-
.. . . .
.. ~ .. . .
"',. , `,
. .
- .:
,'`'' ' ' ,:~
1 FIG. 30 is a table showing the sequence of current
2 pulses in appropriate conductors which is used during the
3 movement of interactive elements into and out of a lattice
4 array in accordance with the structure of FIG. 29.
FIG. 31 is a table showing the various positions of
6 interactive elements for different times corresponding to the
7 sequence of current pulses applied to the conductors forming
8 the structure of FIG. 29.- ,
9 FIG. 32 shows structure for coding magnetic bubble
domain interactive elements for use as information states
11 within a lattice array, the coding being in accordance with
12 the hard and soft properties of magnetic bubble domains.
13 FIG. 33 shows a read means for dstectin~ the hard
14 and soft properties of magnetic bubble domains which have been -.'!.. -.. `"`,.. .'
written in accordance with the structure of FIG. 32.
16 FIG. 34 shows a structure for coding magnetic bubble
17 domain interactive elements in terms of their deflection
18 properties, for use as information'bearing element~ within ,
19 a lattice array. -
FIG. 35 shows a read means for determining the
21 information states of magnetic bubble domains coded in terms
22 of their deflection properties, by structure such as that
23 Of FIG. 34.
24 FIG. 36 shows a structure for coding magnetic
bubble domain interactive elements in terms of the size of
26 the bubble domains. This figure also shows ~he means for
. , ..; .
27 reading bubble domains which have differing sizes representative
28 of information.
YO972-063 -18-
''' ~
',..,'' ~.
: :' '','.
.',' ~`' .
3f~8
1 FIGS. 37A-37B show bubble domains having different -
2 chirality, wk.ich can be used for coding information in a
3 bubble domain lattice array.
4 FIa. 38 shows a plot of the amplitude of an in-plane
magnetic field pulse measured against time, which can be
6 used to detect domains having different chirality.
7 FIGS. 39 and 40 are structures for detecting magnetic
8 bubble domàins having different chirality. -
9 FIG. 41 shows a lattice arrangement of interactive
10 elements which are free to move in a supporting medium. i-,
11 FIG. 42 is a detailed illustration of suitable
12 magnetic interactive elements whi~h can be used in the lattice
13 structure of FIG. 41.
14 FIG. 43 is a schematic diagram of the circuit
arrangement for the lattice structure of P'IG. 41, illustrating
16 a double lat~ice structure in combination with shift registers
17 for transferring interactive elements between the lattices. -
18 FIG. 44 is a detailed dia~ram of the input and
19 output means used to move interactive elements into and out
-20 of the lattices of FIG. 41.
21 FIGS. 45A-45C illustrate the operation of the
22 Btructure of FIG. 44 for removing interactive elements from -; ;
23 a lattice array.
: . . ,
24 FIGS. 46A-46C illustrate the operation of the
structure of FIG. 44 for injecting interactive elements into
26 a lattice array.
27 FIG. 47 is a table showing the presence and absence
28 of currents in ~he conductors comprising the struature of ~ -
~ ', '~ ~ ' " "
YO972-063 -19- ~
... . :-- ~ . . - - -
1 FIG. 44 dur..ng the operations of shifting and input/output
2 to and from the lattices. .
3 FIG. 48 shows a detailed diagram of circuitry used
4 to shift information from one lattice to another in the structure
5 of FIG. 41. .
6 FIG. 49 is a table illustrating the position of : :
7 interactive elements as they move along the shift regiater
8 of FIG. 48, for various currents in the conductors comprising
9 the shift register of FIG. 48.
,,, '
10 DETAILED DESCRIPTION OF THE PREFERRED ENBODIMENTS ~ :
.. , ... . __
11 The following description will be concerned with
12 three major areas~
13 I. La~tice Information Systems - General ~- -
14 Description;
II. Bubble Domain Lattice Information Systems; :. .
16 III. Other types of interactive Element Lattice
17 Systems. . .:
18 In all of the embodiments to be described, elements .~ ..
19 which can interact with each other are brought into a confined ..
20 area or removed from this area, usually in groups of more than .` . .
21 one element. Within the confined area, there is generally no .~
22 structure which will establish the positions of the elements ~.
23 with re~pect to one another. If the elements are close enough
24 to one another they will interact with one another and these `. -
. ; . : .
25 interactions will substantially determine the positions of ~. .
26 the elements in the confined area. The arrangement of elements ; ~
..... .
~ 'A ' ;
. ", . ' . .
YO972-063 -20- .... :
.
,:
' ' ~ ".
1~ 3~
1 in the confined area can form a lattice array of elements,
2 which is particularly useful in various information systems.
3 The first major topic concerns the general principles
4 of systems using confined arrays (such as lattices) of inter- ~-
active elements, while topics II and III describe lattice
6 information systems using particular types of magnetic inter-
7 active elements. In topic II, the magnetic elements are
8 magnetic bubble domain~ while in topic III, the magnetic
9 elements used are, for instance, magnetic dipole elements
supported by a medium in which they are free to move.
11 I. LATTICE INFORMATION SYSTEMS
..
12 FIGS. 1-5 show a lattice array of magnetic elements
13 and various systems utilizing lattice arrays. The systems of
14 FIGS. 2-5 utilize any type of elements where there are
interactions between the elements which establish the positions
16 of the elements with respect to ona another.
17 The discussion in this se~tion will concern lattices
18 in general and the systems of FIGS. 2-5. The later sections
19 will deal more particularly with embodiments to realize these -~
general systems. In particular, discussions concerning bubble
21 domain lattiae information systems (II) will di~cuss many of
22 the parameters of the~e systems in a general way. Additionally,
23 the effect o~ other parameters where magnetic bubble domains -~
24 are utilized will also be discussed.
... .-
YO972-063 -21- -
' .
.. ...
.
. ~. .
.-:
:-.~ :...-
: . . . .
,~
3f~
FIG. 1
2 FIG. l shows a lattice 30 which is comprised of a
3 plurality of interacting elements 32. These elements
4 interact with one another in a manner which detèrmines the
positions of the elements with respect to one another.
6 The elements are shown in this diagram as being round elements
7 having a diameter d, and a center-to-center spacing aO, which
is the lattice constant. In FIG. 1, the elements are in a
9 hexagonal packing, but it should be understood that ~quare
lattices can also be used in the present invention.
11 The element~ 32 are free to interact with
12 one another and to move such that their positions are determined
13 by the interactive forces existing between them. In the CaJe
14 of repulsive interactive elements and a fixed number of
elements in a given area, a hexagonal arrangement leads to
16 a maximum separtion between elements. This is illustrated by
17 the arrangement of the shaded elements in FIG. 1. Each magnetic
18 element has six nearest neighbors arranged in a hexagonal -
19 ;
structure. In this manner, the lattice array 30 is similar
to an array of atoms in an atomic lattice.
21 The lattice is characterized in that the forces on
22 any interactive element~32 are primarily those due to its
23 nearest neigh~ors. Of course, the elements around the
24 periphery of the lattice array will have the position shown ~ ~-
25 in FIG. 1 only if there:are restraining forces provided on ~ -
6 them in order to compensate for repulsive forces from the other
27 elements in the lattice. That is, confinement forces are ~.
- - ' '' ~.
YO972-063 -22-
,
: ' '
3~8
required to ensure that the interactive elements on the
2 periphery of the lattice are not expelled from the lattice
3 due to forces from elements within the lattice.
4 - The arrangement of FIG. 1 is capable of providing
very high packing density, since the interactive elements can
6 have a close-packed structure in which the lattice constant
7 aO is very small. For instance, a lattice array of inter-
8 active bubble domains in a material with applied bias field
9 Hb=O can have a lattice aO slightly greater than the domain ~ -
diameter (aO=1.35d). This provides extremely high packing
11 densities and, if information is stored in the properties of
12 the interactive elements themselves, extremely high storage -
13 densities can be provided. Additionally, since the system
14 will operate without propagation structures within the lattice
array, ease of manufacturing and utilization results.
16 The shape of the lattice array can be varied greatly
17 as will be more fully appreciated in further discussions. The
18 particular shape of the lattice in FIG. 1 uses the sylNnetry ~;
19 planes of ths lattice such that columns of intera¢tive elements
32 are at an angle of approximately 60 with horizontal rows
21 within the lattice. This is a particularly useful arrangement
22 since it provides for more direct accessing of~elements within
2~3 the lattice. However, other combinations exist, it being only ~ ~ -
24 important that the interactive- elements be contained in an
area having substantially no position-defining structure for the
26 elements and where they can be selectively addres~ed. When
27 the elements are close enough their- positions are locally
,
YO972-063 - -23- ~
,~- ..
',' ~
.
.
. . .
. . . : .
3~ ~
1 determined primarily by interactions with other elements,
2 since position-determining structure i9 sub~tantially absent
3 from the area of confinement.
4 As the discussion proceeds, lt will be`apparent how
information can be coded in the elements 32, how these
6 elements or groups of such elements can be moved into and
7 out of the lattice array 30, how such elements can
8 be detected, and how very dense storage systems can be provided.
,~, ,
g FIG. 2
This figure shows a system using a cenfined array
11 30 of interactive elements 32. In this figure, array 30 is
12 a lattice which is maintained by confinement means 34 which ; -
13 provides a force on the peripheral elements 32 within the
14 lattice, thereby-maintaining the lattice. The size of
, .~:: .
lS lattice array 30 is immaterial; it could be, for example,
16 a 2 x 2 lattice, or 1000 x 1000 etc. In principle, any
portion of the lattice looks as th~ugh it were part of an
18 infinite lattice of elements, since the primary contribution ;;~
;. . . . .
19 of forces on any elemènt within the lattice is that due to s
.. ~ .
its nearest neighbors.
21 A write means 36 is used to provide interactive ~ i
22 elements 32 to the lattice. Additionally, this write means
23 could include further means for coding the elements 32 so that
24 different properties are obtained. In this regard, the -
elements 32 in the lat~ice will themselves be the carriers of
2~ information.
27 Input~means 38 receives the interactive elements from
28 write means 36 and enters them into lattice 30. Input means
~'..': .
Y0972-062 -24- ~
L~ 3f~
1 38 provide~ sufficient force to the input elements that
2 they are able to overcome the force produced by confinement
3 means 34 in order to enter the lattice array.
4 An output means 40 i8 used to remove elements
32 from lattice 30. Output means 40 i9 similar to input means
6 38 in that it provides sufficient forces to overcome the retain-
7 ing forces provided by confinement means 34. This enables
8 elements 32 within lattice 30 to be withdrawn from the lattice.
9 As will be more apparent later, providing elements 32 at
the input of the lattice can be a technique for removing
11 other elements from the lattice. That is, elements 32 already
12 within the lattice have forces exerted on them by the elements ~ - -
13 being entered into the lattice and these force~ are transmitted
14 through the lattice causing the elements at the output end
lS of the lattice to be expelled from the lattice. -~
16 A read means 42 receives elements 32 from the -
17 output means 40. The read means is used to detect the elements
18 32 from the lattice~and in particular is used to-~detect
19 different information as50ciated with the elements. Of --
course, a read means is not required when the lattice elements
21 32 are to be utLlized elsewhere. Utilization means 44, such
22 as a computer or other apparatus, can be means responsive
23 to the signal produced by the read means or can be means for
24 utilizing domains previously stored within lattice array 30.
The system of FIG. 2 operate~ under control of
26 control means 46, which provLdes clock inputs to input means
27 38 and to output means 40. Thus, Yynchronization is provided. ~ -
' . ' . ':
Y0972-063 -25- - -
' ~, ',
'. ' : .:
, :-, - : -
3~
1 FIG. 3
2 FIG. 3 is an information system using a lattice array
3 which is a modification of the system of FIG. 2. Accordingly,
4 in this figure and in other-figures, the same reference ~- -
numerals will be used whenever possible to identify components
6 having the same or similar functions. In FIG. 3, the system ~-
7 is comprised of a write means 36 for producing interactive
8 elements to be stored in lattice area 30. Input means 38 is
connected between write means 36 and lattice 30 and provides
accessing of interactive elements into lattice 30. An output
11 means 40 removes the interactive elements from the iattice
12 area 30 and transmits them to the read means 42, where their ~-
13 information content is determined. If desired, a utilization
14 means is provided for using the output of the read means for ;-~
15 other purposes. Synchronization of the overall system is ~ i
16 provided by control means 46 which produces clock pulses to
17 synchronize the input and output means, as well as any other
18 means requiring control.
19 The system of FIG. 3 diff-rs from that of FIG. 2 in
that a path 48 is provided for returning the interactive magnetic
21 elements from the read means 42 to the input means 38. In this
22 manner, nondestructive readout occurs and the interactive
23 elements can be returned to the lattice area 30. As will be ~, -
24 seen later, these elements can be replaced in corresponding
positions in the lattice or can be intermingIed with new
26 elements from write means 36. Thus, it is possible to provide
27 entirely new information in the lattice, to replace only part
28 of the information in the lattice, or to restore the original
29 information to the lattice area. This feature is particularly -
advantageous in large capacity storage systems where nondestructive
31 readout is a preferred mode of operation.
, .~ ., , .:
YO972-063 ~ -26- . ~
- ` ~
~O9~3~i8
1 FIG 4
FIG. 4 is a schematic diagram of another information system using
a lattice area as an integral part thereof. However, the system of
FIG. 4 utilizes two lattice areas 30A and 30B, wherein the magnetic ;
elements move back and forth between the two lattice areas. Thus, in-
formation read from one lattice is transferred to another for storage -
therein. This is a nondestructive readout scheme which is particularly
suitable for use as a high-capacity storage system. That is, the
interactive elements within a lattice area are very closely packed
to provide high density storage. These elements are removed from the
lattice for the reading operation and are returned to another lattice
for storage. If desired, the information taken from lattice 30A can
be recoded before being put into lattice 30B. Consequently, information
is retained and a very efficient high-capacity system is provided.
In FIG. 4, the same reference numerals are used as were used in
the previous figures. The first lattice area and its associated compo-
nents are identified by the suffix A while the second lattice area and
its associated components are generally identified by the suffix B. ~
In more detail, a write means 36A provides interactive elements to ~ ;
input means 38A. These elements are placed in lattice 30A by the input
means 38A. and can be removed from lattice 30A by output means 40A.
Control of the input means 38A and output means 40A is under the : ^ ~
direction of pulses provided from control means 46A. Elements ta~en ;
from 1attice 30A are read by read means 42A, after which they are ~
directed to lattice 30B. -
'' '~
-
., ~ ''. ,
Y09-72-063 - 27 - ~
~ . ',- ',':
': ` -' - -
1 Input means 38B directs the output of lattice 30A to lattice 30B
- where they are entered therein. If desired, new information can be -
provided by write means 36B, the information from lattice 30A being
detoured or destroyed by means 38B under control of unit 46B. The
elements in lattice 30B can be removed therefrom by output means 40B
and then read by read means 42B. The input and output operation for
lattice 30B is controlled by control means 46B. The overall control
of the system comprising both lattices 30A and 30B is synchronized by
syncronization means 50 which provides inputs to control means 46A and
46B. This ensures that information moves smoothly from one lattice to
another and that the operations concerning each particular lattice are
timed properly.
The output of read means 42B is directed to input means 38A asso-
ciated with lattice 30A. If desired, the output of lattice 30B can
be directly entered into lattice 30A. However, the input means 38A can
include means for destroying or rerouting interactive elements from
lattice 30B in order to be able to write new elements into lattice
30A. In this case, the write means 36A will produce the elements which ;~
then are entered into lattice 30A.
Thus, the system of FIG. 4 provides a circulation of elements from
one lattice area to another under control of associated components.
.. ...
This is a particularly advantageous systems approach and one which can ^
be easily modified or upgraded, for instance by the proYision of addi-
tional lattice areas. Further, the size of the two lattice areas shown
does not have to be the same, and asynchronous operation can be
utilized. The principle is that information`from a lattice need not be
.
,
~I '
`;::
Y09-72-063 - 28 -
, .
.. . .:
JL~ 3ti8
1 returned to that particular lattice or be destroyed; rather, it can
be moved from one lattice area to another in order to retain the infor-
mation in a more economical and efficient manner.
FIG. 5
FIG. 5 illustrates that optical readout can be used in conjunction
with information contained in the lattice area. Further, the lattice
area can be used for display of information in the form of the different
interactive elements within lattice 30.
In more detail, lattice 30 is located between a light means gen-
erally designated 52 and a read means, generally designated 54. Light
means 52 is comprised of a light source, such as a laser, and a polari-
zing element 56. Of course, for various systems the light source need
not be a source of coherent light and a polarizing element need not
be used. In the particular case where the elements 32 in lattice 30
are magnetic bubble domains, the use of a polarizer in combination with ;
a light source is particularly advantageous.
Read means 54 is comprised of means for detecting the light which
is transmitted through the lattice or reflected from the interactive - ~;
elements within the lattice. In FIG. 5, read means 54 comprises an
analyzer 58 and a suitable light detector 60. In some cases, the use ~;
~f polarized light is not required and consequently analyzer 58 would
be unnecessary. Additionally, the detector could be any sort of light
responsive mechanism and in some cases will be the human eye viewing
the arrangement of interactive elements within lattice 30. This will
''''" '; ' ''
-'". ~ ' ,:'
~'''';~' .''
~ .
.,~'.'
Y09-72-063 - 29 - ~
.
1 be more apparent in the description of the embodiments which follow.
Although the read means is shown being outside the lattice area, it
could be positioned to read information associated with interactive
elements while they are within the lattice area.
The description of FIGS. 1-5 has been a general description for
systems using a confined array (which could be a lattice~ of inter-
active elements where the interactions between the elements are the
primary determinents of the position of the elements within the area
of confinement. While some general considerations have been discussed
concerning lattice properties, accessing of information within the -
lattice, and establishment of interactive elements having different
properties, these concepts will be more fully explained in the following -
description, which describes particular embodiments for achieving in-
formation systems using lattices of interactive elements. '
::~. II. BUBeLE DOMAIN LATTICE INFORMATION SYSTEMS
This topical portion will be concerned with information systems
using confined arrays (lattices) where the interactive elements are -
magnetic bubble domains that are free to move in a magnetic medium --~ ~
which supports them. Such a medium is well known in the art, and -
includes orthoferrites, garnets, amorphous magnetic materials, and , -
~any other magnetic medium capable of supporting magnetic bubble do-
mains. In the discussion to follow, many of the aspects of the physics -
and mathematics
-
YO9-72-063 - 30 - i-~
~ 3t~
1 concerning lattice arrays will be applicable to systems using ;nter-
active elements other than magnetic bubble domains. In the particular
case of magnetic bubble domains, other parameters, such as a magnetic
bias field, can have some influence on the physics of the system.
These particular influences will be described in detail. r
FIG. 6 shows a representative system using a lattice array 30
where the interactive elements 32 are magnetic bubble domains existing
within the magnetic medium 62. The confinement means 34 is used to
control the shape of the lattice 30 and to retain magnetic elements
32 located along the periphery of the lattice. As described prev;ously,
a write means 36 produces magnetic bubble domains for entry into
lattice 30 via the input means 38 and output means 40 can be used to -remove bubble domains from lattice 30, which domains are then read
by read means 42. This embodiment assumes that the bubble domains
are coded to have different properties indicative of various infor-
mation states. The output of read means 42 is applied to utilization ;
means 44. Control means 46 provides inputs to write means 36, input
means 38, output means 40, and read means 42 in order to synchronize -
operation of the system.
The bubble domain lattice system of FIG. 6 also includes a bias ~
field means 64 for producing a magnetic bias field, generally but ~ ;
not necessarily directed substant;ally parallel to an easy d;rection ~ -
of magnetization of the magnetic medium in which the bubble domains ~ ~-
ex;st. As w;ll be more apparent later, the magnetic bias field can
be uniform throughout the magnetic medium or may have differing
values in -i
,~, . '
., . "
'-'' . ,,: ',:
~
;' ' .
Y09-72-063 - 31 -
.~
1C~ 3~
1 different regions of the magnetic medium. For instance, a very small
(or zero) bias field may exist in the lattice area while a larger bias
field may exist in areas of the magnetic sheet surrounding the
lattice area.
A propagation field means 66 is also provided which is generally
used to provide magnetic fields for moving the magnetic bubble domains.
The propagation field means can include many different structures in-
cluding current carrying conductors or elements of magnetically soft ,~
material located adjacent to the magnetic medium in which the bubble ~ -
domains exist. The propagation fields are used to move domains both
..... .
within lattice area 30 and in areas of the magnetic sheet 62 surroun- -~
ding the lattice area.
The following will be a description of the various operating
parameters required for providing systems using lattices of inter-
active elements. ~ ~ -
A. Bias Field Conditions
For a lattice comprising magnetic bubble domains, a parameter
which can be adjusted for design purposes is the applied magnetic
bias field Hb parallel to an easy direction of magnetization of the
bubble material 62. In general, various bias field arrangements are
possible, such as
1. Bias field Hb ~ or a small value within the lattice area
but has a larger value outside the lattice area. Its value outside
the lattice area is approximately that used in the case of is~lated
bubble domain devices (i.e., (4~rMs)/3, where Ms is the saturation
magnetization of the -~ -
, .
, ` ', " ':
- ~
~ Y09-72-063 - 32 -
- .
' ' ' - .
..... , . . . ~ . , . ~, = .. , , ., ~ . . ......... . , . - . ,.. , . ,, . .. -... .. , . . . - .. .
~L~J~ 3
1 magnetic bubble domain material).
2. A small uniform bias field Hb can be used for areas inside
and outside the lattice. In this case, the bias field is uniform
over the entire magnetic bubble domain material, suitable value being
approximately 1/4 (4~rMs). If there is an array of bubble domains
surrounding the storage lattice array, the applied bias field Hb can -
be uniform with a zero value both inside and outside the storage lattice.
Using different ranges of bias field is useful as a design para-
meter depending upon the application desired. For instance, a uniform,
small applied bias field throughout the magnetic material is helpful
~, !. ~ , ''
in providing a lattice which is relatively insensitive to variationsin bias field and in which the magnetic bubble domains can be more
easily moved. Since the interaction force between bubble domains is
proportional to d4/aO4j a change in bias field which changes these
parameters will alter the force existing between magnetic elements.
Since the ability to easily move magnetic elements within the lattice
is a function of the force existing between the elements, varying the
. . -.
bias field in a particular design is a useful parameter.
In addition to these bias field considerations, it is possible
to use an ac tickling field produced by a current-carrying coil
surrounding the magnetic medium. An ac field or a pulsed bias field
tends to reduce damping caused by coercivity Hc in the bubble domain
material which in turn enables the ~ ; ~
.- - , . . .
", , . ~ .
',~ '`'``~ .
: . .
. .,
Y09-72-063 - 33 -
;,
3~8
1 domains withi~ the lattice to move more freely. The frequency
2 of the ac field is quch that a few cycles of thi~ field will
3 occur during lattice shifting. For instance, pulse^q of 2-3 Mc
4 frequency having a width of about 1 microsecond~are suitable.
For some magnetic element~, uniform bias fields will
6 have no effect. For instance, in the embodiment to be shown
7 using styrofoam balls floating on a liquid and having magnetic
8 elements therein, the presence of a bias field has no effect.
9 Therefore, interactive elements can be provided which may or ~`
may not be influenced by a magnetic bias field substantially
11 normal to the medium in which they exist. Generally, when
. .
12 the magnetic moments of the magnetic elements are a function
13 of bias field, the bias field will have an effect due to the
- i
14 magnetic energy term which it introduces.
~attice PropertieS
16 In this subsection a lattice, such as that shown in
17 FIG. 1, will he discussed in more detail. The lattice of FIG. 1 ;
18 has been described as being made up of a plurality of interactive
19 elements 32 located in an arrangement where there is equal
spacing between the elements, the spacing being defined by
21 a lattice constant aO. The interactive elements have a
22 diameter d.
23 Generally, interactive elements will arrange them-
24 selves such ~hat the total energy of the system is minimized.
For instance, in a lattice comprised of magnetic bubble domains
26 where Hb=0, the magnetic medium in the lattice area will comprise
27 roughly equal areas of up and do~n magnetization. As additional ~ -
28 bubble domain~ are put into the lattice area, the size of bubble
YO972-063 -34- ~
: ' .
,: ' ' .
': ' .''. ' .
....
3~
1 domains within the lattice will change in order to have equal
2 areas of up and down magnetization. Further, the total
3 magnetostatic energy plus the domain wall energy is minimized
4 for a particular ~ize bubbl~ domain when in a lattice , ~-
arrangement~ Therefore, for a given number of bubble domains,
6 the bubble domains will position themselve~ ~o as to minimize -~ -
7 the total energy of the system. ~ ~
8 A ]attice characterized by equal areas of up and ~ ;
9 down magnetization is termed a demagnitized lattice. This -
10 lattice can be affected to change somewhat the distance --
11 between domains within the lattice in order to achieve a ~ -
12 more dense lattice with smaller bu~ble domains. Within the
13 bubble domain lattice there i9 a lowest energy configuration ;
14 for a particular bias field. For instance, at Hb = r the
spacing aO between bubble domains is 1.35d. This spacing
16 can be varied by use of an applied bias field Hb. In a manner .
17 analogous to friction, the coercivity Hc of the medium may
18 cause the lattice constant aO to locally deviate from this -
19 value.
A lattice has an inherent stability which is greater
21 than the stability for isolated bubble domains ~bubble domains -
22 which do not substantially interact with one another~.
23 This i~ illustrated by FIG. 7 which is a plot of lattice
24 constant aO and bubble domain diameter d as a function of
25 applied bias field Hb; The bubble domain material iS ~-
26 (YEu)3~FeGa)5012. From these curves, it is readily apparent
27 that there is a fairly wide range of bias field over which the ~ -
Yo972-063 ~35-
, . . .
.. ~
"-. . .
~-" "' ' .',' .
... ' .'. ~' .
':
3~
l quantity aO changes very little. During this same range
2 Of applied bias field, the diameter of the domains will
3 change slightly but not by a great amount. As the bias field
4 becomes large~, the distance between the domains increases at
a greater rate until the domains become isolated domains.
6 The diameter of the domains begins to decrease more rapidly
7 also. The bias field Hb can range from a negative ~alue to
8 approximately one half the value for isolated bubble -~
9 domains without substantial changes in the lattice
constant aO. However, there will be a change in the
ll total area of up magnetization of the magnetic medium
12 versus down ~agnetization of the magnetic medium in the
13 lattice area, due to a change in diameter of the bubble
14 domains.
15 The total bias field on domains within the lattice ~
16 is comprised of the applied bias field Hb and the bias -
17 field due to interacting stray magnetic fields of the domains.
18 As the applied bias field is increased, domains within the
l9 lattice will collapse at values of applied bias field less
than that which would cause collapse of domains isolated from
21 one another. This is because the total bias field on domains
22 within the lattice is a combination of the applied bias field ~; ~
23 and the interactive magnetic field produced betwcen interacting - -
24 domains, as explained previously. As the applied bias field
decreases, there is a range over which the domains in
26 the lattice won't be converted into stripe domains. This is `
27 due to the bias field arising from the interacting magnetic
28 fields between the domains in the lattice. When the applied
Y0972-063 ` -36-
` `'
3~i~
1 bias field decreases such that the center-to-center spacing
2 (aO) of domains within the lattice is approximately 1.25d,
3 the bubble domain shape changes, although a lattice arrangement
4 still exists. This is illustrated by the hexagonal bubble
5 domains of FIG. 8. ~ -~
6 As the value of applied bias field exceeds a certain
7 negative vallle, domains may start to combine and the lattice
8 structure no longer exists. When this type of domain combina- ;
9 tion occurs, it is not possible to reobtain the lattice ;
arrangement by merely increasing Hb.
11 If there are constraints on the lattice boundary ~-
12 (as for instance by structure which exerts forces to confine ~-
,....: . :,:
13 the lattice domains) the curves of FIG. 7 remain flat over a -~
14 greater range of applied bias field Hb. The bubble domains
will shrink in diameter but the lattice constant aO stays
16 approximately the same until the total bias field (applied field
17 Hb plus interaction field Hi~ becomes sufficient to collapse
18 the domains.
19 The range of bias field over which the lattice `;-
of bubble domains is stable is a function of the properties of
21 the material, including magnetization, anisotropy, and thickness,
22 as well as the strength of the exchange interaction of the
23 material. The range varies from small negative values to values
24 comparable to that which are the criteria for stable isolated
bubble domains. (Thus`for a typical 5 micron rare earth iron
26 garnet bubble domain-material, the total bias field ~Hb + Hi) i
27 is
YO972-063 ~37~
,''"',''''~'~
'.' . ~' ':'
~,, ~,,
., ~.;: .
. .: :
~Q4-~3~8
1 102 Oe > Hz > -25 Oe
2 The lower bound of total bias field is that total bias field
3 which will prevent domains from co~ining and th reby
4 destroying the lattice. The high value of suitable bias
field is a value less than the value which would cause isolated
6 regions of bubble domain collapse within the lattice. That
7 is, spontaneous collapse of bubble domains at areas within
8 the lattice should not occur, if it is important to have a lattice
9 with all positlons filled. Stated differently, the high and
low ends of a~plied bias field Hb are chosen such that the
11 lattice does not disappear, whether by reason of the domains
12 becoming s~ripe domains or by reason of the domain~ collapsing
13 within the lattice. ~
~ . -
14 Lactice Shape - Information Use
For a hexagonal lattice, the interactiYe elements
16 in the lattice arrange themselves in a hexagonal close packing
17 arrangement with each element having six nearest neighbors
18 arranged at the corners of a regular hexagon. That is, the
19 symmetry lines and planes of interactive elements in this type
20 of lattice a~e at 60 with one another. In the case of a .~
21 square lattice, the symmetry lines and planes will be at 90 `
22 with respect to each other. As will be more apparent later, ~
23 in order to provide a regular lattice (i.e., one in which -
24 all positions of the làttice are filled) the shape of the lattice
must be alon~ the symmetry lines and planes of the interactive
- `', ' .' ~:'` ,
YO972-063 -38- -~-
" ''. '"' ', '.,
,'
' ,'' ~` ' '
-: .
i8
1 elements.
2 FIGS. 9A, 9B, and 9C show three possible hexagonal
3 lattice shapes which will provide regular lattices with all
4 positions within the lattice filled by interactive elements 32.
5 FIG. 9A shows a hexagon-shaped lattice, FIG. 9B shows an equilat- -
6 eral triang.e-shaped lattice, while FIG. 9C shows a parallel ~
7 piped-shaped lattice. The symmetry lines and planes for each- ~ ;
8 of these lattices are arranged at 60 with respect to one
9 another. '
10 The lattice of F}G. 9D has a circular shape and is -~ -~
11 not a regular lattice since dislocations (that is, vaaancies
12 and misplaced elements 32) will exist within the lattice which
13 is bounded by a confinement means having circular shape.
14 Although the lattice of FIG. 9D is a lattice which could be
used in an information system, lattices having shapes determined
16 by symmetry lines and planes of the interactive elements
17 are easier to use. In particular, it is easier to move inter-
18 active elements into and out of these regular lattices, and
19 each position within the lattice will be filled. If the
interactive elements are coded to represent information, use
21- of a regular lattice will ensure that no information is lost.
22 If _he geometry of the lattice area, and the lattice
23 constant a~ are initially determined, a certain number of ,~
24 interactive elements will be required within the lattice ; ;
area in order to provide a regular lattice having all positions
26 filled. However, the lattice has some flexibility and it is
27 possible to insert additional interactive elements without
28 causing severe disturbances in the lattice. For instance,
'' '-
Yo972-063 ~39~ ~ ~
''.''",''~
-.' '' '
, ,. `'' '' .
1 FIG. lOA shows a lattice 30 which is assumed to be completely
2 filled with in~eractive elements 32, thereby farming a
3 regular lattice. It is desired to push new interactive elements
4 32A into th~ lattice and to have all elements 32 already in
the lattice remain there. This will cause a crowding of the
6 lattice at ~he edge where the element~ 32A are entered and
7 an adjustment of the lattice constant aO at that edge will
8 result. ~ -
9 FIG. lOB illustrates the case where only two interactive
elements 32A are to be entered into a regular lattice 30 having
11 interactive elements 32 at each position. When elements 32A
12 are entered into the lattice, vacancies corresponding to the
13 missing interactive elements in the column to be entered will
14 be moved into the lattice. These vacancies can be propagated
through the lattice and represent lattice dislocation~. In
16 general usage, it is prefer~ble not to have vacancies in the
17 lattice as information will be lost or the properties of the
18 lattice will not be maintained in a regular fashion. That is
19 the interacti~e elements within the lattice will rearrange them-
20 selves to compensate for the vacancies (or looked at another -
21 way, to compensate for the additional two interactive elements ~ -
.
22 32A). This will cause localized adjustments in the lattice
23 constant and the uniformity of the lattice will be perturbed.
24 AnoLher aspect of tolerance for successful accessing
of domalns into and out of the lattice concerns the dimensions
26 of the lattice. In particular, the dimensions of the left
27 and right hana edges 68L and 68R (along whi¢h domains enter - ~-
,: . . .
: ~:
.Yos72-063 --o~
~' ': . ..
''', ~ ''
-~ \
1 and leave the lattice) of the lattice are more important
2 than the other dimensions of the lattice. Any variation in : ~ -
3 lattice genera~ing will cause a local variation in lattice
4 spacing. Whatever variation occurs in aO should fall within
the limits of stability, as illustrated with respect to
6 FIG. 7 and must be gradual enough so as not to cause dislo-
7 cations in the lattice. As an example, a tolerance on the
8 left hand an~ right hand edges of the lattice of approximately
9 +1/2 aO is suitable.
Another consideration for the lattice concerns the
11 angle at which the input elements 32A enter the lattice.
12 Generally, the direction of the input elements 32A is`about
13 60 with respect to the edge 68L of the hexagonal lattice,
14 although vari~tions of this angle are allowed. For instance,
15 a variation Gf about ~2 is a suitable example. For a square - :
16 lattice the input angle is approximately 90 with respect to
17 the edge of the lattice, and can also be varied. This angle ~ ;
18 is not highly critical since the interactions between elements ,
19 32A and elements 32 within the lattice will have a
stabilizing effect on elements entering the lattice, thereby
21 maintaining the proper entry and exit directions. `
.'.'''.,
22 Confinement Force -
':
23 A lattice arrangement of interactive elements is
24 stable due to the interaction forces between the elements : ~-
. . .:
25 themselves. Ho~ever, for repulsive interactive forces, the : ;
26 elements located at the periphery of a lattice arrangement `
27 undergo forc~s which are not balanced by additional elements
,:
YO972-063 -41- ~ ~
. .
1 outside the lattice area. For instance, in FIG. 1 the
2 left-most u~per element will have forces exerted on i~ by other
3 elements within the lattice. If these are repelling forces,
4 this corner element will be pushed away from the rest of the
lattice arrangement. Therefore, a confinement means (generally
6 designated 34 in FIG. 2) i~ used to maintain the shape of the
7 lattice and to ensure that information in the form of inter-
8 active elemqnts is not lost from the lattice. This ~ubsection
9 will deal with the force required to maintain the lattice
while the following section will describe specific structures
11 for confinement of the lattice.
12 In general, the confinement means provides forces
13 which locally change the spaces between interactive elements.
. .
14 It should be remembered that the effects on any element within
the lattice æe primarily due to its nearest neighbors.
16 Conse~uently, if a confinement means supplies a force around -~-
17 the periphery of the lattice, the lattice will appear as an
18 infinite lattice to any element within it. In thi~ regard, -
19 the confinement force could be provided by elements outside ;~;
the confined storage area and these could be a latt1ce of
21 elements outs de the ~torage lattice area. -
22 Interactive elements at the edges of the lattice
23 can move somewhat so that the distance from these domains to -
24 the confinement means may vary. As illustrated in FIG. 11,
25 a variation of +20% aO is generally permis~ible, although a `
26 larger tolerance may be suitable for some applications. If
27 the confinement force is very strong, elements along the edges
28 of the lattice area wiIl be pushed toward the center of the
29 lattice while if the conf1nement force is weak, the outer
:
Y0972-063 -42-
.
.1 ' . ~ ~ .
, ~. ~ .. :
:J
!. :
1 rows of in~eractive elements will move closer to the confine-
2 ment means.
3 The confinement force can be either an attractive
4 force which tends to pin elements along the periphery of the
lattice or a repulsive force which tends to push peripheral
6 elements into the lattice area.
7 The confinement force is used to separate (by an
8 amount greater than aO) rows of interactive elements in order
9 to define a confined storage array (lattice). If there are
10 no interacti~e elements outside the storage area, the con- -
11 finement force is approximately equal to the interaction force
12 Fi acting on any interactive element due to its neighbors.
13 However, if there are interactive elements outside of the
14 intended storage area, the confinement force - which is just
the force required to provide separation (>aO~ between elements
16 inside and outside the storage area - can be less than Fi.
17 The amount of separation (in excess of aQ) to be
18 achieved between rows of interactive elements determines the
19 magnitude or the required confinement force. If the distance
between the roT~S iS aO, then there is no "separation"
21 due to a confinement force. Generally, for generat~on of lattice
22 arrays where elements are moved into and out of the storage --
23 area a separation in excess of aO by an amount aO or less is
24 sufficient, and the magnitude of the confinement force is
chosen to provide this. For instance, a center to center
26 spacing between row~ of (aO + aO/2~ is a suitable distance.
27 Of course, the separation could be greater, which in turn would
28 require larger confinement forces.
Yo972-o63 -43-
'~ '; '
3~i8
1 In the case of magnetic bubble domains, the confine-
2 ment force is conveniently provided by localized variations in
3 bias field in different regions of the magnetic~material. For
4 instance, in a system such as that shown in FIG~ 4, the
applied bias within the storage lattice areas could
6 be 0.1(41rMs) where MS is the saturation magnetization of ~ :
7 the material, while the applied bias in regions where bubble ::
8 domains are not used could be the saturation value. For the -~
9 shift registers which move bubble domains between the lattices,
10 the applied bias could be a value between 0.1(4~S) and the -~
11 saturation field, in order to stabilize bubbles in the shift . ~ .
12 register~ Of course, the value of Hb (applied bias field) .
13 in the lattice can be adjusted in order to obtain a desired . ~.
14 density of storage (aO). - ~ :
15 For magnetic bubble domains, the interactive force :::
16 Fi between isolated domains can be calculated as dipole forces ~ .
17 existing between the domains. Thi~s calculation establishes
18 that the interactive force is given by the following expression~
~ .:
19 ~1) Fi (2~r2h MS)2 . ; .
aO4 ~ ~. . ' ,
20 where : -
21 Ms is the magnetization of the magnetic medium in - -
22 which the bubble domains exist,
23 r is the radius of the bubble domain, ..
24 h is the height of the bubble domain,
a~ is the center-to-center distance between domain~
,~, ' . . '
Yo972-063 -44- . :-
A. ',
' ' ' '~,
3Ç~ :
1 Thus far, the discussion concerning confinement forces
2 has been for the case where the total bias field Hz = Hb + Hi i8
3 the same inside and outside the storage lattice area. However,
4 if the tota~ bias field within the lattice area is different ~-
from that ou~-side the lattice area, a gradient in total bias
6 will exist in the region of the lattice boundary. This
7 gradient in total bias field will provide an additional
8 force on bubble domains within the lattice for which account
9 must be made in determining the confinement force required.
The additional force term due to any gradient in
11 applied bias ~ield is given by dVHb, where d is the bubble domain
12 diameter and VHb is the gradient across the bubble domain.
13 FI~. 12 illustrates a situation where the applied
14 bias field Hb out9ide the lattice i9 different from that within
the lattice. However, the lattice area 1 may be such that
16 there is a gradient VHb which extend~ into the lattice area.
17 The gradient will lead to a force on those bubble domains
18 which experience the gradient and the presence of this force `
19 may cause an adjustment of lattice constant aO in localized
areas of the lattice. Consequently, the applied bias field
21 and the gradient should be adjusted so that the total bias
22 field on the magnetic bubbles will not be sufficient to
23 collapge the bubbles.
24 If the gradient in Hb extends into the lattice area
over a distance about aO or less, lattice area 1 (FIG. 12)
26 can be used. In that cas~, only one row of the lattice
27 would be perturbed by VHb and the resulting force would aid
28 in confining the lattice.
Y0972-063 -45-
..
': "';~"
d ..
If the gradient in Hb extends into the lattice area
2 over a distar.ce of several aO, several rows of domains in ~-
3 the lattice would be perturbed. In this case it may be ~ -
4 advisable to use lattice area 2 (FIG. 12) for the storage area.
The distance into the lattice which the gradient
6 can extend is quite flexible and depends upon the amplitude of
7 the bias fields, the slope of the gradient, and the thickness of
8 the bubble domain material. In order to minlmize the extent of
9 the gradient in the lattice, a small applied bias is preferable
10 in the lattice area. In accordance with FIG. 7, the lattice --
11 constant and bubble domain diameter will change very little if
12 a small field ~b exists in the lattice area. This has the
13 added advantage that the interaction force Fi is slightly
14 smaller (du-e to smaller diamseter); accordingly, it is easier
to move bubbles within the lattice.
16 A sharp slope in bias field at the edge of the
17 lattice is ~efined as one where most of the slope occurs over
18 one lattice constant aO. That i8, two rows of ele~sents are
19 separated by an amount greater than aO, while the rest of
the lattice is of uniform spacing (aO). A gradual slope is
21 one which occurs over several lattice constants, and which
22 stresses the lattice.
23 Generally, the lattice can be stressed up to a
24 point where it deforms plastically. That is, the stresses
should be less than those which would cause irreversible
26 changes in the lattice. For instance, a regular hexag~nal ~ -
~: , . . ...
27 làttice is one in which each element in the lattice has six ~
.
28 nearest neigh~ors each at a uniform distance of aO from it. ;;
.
..;
Y0972-063 -46~
: ~ . : .
.''",'. ~'' :,
j~ "~ . ., ; .: ~,: .:. '
~ , ,
368
1 If stresses are put on thi~ regular lattice, elements will
2 change their regular spacing and the lattice will not appear
3 to be uniformiy hexagonal. When the stresses reach the
4 ela~tic limit of the lattice, it will deform plastically,
and not retain its original uniformity when the stress is
6 removed.
7 T~e elastic limit of a bubble domain lattice is
8 dependent on factors such as the applied bias field Hb.
9 AS an example, local modulation of 30-40% can be used
10 without exceeding the elastic limit for Hb ~ 0.1(4~MS). -
11 As Hb increases, aO becomes larger and the interaction
12 force Fi bPtween domains decreases. This means that the -- -
13 lattice positions are less well defined, and the lattice ~ ;
14 is more susceptible to deformation. As long as the gradient
15 in bias field causes stresses within the elastic limit for --
16 any given lattice, the uniformity of the lattice will be
17 maintained.
18 As an aid to designing systems using lattice arrays,
19 the gradient should be chosen 80 that localized collapse of
bubble domains will not occur due to bias field amplitudes which
21 become too great. Additionally, the bias field should not become
22 so small that run-out of domains into stripes occurs outside
23 the lattice area. If an infinite slope gradient can be
24 provided, no additional force will exi~t on domains within the
lattice. Use of grooves in the magnetic material may lead to
26 very sharp graaients in bias field, as will be explained later. ~-
27 The gradient can extend into the lattice to any extent as long
28 as the local symmetry of the lattice is substantially maintained,
;.:.~, ,' .:
Y0972-063 -47- ` ~ ;
'",.
. ,
,~' " . .
;,; ~ . , . . - ~ : ; ; - -
1 that is, as long as each bubble domain sees a fairly uniform
2 pattern of bubble domains surrounding it. Of course, this
3 criterion is based on the desirability of using a regular lattice
4 having substantially uniform lattice constant throughout. For
certain applications, such a regular lattice-may not be required
6 in which case the gradients can be varied.
7 It should be remembered that the confinement force
8 is the same for a lattice comprising a small number of interactive
9 elements as well as for a lattice having a very large number of
interactive elements, since the interaction force is based on a
11 nearest neigh~or consideration.
.~, . .
12 Repulsive Boundary
- . .
13 A repulsive boundary is one which provides forces
14 tending to repel the interactive element~ 32. For inter- -
active elements w~ich have repelling forces existing between~
16 them, the forces provided by a repulsive boundary will be
17 directed into the lattice area. These forces will generally
. . .
18 be about equal on all sides of the lattice to within about ; ~
- 8HC/~
20~ Structures to provide repulsive boundaries can be ~ -
21~ fabricated from current-carrying conductors, and magnetic
22 materials. Also, changes in the magnetic properties o~ the
23 ~ bubble domain~material can be used. Such changes include
Z4 ~thickness changes and changes brought about by ion implantation,
5~ diffusion, etc. Thus, the anisotropy or magnetization of a
26 magnetic material can be locally altered to provide repulsive
27 forces on magnetic interactive elements supportod by the
28 magnetic~material.
YO972-063 ~ ~ -48
:' . . ,~ -.,
- . ~ - ,.
r ~ . '
~ ~ ' ' " . .. .
1 ~IGS. 13A-13D show some structures which can be
2 used to prov de repulsive ¢onfinement forces. Although each
3 of these structures i_ shaped to confine a parallel-piped
4 lattice, it should be recognized that lattices having any
structures ~6hape) can be confined using these ~ame principles.
6 The lattice area chosen for illustration is that which is
7 easiest to utilize in a practical system. Accordingly, the
8 shapes chosen have advantages of ease of fabrication and
9 ease of access of interactive elements into and out of the ~ -
lattice areas.
11 FIG. 13A shows a ba-qic conductor loop 70 which is
12 - directed along the symmetry planes of interactive element~
13 32 which folm a lattice. Current Ic in conductor 70 will
14 produ~e a magnetic field which exert~ a force to repel inter-
active elements 32 located within conductor 70.
16 In order to compensate for any local variations of
17 of magnetic field produced by current Ic in the regionS72
18 where conductor 70 ha~ its terminàls, an auxiliary conductor 74
19 is provided. Conductor 74 i9 insulated from conductor 70 and
merely serves to provide a uniform magnetic ield along that
:: .
21 edge of lattice area 30 where electrical connections are made
22 to conductor 70.
In ~IG. 13B, a plurality of conductors is utilized
24 which are all on the same fabrication level. In this confinement -~-
structure, oonductors 76A and 76B provide confinement forces for
26 the top and bo*tom of lattice area 30, while conductors 78A and
27 78B provide confinement forces exerted on the lef~ hand and~
28 right hand sides of lattice 30. Becau_e magnetic discontinuities
~ , .
~ ~ Yo972-063 -49~
-', ~ ''
. ~
~,o~L~3~8
1 may exist ir. the four corner_ of lattice area 30, magnetic
2 elements 80 are provided to compensate for these discontinuities - ~'
3 and to ensure that interactive elementq 32 will not be lost from
4 lattice area 30 at the corners of the lattice. Such magnetic
,elements may include, for example, hard magnetic materials which ' '
6 provide forces to repel interactive elements within the lattice. `
7 In FIG. 13B, the arrows 82 represent possible
8 directions of movement of interactive elements into and out
9 of the latti~e area 30.
FIG. 13C shows'another conductor confinement means
11 in which two levels of conductors are utilized. For instance,
12 conductors 84A and 84B are located on the first fabrication , `~' -'
13 level while conductors 86A and 86B are located on a second level
14 of fabrication. Generally, an insulating layer i3 provided
15 between the ~arious conductor levels. ~s with FIG. 13B, the ;-,-
16 arrows 82 represent possible directions of ~ovement of the ,,
17 interactive elements into and out of lattice area 30. '~
18 FIG. 13D represents another confinement structure - ;'
19 which isrparticularly suited for,providing repulsi~e forces ', ''~
20 on magnetic interactive elements such as magnetic domains in , ~;
21 magnetic medium 62. In this embodiment, current-carrying ' '
qq . . ..
-~ conductors 88A and 88B provide repulsive confinement forces - -
,' 23 along the top and bottom edges of lattice area 30. The '',~, ,
, 24 confinement forces along the left and right hand edges of lattice
25 area 30 are provided by magnetically destroying the properties ,'
26 of the material in which the magnetic interactive elements -
27 exist,. That is, shaded areas 90A and 90B are areas of material ,''~,
28 6'2 in whi~ch the magnetic properties of the material required ~ '
:.: - . . .
Yos72-o63 ~ -50-
' '' ' ~ ~ ;','-,'.', '
S ~
~ - : . ~ - .
~Q ~ 3~ :
1 to sustain magnetic elements have been destroyed. Areas 9~A
2 and 90B can extend where conductor 88A and below conductor 88B, '-,
3 if desired. This means that the magnetic elements will not
4 drift into areas 90A and 90B, thereby effecting a repelling
confinement force. In this structure, the magnetic elements
6 move into and out of lattice area 30 in the dlrections indicated
7 by arrows 82.
8 Alteration of magnetic properties in areas ~OA and
9 9 OB can be provided by techniques such as ion implantation ~' ,
10 and diffusion. For instance, such a technique would destroy '~
11 the perpendicular anisotropy of a bubble domain medium where
12 the magnetic elements are magnetic bubble domains. Additionally,
13 the magnetic material can be removed from those areas so that
14 the magnetic elements cannot be sustained there. '
15 In the confinement structures illustrated, there '~
16 may be slight magnetic discontinuities at the corners of
17 the structures which lead to slight changes in confinement
18 force. However, the interactive elements will adjust their '"'~
19 diameters and/or spacing from one another in the area of the
20 confinemenS corners. These elements later will correct ~
21 their relative positions and sizes as they move away from -,
22, the corners.
.
23 Attractive Boundary:
24 FIGS. 14A-14D show various structures for providing - ;- ''-
25 attractive boundaries for use as confinement,means. In these '~
26 examples, the attractive boundary will hold interactive elements
27 32 and the held elements will in turn interact with other ;
28 elements within the lattice to esta~lish a confinement means.
Y0972-063 -51-
. -
..
~ . . .: ~ . . . .
3t~
1 For nstance, FIG. 14A shows a confinement means
2 which is used to retain interactive elements 32 within the
3 lattice area 30. In this case, the confinement means is
4 comprised of conductors,92A and 92B, as well as magnetic ;;
S elements 94. Interactive elements 32 can move into and out of
6 the lattice area 30 by movement in the directions indicated by
7 arrows 82. Movement of the elements across the edges defined
8 by magnetic pieces 94 is also possible.
9 The conductors 92A and 92~ have currents therein
.- .. . . .
10 which establish fields tending to exert forces to hold the -~
11 interactive eIements within lattice array 30, in a manner
12 previously described. The magnetic elements 94~ attract inter- -
13 active elements 32 and hold these elements along the edges of
14 lattice 30 defined by elements 94. The interaction of held
. .
15 elements 32 with other elements within the lattice area serves -~-
16 to confine elements within the lattice. ~-
17 Of course, it should be understood that elements 94
18 need not be m~gnetic if the interactive elements 32 are not
19 magnetic. For-instance, in the case of interactive elements
Z 32 having electrostatic fields associated therewith, app~o~
21 priately charged conductive elements 94 ~an be used to hold
22 the electrostatic elements 32.
.
~3 FIG. 14B shows a confinement structure serving as
24 a boundary of the lattice area 30 which iq entirely comprised ~ -
of discrete elements 94. As previously mentioned, these element~
26 are suitably chosen to be magnetic elements if the interàctive
27 elements 32 are~magnetic elements. In the structure of FIG. 14B,
28 interactive elements 32 can be moved into and out of lattice -; ~
. :
YO972-063 -52-
. . .
. - ,. .:
;
., ,:
' :, :
3~
1 area 30 in a direction substantially transverse to any of the
2 edges defined by the.elements 94.
3 FIG. 14C shows a confinement means using discrete
4 elements 94, as well as continuous elements 96. The discrete
elements 94 serve to hold interactive elements 32 ln the manner
6 previously described. The continuous elements 96 form an entire
7 edge of the lattice area 30 and are used to hold elongated
8 interactive elements 98. As an example, continuous elements
9 96 can be magnetic materials and interactive elements 98 can . .:~
10 be stripe magnetic domains in a magnetic material which are . ,~.
11 attracted to the magnetic elements 96. Stripe domains 98
12 will in tur~ exert repulsive forces on interactive magnetic .
13 domain elements 32 within the lattice, thus providing the
14 confinement function. ;: -~
In FIG. 14C, magnetic interactive elements are more
16 easily moved into and out of the lattice area across the
17 edges defined by discrete elements 94. Therefore, arrows 82 .- -~
~18 indicate the preferred direction of movement of interactive
19 elements 32 into and out of lattice area 30.
. .
20 In the structures of FIGS. 14A-14C, as well as that ~ .
21 : of FIG. 14D, interactive elements 32 which are held by the :.:
. - ,, .
22 attractive confinement means can be made to move away from~:
23 this confinement means when suitable accessing forces
24 are provLded to the interactive elements. This will be
25 appreciated re fully when the accessing structures are ;
: 26 di~cussed. ~
:.
27 FIG. 14D shows another confinement structure,.which
2B in this case uses conductors located on two levels of fabrica~
- 29 tion, similar to tho~e shown in FIG. 13C. Conductors lOOA and
! . . . .
~ YO972-063 -53-
. .
.:
J .~
1 lOOB are located on the first fabrication level while conductors ~ ~
2 102A and 102B are located on the second level of fabrication. - -
3 Currents in these conductors are used to provide attractive
4 magnetic fields for magnetic interactive elements 32.
FIG. 14E is used to explain the opera`tion of conductor
6 confinement means, such as those shown in FIGS. 13C and 14D.
7 FIG. 14E i~ strates a current-carrying conductor 104 having - ;
8 a current I directed into the conductor. This current ~;
9 establishes a magnetic field surrounding the conductor which
has components Hx and Hy. The component Hy is a component
11 directed substantially normal to the medium in which magnetic
12 elements 3~ exist. In the case of bubble domain interactive
13 elements, th3 component Hy would be directed parallel to an
14 easy direction of magnetization of the magnetic material in ;~
which the domains exist (that i9, Hy is directed along the
16 direction of magnetization of the bubble domains).
::.: . . .
17 The plot of Hy versus distance x along the conductor ; ~ ~-
la illustrate~ that the component Hy is positi~e at one side of
19 the conductor and negative at the other. Conse~uently, a
20 bubble domain located near conductor 104 will experience an ;
21 attractive force or a repulsive force depending on its location ~;
22 with respect to conductor 104. If the bubble domain experiences
23 a gradient in the field H , a force will be exerted on the
. Y
24 domain tending to move it. Domains will move in the direction
25 of decreasing net bias field. ~:
26 For instance, if the bubble domains have ma~netization
27 directed upwardly (in the direction of + H ) and are located ~`
y . . :,~.. ,.. ,,.,.. ;. .-
28 to the right of the positive peak of Hy they will be m~ved - ~
: - ': ' '' ~ . ." ' , ,
YO972-063 ~ -54-
. ~ ' ''~ '
: ,, - -
1 further to the right when the current I flows through con- ;
2 ductor 104. If these same domains are located in po~itions be-
3 tween the positive and negative peaks of the field Hy they
4 will move to the left when current I i9 present. Further, if
5 these domains are located to the left of the left-hand edge ,~
6 of conductor 104, they will be attracted to the left-hand ;
7 edge of the conductor when current I flows in conductor 104~
8 Thus, establishment of proper current direction in
9 conductors 1~0 and 102 (FIG. 14D) will cause attractive
magnetic forces for holding elements 32 along the outside
11 edge of lattice area 30. These pinned outer elements 32
12 will provide the necessary forces to contain other eiements
13 32 in the interior of lattice area 30. ``
14 Various confinement means have been shown which
utilize conductors, magnetic materials, and regions of the
16 supporting medium which have their properties locally altered. ;~
17 These various means can be used together, as is illustrated by :~
18 FIG. 15. Here, an embodiment particularly suitable for magnetic
19 bubble domains is shown in which the magnetic medium 62 (FIG. 6)
has a groove 106 in it. A spacer 108 separates bubble domain
21 medium 62 and the overlying conductor 110 which i~ used to
22 provide confinement forces. Spacing the conductor from medium
. - . ~ - -.
23 ~ 62 provides a more uniform magnetic field distribution. As is ;
24 apparent, the peripheral domains 32A are magnetically attracted ~l
to the groove 106 while the domainY 32B within the lattice
,
26 are further confined by the action of domains 32A. Of cours~,
27 the groove itself may have sufficient attractivc properties to
28 provide the ~onfinement means while the conductor 110 i~ used
1' - :
i ,
YO972-063 -55-
,. . . '. '
~o~
1 to access domains into the lattice from the area of the bubble
2 domain medium 62 to the left of groove 106.
3 T~e thicknes~ of the magnetic elements used for
4 attractive boundaries is quite arbitrary; therefore, a magnetic -
element can be very thick.` When magnetic elements are located
6 on a boundary of the lattice across which interactive elements ~;
7 are to be moved, its thickness is chosen such that the attractive '
8 force exerted for confinement will not be greater than that
9 which can be overcome by the input means 38 (FIG. 2~ used to
move elements ~2 into and out of the lattice.
11 Uniform Bias Throughout-the Magnetic Material `
12 The bias field Hb can be applied uniformly through-
13 out the magnetic bubble domain material. However, it must be ~
14 remembered that Hb adds to the interaction field Hi within ~ -
lS the lattice, so that the net bias field within the lattice will
16 be greater than that outside the lattice. The net bias field
17 ~ithin the lattice should not be so great that collapæe of
18 bubble domains within the lattlce begins to occur. Additionally,
19 Hb should not be so small that bubble domains outside the lattice O -
20 tend to run out into stripe domains. In general, Hb is chosen -~
.
21 so that bubble domains outside the lattice are close to run-out
22 width and that operation within the lattice will not cause `
23 localized collapse. -
.. , ~ .
24 Generally, ~he applied bias Hb is approximately
25 ~ N2 + 1/4 (Ho - H2),~where
26 H2 is the run-out magnetic field,
27 Ho i~ the collapse magnetic fiald.
YO972 063 56
.:
: :
- " . .
:. . .;: .
: : .. ,,.. '
. :
~t~ 8
1 For very high density packing in the lattice, Hb is approximately
2 H2 (this allows Hi to be large in the lattice without causing
3 localized collapse of domains).
4 Having a uniform bias field across the entire
bubble domain material is easily achieved with simple structures.
6 In addition, the interaction forces Fi between bubble domains ~ -
7 are less when an applied bias field is present. This in turn
8 makes it easier to move domains into/out of the lattice. -~;
9 In FIG. 16A, bubble domain material 62 has located
adjacent to i~ permanent magnets 112 and magnetically soft
11 members 114 (such a~ permalloy) which extend over the
12 entire area of material 62. Members 114 provide praferred
13 paths for the magnetic flux from magnet~ 112, and cause the
14 magnetic field lines normal to material 62 to be uniform
throughout material 62.
16 In ~IG. 16~, an exchange-coupled layer 116 is in
17 contact with the entire surface of bubble domain material 62.
18 Layer 116 is comprised of a magnetically hard material and
19 acts to provide a uniform bias field over the material 62.
Use of exchange coupled layers as bias layers is described
21 in U. S. patent 3,529,303. A~ an example, SmCo5 is a suitable
22 exchange-coupled layer on orthof~errite materials. When the
23 bubble domain material is a garnet film, a 5pun garnet film
24 can be used as an e~xchange-coupled layer. For instance, spun ~ -
Gd3Fe5012 can be~uJed as a exchange-coupled material on
26 (EuY)3(GaFe)50I2 bubble domain films.
27 In FIG. 16C a current-carrying coil 118 surrounds the ~ -
,.
28 bubble domain material~62. As is well-known, this coil
Yo972-063 ~57~
.. . ~
- ',.:. ::~ :
.. .. .
' ~ ' ''. :-',
,
produces a ~niform magnetic bias field across material 62.
These bias structures can be used in various combinations
and can be used with the mean~ shown previously for confining
domain~ within the lattice area 30.
Maqnetic Bias Inside and Outside Lattice Area
As stated earlier, various magnetic bias conditions
can be utilize~ when the interactive elements 32 are magnetic
bubble domains. The applied bias field Hb can be zero or a
small value within the lattice tincluding even negative values),
while outside the lattice the applied bias field is adjusted
to prevent bubbles from running out into stripe domains.
Generally, the bias field outside the lattice area need
only be present where it is desired to perform sy~tem
functions, such as writing, reading, etc. Various structures
may be employed to obtain greater applied bias fields outslde
the lattice area 30 than within the lattice area. Interacting
bubble domains within the lattice tend to bias one another
to provide ~tability.
FIGS. 17A and 17B show structures for ~uitably
providing magnetio bias fields outside the lattice. In
FIG. 17A,` the bubbie domain medium 62 has located thereon
exchange-coupled layers 120 which are comprised of magnetically
hard material. Thus, layers 120 act as permanent magnets to
provide bias in the regions of bubble domain material 62
outside of lattice area 30, in t~he same manner as previously
described with respect to FIG. 16B.
Another suitable structure for providing magnetic
bias outside the lattice is shown in FIG. 17B The bubble
~,~3~ 3~
1 domain material 62 has permanent magnets 112 located around
2 it. Additionally, magnetically soft members 122 (~uch as
3 permalloy) are used to provide preferred paths for the
4 magnetic flux from magnets 112. Members 122 extend to the -
area 30 where the lattice is located, so that an applied
6 bias field H~ exists outside the lattice but not within it.
7 Of course, there may be some gradient of field Hb within portions
8 of the lattice area, but this gradient can be utilized in the
9 design of confinement structures as was indicated earlier.
Also, if the ield Hb extends somewhat into lattice area 30,
11 the lattice constant aO will not change greatly if the gradient
12 is not steep.
13 In additi~n to the use of exchange-coupled layers ~$ r
14 and permanent magnet~, current-carrying conductors can be used
15 to provide the field Hb outside the lattice area. Design ~ -
16 of useful conductor patterns for this purpose i8 known to
17 those of skill in the art.
' '' ' '
18 Accessing Elements Within the Lattice
19 As~explained previously with reference to FIGS. lOA `~
and lOB, interactive elements 32 are generally (but not necess-
21 arily) moved into and out of the lattice area in amounts corres- - ;
22 ponding to full rows or columns. Of cour e, for a lattice
23 comprising a single row or column of interactive elements,
24 only a single element need be moved into and out of the
....
lattice at any one time.
26 Th~ force required to move an interactive element
?7 ~ into the lattice area is that which overcomes the repulsive
28 force of interactive elements within the lattice. If
.
. - ' . ~ :
YO972-063 ~ -59-
. :.
..
. ~:
~ , ."~ '
~3~
1 there are no elements within the lattice those elements
2 entered into the lattice will spread out in order to minimize
3 the energy of the lattice. Consequently, interactive elements
4 are continually loaded into the lattice until a number is ''
reached which will provide a regular lattice having a given
6 lattice spacing aO. For instance, m columns having n elements
7 in a column may be placed in the lattice. After this, the
8, lattice can be perturbed by further input elements in order ,, ;
9 to remove any dislocations or vacancies from the initially
10 formed lattice. That is, after the lattice is initially formed, "' -~'
11 new rows or column-c of interactive elements are entered into ~
12 the lattice and a corresponding number removed from ,the lattice. ;,-
13 This ensures that all dislocations and vacancies will have
14 translated through the lattice area and will be removed.
15 This operation may take one or more cycles in which the lattice / ' ,
16 is totally recycled. , ,~
17 An alternate technique for achieving an initial ,`,~' , ''
18 lattice of magnetic bubble domainR is to first apply a large
19 in-plane magne-ic field to saturate the magnetic medium. '~
20 After this, the magnetic field is released to obtain a dense ' ''
21 random array of bubble domains. The lattice lS then magnetically
22 annealed by a time modulated magnetic field normal to the bubble
23 domain material to obtain a regular lattice. , , ',
24 Another technique for providing an initial lattice '" ,', '
25 generates bubble domains at selected locations in a magnetic ' ''~
26 sheet. For instance, a permanent magnet having'a pattern of
27 apertures in it can be brought into close proximity to the
2~3 magnetic sheet, after the sheet has been heated to above its '
Y0972-063 -60~
,.'":' ~'.~
. ~ -
. .
.
3~
l Curie temperature Tc. This will cause nucleation of bubble
2 domains in ~he magnetic sheet at locations corresponding to
3 the pattern in the permanent magnet.
- 4 Still another technique for providing~an initial
lattice involves cutting of stripe domains. A pattern of
6 stripe domains is produ~-ed by a magnetic field in the plane
7 of the magnetic sheet. The stripe domains are then cut to
8 provide rows of bubble domains. The cutting device is any `.~"t:
9 device which produces a magnetic field of sufficient amplitude
in a direction sub~tantially normal to the magnetic i~heet.
11 As an example, a recording head can be mcved across the stripe
12 pattern in sequential fashion to cut the stripes thereby
13 producing rows of bubble domains.
14 The force required to enter elements into the lattice
or to remove elements from the lattice is a force ~hich over-
16 comes the energy barrier between the lattice and the region
17 outside the lattice. This force depends on the amount of
18 separation~between elements inside and outside o~ the lattice
19 and is chosen so as not to materially disturb the lattice
properties. That i3, the input and output operations
21~ elastically deform the lattice but the amount o~ deformation -
22 is ufficiently small that the lattice can relax to its -
- 23 initial uniformity when the force is removed.
~':
, ~ . : ., .
-. -.
YO972-063 -61- ~ ., ~.: '' :
, : ,". ~
- , .: .
.~ . ~ , . : : .
.
~, ~ ' . -
`! , ~ :
, .
3~
.,.
1 In the case of magnetic bubble domains, the
2 gradient in magnetic field used to provide the input force
3 on the domains to be injected is properly chosen so that ~-
4 these domain~ will not be collapsed before they can be ; -
~ moved into the lattice. Additionally, domains within the
" ~ ~.. .
6 lattice may experience forces due to the input operation
7 and these domains also should not be collapsed by this
8 operation. It must be remembered that the domains within ~ -
9 the lattice are subjected to the applied bias fie?d Hb and
10 the interaction field Hi; therefore, an additional term due ~ -
11 to the magnetic field used to access domaino must not create --
12 forces on domains already in the lattice greater than those
13 forces which would collapse them. ~-~
14 In general,
(~) Hb + ~Hd + Hi ~ Ho~
16 where
17 Hb is the applied bias field,
18 ~d is the drive magnetic field for acces~ing
.. . ..
19 domains into and QUt of the latti¢e area,
Hi is the interaction field existing between
21 bubble doma~ns, and
22 Ho is the collapse magnetic field for the bubble
23 domains.
24 Transposing terms,
? ' .-
(3) ~Hd < Ho - Hb Hi-
,'.,'''' :'
YO972-063 -62~
.j.
: -:
' ' ' ""
36~ :
1 From these equations, it is apparent why it is desirable
2 to operate close to run-out when isolated bubble domains
3 are used outside the lattice area.
4 Removing interactive elements from the lattice area
is similar to the input operation where elements are moved into
6 the lattice area. Basically, the operation is the reverse of
7 the input operation and elements within the lattice are moved
8 over the energy barrier which confines the lattice. -
9 For domains within the lattice, the total z-field
Ht acting on them must be greater than run-out and less than
11 that which would cause collapse. The driving force to move
12 domains is the gradient in z-field across the bubble domains.
13 This force must be sufficient to overcome coercivity effects.
14 The total z-field Ht is given by -
(4) Ht = Hb + Hd + Hi
16 whers Hb = applied field ;
17 Hd = drive field
18 Hi = interaction field
19 ~ If x is a di~tance measured along the direction of motion of
a domain and d i~ the domain diameter, then the driving force -~
.
21 on a bubble domain is ~ ~
. .
.
22 ~ d ~ ~ - Hc
. .
23 in order to be able to move the domain. With a uniform field
, . :.
' YO972-063 -63- -~
,' ., . -
s . . . ., ' ~: '
;, ~- ` :
. . -
,, , . ~ ,
~': ~ ' , ~.,
,
.. ~ .
1 Hbr axb = 0. The derivative axl always produces forces `-
2 tending to push bubble domains out of the lattice. Hence
aH -
3 aXl aids in removal of bubble domains from the lattice and -
4 hinders injection of bubble domains into the lattice.
As will be re fully apparent, the ~ame kind of
6 structures can be used to move interactive elements into the
7 lattice and to remove interactive elements from the lattice.
8 Although it may be preferred to provide structure for pushing
9 elements into the lattice and other structures for pulling
elements out of the lattice, such push-pull operation is
11 not the only mode of operation which can be used. The timing
12 of the~e two operations i~ not critical and they need not occur
, . .. .. ..
13 simultaneously. However, the input and output operations can
14 be suitably accomplished at the same time.
lS When a reasonably large lattice is used, an extra
16 column or rcw of interactive elements can be accommodated
- 17 in the lattice. However, if the lattice constant
-.
18 between columns or rows changes by`greater than about 10~
19 some perturbation of the positions of the interactive elements
within the lattice may occur which may be trouble ome in a
21 practical operation. As long as the lattice constant does
..
22 not change appreciably, timing of the input and output operations
23 is not critical. The timing is generally a function of the
24 size of tho lattice and the elasticity of movement of interactive
elements within the lattice. It should be remembered that
26 only one lattice shift (either a column or a row) occurs for
.
,' :'.-.
Y0972-063 -64-
.
:
.
~ ~ .
1,Z~ L,~
1 many time c~cles of movement of the elements by the
2 propagation structure outside the lattice area. Consequently,
3 the speed of movement of elements within the lattice does
- 4 not have to be fast.
Depellding upon the size of the lattice, coercivity,
6 etc., it may be possible to insert a row or columZn of inter-
7 active elements into the lattice and have the disturbance
thereby produced transmitted through the lattice to e,Zllect - -
g an output row or column of interactive elements from the
lattice. The size of a lattice through which sufficient
11 forces can be transmitted to cause a group of elements
12 within the lattice to be ejected from the lattice is limited ~ -
13 by damping processes and terms such as the coercivity of the
14 material (if magnetic domains are used). The force required
to overcome damping should not be so great as to cause collapse
16 of magnetic interactive elements along the input edge of the ~ ;
- 17 lattice array. ~;
18 Due to damping and the other quantities mentioned,
19 the force propagated in the lattice when interactive elements
are entered into the lattice may decrea~e with distance into ;~
21 the lattice array. When the input force has diminished
22 such that it is less than the coercive term in equation 1,
23 the next column of elements in the lattice will not be moved
24 since the remaining force cannot overcome coercivity teZnding-
to prevent mo~ement of these elements. The energy will then
26 be stored in a distortion of the lattice.
. .
27 To estimate the number of rows n which can be moved
- :"
2~ by the input force, consider that the gradient required to
YO972-063~ -65- ;~
.
1 overcome the coercivity of a simple isolated bubble is HC/d,
2 where d is the bubble diameter.
3 Now, suppose that this bubble interacts with (n~
4 other bubbles, all in a linear chain, in such a~way that
they all m~ve simultaneously. If the field gradient aH is
6 applied only to the first bubble in the chain, its value
-7 must be
8 d~ = n HC/d
9 However, the total difference in H across the bubkle
domain diameter cannot exceed (Ho - H2) where Ho is the
11 collapse field and H2 i9 the run-out field. Therefore:
, - . :.
12 dH ~ Ho H2 nHc
:.~'- ' - .
13 Hence
n ~ (Ho ~ H2)/Hc
~ . ;:
FIG. 18 ~hows an end view of a structure suitable
16 for confining interactive elements within the lattice and for
17 moving interactive elements into the lattice area. In this
18 drawing, the interactive elements 32 are illustratively magnetic
.
19 bubble domains, although other types of interactive elements
could be used as well. Additionally, the structure shown here
21 can be used to remove domains from a lattice by reversing the
-.
~ 22 operation to be described.
-
i YO972-063 -66-
~ '' ' ~ . .
.. . . .
' :.
~g~3~8
1 In more detail, the magnetic bubble domain material
2 62 has on one surface thereof a spacing layer 124 (~uch as an
3 insulator) over which is located conductors 126 and 128. The
4 spacer is used to make the field from conductors 126 and 128 ~ -
more uniform, and may not be essential for operation. Current
6 in conductor 126 establishes a confinement force for magnetic -
7 bubble domains within the lattice, as described previously. -
,
8 Additionally, it is to be a portion of the input means 38
9 tFI~. 6) used to inject bubble domains into the lattice array. ~;
In this drawing, the lattice 30 comprises an area of the
11 magnetic medium 62 to the left of conductor 126. FIGS. l9A -
12 l9E show the operation of input means 38 for moving magnetic
13 bubble domains 32 into the lattice. From these drawings, it ~ -
14 will be readily apparent that reversal of current polarities
in conductors 126 and 128 will move bubble domains 32 in the ;
16 opposite direction, thereby achieving the output operation from
17 the lattice array 30. Since FIG. 18 and FIGS. l9A - 19E are
. .
18 views which are rotated with reference to each other, move-
19 ment of domains 32 upwardly in FIGS. l9A - l9E corresponds to
movement of domains into the lattice area.
21 In FIG. 18, and FIGS. 19A - l9E, the domains which
22 are within the lattice area are numbered 32A, while domains which `
23 are going to be moved into the lattice area are labeled 32B and
24 - 32C. still further, it should be recognised that a plurality of
domains can be entered into the lattice~area at the same time
26 or removed from the lattice area at the same time. For in~tance,
27 FIG. l9A show~ domains 32A and 32A' which are in the lattice and
YO972-063 ~ -67- ~
1 -' " '
1 ~ -
I ~
,, , : ~. ~.-
.~
.. . . .
&i8
. .. :.
1 domains 32B, 32B' which mo,ve in unison. Domains 32C and
2 32C' also move in unison. The second column of domains
3 (32A', 32B', 32C'~ is not shown in FIGS. l9B - l9E, for ease
4 of illustration. The current pul~es in conductors 126 and
128 are indicated as Il and I2, respectively, and the direction - -
6 of movement of the domains i3 indicated by arrow 130. For
7 ease of illustration, the column of domains 32A', 32B', and
8 32C' are not shown in FIGS. l9B - l9E.
9 To aid in understanding the mc~vement of domains as
illustrated in FIGS. l9A - l9E, reference is made to FIG. 14E,
- 11 and the descriptive matter concerning that figure. Currents
-12 in the conductors 126 and 128 will produce magnetic field gradi- ~
3~ ents acting on the domains which will move the domains in desired ~-
14 directions. Additionally, the domains may exert interaction
forces on one another which also aid movement in de~ired direc~
16 tions. -
17 FIG. 19A illustrates the positions of domains 32A,
18 32B, and 32C at time T = 0 when currents Il and I2 are present
19 in conductors 126, 128 respectively. At this time, domain
32A is stationed at the left hand (upper) edge of conductor
21 126 in a position bordering the domains in the lattice area.
22 At time T = 1, currents Il and I2 exist in conductors 126 and
23 - 128, respectively. The~-~e currents cause a combined magnetic
24 field~between the conductor~ causing domain 32B to center
between the conductors. The movement of domain 32B exerts a
26 repulsive force on 32A, causing it to move into the gradient
27 field outside (above) conductor 126. This gradient field causes
28 domain 32A to move further into the lattice.
` YO972-063 ~ ~ -68-
,.
., ~ . . :
., .
:ii .
.i ~ . ... .
1 At time T = 2 (FIG.19C) the current I2 reverses
2 and domain 32C moves upwardly due to an attractive gradient
3 field produced by current I2 in conductor 128. This creates
4 an interaction force on 32B, causing it to move partially
under conductor 126. At the same time, domain 32A moves
6 further into the lattice area due to the gradient field
7 produced by current Il in conductor 126.
8 F~G. l9D shows the position of the interactive . ,
9 elements at t'me T = 3. The direction of current Il has re- ~:
' versed, creating a n attractive gradient magnetic field for
11 domain 32B. When domain 32C moves further under conductor
12 128 due to a repulsive force from the domain following it ,~ '.
13 (not shown), it causes domain 32B to move into the attractive
14 gradient car,sed'by current Il in conductor 126. Hence 32B
moves under conductor :126 to a position centered on the top , :~
16 edge of this conductor (FIG. 19D). ~ : ,
17 ~ FIG.. 19E illustrates .the position of the interactive
18 elements~32aat.time T = 4. The,direction of current I2 has ' :
19 . again.reversèd~and current Il i~ in the same~direction. Intcr-
active element 32C now experiences an attractive`force pulling I ' '
21 it in the direction of arrow 130 and moves to a position
22 ~under the top edge of conductor 128. Int-ractive element~
23 32B and 32C do not cxperience a push from element 32C or a '.~
24 gradi.ent.magnetic field so they remain approximately in :-. ~:'
.
25 : thelr same p~sitions. These are the position~ corre3ponding .;~::
Z6 to elements'32A and 32B of FIGS. 18 and ,19A. That is, the :.
Z7 situation at time T = 0 is oncé-again obtained. During the .,~
'28 next'cydle o'$ operation, an additional row of domains will .
YO972-063 -68A-
~ .
1 be entered into the lattice area, in the same manner.
2 An example of the accessing operation described
3 with respect to FIGS. 18 and l9A - l9E i~ the folIowing~
4 Bubble domain~ of diameter d can be used as interactive ,~
elements 32 for movement into a lattice area by conductors
6 having a width of about aO/2, center to center spacing of about
7 aO, and a thickness of about 1/2-1 micron. The current '
8 amplitudes in these conductors can be about 30-50 milliamps
9 and the pulse duration of these currents can be about 0.5
microsecond~. These magnitudes provide magnetic field~ which
11 are sufficient to overcome the coercivity of the bubble
12 domain medium. '
13 FIG. 20 shows another ~tructure ~uitable a~ an
14 input means 38 for magnetic bubble domains used as interactive
elements 32 in lattice array 30.
16 The magnetic bubble domain sheet 62 haæ a coating
17 of insulation 124 thereon, over which is located conductors
18 126 and 128. The magnatic medium 62 has a groove 132 which ~ ' '
19 serves as the confinement means for the magnetlc domains 32
20 ' within lattice area 30. Consequently, domains moving into and ~,
21 out of the lattice area 30 pass under the groove 132 and have
2'2 reduced height while under the groove. The groova can be a -
23 physical groove in the surface of the material 62 or can be
24 ~ a magneticaIly altered region of material 162. As mentioned,
this can be achieved by techni~ues such ,as ion implantation
26 ~and diffu~ion.
27 The operation of input means 38 in FIG. 20 is the, ' ;'
. . -: -
, 28 same as that illu~trated by FIGS. 18 and l9A - l~E. That is, , - ~'
. .
- ~ . . . .
, ~ Y0972-063 ~ '-69~
.
- . .
, ' ~ .",. '. ' :'. .: '
~_~L~ 3~
1 suitably directed currents in conductors 126 and 128 cause
2 domains 32' to be moved into lattice area 30. If these ;
3 current sequences are reversed in polarity, domains 32 within
4 lattice 30 will be moved to the right, out of the lattice area.
FIG. 21 shows a structure which was used to move
6 magnetic bubble domains 32 into and out of a lattice area 30,
7 thereby illustrating the principle of access of elements to
8 and from a lattice. The lattice area 30 is illustratively a
9 regular lattice bounded by confinement means 34, which in
this case is ~ conductor having currents flowing in the -
11 directions indicated. Bubble domains 32 were moved in rows
'. - - . ~ .
12 into and out of the lattice with the same relative positions,
13 thereby assuring the preservation of information by the system,
14 - and further demonstrating the principles of l~attice operation.
15 - ~Three conductors A, B, and C are located~at the
16 top of~the lattice while corresponding conductors A', B' and
17 C' are located at the bottom of lattice 30. The~channel between
1~ conductors A and B is shaded to indicate that this ~hannel is
19 a shift register for movement of domains along the dhannel
prior to their entry into the lattice area or exit from the
21 lattice area. In the same manner, the channel between conductors
22 A' and B' defines a shift register in which the bubble domains
.
23 can be moved before or aftor accessing into or out of the lattice ;~;
24 30.
FIGS. 22A - 22G illustrate the oporations of injecting
¦ 26 a bubble domain into lattice 30 and removing a bubble domain
27 ~rom the lattice 30. A plurallty of time cycles T = 1, 2, ...... , 7 ;;
. ..
28 is shown in which the current directions in the cond~ctors A,-
29 B, C, A', B', and C' are indicated by the arrows on the
conductors.
For instance, at time T = 1 (FIG. 22A) the bubble ~
. . ' ' :.': .
: ' '. ':' "'
YO972-063 ~7~
- ; ,
L3~Y~
domain 32 is to be injected into lattice 30 while the bubble
2 domain 32 ' is to be removed from lattice 30. Currents exist
3 in conductors A and B in the direction indicated by the arrows.
4 Additionally, currents exist in conductors A' and C'.
A~ time T = 2 (FIG. 22B) a current is present in
6 conductor C and also in conductor A. This causes bubble domain
7 32 to move to the edge of conductor C. Ouring time T = 2,
8 currents exist in conductors A' and C'. Thus, bubble domain
9 32 ' moves to the bottom edge of conductor C'.
FIGS. 22C and 22D show the next sequence of pulses
ll in the conductors. These pulses create magnetic fields which
12 move bubble domain 32 into the lattice and bubble domain 32' -
13 out of the lattice into the shift register area between conductors
14 A' and B'. Thus, FIGS. 22A - 22D illustrate the injection of a
bubble domain into lattice area 30 and the ejection of bubble
16 domain 32 ' from the lattice area 30. ~;
17 FIGS. 22E - 22G, together with FIG. 22D, illustrate -
~:
18 the reverse operation of the structure of FIG. 21. In these
19 figures, bubble domain 32 is to be moved from a position within
the lattice to the shift register area between conductors A and B.
i ~-
- - 21 Also, bubble domain 32 ' is to be moved from the shift register
; 22 area between`conductors A' and B' to a position within the lattice
23 30. Operation in the time period T = 4 to T = 7 inclusive is
- .
24 shown in these~figures and is readily understood from the prsvious
discussion.
26 The discussion of FIGS. 21 and 22A - 22G has assumed
27 that the interactive elements 32 are magnetic bubble domains.
28 , However, this structurs can be used on any type of interactive
: . :,
~, :
: .
~ ~ ~ YO972-063 -71- ~
, ~.: ~ , , . :
. J ~ ' ' . .:
~ g.~L3~
1 elements and illustrates the movement of such elements into and
2 out of the lattice area 30. In the particular case where the -
3 interactive elements 32 are magnetic bubble domains, operation
4 of the structure of FIG. 21 has been demonstrated. For instance,
the operation denoted in FIGS. 22A - 22G was demonstrated
6 using as a garnet magnetic bubble domain film
7 Y2 35Euo 65&al 08Fe3 92~2' which was grown by liquid phase
8 epitaxy on a suitable garnet substrate. The bubble domain
9 garnet had a thickness of 4.8 microns and the bubble domain
diameter wa-~ about 5 microns. A uniform bias field Hz of
11 approximately 80 Oe existed over the entire garnet film. The
12 conductors A, B, C, A', B', and C' were 4 microns wide and
13 1.5 microns thick. A current of 20 milliamps flowed through
14 the confinement means 34. The amplitude of currents fiowing
in the transfer conductors A, B, C, A', 3', and C' were
16 as follows:
17 Current in conductors B, C, B', C' had
18 amplitudes of 50 milliamps.
19 Current in conductors A and A' had amplitudes
of 25 milliamps. The bubble domain magnetization was directed
21 upwardly out of the paper in the illustrated operation.
22 Additionally, the dimensions of the lattice array and spacing `;
23 between conductors-for this operation are noted in FIG. 21.
24 The pulse durations for movement of the domains are ~
not critical, and depend on how fast domains will move in the - -
26 bubble domain material. For instance, pulses of duration
.. ..
27 about 0.3 microseconds or greater are suitable for many rare- -
28 earth iron garnet materials.
': '
YO972-063 -72- ;
.
.
. . .
~: .
! ~
~ . . , , `
1 As mentioned, coercive force exists in bubble
2 domain materials tending to resist movement of magnetic
3 domains. Various techniques exist for overcoming the coercive
4 effects within the material to provide more mobile magnetic
bubble domaino-. In particular, this will increase the value
6 of n (number of bubble doamins which are moved by an input
7 force) which was derived previously. FIGS. 23 and 24
8 illustrate techniques which can be used to aid the movement of
9 magnetic bubble domains.
In FIG. 23, the magnetic bubble domain medium 62 is
11 surrounded by a current-carrying coil 134. Current pulses are ~-
12 produced in coil 134 which provide an ac bias field substantially
13 parallel to an easy direction of magnetization in magnetic
14 medium 62. If desired, pulsed dc or rectified ac current
pulses can be-applied to coil 134. These current pulses in
16 coil 134 produce a magnetic field which tends to oscillate
17 magnetic domains in material 62 without collapsing them. ~ -
18 oscillation of these domains enables them to go into a minimum
19 -energy configuration which is the regular lattice previously
20 ~ defined.
21 me oscillating bias field is a field approximately - ~
22 equal to the coercive force term in force equation 1 previously ~ -
23 discussed. That i8, the force exerted on the bubble domains by
24 ~thè oscillating fleld is sufficient to overcome coercivity
- ~25 to cauee a ~mall oscillation in bubble domain diameter. An
26~ ~ oscillating field of amplitude about equal to ~ is sufficient.
, ., ., . c
~: .
1-
~ Y0972-063 ~ -73- -
!
- ,, ~ .. . ; ~ :
,- . , .: . : - - ,. : -~
- . .- :
J ~-. , . . - .
-
~l~f~3~
1 The ac bias field is generally designed to be uniform across
2 the lat~ice area.
3 FI~. 24 shows another technique for facilitating
4 movement of magnetic bubble domains in magnetic material 62.
In particular, this means i8 suitable for aiding movement of
6 magnetic bubble domains in the lattice area 30.
7 In more detail, a lattice "sweeper" is comprised of
8 a conductor, such as 136 and/or 138, whose width is generally
9 at least as great as the bubble domain diameter and up to sev- -
eral bubble diameters. Current Is through these conductors
11 will create magnetic fields which aid the movement of
12 interactive elements 32 in magnetic medium 62. The magnetic
.
13 fields established by current Is serve to break up the lattice
14 area into smaller portions so that the elements in those
portions will move more easily. The maximum magnetic field i
16 produced by conductors 136 and 138 should not be great enough~
17 to cause collapse of any of the domains in the lattice. A
18 magnetic field approximately equal to the magnetic field used
19~ for providing the injection forces to move domains into the ;
lattice is sufficient. In general, the limits on the magnetic
- : . .....
21 fields produced by conductors 136 and 138 is the same as the
22 limits previously expressed. That is, the magnetic fields
23 produced by~these conduc~ors preferably should not lead to
24 serious per~urbations in he lattice (in order to retain
information in correct positions) or to collapse of any domains
26 within the lattice).
, .::
27 The discussion concerning ac~ess of information to
YO972-063 -74-
~j . .. ..
.t, -' . : ' ', ' ' ~, ' ' ~,;
3~
l and from the lattice area i9 general for any type of interactive
2 element 32. When considerations of bias field are discussed
3 in terms of changing diameter size, the discussion is directed
4 to magnetic bubble domain interactive elements.~ However, the
principles directed to the forces required to move interactive
6 elements and the structures provided for so moving these elements
7 can be utilized with o her types of interactive elements, as
8 will be more particularly apparent later. For instance, a -~
9 lattice sweeper a~ shown in FIG. 24 can be used with any type
- lO of magnetic interactive elements, and can be modified for use
ll with other types of interactive elements.
12 Transfe~:Isolated Elements--Interactin~ Elements
13 This section will be particularly concerned with
14 movement of interactive elements into and out of the lattice -
in a manner which takes into account the change in interactive
16 force between the elements. In particular, problems associated
17 ~wi~th bubble domain interactive elements will be discussed. `
18 For Lnstance, operations on bubble domains outside of the
-l9 lattice area may require that the bubbles be isolated, i.e.,
20 ~ 1solated in the ssnse that interactions between the bubbles do ~ -
21 ~not materially affect their positions with respect to one
22 another. In contrast with this, bubble domains within the
23 lattice have positions determined subsitantially by the inter-
24 active forces existing between them.
Another related problem where bubble domains are the
26 interactive elements within the lattice concerns the-bia~ field
YO972-053 ~75-
,
, . . - ;
- ~ .
~, '~' ' , ':
.
3~
1 which is exerted on the bubble domains. AS mentioned previously,
Z outside of the lattice the applied bias Hb generally increases
3 to support isolated bubbles while in the lattice the applied
4 bias is minimal since there is an interactive magnetic field
bias Hi due to the bubbles themselves. Of course, a uniform
6 applied bias field can be applied over the entire magnetic
7 medium. Additionally, the magnitude of the applied bias field
8 Hb can be greater outside the lattice area than within the lattice ~ -~
9 area such that the net bias field Hz = ~ + Hi is approximately
uniform over the entire magnetic material. ;`
11 Various structures can be provided to move the
12 interactive el,ements from positions where they are essentially
13 isolated to positions where they interact strongly. FIGS. 25 ~ ~
14 and 26 illustrate one particular structure which utilizes ~ -
geometric f an-in and fan-out. These structures will be explained
16 particularly with respect to bubble domain interactive elements,
17 but i* should be understood that other types of interactive
18 elements can be brought into the l~ttice area and removed from
19 the lattice area using the same basic structure.
FIG~ 25 illustrates the principle involved. The
21 ~ interactive elements 32 are spaced at a distance 4d when they
22 are iso}ated from one another. These interactive elements are
23 ~ to be brought to lattice area 30 where their spacing is going
24~ to be approximately 2d. The total amount of force which is
required to bring the elements 32 from positions where they are
26 isolated to positions within the lattice is approximateIy the
27 amount of the force required to overcome the barrier forces
Yo9?2-063 -76- -
,
.
-
, ~ . . ,., :
LL~3b~
1 - when providing input and output of elements from the lattice~
2 However, the transition from isolated positions to interactive
3 positions occurs over a length of space S so the change is more
g gradual and the holding forces on the interactive elements as
they move through the distance S are only fractions of the
6 amounts needed to move the elements into the lattice area.
7 This distance S is arbitrary and is determined merely'by the
8 -smoothness of the transition desired.
g If bubble domains are the interactive elements, their
size will change as they move toward the lattice 30 unless the
11 net bias field Hz remains relatively constant as they move
12 toward the lattice area. FIG. 26 shows the magnetic fields
13 ~ which approximate this. The applied bias field Hb decreases
14 as the bubble domains get closer to the lattice area. ~his
.- :
compensates for the increase in interactive magnetic bias ` --
16 field Hi, caused by the bubble domains getting closer together
17 as they move toward the lattice. ~If desired, an applied
' 18 bias field Hb in the area outside the 'lattice can be provided ~ -
, ~ .
19 as long as the net bias field on the domains is not sufficiently
- great to cause coIlapse of any of the domains~. Also, the bias
2} ~ field should be great enough that run-out of domains will
22 not occur when the domains have maximum spacing from o~e
23 another ~isolated domains)'. '
2~ As stated previously, various structures can be
2~ used to provide a field Hb outside the lattice area. One
"
:26 structure is comprised of permanent magnets and magnetically
'l`27 soft elements, where the distance of the soft elements from ` `
i- ' . . ' .
`2'8 the bubble d~main material becomes greater as the lattice area '~
,,i - ;
~ Y0972-063 ` ' -77~ ~ -
1 30 is approached. Another suitable structure can utilize ~ -~
2 a permanent magnet of variable thickness. Still another
3 structure can utilize conductors which follow the directions
4 of the domains ~oward the lattice, as will be described
with respect to FIG. 27.
6 These prineiples apply to any type of interactive -~
7 elements 32. As the elements approach the lattice area, they
8 get closer ~ogether and their interaetive forces increase.
9 Therefore, structure must be provided to confine them within
the restricted area as they move toward the lattice. This
11 barrier force is similar to the barrier force which is exerted
12 in order to retain the lattice, as explained previously.
13 Ideally, the diameter of the magnetic bubble domain
14 interactive elements is such that collapse and run-out do - -
not occur throughout the magnetic medium. FIG. 26 shows a -
16 bias field whlch achieves this. The applied bias field Hb -
17 is large outside the lattice area and small within the lattice
~18 area, while tha interactive bias field Hi is small outside
19 the lattice area but large within the lattice area. The h ,,,.,. ;,'"':,.. .
combination of Hb and Hl is therefore within the range of - _
21 aeceptable values over the entire magnetic medium.
22 FI~.~27 shows a structure for moving interactive
23 elements from a~write means 36 into a lattice area 30 and out
24 of the lattice 30 to a read means 42. The input means 38 and
output means 40 utilize conduetor patterns and ean for instance
26 be the structures des~ribed in the previous seetion.
27 In more detail, geometrie fan-in and fan-out
28 are provided in whieh interaetive elements 32 move from the
YO972-063 - -78-
'
: '.
.
. . .
! . : ::
. . . . .
' ~ '.
' , ' ' " ," ','', ~'
3~
1 left into the lattice area 30 and then to the right to the
2 read means 42. Interactive elements 32 move from the write
3 means 36 in the direction of arrows 140, using propagation
4 structure illustratively shown as T and I bars 142. If these
are bubble domains, they will be isolated domains in this
6 region and an applied bias field Hb is provided. Propagation
7 using structure such as 142 is well known in the art and a .
8 rotating in-plane magnetic field will be provided. :
9 '
11 Domains 32 which arrive at the right-most pole
12 position of T bars 144 are separated by a distance 4d and are
13 now ready to be moved gradually closer together to provide access
. .: .
-; 1.4 to the.lattice area 30. The structure for achieving this is
- 15 conveniently a plurality of propagation conductors P1, P2, P3r t
1 16 P4, P5, and P6. Conductor P6 can also be part of the confinement
' 17 means for the lattice array and part of the input means 38 for
18 moving elementæ 32 into the.lattice area 30. These struc~ures
19 and their operation have been pre~iously described. The
l 20 -propagation structure also comprises means 146 for maintaining
:~ 21 the paths of the interactive elements as they move toward the
.. 22 la~tice area 30. In the case of magnetic bubble domains, :.
23 means 146 is conveniently provided by grooves in the magnetic
-24 material, or ion implanted regions which provide selective
~:25 channals for movement of the bubble domains as they proceed
, 26 -to the lattic~:area. ~ :
. 27 ~ confinement structure 34 is used to maintain the:-.
~ -' , .
28 shape of~the lattice array 30 and also to provide barrier forces .. ~
~ .. .
- .
. YO972-063 ~ 79 ~
r .
r . . .
3~
1 along the fan- in and fan-out portions. Conveniently, means 34
2 is provided by conductors carrying currents in the direction
3 indicated by the arrows on the conductors.
4 Propagation of interactive elements to the lattice
area occurs by providing sequential current pulses in the
6 conductors Pl-P6. Movement of the elements occurs identically
7 as illustrated in FIGS. l9A-19E. The channel means 146 en-
; 8 sures that the interactive elements will stay on the appropriate
9 path as ~hey get closer to the lattice, rather than separating
in order to balance the increasing interactive forces as they
11 move toward ths lattice. However, the confinement forces
~l2 provided by currents in conductors 34 can also maintain the
13 relative positions of the interactive elements without need l`
~14 for means 146. In this case, currents in the conductors Pl~
P6 prevent the domains from moving away from the lattice area
. ~7
~16 and, coupled with the forces provided by currents in conductors
17 34, will move the domains from isolated positions (spacing = 4d)
~18 to interactive positions (spacing = 2d) at the input area of
19 the lattice. At this time, the input means 38 is operative to
insert a column of elements 32 into the lattice area in the
21 manner previously explained.
22 Output from the lattice is provided in a manner
23 entirely similar to the input operation. The output means
24 40 is as illustrated previously and the conductors utilized i
~25 to take elements 32 from the lattice area to positions where
26 they are isolated are designated P'l, P'2, P'3, P'4, P'5, and -
27 P'6. Again! means 146 can be used to channel the elements 42
28 as they ve away from the lattice 30.
j . I
~ Y0972-063 ~ -80~
. -
1,:~"~` '' `'
,. ,: ~
5~
1 When elements 32 reach the left-hand pole positions
2 of the T bars 148, they are separated by a distance of
3 approximately 4d and can then be propagated to the right in the
4 direction of arrows 150, using the propagation structure 152,
shown here as being comprised of T and I bar elements.
6 Elements 32 which are moved by propagation means 152
7 can be brought to a read means 42 for determination of their
8 properties, in the case where the elements have information
~ g associated with them. Additionally, the elements can be
- 10 utilized for other purposes, depending upon the application to
11 which the lattice structure i9 to be applied.
12 As is the case with the input side of the Lattice,
13 an applied bias field Hb is used when the interactive elements
14 are magnetic bubble domains. The same considerations apply as
were discussed previously.
16 Another structure for moving interactive elements
17 from the lattice area 30 to the read means 42, or from write
18 means 36 to the lattice area 30 is shown schematically in
19 FIG. 28 and in detail in FIG. 29. FIGS. 30 and 31 show the
applied current sequence in the conductors of FIG. 29 used to
21 move the interactive elements, while FIG. 31 shows the positions
22 of the interactive elements (for instance, bubble domains) at
23 times corresponding to the application of different current ~ ~-
?4 pulses in the applied sequence of current pulses. ~ ;
Referring more particularly to FIG. 28, a block ~-~
26 diagram is shown which illustrates the reversible mode of
27 operation of the structure of FIG. 29. The structure basically
28 comprises the lattice 30, a transfer register 154, a shift
. . ' .
YO972-063 -81- -
.:
~ ' ' . ''.
1 register 156, and the write means 36 and read means 42.
2 Information can flow from the write means 36 to register 156
3 then to transfer register 154 before entering the lattice 30.
4 In addition, interactive elements 32 can move from the lattice
- 5 30 to the transfer register 154, and then to the shift register
6 156 for movement to read means 42. Thus, depending upon the
7 sequence of applied current pulses in the registers 154 and 156
8 bidirectional movement of elements 32 is obtained.
~- 9 Shift register 156 is a register comprised of m bit
positions where the ~egister is loaded with m/2 interactive
11 elements. ~hese are isolated elements where the interactions
12 between elements are minimàl. The spacing between elements is ~ ~ -
13 illustratively 4d, and the total spacing is 2dm, although
14 other spacings can be utilized.
The transfer register 154 is a two stage structure
16 which links the shift register 156 and the lattice 30. In the
'~17 stage closest to the lattice, it contains m elements with
,~18 2d spacing between elements, while in the second stage it con-
-,19 tains m/2 isolated elements with 4d spacing between elements.
~,20 The shift register i56 contains m/2 isolated elements, having ~ ~-
21 4d spacing between them. That is~alternate bit positions
;22 are filled by the elements in register 156.
23 ~ FIG. 29 illustrates the structure of the transfer
24 register 154 and shift register 156. In the example used to
illustrate operation of this structure, it will be assumed
26 that the interactive elements 32 are magnetic bubble domains.
'27 However, it should be recognized that other interactive elements ~
.. . .. .
YO972-063 -82- ;~ ~
' ' ' , ':
,: . - ' '~ :~
~ . . - ::
~ . ;
.
,
~ .
~'~3~3~'~
1. can be utilized. The la-ttice 30 contains interactive elements
2 32 included within the confinement means 34. The bubble domains
3 in the lattice are indicated by the lower case letters a, b,
. 4 c, .... n, o, p. The transfer register 154 is comprised of a
plurality of conductors A, B, C, and D, together with bubble ,
6 domain defining positions -- (permalloy dots) -- indicated as l.l, 1.2,
7 1.3; 2.1, 2.2, 3.1, 3.2, 3.3, 4.1, 4.2, 5.1, 5.2, 5.3, 6.1, fi.2, 7.1, 7.2,
8 7.3, and 8.1, 8.2. These permalloy dots could be replaced
9 by grooves or by ion implanted regions in the magnetic material
62. Their only function is to provide preferred locations for
11 defining the-paths that the domains follow in going from the
12 lattice 30 or to tha lattice 30. The paths taken by bubble
13 ~omains in shift register 154 are indicated by the double
-.14 headed arrows in this register. Th~se arrows represent that
the bubble domains can be taken from the lattice or sent to
~716 the lattice along the same pat.h~
. 17 Shift register 156 has a preferred path for domain :
18 movement which is in the direction~indicated by arrow 158.
19 Movement-of bubble domains in the direction of arrow 158 occurs
. ~
'20 due to the permalloy structure 160 that is located between ~.
:21- conductors C and D. Permalloy 160 has varying width alon~
~22 its length and.in this way acts.as a guide to move bubble ..
:. .,:
23 domains in the direction of arrow 158. If desired, structure~ ~
24 160 could be a groove in the bubble domain material 162 which `.:.
has variable width as indicated in FIG. 29. Additionally,
26 this structure 160 could be replaced by permalloy triangles with .- :
~27 the apexes pointed downwardly for movement of bubble domains in
.2~8 the direction of arrow 158. In general, domains in shift register
.- .
YO972-~63 -83-
~': ' :.'.',
... ``~ '
, .' :' ~' '
. :..
." ,...
156 are isolated from one another and any type o~ propagation
2 structure is suitable. ....
3 Before describing the applied pulse sequences in the
4 cond~lctors A-D, it should be noted that bubble domains taken
from lattice 30 travel in paths (indicated by arrows in register
6 154) which bring them to the shift register 156. For instance,
7 . bubble domain a follows the path 1.1, 1.2, 1.3, and 1.4 in
8 order to get to shift register 156. Bubble B will follow a
path 2.1, 2.2, 1.2, 1.3, and 1.4 in order to get to shift
register 156. Thus, it will be seen that alternate bubble
11 domains a, c, e and g proceed in generally straight paths
12 from their positions in lattice 30 to their respective positions
13 in shift register 156. On the other hand, bubble domains b, d, .
14 f, and h proceed in paths which are not straight and which are .
partially coincident with the paths of the first-mentioned . .:
6 . domains in going to shift register 156. ...
; 17 Preferred paths for the bubble domains can be : ::
18 provided in a variety of known ways, such as by etching grooves . .
! in the magnetic bubble domain material 62. Additionally, suitable
- deposits of- magnetically soft material (such as permalloy~ can ~ ~
: 21 be located adjacent.to the bubble domain material 62. Also, : :
22 the properties of the bubble domain material can be locally
.- 23 modified using, for example, ion implantation. The bubble domains .: :~
24 will be movod along these tracks by magnetic f1eld gradients -~
25 produced by current pulses in conductors A-D. In a known manner, ~-
26 the preferred bubble domain paths can be shaped to allow bubble
. 27 domain motion in only a single direction or in two directions.
2a FIGS. 30 and 31 indicate the sequence of applied
' :.' .
YO972-063 -84-
:
,. :
,,, . : .. ..
1 current pulses and corresponding bubble domain positions during
2 the transfer and shifting operations. An arbitrary scheme is
3 used to designate the polarity of these current pulses. L
4 indicates that the left edge of the conductor A, D attracts
bubble domains while the right edge of the conductor repels
6 bubble domains. R is a designation for the opposite effect,
7 i.e., the right-hand edge of the conductor attracts bubble domains
8 while the left-hand edge repels them. The actual polarities
g used depend ~n where the pulses are injected into the conductors
and on the directions of magnetizations in the bubble domains.
-11 Consequently, the conductors A-D in FIG. 29 have
12 widths and spacing such that the magnetic fields produced by
13 currents in these conductors will move the bubble domains
14 through the transfer register in a manner similar to that
previously described with respect to FIGS. l9A-19E. For
16 instance, assume that a bubble domain has been attracted to the
17 left edge of a conductor and then a R pul6e is applied. This ; ~ -
18 type of pulse means that the right edge of the conductor will ;-
19 attract the bubble while the left edge of the conductor will ~-
~20 repel the bubble. If this bubble domain were near the left-hand
21 edge of the conductor but not under the conductor, it will
- 22 experience a greater force from the left-hand edge of the
~23 conductor and will be pushed to the left away from that
24 conductor. On the other hand, if the bubble domain were initially ~
under the left edge of the conductor it will be attracted ;
26 to the right (and pushed from the left) and will consequently
; . . .
~ 27 move under the conduc~or to the right-hand edge of the
.
YO972-063 -85- -
; ~ :,
;.
.
~, :
,. . .
,, ,. ~,
. .
: ','- ,.
3~
1 conductor. Since the magnetic field gradients produced by
2 currents in the conductors extend a considerable distance
3 from the conductors, the bubble domains can be suitably
4 positioned with respect to the conductors in order to obtain
either one of these two situations. This positioning is easily
6 obtained by applying current pulses to adjacent conductors.
7 In FIG. 29, the conductors A-D are conveniently
8 shown as single lines. However, in an actual device the
9 conductor width would be comparable to or larger than a bubble
domain diameter and the bubble domain position would be suitably -~-
11 located to produce the desired motion. That i3, the principles
12 described with respect to FIGS. 14E and l9A - l9E apply here
13 alS-
14 FIG. 30 is a table showing the 16 current pulse ,
sequences which constitute the basic shift cycle. FIG. 31 is
16 another table which lists the positions of 16 bubbles a-p
17 following each step of the first two cycles of shift operation.
18 From position zero in shift regist~er 156, the bubbles move
19 downwardly in the direction of arrow 158. Position 84 of
shift register 156 i8 used if the shift register is connected
21 to another array ~lattice) or to a write m2ans 36.
22 The transfer register 154 and shift register 156 ,
23 space the bubble domains in the shift register 156 twice as far `~
24 apart as the bubble domains in the lattice 30. If greater ;~-
;.
separation is desired, it can be easily achieved by providing an
26 additional conductor to provide another 2:1 fan-in or fan-out ~ -
27 steP-
28 In FIG. 29, if either conductor A or B is not activated
YO972-063 -86-
'' . ~ '' ~'
, ,
. .
\
G~
1 by a current pulse, the transfer process does not occur. If
2 conductor A is not activated, a force produced by current in B
3 is insufficient to attract the bubble domains past the confinement
4 barrier 34. If conductor A alone is activated by a current pulse,
S then at step 9 of the applied pulse sequence the bubbles
6 are repelled back to or through confinement barrier 34. Thus,
7 conductors A and B can also function as input/output control .~ :
8 gates as described previously.
9 AS another alternative, the conductors A, B, C,
and D can be used to provide transfer operations to and from :
.11 a plurality of lattice arrays on the same magnetic medium 62. .:~
12 If these conductors are connected in groups, decoding can be
13 provided for the selection of any of the lattice arrays.
-14 As will be more apparent from FIGS. 30 and 31, the
transfer process automatically synchronizes with the current
16 pulse sequence in conductors A-B. If the pulse sequence starts
17 at some step other than number one, then nothing happens during
.
18 the remainder of the partial current pulse cycle and correct
.,19 operation of transfer begins with step one of the following ::
.~20 cycle.
-:~21 Reference will now be made to the tables comprising : .
~ ~
22 FIGS. 30 and 31. In this operation, bubble domains a, b, ..... h :
~23 are to be removed from lattice area 30 and brought to shift
24 register 156. After this, bubble domains i, j, ..... p will be
re ved from lattice area 30 and brought to shift register 156.
,Y .26 At time 1, currents exist in conductors A and D which are
5 27 attractive for bubble domains a, b, .... h. Therefore, these .~ -
~ .28 domains will be moved into positions 1.1, 2.1, 3.1, .... 8.1, `~ -
5 , .
. . .....
. YO972-063 -87~
- :
.
, : .
.
~ ' . , ', . '~,'.
~, : ' .: ~
1 respectively. At this time, currents in conductors C and D do
2 not have an affect on the bubble domains a-h coming out of ~ -
3 lattice 30. ;
4 At time 2, the current direction in conductor B
. . ...reverses thus making the left-hand edge of conductor B repulsive.
6 Consequently, bubble domains a-h retain their respective
7 positions. -
8 During times 3-8, currents exi~t in conductors C and D
9 which are used to complete an operation for domains located
within transfer register 154 and shift register 156. From step -;
11 3 through step 8, currents do not exist in conductors A and B.
12 At step 9, the current directions in conductors A
13 and B reverse from what they were in steps 1 and 2 respectively.
14 At this time, there is a current in conductor A which makes
the right-hand edge of A attractive and left-hand edge -
16 repulsive. At the same time, the current pulse~in conductor
17 B make~ its left-hand edge attractive and its right-hand edge
18 repulsive. During this step, the bubble domains a-h move ~ -~
19 one step. For instance, bubble domain a moves from position
1.1 to position 1.2 while bubble domain b moves from position
21 2.1 to position 2.2. Consequently, all bubble domains a-h
22 move one step during the application of the current pulses
23 in conductors A and B in step 9 in the sequence of current
24 pulses.
At step 10 of the sequence, no current pulse existæ
26 in conductor A and the current pulse in conductor B has
27 changed its direction. At this time, conductor C has a
. . .
28 current which makes its left-hand edge attractive. Consequently, ~--
~. 1 . ~
, YO972-063 -88-
'. 'i ~ . '' ,:
~ . . .
:.: .. .
;.' . : . ~':~'
,. ,~- . ::
,. ... . - . : . ~
, . . . -
bubble domains close to conductor B see a force due to
2 conductors B and C~ Therefore, bubble domains a, c, e,
3 and g move. For instance, bubble domain a moves from position
4 1.2 to position 1.3 during this time. Bubble domains b, d, f,
and h retain their positions. ~;
6 At time sequence 11, pulses in conductors C and D
7 make their left-hand edges repulsive to bubble domains. Con-
8 sequently, bubble domains a, c, e, and g move to the next
9 position. For instance, bubble domain a moves from position
1.3 to position 1.4 while bubble domain c moves from position
11 3.3 to position 3.4. Thi~ movement is due to the fact that
.
12 these bubble domains are sufficiently under conductor C to
13 be able to experienoe the attractive force on the right-hand
14 edge of this conductor rather than the repulsive force from -
lS the left-hand side of conductor C. ~
16 At time step 12, the current directions have reversed -
17 in both conductors C and D. This means that the left-hand edges
18 of these conductors will be attractive to bubble domains
19 while the right-hand edges will be repulsive to the bubble
domains. During this time step, bubble domains a, c, e, and
21 g move. For instance, bubble domain a moves from position
22 1.4 to position zero in shift register 156, while bubble domain ~
~$
23 c moves fro~ position 3.4 to position 2.4. Thus, bubble
24 domains a, c, e, and g, which were moved into shift register
156 by the~ previous time step, now begin ~o move in shift -
.. .. .
26 register 156 in the direction of arrow 158. These bubble domains
27 can then be read and returned to the lattice or removed to ~ -~
28 different areas of the magnetic sheet. As is apparent from ;
.~ . '~ .. .
..
YO972-063 -89~
.,' ' ' ,'' ~ ', . '
., - .
1 FIG. 31, bubble domains a, d, f, and h do not move from
2 their second positions until time step 16. At this time, ~-
3 they begin moving in paths previously travelled by bubble
4 domains a, c, e, and g in order to propagate to shift
register 156. When they reach register 156, they move
6 down shift register 156 in the direction of arrow 158. -~
. .
7 The movement of the next column of bubble domains,
8 i, j, .... p of lattice area 30 follows in a similar manner.
9 These bubble domains i - p begin to move on time step 17 and
-10 thereafter, starting with time step 24, begin moving toward
11 shift register 156. As with the bubble domains a - h,
12 alternate bubble domains in the column i - p move first,
13 followed by the movement of the remaining bubble domains in -
14 that column. For instance, bubble domains i, k, m, and o
move to shift register 156 before bubble domains j, 1, n,
16 and p. This is because the latter group of bubble domains
'17 j, 1, n, and p have to travel partially on the same path as
18 was used by bubble domains i, k, m, and o. The second
~19 column of bubble domains uses the same paths as were used
;20 by bubble domains in the first column. For example, domain
21 i follows the same path as domain a, and domain j follows
22 the same path as domain b. ;
23 Thus, it is apparent from FIGS. 30 and 31 that groups
~,, , ~,
24 of m (where m is the total of bubble domains in a column) are ;`
: - . .. ~ ,
shifted at a single time by the transfer register 154 to the
26 shift register 156. In reverse, bubble domains from the shift ~;
27 register 156 can be moved to the lattice area in the same fashion
-28 by changing the polarity of currents in the sequence of pulses
' ,~ ' ' .,
YO972-063 -90-
i . '
:
f . . ' ' .~, ' '
. ... : .. . .. ... ... ..... .. .. .. .. ~.. . . - . . .... . .. .. : . - . ~ . . . .
13~ :
l applied to conductors A-D.
2 As mentioned previously, the width of conductors A-D ~
3 can be suitably chosen to be approximately the bubble domain -
4 diameters. The center-center spacing of the conductors is
conveniently twice the width of ~he conductors.
6 It should be easily recogni~ed that these current
7 sequences can be rearranged using external buffers. Also,
8 various modifications of this scheme can be utilized to move
9 bubble domains from the lattice area to isolated bubble areas,
~10 and from isolated bubble areas to the lattice area. Additionally, ;
11 greater fan-in or fan-out can be provided.
~12 This basic scheme will also work when the interactive
13 elements are not magnetic bubbla domains~ In some cases, the
14 interactive elements are not bias field ~ensitive as are bubble
domain elements so the provision of a technique for changing
16 the spacing between the interactive elements may not always be
~17 necessary. Further, even in the case of magnetic bubble domains,
-~18 it may not be necessary to change the spacing when the domains ;;
19 are taken from the lattice area, or placed into the lattice area
~20 from an external area.
,
21 Coded Interactive Elements
22 ~ Information can be associated with the interactive -~
23 elements 32. In the case of magnetic bubble domain elements,
;24 various properties of the bubble domains oan be used to
distinguish one type of bubble domain from-another, thereby
26 enabling coding o information in terms of bubble domain
27 properties. As will be ~een, the wall properties of different
2a typ-~ of magn-tic bubbl- domains are partioularly useful for
~, Y0972-063 -91- -
, ~, . . . :
,
' ~'':
~; :
1 the coding of information. In the case of other types of
2 magnetic interactive elements coding in accordance with
3 physical appearance is suitable.
4 The following sections under ~his sub-heading will
illustrate the different types of coding that can be utilized
6 when information is to be associated with the interactive
7 elements. Although several different types of coding will
8 be illustrated, it ahould be remembered that information can
9 be coded in the interactive elements in any way; therefore,
the present invention can be utilized with new discoveries
11 pertaining to different properties of interactive elements,
-12 such as bubble domains. The main teaching of the invention,
13 utilization of c~nfined arrays (such as lattices) in information
14 systems, can be exploited whether or not the interactive -
elements are themselves associated with information.
- . ' , :,:
16 Bubble Domain Coding:Hard/Soft Bubbles:
,
-~17 In this first type of coding, the teaching of U.S. patent
18 application,.~;erial No. 375.,2~, flled June.29, 1973 now U.S. patent 3,899,779.
19 is utilized. In this copending applicatlon, the existence of
so called "hard" and "soft" magnetic bubble domains is utilized
21 to provide ~ystems in which information is conveyed by different
22 domain properties, rather than by the presence and absence
-~23 lof domains. A "hard" domain is one which has a large number i-
24 of vertical Bloch lines in its domain wall and which collapses
at higher bias fields than a "soft" domain, which has a small
.
26 -number tor Zero) vertical 210ch lines in its domain wall.
27 These different types of domains are described more explicitly
;Y0972-063 -92- ;
: : ~: .
' `, ,' ,
: .
, ~3~ ' ',': ,
1 in a paper by A. P. Malozemoff, Applied Physics Letters, 21,
2 149 (1972).
3 FIG. 32 shows a means for writing information in
4 terms of hard/soft coding of magnetic bubble domain interactive
elements, while FIG. 33 shows the read operation when
6 information is stored in the terms of hard/soft bubble domains.
7 In more detail, the structure of FIG. 32 provides
8 patterns of hard and soft magnetic bubble domains for coding
9 of information (such as binary information). If desired,
multilevels of information in addition to two levels can be
-11 provided in accordance with the teaching of aforementioned
12 United States patent 3,899,779.
13 Magnetic medium 62 supports magnetic bubble domains. ~ -
14 A hard bubble generator 160 is comprised of a current-carrying -
coil 162 connected to a dc bias source 164 and to a pulse
i 16 current source 166 which can be selectively connected in
~17 parallel with source 164 via switch 168. Within coil 162 there
18 are provided a plurality of current-carrying conductors 170A,
- 19 170B, and 170C. These conductors are connected to current
sources (not shown) which provide currents IA, IB, and IC
21 through conductors 170A - 170C, respectively.
22 Also located within coil 162 is a propagation means
23 172 which in this case comprises T and I bar patterns of
24 magnetically soft material such as permalloy. Domains brought
;~5 to propagation pattern 172 will move in the direction of arrow
26 174 in response to rotation of magnetic fields A in the plane
27 of magnetic sheet 62, as is well known. Further, propagation
'~ ' ' ' ; ~ '` ', ' .
~ YO972-063 -93- ~
" ;"' ,.
- . ~,.'.'
,, .. , . , . . , . , , . .-
p
means 172 can be comprised of conductor patterns rather than
2 magnetically soft material.
3 The operation of hard bubble generator 160 depends
4 upon the provision of varying magnetic fields within the area ~ -
of coil 162. These magnetic fields operate on stripe domains
6 176 in order to chop these domains into smaller domains containing
7 vertical Bloch lines.
- 8 The dc level of the bias field Hz within coil 162 is
9 provided by dc source 164. Fluctuations in the net bias field
L0 within coil 162 are provided by current pulses generated by pulse - ~
!1 source 166. Initially, a negative pulse is provided by source ~ -
L2 166 in order to lower the net bias field in the area of coil 162. ~-
-L3 This creates an attractive region for magnetic stripe domains
L4 176 such that these domains will move within the area of coil ;
,15 162. At this time, a positive pulse is produced by source 166
-L6 to raise the level of the bias field above the dc level. For
~7 sufficiently short and strong pulses, this will cause a chopping ~-
:L8 of the stripe domains 176. Generally, the number of hard domain~
- .:
~9 produced by this chopping action increases with the nu~er of ~` -
~20 pulses applied. After this, the net bias field is increased
~1 greatly in the area of coil 36 in order to collapse all domains
22 other than hard domains. Consequently, the only domains remaining
23 in the area of coil 162 after the chopping and discrimination
24 steps are hard magnetic domains.
These hard domains are then moved to the vicinity
26 of propagation means 172 by applying current pulses in the -
27 conductors 170A - 170C. The magnetic fields produced by currents
28 in these conductors crea~e bias field gradients which will ~ '
' . ' ,.~, ,' .
,. ...
~ YO972-063 -94- ~
.
, . ,. ,' ~'' :'
.. ... . .
,, . ~ . .
:~3 ~ 8
1 attract the hard domains to the vicinity of propagation means
2 172. Once they are in the vicinity of means 172, the hard
3 domains will be attracted to magnetic polas created on the T and
4 I bars when the propagation field H i9 rotated. These hard
domains will then propagate in the direction of arrow 174 in
6 a well known manner.
7 As a representative example, hard domains were
8 created in a magnetic sheet of 5.25 microns thickness having
9 the compOsition (Tbo o4EUo.66y2.3)Fe3-85Gal-l5ol2 ~ t
. .~ , . . .
field pulses applied to magnetic sheet 62 generally ranged from
11 about 10 Oe. to 50 Oe. and had a duration from about 0.2 micro-
12 seconds to aboùt 10 microseconds. The number of pulses applied
`13 by source 166 can be varied from 1 to practially any number.
14 The number of applied pulses will generally depend upon the
distribution of the various types of domains desired to be
~16 produced. Generally, as the number of applied current pulses
17 increase~, there is a greater li~elihood to create domains having
- 18 larger numbers of vertical Bloch lines. Correspondingly, the
~19 longer the duration of the applied current pulses the greater
~20 the likelihood to collapse domains other than those ith large
21 numbers of ver-tical Bloch lines. Of course, the final current
22 pulse is chosen to be of such magnitude as to collapse all
-23 domains within coil 162 other than those domains having the
24 minimum desired number of vertical Bloch lines in their domain
~25 walls. This ensures that sufficiently hard bubble domains are
26 obtained for the particular operation desired.
27 Generally, the magnitude of the applied current pulses
28 depends to some degree on the magnetization 4~M9 of the magnetic
~ YO972-063 -95-
., . "'',:
.
" , . .~ ' ' '
~ ,
3~
1 sheet 62. As 4~Ms increases, higher values of magnetic bias
2 pulses will be required to produce the hard domains. Generally,
3 up to 50% of 4~Ms is a reasonable range for the magnitude of the
4 applied bias field pulses.
The longer the duration of the applied chopping pulses,
6 the better is the chance that chopping will occur. After this,
7 the longer the duration of the applied current pulse, the greater
..... ::.
8 the likelihood that the domains within coil 162 will collapse. -
9 Since the hard bubble generator can be part of a
separate structure which is brought into proximity to magnetic
11 sheet 62, the provision of hard domains within medium 62 can
12 easily be done at the time the magnetic bubble domain system
13 utilizing a lattice is being fabricated. Consequently, a res- -
., ~
14 - ervoir of hard and soft domains can be provided for use by the
user of the systems (FIG.-2). In this case, the hard domain
16 generator 160 need not be a portion of the system which is
17 delivered but instead could be utilized in the manufacturing
18 facility.
; . .
- 19 In FIG. 32, the hard bubble generator 160 is part of .. ..
the overall write means 36 (FIG. 6) which is used to provide a
21 pattern of coded information. Consequently, the hard domains
22 are combined with soft domains from a normal bubble domain ~
23 generator 178. The final output from write means 36 is a - ;
24 pattern of hard and soft domains movlng in the direction of ;
arrow 180 toward the input means 38. In the embodiment of
26 FIG. 32, the hard domains affect the delivery of soft domains
27 to the information stream applied to input means 38.
28 As mentioned, hard domain generator 160 produces a -
YO972-063 -96- ~
", ' ' ' ', '
~ ~:,
1 pattern of hard domains moving in the direction of arrow 174.
2 These hard domains are further propagated by propagation means
3 182, also comprised of T and I bar patterns. A current-carrying
4 coil 184 provides a magnetic field in the direction of the bias -
magnetic field Hb at pols position 1 of T bar 186.
6 Normal bubble domain generator 178 is of the type
7 illustrated in U.S. application, Serial No. 266,758, filed June 27, 1972 now
8 U.S. patent 3,825,885. Generator 178 is comprised of a disc 188 of magnetically
g soft material such as permalloy and an additional layer 190 of
magnetically soft material, such as permalloy which is exchange-
11 coupled to medium 62. Layer I90 acts to suppress any hard
~12 domains produced by generator 178. Soft domains are provided
13 each cycle of rotation of the drive field H. The e soft domains
14 travel downwardly to the propagation mQans 182, following
-~15 repetitive pole patterns 2, 3, and 4 on T bar 192. Associated :
16 with the soft bubble generator 178 is an L-bar 194 which serves
17 as an annihi'lator for the ~oft domains produced by-generator
- 18 178. For certain circumstance~, domains produced by generator. ~ :
.19 178 are deflected to annihilator 194 and do not enter the ~ .
.
information pattern travelling to the right along propagation
~21 means 182, as will be explained in more detail later. . . :. -
22 The final information pattern of hard and soft domains ~ .
23 continues in the direction of arrow 180 and enter~ the
24 input means 38. In operation, hard domains enter propagation .
means 182 and move to pole position 4 on I bar 196. If a
26 current Ig exist~ in.loop 184 at this time, the hard domains at .
27 the end of I bar 196 will not see an attractive pole at pole
.
~ Y0972-063 -9?- .
'' ' . ' ' .
r ; ~ ~
position 1 of T bar 186. Consequently, they will remain at
2 pole position 4 of I bar 196. As the propagation field H
3 continues to rotate, the domains will be attracted to pole
. . .- - . .
4 position 2 on T bar 198. After this, the domains will continue
to the annihilator A. If desired, the hard domains can be
- 6 redirected to other circuitry for further use within an informa-
7 tion system. Thus, hard domains at pole position 4 of I bar 196
8 will be allowed to pass further to the right depending upon the
9 presence and absence of current Ig in loop 184. In this manner,
a gate is provided for the passage of hard domains.
11 A soft domain is produced by generator 178 during
12 each cycle of field H. The soft domains propagate to T bar 192
13 and follow successive pole positions 2, 3, and 4 to the propaga-
14 tion means 1a2, after which they travel to the right along the
direction indicated by arrow 180 in response to rotation of
16 field H. However, if a hard domain passes through the successive - -
17 pole positions 1, 2 and 3 of T bar 186, a soft domain from
18 generator 178 will not be able to move from pole position 3 to
19 pole position 4 on T bar 192. Consequently, on the next ~ ;
rotation of field H, the soft domain will move from pole ~-
21 position 3 on T bar 192 to pole position 4 (elbow) on L bar 194. ~
22 When field H rotates to position 1, the soft domain will ~ ~ -
J 23 continue to be trapped at the elbow of L bar 194. When field
; 24 H continues to direction 2, a negative pole will be produced
` 25 at the elbow of bar 194 which will collapse the domain located
26 there. This collapse is then enhanced when field ~I rotates
27 to position 3. Consequently, the presence of hard domains on `
~i28 T bar 186 influences the entry of soft domains from generator
:. ,.~ ' ~, ' :':
YO972-063 -98- ~
,. . . .
~'''` ,.
'. , ' .,
, ., ,~
3~
1 178 to propagation means 182. In this manner, an information
2 pattern wili be sent to input means 38.
3 - Alternate structures can be utilized to provide
4 patterns of hard and soft domains for coding of information in
a lattice array. For instance, the propagation paths comprising
6 magnetically soft elements can be replaced by conductor patterns,
7 and other techniques can be used for chopping stripe domains
8 to produce domains having varying numbers of vertical Bloch
g lines (i.e., domains of varying hardness). Of course, provision
~, 10 of domains of varying hardness means that multilevel information
11 - storage i8 possible.
-12 FIG. 33 shows a technique for reading information
13 which has been coded in terms of hard and soft magnetic domains. ~ -
14 In this figure, a group of hard and soft domains has been removed ~ -
from lattice 30 by the output means 40. This pattern of ~nforma-
16 tion-containing domains propagates in the direction of arrow
,17 198 to a hard bubble discriminator 200. Discriminator 200 is
18 used to collapse all soft domains in the information pattern, ~
..
19 thereby allowing only hard domains to pass. These hard domains ~
: . . -:
are then sensed by any type of bubble domain sensor, such as a
21 magnetoresistive sensor which is shown more particularly in -
~22 U.S. 3,691,540. After the information is sensed, the soft
23 domains have to be reestablished in the information patter~
24 if nondestructive read-out is desired. Consequently, a soft
domain generator is used to reestablish the original informa-
.
26 tion pattern. ;
27 In more detail, the pattern of hard and soft domains - -
28 enters hard bubble discriminator 200 in the direction of arrow
.: ,.
- .
YO972-063 -99-
~' ', :~ . ' . .
~ -. , : .
r ~
3 ~
1 198. Discriminator 200 is comprised of a current-carrying
2 coil 202 which is connected to a current source providing a
3 current Ic. Current Ic in coil 202 produces a magnetic field
4 in the same direction as the bias field Hb. This increases
the bias field at pole position 4 of I-bar 204, thus collapsing
6 all soft domain 3 which appear at this location. This means
7 that only hard domains will continue propagating to the right ''
8 along the T and I bar pattern 206.
': 9 Domains propagating along pattern 206 will pass a
sensing means 208, which is shown as a magnetoresistive sensor.' '
-' 11 Means 208 illustratively includes a magnetoresistive sensing
12 element 210 which is connected to a current source 212. Source ~,
; 13 212 produces a measuring current Is through sensing element 210. '
14 When a domain passes element 210, the magnetization vector of '~
3 15 the element will be rotated, causing a resistance change. This
~ 16 resistance change is manifested as a voltage change Vs indicative ;','
''~ 17 of the presence of a hard domain in flux-coupling proximity to '
-f 18 element 210. If no domain passes s~ensing element 210 during a
f 19 cycle of drive field H, this wili indicate that a soft domain '~
, ,~, 20 was originally present at that cycle time.
~ 21 After being sensed, domains propagate further to the '
,3 22 right along the direction indicated by arrow 214. The domains
23 pass a structure which is a soft bubble replacer 216. This "~'-
,i, 24 structure is similar to that used in FIG. 32 for providing
soft bubble domains in the information pattern. The soft ~ '
26' bubble replacer is comprised of a soft domain generator 218
~33 27 together,with a layer of magnetically soft material 220 ' '~
'l~8 for supDression of hard domains. Propagation means 222
'! YO972-063 -100-
::', . , ., :'
, ~ '
. j .
.. : . :
: . . .
3~i~
1 carries the soft domains to the information pattern stream.
2 Additionally, annihilator 224 is provided.
3 During each cycle of field H, a single soft domain
4- is produced by generator 218 and propagates along T-bar 222.
However, if a hard domain is at pole position 3 of T-bar 226,
6 domains from generator 216 will be deflected from element 136
7 to the elbow of annihilator 224 where they will be subsequently
8 annihilated as field H rotates. However, if hard domains are
9 not present at pole position 3 of T-bar 226 at this time,
domains produced by generator 218 will propagate to the horizontal -~
11 propagation means 206 and will continue to the right in the
12 direction of arrow 214. Thus, the original combination of hard ;
13 and soft domains from the lattice area will be re-established.
14 This re-established information pattern can be sent to an input
lS means 38 for entry into the same lattice array or a different ~ -
16 lattice array ~FIG. 4), or can be sent to an annihilator for
17 destruction of information. Additionally, the domains can be
18 used for other types of circuitry. Thus, coding of domains in
19 terms of the number of vertical Bloch lines existing in the
domain walls is a suitable technique for establishment of
21 information within the lattice array. However, ince soft
22 domains have greater mobilities in magnetic medium 62 than hard
,. :
23 domains, the speed of operation of the system may be limited to
24 that of the hard domains, in order to have synchronized domain
2~ movement.
':26 r This type of coding can be used to provide levels of
.. . ~. - : .
. 27 in~ormation ~reater than two, depending upon the hardness of ~`
28 the domains utilized. This is readily achieved since the
~29 domains will collapse at magnetic fields dependent upon their
.:
YO972-063 -101- ;
:.:
3~
1 hardness and can be detected in terms of the degree of
2 hardness, as explained in aforementioned S.N. 375,285.
.. . .
:. . . . .
3 Bubble Domain Coding: Deflection Properties -
. . .
4 Coded magnetic bubble domains using deflection
properties are described in U.S. application, Serial No. 375,289, filed
6 June 29, 1973 now U.S. patent 3,890,605. That, application describes magnetic ~ .
bubble domain systems using bubble domains for information,
8 where different information states are represented by
9 different properties of the domains. The deflection of a
~10 magnetic bubble domain in a gradient magnetic field depends
11 upon the number of rotations of the magnetization vectors
12 around the periphery of the domain wall. Thus, a domain
13 having zero vertical Bloch lines will undergo a deflection
~4 in the gradient field while a domain having a pair of verti-
cal Bloch lines may or may not undergo deflection, depending -
a6 ~ upon the sign of the ~loch line pair.
:17 In the present application, information can be ~ - -
.
~8 coded in the bubble domains in a lattice array using their
~9 deflection properties. For instance, domains representing
"1 bits" may be represented by a domain which has a particu- -
21 lar deflection in a gradient magnetlc field, while a differ-
22 ent information state (0-bit) is represented by a domain having
.. . .
23 a different dsflection in the gradient magnetic field.
24 Figure 34 shows a structure for generating bubble
. ' ~
~25 domains having-different deflections. Figure 35 shows an
~26 apparatus for sensing domains having different deflections
~7 which have been taken from the lattice area.
.: . .
~ ~ .
YO972-063 -102- ; ~
,. i . . ~
1~ 43;6,8 , .
1 In more detail, FIG. 34 shows a generator designated
2 228 which is used to provide domains having different deflec-
3 tions in a gradient magnetic field. The remaining structure
4 in this figure is a write means generally designated 230,
which separates domains produced by genexator 228 in accordance
6 with their deflection properties in a gradient magnetic field.
7 The domains from write means 230 can be directly sent to input
means 38 for entry into a lattice array 30.
9 Generator 228 is similar to the generator shown
in FIG. 32 for producing hard and soft magnetic bubble domains.
11 It comprises a current carrying coil 232 connected to a current
12 source 234 which is used to chop a stripe domain 236. The
13 rest of the generator 228 i8 comprised of a current carrying
14 coil 238 connected to a d.c. source 240 and a pulse source
242. Relay 244 is used to selectively connect ~ource 242 to
16 the circui~. Additionally, conductors 246A, 246B and 246C are
17 provided.
A~ 18 In operation, current pulses Il, I2, and I3 in
19 conductors 246A - 246C respectively provide attractive magnetic ~-
fields to move domains into the area within coil 238. Current
21 in coil 238 reduce~ the net ~ias field within this loop causing ` ~`
. .
-' 22 a domain therein to become stripe domain 236. After stripe
23 236 is created in loop 238, a current pulse I is produced in
24 conductor 232. This splits domain 236 since different magnetic
fields are exerted on both sides of domain 236. The split domain -
~ 26 then propagates to the right under control of propagation means -~ -
- 27 ?48, shown here as a circuit comprising T and I bars. -
28 Domains propagating to write means 230 are separated
YO972-063 -103-
'".'' ' ."~. .. '
1 in accordance with their deflection in a gradient magnetic field
2 and applied to various storage blns for use as inputs to input
3 means 38.
4 Generator control 250 provides current pulses in
5 conductor 252 which produce magnetic fields for selectively ;
6 collapsing domains from generator 228.
7 A deflection means, generally designated 254, com-
8 prises a pair of current-carrying conductors 256A and 256
9 which are connected to a current source 258 through current
limiting, variable resistors RA and RB. Source 258 is con-
11 trollably operative by gradient control unit 260.
12 Deflection means 254 can be provided by a plurality
13 of structures. Its function is to produce a magnetic field -
, 14 gradient and bias field which will deflect certain domains
,
moving into the region where the gradient exists. The gradient
16 producing means is conveniently shown as current carrying
17 conductors but can comprise structures such as permanent
18 magnetg selected to provide different magnetic biases, or
19 layers of magnetic material exchange-coupled to magnetic sheet -
62 having graded properties to provide the gradient field.
21 Additionally, the magnetic properties of material 62 may be
22 locally altered to provide the gradient field. -
23 Domains produced by generator 228 propogate in tl.e -~
,
~, 24 direction of arrow 262 until they reach location A. The net
bias field at A is different than at point B and the domains
26 ~ill be deflected in accordance with their rotational proper-
27 ties. In this drawing, domains having ~1 rotations of their
28 wall magnetization will be deflected through the angle p and
.. ~ ` - .
i Y0972-063 -104-
. ........................................................................ - , .
, '. , ~:
,~:.. , .: :
.i . . . . .
.. : ~^. . . . - - . . . . . , . .-
3t~
1 will be sent to a storage location defined as the +1 bin. Domains
2 having 0 rotations of their wall magnetization will not be de-
3 flected and will be propogated to a storage location designated
4 O bin. Domains having -1 rotation of their wall magnetization
will be deflected through an angle -p and will propagate to
6 the storage area labeled -1 bin. These bins are conventional
- 7 storage locations and can be, for instance, closed loop shift
8 register~ around which the various domains continuously tra-~e'.
9 Generally, it is desirable that all the bins contain
the same number of domains and that information be selectively
11 switched from the bins as desired. However, because the generator
- .12 may produce a statistical distribution of domains ha~ing various
13 deflection properties, circuitry is provided to keep track
14 of when eacih bin is fully loaded. When a bin is fully loaded,
means are provided to collapse other domains havin~ similar
16 properties which would normally enter that bin. This continues ~ -
17 until all bins have a full supply of the domalns which they are
18 storing. Of course, other design considerations may be readily
., ~ .
19 available to those skilled in the-art. ;
A~soicated with each bin is a switch SW +1, SW 0, ! . ~ .
-21 SW -1, respectively. These are switches operable under control
22 of the decoder s~itch control unit 264. These switches pass
23 domains in one of two directions depending upon whether domains ~ ~-
24 are to be taken from the bin or allowed to recirculate in the
bin. Standard current controlled switches of the type shown in :-
26 U. S. 3,689,902 or U. S. 3,701,125 are suitable.
.
Z7 : Each storage bin has associated therewith circuitry ;~ ~
. ~ . - .
28 for counting the domains which enter the bin and circuitry for
YO972-063 ~ -105- ; ~-
.: ~
~' ~
1 collapsing domains which try to enter the bin after the bin is
2 fully loaded. Generally this circuitry comprises a counter
3 which merely counts the number of domains entering the bin and
4 a current source for producing a current whose magnetic field is
sufficient to collapse unwanted domains which propagate in the
6 direction of the bin. For instance, +l counter in collapse
7 current generator 266 ~1 is associated with the +1 bin. Counter -
8 266 +1 detects domains having +p deflection via conductor
9 loop 268 +1. After the number of domains required to fill the
+1 bin have been counted, unit 266 +1 provides a current pulse
il in conductor 268 +1 which then destroys any other domain trying
12 to enter the +1 bin. In a similar manner the 0 bin has associated
13 therewith a 0 counter and collapse current generator 266 -0
14 which is coupled to the 0 bin propagation path by conductor
lS 268-0. In the same manner, the -1 bin has associated therewith
16 a -1 counter and collapse current generator 266 -1, and con-
17 ductor 268 -1. These countsr and collapse current generators
18 provide inputs to the AND circuit 270 which upon coincidence
19 of all inputs, provides a signal to generator control 250.
This signal causes generator control 250 to produce a current
21 pulse in conductor 252 which stops the further passage of ~,~
22 domains in the direction of arrow 262.
23 The counter and collapse current generators also
24 provide inputs to a unit entitled Control and Clocking 272
which in turn controllably activates the decoder switch control
26 264. Thus, control unit 272 provides a signal to the decoder
27 switch control 264 after receiving signals from all of the
28 control and clocking units 266, in order to signal decoder
YO972-0~3 -106-
~ -, ;' .
- .
i: ;` . , ~ ,.
l switch control 264 that remo~al of domains from the storage
2 bins can occur. In this manner, it is possible to selectively
3 remove domains from the storage bins for propagation along
4 the direction of arrows 274. ~ --
Thus, FIG. 34 shows circuitry for providing in a
6 reproducible manner domains having a plurality of different
7 properties for transmission by known means to input means
8 38. In FIG. 34, domains having deflections other than the
9 deflections utilized for information states are deflected
to annihilators 276 where they are destroyed.
ll As is apparant from FIG. 34, conventional propaga-
12 tion circuits can be used to move domains having different
13 deflection properties. Additionally, the propagation structure~
14 can be combined with path defining grooves in the magnetic
material if additional control is desired. Once within the
..
16 lattice array, domains having different deflection properties
17 interact with one another to provide stable paths of movement.
18 Thus, the coded domains behave in the same manner as uncoded
.. ~ , . .
19 domains within lattice 30.
Aforementioned United States patent 3,890,605 ~ . ~:
21 describes how to controllably make domains having desired-wall
. .
22 magnetization rotation. Consequently, it is possible to
23 - eliminate many of the components shown in FIG. 34, in an -~
24 actual circuit.
.. . .
FIG. 35 shows a device for reading bubble domains
26 using their deflection properties. In this structure, a
27 patter~ of bubble domains is provided from the output m~ans
28 40 of the system of FIG. 2. This output pattern propagates to
.
-
' ; YO972-063 -107-
- , , - .-
- - . ~
: . :
~s .. ..
I{~
3~
1 a deflection means 254 which is identical with that shown
2 in FIG. 34. Deflection means 234 comprises conductors 256A
3 and 256B connected, via resistors RA and RB, to current sou;ce
4 258. Gradient control 260 provides an input to source 258 to
determine the strength of the gradient provided by current in
6 conductors 256A and 256B.
7 Domains coming from the lattice area have deflections
8 previously determined by the structure of FIG. 34. Therefore,
9 these domains will be separated by the gradient field produced
by means 254 in accordance with their deflection properties.
11 Various propagation circuits 276 +1, 276-0, and 276 -1 move
12 the domains to the sensing devlce, generally designated 278.
13 Sensor 27a is comprised of sensing elements 280 +1, ~-
14 280-0, and 280 -1. Illustratively these are magneto resistive
sensing elements which are spatially stagered from one another
16 so as to indicate the presence of domains traveling in the -
17 separate propagation paths at diffe~rent times. Current source
18 282 provides a series of electrical currents through the sensing
19 elements. Depending upon the presence and absence of domains
in flux coupling proximity to the sensing elements, different
21 voltage signals Vs will be provided which can be sent to the
22 utilization means 44 (FIG. 6).
23 After being sensed the bubble domains can be des-
24 troyed or returned to the same or a different lattice array,
if information is to be retained. Additionally, the domains can
26 be destroyed, or directed elsewhere and new information written
27 by the generator. Consequently, utilization of the deflection `
28 properties of magnetic bubble domains is a very convenient
:
YO972-063 -108-
. . '.
.`, ~ ~ ' ; ,
; ~ '
.. ~ . ' .~. 1 '
. .
3~
1 means of coding information in the form of bubble domains
2 within a lattice array. Since domains having small numbers
3 of Bloch lines can be used, the problem of different mobilitie~
4 (mentioned for hard/soft domain coding) is not present when
coding is in terms of deflection properties.
. ~ . .
6 Coding With Dual Size Bubble Domains:
7 Bubble domains having different sizes can exist
8 in the magnetic medium at the same time, as is more fully
g described in U.S. application, Serial No. 319,130 filed December 29, ~;
1972 now patent 3,911,411. These different type domains can be
11 produced by a suitable generator and in addition each type ''`~;
12 of domain can be changed to the other type and back again
13 in a very simple manner. '~'~
14 FIG. 36 illustrates a generator 284 which can be
used to produce dual size magnetic bubble domains in magnetic
16 medium 62. These domains of different size will be designated
17 type A domains-and type B domains, where the type,A domains
18 are those which a,ppear larger when viewed optically. Generator
19 284 is comprised of a current-carrying coil 286 connected to
a pulse current source 288 through a variable resistor R. The
21 function of conductor 286 is to provide a localized magnetic "
22 field in a direction substantially parallel to a~ easy direction ~ '
23 of magnetization of magnetic medium 62. Depending upon the ~
24 polarity of current in conductor 286, the magnetic field '-
produced within this conductor will be either parallel or
,~26 antlparallel to the applied magnetic bias field Hb produced by ~i,
27 source,290. A propagation field source 292 is used to provide '' -,
~, ,.
:,:: -
~- 'Y0972-D6'3 -109- ~' ~
-
','',
-: , . ,. ,: . ", ... , . - .. .. :-. - , .. -; , . - .,- ~ -
1 a rotating, in-plane magnetic field H which is used to move
2 domains in conjunction with propagation means 294, in a manner
3 which is well known. A control means 296 provides control
4 signals to pulse current source 288 and to magnetic field
sources 290 and 292.
6 In operation, stripe domains 298 exist throughout
7 the magnetic medium 62 when it is in a demagnetized state. ,
8 A current pulse applied to coil 286 will chop the stripe dc~3i~s ~ -
9 298 into segments, some of which will be type A domains while ~ -
others are type B domains. Under the action of the propagation ~ ~
11 field H and propagation means 294, these two types of domains ~ -
12 will move in the direction of arrow 300.
13 In order to determine the type of domains taken
~ 14 from generator 284, the type A domains and the type B domains
z 15 can be interchanged into one another. To achieve this, a
16 current pulse is applied in coil 286. This current pulse has
17 a polarity which will produce a magnetic field which will switch
~ 18 the magnetization direction of a portion of a type A domain in
I 19 the reverse direction (i.e., into the direction of magnetization :
of magnetic medium 62). Thus, the type A domain will be con-
21 verted into a type B domain which does not extend throughout the
.; .
22 entire magnetic sheet, or which has a different wall configuration
23 than the type A domain.
s 24 ~he current pulse amplitude for switching domain type
is chosen to produce a peak magnetic field suitable for this
26 conversion. In the case of typical garnet bubble domain materials i~t
27 an amplitude of approximately 50-100 Oe is suitable. Either a
i 28 single pulse of current or a pulse train is useable, where the
l YO972-063 -110-
.',', ' ' .
. , .
'' ~
. ' ~
3tj~
1 pulse duration is about 10 milliseconds. If a pulse train is
2 used, the frequency of the pulse train can be in the approximate
3 range 10-100 cycles/second. The idea is that chopping up
4 the stripe domains will produce the type A and type B domains
within the area of coil 286 and that changing the polarity of
6 a pulse of suitable amplitude will produce the conversion of
7 one type of the domains to the other type, and vice versa.
8 In order to convert a type B domain to a type A domain, t~
9 a current pulse is applied in coil 286 having a polarity to
produce a magnetic field oppositely directed to the direction
11 of the magnetic bias field Hb. The pulse magnetic field pro~uced ~-
12 by current in coil 286 rotates the direction of magnetization of -~
13 medium 62 in the area under a type B domain. This will produce
14 a type A domain which extends throughout magnetic medi~m 62. As
in th case of switching type A domains to type B domains, the
16 same current pulse values are suitable. This applies to the ~ -
17 amplitude, duration, frequency, etc. parameters which were
18 previously mentioned.
19 A sensing means generally designated 302 is used ~-
to determine whether a domain passing thereby is a type A
21 or a type ~ domain. In this case sensing means 302 is comprised
22 of a sensing element 304 which could be, for instance, a
23 magnetoresistive sensor. A current source 306 provides measur-
24 ing current through element 304 and a voltage signal Vs is
generated across element 304 when a domain passes thereby.
26 The strength of this signal will vary depending upon whether -
27 a type A or type B domain is being sensed. This distinction
28 is noted by control 308 which selectively provides current Ic
:' ' ' , '
YO972-063 -111-
- . . -: :
-
~;;~ "' ' . ''
~L~ 3~
1 in conductor 310 to collapse domains which are of the un-
2 desired type. Therefore, a pattern of domains will move in the
3 direction of arrow 312 to an input means 38, or to other
4 circuitry for storing different types of domains, as was
S illustrated in FIG. 34 for domains having different deflection
6 properties.
7 FIG. 36 also illustrates a technique for reading
8 different size domains which can be used to determine informa-
9 tion being taken from the lattice 30. For instance, if the
pattern of domains moving in the direction of arrow 300 is the
11 pattern of domains from a column in the lattice which has
12 been removed from the lattice by the transfer register 154 ;
13 and shift register 156 (FIG. 29), a sensor such as 302 can
14 easily be used. In addition, the type A and type B domains
lS appear to have different sizes when viewea optically so that
16 standard polarized light techniques (Xerr effect and Faraday ~ -;
17 effect) can be used for sensing.
18 Another suitable sensor is a conductor loop past
19 which the different type domains move. The conductor loop
will detect a 1ux change due to the large type A domain which
21 is different than that due to the smaller type B domain. -
22 This type of sensing is also described in aforementioned United ; ~ -
~23 States patent, Serial No. 3,911,411. Also, the oscillating sense technique
24 described in United States application in Serial No. 267,877, filed June 30,
1972 now United States patent 3,842,407, can aiso be used.
Coding by dual size bubbles may cause some problems
`27 in forming a suitably regular lattice, unless the sizes of the
., .
28 bubbles are reasonably close to one another. However, these
~29 types o bubbles should be useable in a variety of lattice
~ systems~
- .
-? : :
~. :
~ YO972-063 ~ -112-
1 Codinq by Chiral State: ~
2 The IBM Technical Disclosure Bulletin, Vol. 13, , i
3 No. 10, P. 3021, March 1971 contains a paper by G. R. Henry,
4 entitled "Magnetic Domain Wall Information Storage". In that
paper, the author states that the different chiral states which
6 appear in the domain walls of bubble domains can be used for
7 information storage. That is, rather than using the presence
8 and absence of a bubble domain for information, it is proposed
g to use the right-handed chiral stat~ and left-handed chiral state
for information. Read-out of information is accompl1shed by
11 forcing an unknown-domain into a high-speed collision with a ;
12 "reference domain" of known chiral state. The domain wall -'
13 inertia drives the domains into intimate contact and, if they
14 have the same chiral state, they will rebound. If they have
i 15 opposite chiral states, they will coalesce. Therefore, the
16 existence of one or two domains after the collision is an -
17 indication of the chiral state of the unknown domain.
18 The technique taùght by G. R. Henry can be used to `~
: . . .. ~
19provide information in the form of bubble domains in the ~-
lattice array 30. An absolutely stable bubble domain state
21 can have a domain wall configuration having pure right-handed
22 or left-handed chirality as shown in FIGS. 37A and 37B. Here,
23 the bubble domains BD have domain walls 313 where the --
24 magnetization in the walls is represented by the arrows 314.
A right-handed chiral state is shown in FIG. 37A while a
26 left-handed chiral state is shown in FIG. 37B.
27 Another way to distinguish magnetic bubble domains
28 having pure chiral states, where the states are either right
. ~ , , , ~
.
~ YO972-063 -113-
: .'. . ': '
'; ~ ~ , - ,
1 handed or left handed, utilizes an in-plane magnetic field,
2 shown in FIGS. 37A and 37B as the magnetic field Hx. This
3 magnetic field causes precession of the magnetization vectors
4 314 in the walls of the domains BD, causing these domains to
move in a direction parallel or antiparallel to magnetic field
6 Hx. For instance, the domain in FIG. 37A will move in the
7 direction of arrow 316 when the pulse Hx is applied while the
8 domain in FIG. 37B will move in the direction of arrow 318
9 when pulse Hx is applied. Thus, the pure right handed and
pure left handed chirality domains can be distinguished by
11 observing their displacements in the presence of an in-plane
12 magnetic field pulse.
13 FIG. 38 shows the shape of a suitable magnetic
14 field pulse Hx for aisplacement of pure chirality domains.
There are certain requirements on the rise time and fall time
~3
16 of the pulse Hx for its use when reading domains coded in
17 terms of chirality. For instance, the optimum amplitude of
18 the pulse Hx approximately 8MS, where Ms is the magnetization -
19 of the magnetic domain material. The rise time of the pulse
i 20 Hx should be smaller than the approximate quantity y l(Hc+8aM~) 1
21 where a is the damping coefficient of the bubble domain material,
22 and y is the gyromagnetic constant of the bubble domain
23 material. For a practical low-loss rare earth iron garnet
24 material (y=1.7xlO sec 10e , a=10 , Hc=0.30e, MS=20 gauss),
, 25 this quantity is approximately 30 nanoseconds.
Y097~-063 -114-
-, ~
, ~ ' ,
~. ~
The fall time of the pulse Hx should be several times
2 this maximum allowable rise time. Under these assumptions the
3 bubble displacement produced by Hx is approximately
4 ~/[a ~(HC/8Ms)] or typically several-micrometers for garnets.
A train of such pulses may be used to obtain a cumulative
6 displacement. Thus from these quantities it appears that
7 materials with higher damping coefficients may be more desirable.
8 Due to the velocity saturation effects in materials, such
9 requirement is not an adverse factor in bubble domain velocity.
It has been discovered that a pulse of magnetic bias
11 field Hb normal to the plane of the magnetic material may affect
12 the chiral state of the bubble domain. That is, changes in the
13 magnetic bias field may cause switching of a right handed
14 chiral state to a left handed chiral state, and vice versa. In
general, a change in bias field of the approximate amount
16 2~ Ms~BL/h is sufficient to cause a change in information state
17 of the bubble domain. Here, ~BL is the width of a 13loch line
18 and h is the thickness of the magnetic bubble domain material. ~ -
19 Thus, the apparatus used for the read means must be carefully
r 20 constructed so that gradients in the magnetic bias field normal
21 to the magnetic sheet do not occur. In order to achieve this,
22 the structures shown in FIGS. 39 and 40 are utilized.
23 In FIG. 39, the magnetic bubble domain material 62 is
24 surrounded by a current-carrying coil 320 which serves as a
portion of a read meaAs for determining the chiral state of a
26 domain and, therefore, the information state. Coil 320 is
27 connected to a current source 322 which provides a current Ix. ;
28 Current in coil 320 establishes the magnetic pulse field Hx
- , ., ,:.
; YO972-063 -115-
:'
' ':: ': ~ '
`''' '.~'' ".
,,. '.' '~' '"
., '.. .' ' ~
, : . ", : . -, .- .. . . . . . .
;8
1 in the plane of medium 62.
2 In operation, the magnetic field pulse Hx causes
3 movement of the bubble domains BD in the direction of arrows
4 316 and 318, depending upon the chiral state of the bubble
domains. Detectors 322A and 322B, such as magnetoresistive
6 detectors, are used to determine the chiral state of the domains.
7 These detectors are connected to utilization means 44, such
8 as have been described previously. ~ -
9 As the magnetic pulse Hx decays in amplitude with time,
i- ~
the magnetization vectors 314 will precess in a reverse
11 direction to their original precession and reverse motion of
12 the magnetic bubble domains will occur. Accordingly, a light
13 means, such as a microscope, can be used to observe the different
14 movements of different chiral domains when the magnetic field
lS pulse is applied. ~ ~
16 It is difficult to obtain magnetic pulses having -
17 sufficiently fast rise times using current-carrying coils. --
18 Because of this, the structure of FIG. 40 uses conductor strip ~ ~-
19 lines to provide an in-plane magnetic field having sufficient
rise time. In FIG. 40, a wide conductor 324 is connected to a
21 current source 326 for providing the current Ix. Conductors
22 C and C' are connected to current source 328 via resistances
23 R and R'. Control m~ans 330 is used to activate current sources
24 326 and 328.
The basic in-plane magnetic field Hx is established
26 by current Ix in conductor 324. In order to minimize gradients
27 in the magnetic bias field normal to the plane of magnetic
28 material 62, current in conductors C and C' establish magnetic
',:
YO972-063 -116-
: ': '. '
~ ' .
3~ ~
1 fields tending to cancel any gradient magnetic fields which
2 exist in the area under conductor 324 where the determinations
3 of chirality are to be made. Thus, chirality will be accurately
4 determined and the chirality of a domain will not be switched
to a different state by spurious magnetic pulses in the z
6 direction or by gxadients in the z direction magnetic field.
7 The generating structure of FIG. 34 can be
8 used to provide bubble domains having pure chiral states.
9 Application of magnetic field pulses normal to the bubble
domain material causes splitting of stripe domains to produce
11 some domains which have pure chirality. These can be separated
12 from other domains by observing their deflection in a gradient
13 magnetic field using the apparatus shown in FIG. 34. Additionally, -
14 bubbles of one chirality can be switched to the other chirality
by application of magnetic field pulses normal to the magnetic
16 sheet 62, if this function is desired.
, .
17 III OTHER TYPES OF LATTICE SYSTEMS
18 ~FIGS. 41-48)
19 The concept of lattice arrays for use in information
s 20 handling systems is well demonstrated by the embodiment shown
21 in FIGS. 41-48. Here, the interactive elements 32 are simply
22 made by attaching magnetic elements to styrofoam balls which
23 are free to float on the surface of a liquid, such as water.
24 These magnetic styrofoam balls will interact with one another
and can be moved into a confined array such as a lattice
26 arrangement and taken from the array. Coding is conveniently -~
27 achieved by coloring these styrofoam balls differently or by
'~ , ;.- '
YO972-063 -117-
':~
, ' ,
, ' :- ' '
;". ' - , .:'.
; ~ ' ''
;~3~ 8
1 giving them other differing physical properties.
FIG. 41 shows a general view of confined arrays comprising styro-
foam balls haYing magnetic elements in them. In more detail, a tank
334 has therein a liquid 336 which is able to support the interactive
elements 32. In this case, the interactive elements are elements which
are free to float on liquid 336 and which exert interactive forces on
one another tending to determine the positions of the elements 32 with
respect to one another. In FIG. 41, two lattices 30 and 30' are
shown, each of which is comprised of a plurality of interactive
elements 32.
The interactive elements used in the tank 334 are shown in more
detail in FIG. 42. In particular, each interactive element 32 is made
of a material which will float on the liquid 336. Small permanent
magnets 338 are located in each of the elements 32 in order to give
. them interactive forces. In FIG. 42, one of the interactive elements
'. 32 has its top surface painted black to indicate information different. . - . .
; from that of the other interactive element. The numbers showing the
dimensions of the interactive elements and the dimensions of the
- permanent magnets 338 are dimensions which were used when a lattice
~ 20 file comprising the styrofoam balls was made and operated. In that ~ -
. , .
case, the magnetic elements had diameters of one inch while the per-
manent magnets were 3/16" by 1/2". The liquid 336 was water. Thus,
depending upon the sequence of light and dark styrofoam elements 32
in the lattices 30 and 30', a pattern of information can be obtained
which is directly viewable by an observer or by a scanner 339 (FIG. 41).
'
'~
;~ :
: , "
~ ~
.,
.. . .
Y09-72~063 - 118 -
. .
- ., .. . . . . , . . , .. . . . , . , . , ., , ~ ~ ~, . ; , . . .
3Ei~ : ~
FIG. 43 shows a schematic arrangement of the components
2 of the system of FIG. 41. The two lattices 30 and 30' have
3 shift registers 338A and 338B located on opposite ends of the
4 lattices. These shift register~ are used to convey information
S to the lattices 30 and 30' as well as to receive information which
6 has been removed from the lattices 30 and 30'.
7 The dimensions shown for the size of the lattices and
8 for the distance between the lattices are dimensions whi~h were
9 used in the example previously discussed using the one inch
interactive elements 32. That is, the total structure of FIG. 43
11 measured 36" x 36". Each lattice 30 and 30' contained 16 rows -
12 of eight interactive elements per row.
13 As will be apparent, the structure of FIG. 43 also
14 includes means to bring elements 32 into the lattice and to ~ i
remove elements 32 from the lattice. These means are provided
16 for each of the lattices 30 and 30'. In the example being - ~ -
17 discussed, conductor patterns are used for the input and output
18 means and for the shift register used to move elements from
19 one lattice to another. These cor~auctor patterns can be spatially
located over one another and can be immersed in the fluid 336 used . .
21 in the tank. Particularly if the fluid is water, its conductivity
22 will not affect the operation of conductor input/output means
23 and conductor shift registers.
24 FIG. 44 ahows the input and output means for each of ~;
~25 the lattices 30 and 30'. The direction of movement of the
~-26 elements 32 i5~ indicated by the arrows Al, A2 and Al', A2'.
z7 In this ~igure, the input means for lattice 30 is designated 38
" 28 while the output means iq designated 40. For lattice 30', the
','
.
~: YOg72-063 , -119-
` '
...
.,: . :
:
1 input means is designated 38' while the output means is
2 designated 40'. These input and output means are comprised
3 of current-carrying conductors C2, C3 and C4. As shown in
4 FIG. 4~, the position of conductors C3 and C4 is interchanged
in the input and output means so that the current sequence as
6 shown in FIG. 47 ejects a row of elements along the directions
7 of motion Al and Al' and simultaneously injects a row of
8 elements along the directions of motion indicated by A2 and A2'.
9 Conductor Cl is used to provide confinement for .
lattices 30 and 30' and to keep these lattices separate from
11 one another. In addition, loops of conductors C2, C3 and C4 .
12 are used to provide confinement for lattices 30 and 30'. . .. ~.
13 Conductor Cl is connected to a current.source 340, ;
14 while conductors C2, C3, and C4 are connected to current sources .
342, 344, and 346, respectively. The various current sources
16 340-346 receive inputs from control means 348 in order to .
17 properly synchronize movement of the interactive elements 32.
18 Shift registers 338A and 338B are sho~n as dashed . :.
19 boxes in FIG. 44. Theæe shift registers are also comprised
20 of conductor patterns and can be located in a plane above or ~ -
21 below the conductors Cl-C4 used for confinement and/or input~
22 output. Conductors Cl-C4 can be located in the same plane.
23 The ~hift registers are shown in FIG. 48 which will be described
24 later.
FIGS. 45A-45C show the position of an interactive
26 element 32 which is being ejected from a lattice such as lattice
27 30'. These figures are for times T = 1, 2, and 3, respecti~ely. .~:
28 The presence of currents in the conductor~ is indi.cated by ~
29 arrows on the conductors, where the arrows indi.cate the current
30 directions. For instance, at T = 1 a current exi.sts in conductor ~.
31 C2 but not in conductors C3 and C4. FIGS. 45A-45C sho~ the
, ~ .
YO972-063 -120- ::
.. .
~ j~i, ' . ' ~- :
., :~ ,~. . .
3t~8
1 element 32 moving from within the lattice to a position outside
2 the lattice. When the element is outside the lattice (for
3 instance, in the position shown in FIG. 46A) an underlying
4 shift register is used to move the element 32 to another
lattice or to another location in the array. In the case of
6 FIG. 44, shift register 338B would move the domain 32 to the
7 input means 38 for lattice 30 where it could be injected into
8 lattice 30.
g FIGS. 46A-46C show the operation of an input means
38 for moving the element 32 into the lattice array 30.
11 The currents in conductors C2, C3 and C4 are denoted by arrows,
12 as was the case for FIGS. 45A-45C. When the element 32 moves ~
13 from a position approximately centered on conductor C4 to a ,
14 position within conductor C2 loop it has moved to within ;
lattice 30.
16 For the example-which has been described, the one
17 inch interactive elements are moved by conductors C2-C4 when
18 currents having an amplitude of approximately 30 amps exist in
19 these conductors. That is, current sources 342-346 provide
currents of approximately 30 amps for this type of interactive
1 21 element.
' 22 FIG. 47 shows the presence of currents in the
23 conductors C2, C3, and C4 during the operations of shifting ~-
' 24 elements in registers 338A and 338B, ejecting elements from ~
a lattice, and injecting elements 32 into a lattice. The ~ -
26 large dots indicate that current is present in specified
27 conductors whereas the absence of a dot indicates that current
.,
~ YO972-0~3 -121-
" . :~:
'
.'~ ' ~ .
.. . .
.,j . .
6~
1 is r.ot present in a conductor. For confinement of the array
2 elements 32, current is always present in conductor Cl.
3 FIG. 48 shows one period of a conductor pattern
4 suitable for shift registers 338A and 338B. This register is
comprised of conductors C5, C6, C7, and C8. Conductors C5,
6 C6, and C7 provide a three-phase conductor propagation pattern -~
7 while conductor C8 lS a loop which serves as a guide rail to
8 keep the interactive elements 32 in the proper propagation track.
9 Conductors C5-C8 are connected to current sources
350, 352, 354, and 356, respectlvely. These current sources
11 350-356 receive inputs from the control means 348 sho~wn in
12 FIG. 44.
13 Repetitive sequences of current cycles are applied -
14 during the time Tl to shift an ejected row of elements from ;
the output means 40 and 40' to the input means 38 and 38',
16 respectively. Sequential positions of an element 32 during
17 one cycle of shift register operation are designated by A, B,
18 and C in FIG. 48. FIG. 49 shows the different currents used -
19 during one cycle of shift register operation to move an
! 20 element 32 from position A to position B to position C. In
21 this figure, a + sign indicates that a current is flowing into
22 the indicated conductor. This current divides equally and --
` 23 returns through the propagation conductors designated G. For
24 instance, an element 32 will mDve from position A to position B
of register 338 when conductors C5 and C7 are grounded and
7 26 conductors C6 and C8 have currents in them. ;
Y 27 In the particular example being described using one ~ ~
' 28 inch diameter interactive elements shown in FIG. 42, the ~ -
29 current amplitudes for shifting were approximately six amps.
That is, six-amp currents were provided by the current sources
31 350-356. ~
'~ ' ' .'. . :': ~'
~, YO972-063 -122-
.. . .
,: ~
. i . :. ~ .
., .
3~8
1 Thus, embodiment of FIGS. 41-49 is suitable to
2 demonstrate the use of arrays of confined elements to provide
3 storage of information and also for display purpo~es. For
4 instance, the interactive elements 32 can be coded in ;;
accordance with their color to provide pattern~ of light
6 and dark elements which can be viewed by the individual or
7 viewed by optical apparatus. As another alternative some of
8 the elements can be coded with metallic caps to give them a
9 different electrical resistance, for coding as information.
.'
:-
SUMMARY
11 What has been shown is a new type of information
12 handling apparatus which uses confined arrays (such as lattices)
13 containing interactive elements. The elements within the
14 lattice can be coded to have information associated with them -~
or the lattice can be a part of other apparatus where it i
16 desired to store closely packed interactive elements for use ~ -
17 elsewhere. Many types of interactive elements can be used
:
18 within the lattice, although the lattice is easily illustrated
19 and shown with interactive elements such as magnetic bubble -
domains and elements which are supported in a conveniently selected
21 medium. Additionally, the interactive elements can be used
22 in lattice arrays of many dimensions, for instance three-
23 dimensional lattice arrays where interactive elements (such as
24 monopole electrostatic elements which exert repulsive forces
in 3 dimensions) are maintained within a medium (such as a
26 colloidal solution) in which they can move in three dimensions.
27 As an example, electrostatic or magnetic interactive elements
~, 28 can exist in fluid so that a three-dimensional lattice is
29 provided.
:. , j . ' ~ ~ .
YO972-063 -123-
.
; . . ,:
.- '` .
`
. .
3~i8
1 Many other technique6 can easily be envisioned by
2 those of skill in the art for movement of the interactive elements
3 into and out of the lattice and for realizing the various
4 functions described herein. However, this teaching will
provide sufficient basis for the provision of other apparatus :.
6 and systems incorporating lattice arrays of various sizes for
7 many functions.
8 What is claimed is:
JES/dc/cm/eg , ~:
August 3, 1973
,. . .
~ , ; ~ ~ ' ". '
YO972-063 -124-
.'' '` ;' ~ ' ':
'.:,., ". : ~
~'~ .,~,. .
;,' 'i? ' ' "'' ~''"
. .
