Title:
DOMAIN LOGIC ARRANGEMENT
United States Patent 3676871
Abstract:
A single wall domain logic arrangement is defined by a serpentine propagation conductor in a layer of material in which single wall domains can be moved. The conductor, when energized, provides a drive field for domains in a number of channels between a plurality of inputs and outputs. The arrangement exhibits a signal at outputs determined by the pattern of domains at the inputs.
Application Number:
05/145771
Publication Date:
07/11/1972
Filing Date:
05/21/1971
View Patent Images:
Export Citation:
Assignee:
Bell Telephone Laboratories, Incorporated (Berkeley Heights, NJ)
Primary Class:
Other Classes:
365/37, 365/13, 365/19, 307/407, 307/400, 365/41
International Classes:
G11C19/08; G11C19/00; G11C11/14; G11C19/00
Field of Search:
340/174TF,174SR
Other References:
IBM Technical Disclosure Bulletin Vol. 13 No. 6 Nov. 1970 pg. 1581-1582.
Primary Examiner:
Moffitt, James W.
Claims:
What is claimed is
1. An arrangement comprising a layer of material in which single wall domains can be moved, and means for defining in said layer first, second, and third multistage channels for said domains, said last-mentioned means including an electrical conductor of a serpentine geometry for defining a plurality of stages for each of said channels, said geometry having a modification therein for moving to said second channel a domain in one of said first or third channel in the absence of a domain in a corresponding stage of the other of said first or third channel when said conductor is activated, said arrangement also comprising means for introducing domains selectively to said first and second channels.
2. An arrangement in accordance with claim 1 also including means for detecting the presence or absence of a domain in said second channel.
3. An arrangement in accordance with claim 2 wherein said modification is of a geometry such that the presence of first and second domains in like stages of said first and third channels respectively inhibits movement of either of said first and second domains into said second channel, and means for detecting the presence and absence of a domain in said first and third channels.
4. An arrangement in accordance with claim 3 wherein said modification includes a first portion of said conductor defining a first stage of said channel wherein said first portion includes first and second straight line sections disposed at a first angle with respect to one another so that a signal applied to the conductor generates at said first position a field tending to move a domain in the direction of a line bisecting said first angle.
5. An arrangement in accordance with claim 4 wherein said modification also includes a second portion of said conductor defining a second stage of said channel wherein said second portion also includes first and second straight line sections disposed at a first angle with respect to one another so that a signal applied to the conductor generates at said first portion a field tending to move a domain in the direction of a line bisecting said first angle.
6. An arrangement in accordance with claim 5 also including means for applying a bipolar signal to said conductor for generating drive fields of a first value for moving said domains in said channels.
7. An arrangement in accordance with claim 6 also including means for providing a bias field in said layer of a second value sufficiently greater than said first value for maintaining a constant geometry for said domains when moved.
8. An arrangement comprising a layer of material in which single wall domains can be moved, and means for defining in said layer first, second, and third propagation channels for said domains, said last-mentioned means comprising a conductor arrangement including a conductor of serpentine geometry having in a portion thereof straight line sections disposed at angles to one another in a manner to move domains from said first or third channel to said second when said conductor is pulsed in the absence of a domain moving synchronously in the other of said first or third channel.
1. An arrangement comprising a layer of material in which single wall domains can be moved, and means for defining in said layer first, second, and third multistage channels for said domains, said last-mentioned means including an electrical conductor of a serpentine geometry for defining a plurality of stages for each of said channels, said geometry having a modification therein for moving to said second channel a domain in one of said first or third channel in the absence of a domain in a corresponding stage of the other of said first or third channel when said conductor is activated, said arrangement also comprising means for introducing domains selectively to said first and second channels.
2. An arrangement in accordance with claim 1 also including means for detecting the presence or absence of a domain in said second channel.
3. An arrangement in accordance with claim 2 wherein said modification is of a geometry such that the presence of first and second domains in like stages of said first and third channels respectively inhibits movement of either of said first and second domains into said second channel, and means for detecting the presence and absence of a domain in said first and third channels.
4. An arrangement in accordance with claim 3 wherein said modification includes a first portion of said conductor defining a first stage of said channel wherein said first portion includes first and second straight line sections disposed at a first angle with respect to one another so that a signal applied to the conductor generates at said first position a field tending to move a domain in the direction of a line bisecting said first angle.
5. An arrangement in accordance with claim 4 wherein said modification also includes a second portion of said conductor defining a second stage of said channel wherein said second portion also includes first and second straight line sections disposed at a first angle with respect to one another so that a signal applied to the conductor generates at said first portion a field tending to move a domain in the direction of a line bisecting said first angle.
6. An arrangement in accordance with claim 5 also including means for applying a bipolar signal to said conductor for generating drive fields of a first value for moving said domains in said channels.
7. An arrangement in accordance with claim 6 also including means for providing a bias field in said layer of a second value sufficiently greater than said first value for maintaining a constant geometry for said domains when moved.
8. An arrangement comprising a layer of material in which single wall domains can be moved, and means for defining in said layer first, second, and third propagation channels for said domains, said last-mentioned means comprising a conductor arrangement including a conductor of serpentine geometry having in a portion thereof straight line sections disposed at angles to one another in a manner to move domains from said first or third channel to said second when said conductor is pulsed in the absence of a domain moving synchronously in the other of said first or third channel.
Description:
FIELD OF THE INVENTION
This invention relates to data processing arrangements and more particularly to such arrangements in which information is represented as single wall domains.
BACKGROUND OF THE INVENTION
The term "single wall domain" refers to a magnetic domain which is movable in a layer of a suitable magnetic material and is encompassed by a single domain wall which closes on itself in the plane of that layer.
Propagation arrangements for moving a domain are designed to produce magnetic fields of a geometry determined by the layer in which a domain is moved. Most materials in which single wall domains are moved are characterized by a preferred magnetization direction, for all practical purposes, normal to the plane of the layer. The domain accordingly constitutes a reverse magnetized domain which may be thought of as a dipole oriented transverse, nominally normal to the plane of the layer. Accordingly, the movement of a domain is accomplished by the provision of an attracting magnetic field normal to the layer and at a localized position offset from the position occupied by the domain. A succession of such fields causes successive movements of a domain as is well known.
One propagation arrangement comprises a pattern of electrical conductors each designed to form conductor loops which generate the requisite fields when externally pulsed. The loops are interconnected and pulsed in a three-phase manner to produce shift register operation as disclosed in A. H. Bobeck, U. F. Gianola, R. C. Sherwood, and W. Shockley U.S. Pat. No. 3,460,116 issued Aug. 5, 1969.
In an alternative conductor propagation arrangement, bipolar pulses are applied to a serpentine conductor for displacing a domain pattern therealong between positions defined by magnetically soft overlay elements described, for example, in R. F. Fischer, U.S. Pat. No. 3,564,518 issued Feb. 16, 1971.
BRIEF DESCRIPTION OF THE INVENTION
In accordance with an illustrative embodiment of the present invention, a serpentine conductor defines three parallel domain channels, two side channels and a central channel, between two input and three output positions. Domains are introduced selectively at inputs to the side channels. The serpentine conductor includes a geometric modification which comprises straight line sections at angles to one another and oriented perpendicular with respect to the axes of the channels. The fields generated at the modification, when the conductor is pulsed, tend to move a domain in either side channel into the central of the three channels in the absence of a domain simultaneously introduced at the second input position. In this instance, a signal occurs only at the output position associated with the central channel.
When two inputs are present simultaneously, the repulsion forces between domains tend to separate domains and prevent movement of either into the central channel. When two inputs occur simultaneously, signals are produced at the output positions of the side channels.
The relative values of the bias field for maintaining domain size constant and the drive field for moving domains are adjusted for avoiding domain strip-out during propagation.
BRIEF DESCRIPTION OF THE DRAWING
The FIGURE is a schematic illustration of a single wall domain logic arrangement in accordance with this invention.
DETAILED DESCRIPTION
The figure to be specific, shows a domain logic arrangement 10 including a layer 11 of a material in which single wall domains can be moved. A conductor 13 having a generally serpentine geometry defines a plurality of multistage channels in which domains are moved when the conductor is energized.
Conductor 13 is connected to a propagation signal source represented by block 14 of the figure and is under the control of control circuit 15. Source 14 applies a bipolar current (± I of FIG. 1) to conductor 13 to generate the requisite consecutive field configurations for moving domains from left to right as viewed in the figure for layer 11. The layer, in this connection, is assumed to have a preferred direction of magnetization along an axis normal to the plane of the layer. The resulting displacement of domains is between positions defined by magnetically soft dots (not shown) as is discussed further hereinafter.
Conductor 13 can be seen to include modifications in its geometry illustratively in its second and third legs, l2 and l3, from the left as viewed. A domain which occupies a position tangent to a conductor, when the conductor is pulsed appropriately, moves to the opposite side of the wire where it is again tangent to the wire. Consequently, a domain introduced at an input position represented by arrow A in the figure is tangent to the first leg l1 of conductor 13 occupying a position designated 1A in the figure to designate state one of channel A. When an opposite polarity current pulse is applied to the conductor, the domain moves to position 2A in the figure. Actually, the domains occupy positions slightly beneath conductor 13, as shown, for reasons which will be explained further hereinafter.
At this juncture, the domain encounters the modified portion of conductor 13 at leg l2. The modification comprises two straight line sections l2A and l2B at an angle A with respect to one another rather than a single straight line section. This arrangement functions to cause a domain to move downward and to the right (as viewed) from the axis of a side channel (channel A) to a position designated 3A in the figure. Similarly, the modification of leg l3 of the conductor includes sections l3A and l3B again disposed at an angle to one another to cause the domain to again move downward as viewed to a position designated 4C when the signal polarity in conductor 13 is next alternated. In each instance, the domain responds to a signal pulse polarity change by moving to a position tangent to the side of conductor 13 opposite to that occupied by the domain when the alternation occurs. The conductor geometry is arranged to define a succession of such positions which initiate at a side channel (viz: the A channel) and terminates at a central channel (viz: the C channel) the axis of which corresponds to a line bisecting angle A of the figure.
The next consecutive legs of conductor 13 are of straight line geometry tending to move domains from the left to the right thereof when an alternation occurs. In response, the domain moves consecutively into positions 4C, 5C, and 6C to apply an output signal at the position of arrow C to utilization circuit 17 also under the control of control circuit 15.
An input signal providing a domain in channel B at the position of arrow B in FIG. 1 in the absence of a domain at the input of channel A produces an analogous result. This is apparent in view of the symmetry in the modifications of legs l2 and l3 of conductor 13 about the horizontal broken line terminating to the right in arrow C It should be clear then, that an input resulting in a domain in either of channel A or channel B results in a domain occurring at the output position of channel C thus producing a logical OR operation.
On the other hand, operation of the arrangement is quite different when a domain is introduced simultaneously at the inputs of channels A and B. It is to be recalled that single wall domains repel one another. When two domains occupy first positions 1A and 1B and, then, positions 2A and 2B of the figure simultaneously, the repulsion forces therebetween deny to each domain the next normal position which one domain would occupy in the absence of the second as described above (viz. positions 3A or 3B). Consequently, the domains move to positions 3D and 3E instead when the next alternation of the propagate current in conductor 13 occurs.
Domains, so positioned, encounter the modification of leg l3 of conductor 13 when the next subsequent alternation occurs. The modifications of leg l3 encountered in this instance comprise straight line sections l3C and l3D which are angularly disposed to one another, as above, and cause the domains to move to positions designated 4D and 4E in the figure. For subsequent alternations of the propagate current, the domains initially in channels A and B occupy the consecutive positions 4D, 5D, and 6D and positions 4E, 5E, and 6E to produce signals at output positions represented by arrows D and E of the figure, respectively. It should be clear then that a domain simultaneously at input positions in channel A and channel B produces output signals at output positions at D and E and a null at the output of channel C, thus providing a logical AND operation at the output of either of channel A and B and an exclusive-OR function at the output of channel C.
If a binary zero is represented as the absence of a domain and a binary one by the presence of a domain, the operation of the arrangement of the figure may be summarized by the following truth table:
INPUT OUTPUT A B C D E ____________________________________________________________ ______________ 0 0 0 0 0 0 1 1 0 0 1 0 1 0 0 1 1 0 1 1 ____________________________________________________________ ______________
The outputs of channels D and E (AND functions) may be recognized to provide a "carry" bit for operation of the arrangement as a full adder circuit. The carry bit may be employed in a familiar manner as an input to a following stage.
The domain propagation circuit is shown for simplicity to comprise a single serpentine conductor. A single conductor is suitable for propagation when an array of magnetically soft dots define stable positions for domains in layer 11. The positions so defined are offset typically by one-quarter of a domain diameter to the right as viewed in the figure from those positions to which domains are moved in response to alternations in conductor 13. The use of such dots is well understood in the art and results in this instance in domains assuming positions beneath rather than tangent to conductor 13 as mentioned above. The presence of the dots is assumed herein without further discussion. Alternatively, etched depressions can be used for the same purpose. If the dots (or depressions) are not present, a second serpentine conductor (not shown) offset from the first and of like geometry serves the same purpose, also. If a second conductor is used, the pulses applied to the two conductors (as above) are interleaved to produce unidirectional domain movement.
Domains moved in a manner to provide logic operations as described above originate at the input positions of channels A and B of the figure. It is contemplated that the input positions may constitute the terminations of domain propagation paths also defined in layer 11 by well-known means. Alternatively, domain generators of the type described in U.S. Pat. No. 3,555,527 of A. J. Perneski issued Jan. 12, 1971 and well-known modifications thereof are suitable for generating domains selectively. Such sources are represented in FIG. 1 by block 20 designated input signal source and operate under the control of control circuit 15.
The presence of domains may be detected at the output positions by well-known magneto-resistive devices which apply signals to utilization circuit 17 or, alternatively, the output domain pattern may be propagated into domain channels for utilization in domain circuitry define directly in layer 11.
Block 21 of the figure represents a bias field source, typically a permanent magnet or an exchange coupled magnetic layer, for maintaining constant the domain size in layer 11 during operation. A typical bias field is about 30 oersteds for moving a domain of about 125 microns in a layer about 75 microns thick. A drive field of about 2 oersteds avoids domain strip out in this case.
The illustrative arrangement includes two inputs. It should be clear, on the other hand, that additional inputs can be employed and that the conductor geometry can be adapted to provide channel changes by domain repulsion as described, but in a cascaded arrangement.
Therefore, what has been described is considered only illustrative of the principles of this invention and modifications thereof can be devised by those skilled in the art in accordance with those principles within the spirit and scope of this invention. For example, it may be appreciated that the modification of leg l3 of conductor 13 to include Sections l3C and l3D is not necessary. In the absence of such a modification, the spacing between the axes of the centers of positions 4D, 5D, and 6D and positions 4E, 5E, and 6E will be greater.
This invention relates to data processing arrangements and more particularly to such arrangements in which information is represented as single wall domains.
BACKGROUND OF THE INVENTION
The term "single wall domain" refers to a magnetic domain which is movable in a layer of a suitable magnetic material and is encompassed by a single domain wall which closes on itself in the plane of that layer.
Propagation arrangements for moving a domain are designed to produce magnetic fields of a geometry determined by the layer in which a domain is moved. Most materials in which single wall domains are moved are characterized by a preferred magnetization direction, for all practical purposes, normal to the plane of the layer. The domain accordingly constitutes a reverse magnetized domain which may be thought of as a dipole oriented transverse, nominally normal to the plane of the layer. Accordingly, the movement of a domain is accomplished by the provision of an attracting magnetic field normal to the layer and at a localized position offset from the position occupied by the domain. A succession of such fields causes successive movements of a domain as is well known.
One propagation arrangement comprises a pattern of electrical conductors each designed to form conductor loops which generate the requisite fields when externally pulsed. The loops are interconnected and pulsed in a three-phase manner to produce shift register operation as disclosed in A. H. Bobeck, U. F. Gianola, R. C. Sherwood, and W. Shockley U.S. Pat. No. 3,460,116 issued Aug. 5, 1969.
In an alternative conductor propagation arrangement, bipolar pulses are applied to a serpentine conductor for displacing a domain pattern therealong between positions defined by magnetically soft overlay elements described, for example, in R. F. Fischer, U.S. Pat. No. 3,564,518 issued Feb. 16, 1971.
BRIEF DESCRIPTION OF THE INVENTION
In accordance with an illustrative embodiment of the present invention, a serpentine conductor defines three parallel domain channels, two side channels and a central channel, between two input and three output positions. Domains are introduced selectively at inputs to the side channels. The serpentine conductor includes a geometric modification which comprises straight line sections at angles to one another and oriented perpendicular with respect to the axes of the channels. The fields generated at the modification, when the conductor is pulsed, tend to move a domain in either side channel into the central of the three channels in the absence of a domain simultaneously introduced at the second input position. In this instance, a signal occurs only at the output position associated with the central channel.
When two inputs are present simultaneously, the repulsion forces between domains tend to separate domains and prevent movement of either into the central channel. When two inputs occur simultaneously, signals are produced at the output positions of the side channels.
The relative values of the bias field for maintaining domain size constant and the drive field for moving domains are adjusted for avoiding domain strip-out during propagation.
BRIEF DESCRIPTION OF THE DRAWING
The FIGURE is a schematic illustration of a single wall domain logic arrangement in accordance with this invention.
DETAILED DESCRIPTION
The figure to be specific, shows a domain logic arrangement 10 including a layer 11 of a material in which single wall domains can be moved. A conductor 13 having a generally serpentine geometry defines a plurality of multistage channels in which domains are moved when the conductor is energized.
Conductor 13 is connected to a propagation signal source represented by block 14 of the figure and is under the control of control circuit 15. Source 14 applies a bipolar current (± I of FIG. 1) to conductor 13 to generate the requisite consecutive field configurations for moving domains from left to right as viewed in the figure for layer 11. The layer, in this connection, is assumed to have a preferred direction of magnetization along an axis normal to the plane of the layer. The resulting displacement of domains is between positions defined by magnetically soft dots (not shown) as is discussed further hereinafter.
Conductor 13 can be seen to include modifications in its geometry illustratively in its second and third legs, l2 and l3, from the left as viewed. A domain which occupies a position tangent to a conductor, when the conductor is pulsed appropriately, moves to the opposite side of the wire where it is again tangent to the wire. Consequently, a domain introduced at an input position represented by arrow A in the figure is tangent to the first leg l1 of conductor 13 occupying a position designated 1A in the figure to designate state one of channel A. When an opposite polarity current pulse is applied to the conductor, the domain moves to position 2A in the figure. Actually, the domains occupy positions slightly beneath conductor 13, as shown, for reasons which will be explained further hereinafter.
At this juncture, the domain encounters the modified portion of conductor 13 at leg l2. The modification comprises two straight line sections l2A and l2B at an angle A with respect to one another rather than a single straight line section. This arrangement functions to cause a domain to move downward and to the right (as viewed) from the axis of a side channel (channel A) to a position designated 3A in the figure. Similarly, the modification of leg l3 of the conductor includes sections l3A and l3B again disposed at an angle to one another to cause the domain to again move downward as viewed to a position designated 4C when the signal polarity in conductor 13 is next alternated. In each instance, the domain responds to a signal pulse polarity change by moving to a position tangent to the side of conductor 13 opposite to that occupied by the domain when the alternation occurs. The conductor geometry is arranged to define a succession of such positions which initiate at a side channel (viz: the A channel) and terminates at a central channel (viz: the C channel) the axis of which corresponds to a line bisecting angle A of the figure.
The next consecutive legs of conductor 13 are of straight line geometry tending to move domains from the left to the right thereof when an alternation occurs. In response, the domain moves consecutively into positions 4C, 5C, and 6C to apply an output signal at the position of arrow C to utilization circuit 17 also under the control of control circuit 15.
An input signal providing a domain in channel B at the position of arrow B in FIG. 1 in the absence of a domain at the input of channel A produces an analogous result. This is apparent in view of the symmetry in the modifications of legs l2 and l3 of conductor 13 about the horizontal broken line terminating to the right in arrow C It should be clear then, that an input resulting in a domain in either of channel A or channel B results in a domain occurring at the output position of channel C thus producing a logical OR operation.
On the other hand, operation of the arrangement is quite different when a domain is introduced simultaneously at the inputs of channels A and B. It is to be recalled that single wall domains repel one another. When two domains occupy first positions 1A and 1B and, then, positions 2A and 2B of the figure simultaneously, the repulsion forces therebetween deny to each domain the next normal position which one domain would occupy in the absence of the second as described above (viz. positions 3A or 3B). Consequently, the domains move to positions 3D and 3E instead when the next alternation of the propagate current in conductor 13 occurs.
Domains, so positioned, encounter the modification of leg l3 of conductor 13 when the next subsequent alternation occurs. The modifications of leg l3 encountered in this instance comprise straight line sections l3C and l3D which are angularly disposed to one another, as above, and cause the domains to move to positions designated 4D and 4E in the figure. For subsequent alternations of the propagate current, the domains initially in channels A and B occupy the consecutive positions 4D, 5D, and 6D and positions 4E, 5E, and 6E to produce signals at output positions represented by arrows D and E of the figure, respectively. It should be clear then that a domain simultaneously at input positions in channel A and channel B produces output signals at output positions at D and E and a null at the output of channel C, thus providing a logical AND operation at the output of either of channel A and B and an exclusive-OR function at the output of channel C.
If a binary zero is represented as the absence of a domain and a binary one by the presence of a domain, the operation of the arrangement of the figure may be summarized by the following truth table:
INPUT OUTPUT A B C D E ____________________________________________________________ ______________ 0 0 0 0 0 0 1 1 0 0 1 0 1 0 0 1 1 0 1 1 ____________________________________________________________ ______________
The outputs of channels D and E (AND functions) may be recognized to provide a "carry" bit for operation of the arrangement as a full adder circuit. The carry bit may be employed in a familiar manner as an input to a following stage.
The domain propagation circuit is shown for simplicity to comprise a single serpentine conductor. A single conductor is suitable for propagation when an array of magnetically soft dots define stable positions for domains in layer 11. The positions so defined are offset typically by one-quarter of a domain diameter to the right as viewed in the figure from those positions to which domains are moved in response to alternations in conductor 13. The use of such dots is well understood in the art and results in this instance in domains assuming positions beneath rather than tangent to conductor 13 as mentioned above. The presence of the dots is assumed herein without further discussion. Alternatively, etched depressions can be used for the same purpose. If the dots (or depressions) are not present, a second serpentine conductor (not shown) offset from the first and of like geometry serves the same purpose, also. If a second conductor is used, the pulses applied to the two conductors (as above) are interleaved to produce unidirectional domain movement.
Domains moved in a manner to provide logic operations as described above originate at the input positions of channels A and B of the figure. It is contemplated that the input positions may constitute the terminations of domain propagation paths also defined in layer 11 by well-known means. Alternatively, domain generators of the type described in U.S. Pat. No. 3,555,527 of A. J. Perneski issued Jan. 12, 1971 and well-known modifications thereof are suitable for generating domains selectively. Such sources are represented in FIG. 1 by block 20 designated input signal source and operate under the control of control circuit 15.
The presence of domains may be detected at the output positions by well-known magneto-resistive devices which apply signals to utilization circuit 17 or, alternatively, the output domain pattern may be propagated into domain channels for utilization in domain circuitry define directly in layer 11.
Block 21 of the figure represents a bias field source, typically a permanent magnet or an exchange coupled magnetic layer, for maintaining constant the domain size in layer 11 during operation. A typical bias field is about 30 oersteds for moving a domain of about 125 microns in a layer about 75 microns thick. A drive field of about 2 oersteds avoids domain strip out in this case.
The illustrative arrangement includes two inputs. It should be clear, on the other hand, that additional inputs can be employed and that the conductor geometry can be adapted to provide channel changes by domain repulsion as described, but in a cascaded arrangement.
Therefore, what has been described is considered only illustrative of the principles of this invention and modifications thereof can be devised by those skilled in the art in accordance with those principles within the spirit and scope of this invention. For example, it may be appreciated that the modification of leg l3 of conductor 13 to include Sections l3C and l3D is not necessary. In the absence of such a modification, the spacing between the axes of the centers of positions 4D, 5D, and 6D and positions 4E, 5E, and 6E will be greater.