JPH0219633B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a Josephson junction device,
In particular, the present invention relates to a Josephson track device that can perform various digital functions by using fluxons running on the Josephson track.

In recent years, the development of cryogenic electronics has been remarkable, and Josephson junction devices in particular are highly anticipated as future digital functional devices due to their high speed and low power consumption.

However, the Josephson
The device defines a voltage state in which a voltage is generated across the junction, that is, a state in which the phase difference of the macroscopic wave function continues to rotate between the superconductors forming the junction, as either logic "0" or "1". The other state, that is, the zero voltage state, was associated with the other logical value. Therefore, although their power consumption is certainly lower than that of semiconductor devices, it cannot necessarily be said that they have succeeded in reducing power consumption to the ultimate level.

The present invention has been made in view of this point, and has the possibility of extremely low power consumption due to a new operating principle different from that of conventional Josefson devices, and has a synchronous oscillation function. Frequency division function, counter function, logical product and logical sum function, inverter function, and even full addition function, memory function, etc.
The objective is to provide a Josephson device as a basic device structure that is ideal for assembling devices that realize many digital functions.

When the length of one side of a Josephson junction is longer than the Josephson penetration length λ _J , a vortex-like current state that flows through the junction occurs and exists under certain conditions.
However, it is known that the current state portion can be moved at high speed.

That is, as shown in FIG. 1A, it consists of a pair of superconductors 2 and 3 and a junction 4, and the length l of one side is about four times longer than the Josephson penetration depth λ _J , and the length l of the other side In a Josephson junction whose length (line width) w is smaller than λ _J and which is long in one dimension when viewed as a whole, an eddy current state part a can exist inside, and the magnetic field H localized in this part a is It can exist. Such a Josephson junction is especially called a Josephson line.

FIG. 1B shows the localization of the magnetic field H, with the length direction of the Josephson line taken as the horizontal axis. The spatial spread of the eddy current is also comparable to the spread of the magnetic field, and its magnitude is about λ _J to 4λ _J.

In Figure 1A, if loop b is chosen in which the eddy current is sufficiently attenuated, the total magnetic flux passing through this loop is quantized, and one magnetic flux quantum Φ ₀ (2.07×10 ^-15 Wb) is equal to This quantized eddy current state is called a fluxon because it behaves like a single particle in the Josefson line.

Furthermore, as mentioned above, the fact that the total magnetic flux is equal to Φ ₀ corresponds to the fact that the twist of the phase difference in the part where the fluxon exists is exactly 2π. Therefore, it is suggested that if fluxons, which have an ultimately small phase difference of 2π, are used as information carriers, a device with extremely low power consumption can be constructed.

As mentioned earlier, fluxons can move in the Josefson line, and the voltage generated as they move causes a quasi-particle current to flow, gradually losing energy and slowing down the movement. However, as shown in FIG. 1A, by applying an appropriate bias current along the direction of arrow c, energy can be given to the fluxon, and the moving speed can be intentionally increased. Conversely, due to the Lorentz force generated between the magnetic flux and the bias current, the arrow c
If a current is applied in the opposite direction, this becomes a braking current, and the moving speed of the fluxon can be intentionally reduced.

On the other hand, the upper limit of the movement speed of these fluxons is given by 1/(LC) ^1/2 . Here, L is the inductance per unit length of the Josephson line, and C is the electric capacitance per unit length of the Josephson line. In the Josefson line with the configuration shown in Figure 1A, in which the joint 4 is made of an insulating film, the upper limit of fluxon movement speed is at least several tenths of the speed of light in vacuum (approximately 3.0 x 10 ⁸ m/sec). It is possible to ensure a certain degree of accuracy and is extremely fast.

Furthermore, these characteristics of the Josephson line are based on the first
Even if one superconductor 3 has a larger area than the other as shown in Figure C, or is curved or deformed as a whole as shown in Figure 1D, the expression remains unchanged. do.

The present invention utilizes such a Josephson line as one component, and in order to achieve the above-mentioned object, both longitudinal ends of the Josephson line, which are made up of a pair of superconductors and a joint between them, are electrically connected to each other. a fluxon stop section provided at at least a portion of the Josephson line forming the loop; a current connected to the Josephson line forming the loop and controlling the behavior of the fluxon; and an output circuit connected to the fluxon line and capable of receiving an output current from the flaxon line.

The physical structure of the most basic embodiment of the Josefson line device according to the present invention can be shown in FIG. 2A.

Regarding the physical configuration of this device 10, first of all, a junction 4 between a pair of superconductors 2 and 3 will be explained.
There is a Josefson line 1 sandwiching the two superconductors, and a current source 18 is connected between each end 6, 7 in the longitudinal direction of both superconductors to supply a current Ie for fluxon generation to the line end. On the other hand, the other end portions 11 and 12
An output circuit 17 such as a load resistor that can extract current from the fluxon that has traveled to the end is connected between them.

A feedback resistor 13 is connected between both ends of one superconductor 2, and a conductor 14 is connected between both ends of the other superconductor 3 to form a feedback circuit. When considered together, a kind of loop 5 is formed. In the illustrated case, the part of the Josephson line 1 of the Josephson line loop 5 is shown in a horseshoe shape, but if the above connection relationship is satisfied, the loop 5
Regardless of the geometric shape of the
It may be rectangular or linear.

A fluxon stop section 90, which will be described later, is further provided in at least one part of the loop 5 in the length direction (circumferential direction when viewed as a loop) of the Josefson line 1.
To explain the operation of the Josefson line loop 5 itself without considering the above, when fluxon generated between one end 6 and 7 of the Josephson line 1 travels through the track, reaches the other end 11 and 12 of the line, and disappears. A part of the output is again applied to the line input end via a feedback circuit including a feedback resistor 13 and a conductor 14. Therefore, a new fluxon is generated at the input end again and begins to travel along the track 1 toward the other ends 11 and 12, and the same operation is repeated endlessly thereafter. In other words, the Josefson line loop 5 has an oscillation function.

Incidentally, such Josephson line loop 5 itself has already been disclosed separately by the present applicant in Japanese Patent Application Laid-Open No.

Next, we will focus only on the Josefson track 1 and consider the fluxon stop 90 formed in the middle thereof. In this embodiment, this stop portion 9
0 is formed by arranging a resistance member 9 so as to connect the top and bottom or between the pair of superconductors 2 and 3. In this case, the resistance member 9 has portions 9a and 9c that rest on the main surface of each superconductor, and along one side,
It consists of a portion 9b that extends across the Josefson joint 4, but the portion that exhibits equivalent resistance is the resistive portion 9b, but among these,
Since it is only the portion that passes over the joint portion 4, the resistance portions 9a and 9c on the front and back surfaces may not be necessary in principle.

However, if this continues, the equivalent short-circuit resistance value between both superconductors due to the resistance member 9 will be determined only in the portion of the resistance portion 9b that passes over the extremely thin Josephson junction 4, so the value will be too large. In some cases, it may not be possible to improve the designability or reproducibility of the resistor itself.

Therefore, in such a case, an insulating layer is formed along the side surface of the superconductor across which the resistive portion 9b crosses, and the resistive member goes around this insulating layer to connect the surface of one superconductor 2 with the surface of the other superconductor. By connecting it to the lower surface of the superconductor 3, the geometrical length of the resistor portion 9b can be increased, which not only improves the designability and reproducibility of the equivalent resistance value, but also facilitates manufacturing. become.

When such a resistor portion 90 is provided in the middle of the line, the voltage generated due to the movement of the fluxon traveling on the line is applied to the resistor member forming this resistor portion in addition to the Josephson junction, and energy is generated. As a result, the fluxon loses its kinetic energy and is no longer able to move beyond the resistor portion 90. Therefore, the part attached with such a resistance member is a stopping part 9 that can intentionally stop fluxon.
It can be called 0.

However, as mentioned above, fluxons inherently have a magnetic flux corresponding to one magnetic flux quantum and do not disappear, so even after stopping at the stopping part 90,
The eddy current state is maintained as it is.

Therefore, once the fluxon that is stopped at the stop portion 90 is supplied with the selective driving current I8 from the fluxon driving current source 8, the magnetic flux of the fluxon will be affected by the Lorentz force from this current.
You can leave the stop again and start traveling on the track. Of course, depending on the direction of the current, it is possible to cause the electric current to advance in the same direction in which it has advanced to the stop portion 90, that is, further forward, or it is also possible to cause it to return in the opposite direction. . However, in the application examples to each functional element described later, unless otherwise specified, the selective re-driving current is applied in a direction that causes the fluxon to advance further.

Incidentally, a device formed by forming the above-mentioned fluxon stop portion on the Josefson line has also been separately disclosed by the present applicant in Japanese Patent Application Laid-Open No. 1999-1111.

However, as described above, the device of the present invention having both the configuration of the Josephson line loop 5 and the configuration of the stop section 90 can perform various functions beyond simply combining these functions. Before explaining the embodiments, it is promised that the configuration of the embodiment of the present invention shown in FIG. 2A will be schematically shown as in FIG. 2B for simplicity. That is, the Josephson track 1 is shown as a simple arc-shaped line, and the stop 90 is shown as a simple circle mark placed in the middle of the track.
A feedback circuit consisting of a resistor 13 and a conductor 14 connecting both ends of the line is simply represented by the resistor 13, and each current source and output circuit is shown with only one terminal connected to a corresponding portion.

An example of the operation of the device 10 shown in FIG. 2 will be explained with reference to FIG. As shown at time t1, a pulsed current is superimposed on the current, and one fluxon is placed on one end of the Josephson line 1 in the Josephson line loop 5. Note that the DC bias current Ieo is intended to facilitate the generation of fluxons.
It is desirable to satisfy the following formula.

Ieo<2λ _J・w・Io λ _J : Josephson penetration length, w: line width dimension,
Io: Maximum Josefson current density of the line (1) However, if the current pulse can be made sufficiently large, the DC bias component Ieo may be unnecessary.

In any case, the fluxon generated at one end of the Josefson line 1 as described above eventually reaches the stop portion 90 and stops there. Under this condition, as shown at time t2 in FIG. 3B, when the re-driving current I8 is supplied from the re-driving current source 8, the stopped fluxon starts moving again and reaches the end of the line. A part of the eddy current that has been accompanied is outputted to the output circuit 17 as an output current source Io, as shown in FIG. 3C. At the same time, the remainder of the eddy current is fed back to the input end of the Josephson line 1 via a feedback circuit represented by a resistor 13 as a pulsed feedback current IF (illustrated in FIG. 2B).

Therefore, if the DC bias current Ieo from the fluxon generation current source 18 is still flowing at this point, another fluxon will be generated at the input end of the Josefson line 1, and the generated fluxon will flow through the line. The motor travels until it reaches the stop part 90, but if the drive current I8 from the drive current source 8 has already fallen to zero at this point, it stops there.
Eventually, the same state as the initial state that started this explanation will materialize again.

On the other hand, if the current I8 from the selective drive current source 8 continues to flow as shown in FIG. The fluxon that has returned to the stop part 90 through the stop part passes through this stop part without stopping, and the output current Io is again output from the output terminal.
At the same time, the current component IF is fed back to the input terminal via the feedback circuit. and,
The next fluxon generated by this return will follow exactly the same fate. Therefore, as a result, the output circuit 17 obtains a pulse train defined at a predetermined period, as shown in FIG. 3E.

Here, a little consideration will be given to the output circuit 17. If the reactance component of this circuit is made as small as possible, a larger output current Io can be extracted. In addition, the output end 11 of the Josephson line 1,
12, that is, the sum of the resistance value r13 of the resistor 13 in the feedback circuit and the resistance value r14 of the conductor 14
Parallel combined resistance value of r13+r14 and DC resistance component r17 of the output circuit {[1/(r13+r14)]+(1/r17)}
^-1 is set to have a size on the order of the characteristic impedance of the Josefson line 1, which is particularly advantageous because the fluxon that has reached the output end of the line will go out of the line without being reflected at this end. . Here, the characteristic impedance is (L/C) ^1/2 ,
L and C are as defined above.

As is clear from the above, the conductor 14 in the feedback circuit may be a superconductor (i.e., r14 = 0), or it may have a significant value, for example, the value r13 of the resistor 13.
It is also possible to make it as large as approximately the same. Therefore,
In this specification, the term "conductor" is a comprehensive term that includes not only good conductors with extremely low resistance but also those having a significant resistance component. From this point of view, the resistor 13 may be considered a first conductor, and the conductor 14 may be considered a second conductor.

The embodiment shown in the second figure can perform the following digital functions depending on how it is used, in other words, depending on the required usage.

First, it can function as a non-destructive continuous memory.

That is, if the state in which the fluxon is stopped in the stop section 90 corresponds to logic "1", and the state in which no fluxon is present in the line corresponds to logic "0", then as shown in FIG. 3B to Applying the current I8 as shown in D corresponds to a read command.

When "1" is stored in the memory loop 5, this read command causes an output current Io representing a logic "1" to be outputted to the output circuit 17 in the form shown in FIG. 3C or E. Detected.

On the other hand, if "0" is stored in the memory loop 5, the output current Io does not flow even with this read command, and therefore a logic "0" is displayed on the output circuit 17.

In either case, the same content that was read out via the feedback circuit is stored again in the stop unit 90, and the system waits for the next read command.

The information "1" is erased at least by the feedback current IF via the feedback circuit, at the timing when the fluxon is about to be generated again at the input end of the Josefson line 1. This can be satisfied by excluding the feedback component so that fluxons cannot be generated only by the feedback component.

Conversely, when writing information, a logic "1" can be written by superimposing a pulsed current from a signal source that forms part of the current 18 in a state in which the DC bias component Ieo is applied in advance. If not, a logic "0" can be written.

The embodiment shown in the second figure can also function as a pulse oscillator synchronized to the input signal.

That is, if a pulse train as shown in FIG. 4A is inputted from the drive source 8 as an input signal source while the fluxon is stopped at the stop section 90, a certain input drive current pulse will stop at the stop section 90. The time or period Ts from when the current pulse is applied until the next input driving current pulse is applied, and the fluxon that was present in the stop section begins to move, exits the line output end, and is regenerated at the line input end via the feedback circuit. The relationship Tf<Ts...(2) is satisfied between the time Tf from when the fluxon returns to the stop section 90 (simply put, the time required for the fluxon to go around the Josephson track loop 5). If so, an output current pulse train as shown in FIG. 4B will be obtained in the output circuit 17. Compared to the input current pulse train, this output current pulse train is output with a delay from the input time by the time required for the fluxon to travel from the stop to the line output end for each pulse, but the period is completely different from the input current pulse train. It is synchronized with that of the current pulse train. Of course, as long as the period Ts of the input pulse satisfies equation (2) above, even if it is not constant over time, the output pulse will follow the change.

Furthermore, this second illustrated embodiment can also function as a frequency divider.

For example, the above-mentioned output signal period, that is, the driving current pulse Ts and the fluxon start from the stop section 90 and are regenerated via the feedback circuit, and then again at the stop section 90.
If the relationship Ts<Tf<2Ts (3) is satisfied between Tf and the time taken to return to 0, this functions as a 1/2 frequency divider.

That is, in FIG. 5A, the stop section 90 is stopped at time t1.
If a driving current pulse I8 is applied to the line, and the fluxon that was stopped here starts running, when this fluxon reaches the output end of the line, it will be as shown in Fig. 5B.
As shown in the figure, the output current pulse Io is obtained, but the second fluxon based on the feedback of the fluxon is still at the stop portion 9 when the second drive current pulse I8 is applied after the elapse of time Ts from the beginning.
Since it has not returned to 0, this second drive current pulse is, so to speak, applied to the stop portion in vain. However, the third drive current pulse is at time t1
Since the fluxon has returned to the stop portion 90 before being applied at +2Ts, the output current pulse Io appears in the output circuit 17 again for the third drive current pulse.

As can be seen by considering this third drive current pulse as the first one, the same operation as described above is repeated, and its period becomes smaller than the period Ts of the drive current pulse train I8 as the input signal pulse train. The output circuit 17 can obtain an output pulse train whose frequency has been doubled, such as 2Ts, and whose frequency has therefore been divided by 1/2.

In addition, if we generalize the equation (3) mentioned above, by satisfying the following relationship: (n-1)Ts<Tf<nTs;n=2,3... (4), the frequency division ratio is 1. /n frequency divider can be constructed.

In the above-mentioned case, only one stop section 90 is provided in the Josefson line loop 5, but even so, as described above, by appropriately setting the current value of each current source and its generation sequence, , many functions have been realized.

Therefore, as expected from now on, if a plurality of stop parts 90 are provided in the Josephson line loop 5 and a current source for selectively driving the fluxon is connected to each of them, even more functions can be performed. Can be done.

First, FIG. 6 shows an embodiment in which there are two stop sections 90, and the functions that can be performed will be explained by giving some examples. In addition, each stop part 90
In order to distinguish between..., serial numbers starting from 91 are used for the reference numerals in the drawings for each stop, and similarly, serial numbers starting from 81 are used for the drive current sources associated with each stop. do.

In the embodiment shown in the sixth figure, the same abbreviations as shown in FIG. , two stop portions 91, 9 are provided at appropriate intervals in the circumferential direction in the Josephson line loop 5.
2, and each of them is provided with a dedicated fluxon driving current source 81, 82. The other components may be completely the same as those in the second illustrated embodiment, so their explanation will be omitted.

With this configuration, one fluxon is first generated in the Josefson line 1 by the above-mentioned current supply from the fluxon generation source 18, and this fluxon is delivered to the first stop part 91 by running counterclockwise in the figure. The stopped state is considered to be the initial condition for satisfying each function.

Here, as shown in FIG. 7A, the second stop part 9
The current source 82 connected to the two current sources 8
2a and 82b, and both currents I82a,
If I82b is selectively added so as to be applied to the stop section 92, the current source 81 of the first stop section 91 becomes the first logic signal source, and one current source 8 of the second stop section
By considering 2a as the second logic signal source, the output circuit 1 outputs the logical product of both logic signals.
7 can be obtained.

Now, the current I81 from the current source 81 of the first stop part
Regarding the first logic signal current IA as a current and the second logic signal current IB as a current I82a from one current source 82a of the second stop section 92, the time when both are flowing as current is expressed as It is defined that the logical value is "1", and the same applies to the output current Io.

Therefore, when IA = "1" and IB = "1", we obtain the logical product "1・1=1" at the output, that is, the output current
A time chart explaining the operation in which Io flows is shown in Figure 7B. When IA = "1" and IB = "0", the AND output becomes "1.0 = 0" as specified, that is, the output current A time chart explaining that Io does not flow is shown in Figure 7C.

Application of logic signal current IA is performed at time t1, and application of logic signal current IB is performed at time t2.

In the case shown in FIG. 7B, when a signal current IA of logic "1" is applied at time t1, the fluxon, which was stopped at the first stop section 91 in the previously described initial state, starts moving and moves to the second stop section 92. Come and stop.

When the signal current IB of logic "1" is applied at time t2 after this point, the fluxon that had been stopped at the second stop section 92 starts moving again, and although there is some travel time delay, eventually The output current Io is output to the output circuit 17 as a logic "1". At the same time, if a DC bias current Ieo that satisfies equation (1) is applied to the input end of the Josefson line 1 at this point, the above-mentioned mechanism causes the feedback current to be superimposed through the feedback circuit. One new fluxon is generated at the input end of the line, and is stopped and waited at the first stop section 91.

That is, an initial state is realized for starting the next operation simultaneously with the completion of the operation "1.1=1".

The current I82b from the other current source 82b provided in relation to the second stop portion 92 is
As shown in Figures B and C, this reset pulse is emitted in the negative direction at time t3 after application time t2 of the second logic signal, but in the calculation of "1.1=1", this reset pulse is There is no effect, and the circuit system is auto-reset or self-reset as described above.

In the case of FIG. 7C, the signal current IA of logic "1" is applied to the first stop part 91 at time t1, but as a result, the fluxon that has started traveling and has reached the second stop part 92 For, time t2
When the second logic signal current is applied at , since the signal current IB of logic "0" is applied (that is, the current IB is not applied), the fluxon remains stopped at the second stop portion 92, and therefore Output circuit 1
7, an output current signal of logic "0" (a state in which the output current Io does not flow) is generated, and an initial calculation result of "1.0=0" can be obtained.

However, in the case of this calculation of "1.0=0", the reset pulse I82b from the current source 82b at the subsequent time t3 acts effectively, and the fluxon that remains stopped at the second stop portion 92 is run in the opposite direction toward the first stop section 91 and stopped at the first stop section 91. This state is the initial state of the circuit system and can prepare for the next operation.

When IA="0" and IB="1", and when both are "0", a logic "0" current is applied at time t1,
In other words, since the current IA is not applied, the fluxon stopped at the first stop section 91 remains as it is, and therefore, regardless of whether a current is applied to the second stop section 92 at time t2, and after that, Even if the reset pulse I82b is applied, there is no effect, the circuit system remains in its initial state, and the time
At the calculation result detection timing between t2 and t3, of course, no output current is sent to the output circuit 17, and the desired result of logic "0" is obtained.

The embodiment shown in FIG. 7A can also be used as an inverter to implement the NOT function.

FIG. 7D shows a logic "0" in the output circuit when a negative direction current I82a representing a logic "1" is passed from one current source 82a of the second stop section 92 as an input signal.
A positive output current Io is obtained (i.e., the output current Io
Fig. 7E is a time chart explaining the state in which the input signal current I82a does not flow.
This is a time chart illustrating a state in which the forward output current Io flows into the output circuit 17 as a sign of the logic "1" when the logic is set to "0" (current I82a does not flow).

As before, if the circuit initial state is one in which one fluxon is stopped in the first stop section 91, first, at time t1, the drive current I is supplied from the current source 81.
81 is supplied, and the fluxon in the first stop section is sent to the second stop section 92 and kept on standby.

When a logic "1" signal is applied at the application timing t2 of the input logic signal, that is, when a negative current I82a is supplied to the second stop section 92 as shown at time t2 in FIG. 7D, , this stop 9
2, the fluxon that had stopped at 2 moves in the opposite direction toward the input end of the track, reaches the first stop section 91 again, and stops. On the other hand, when a logic "0" signal is applied, that is, as shown in FIG. As shown at time t2 during E, when the current I82a is not flowing, the fluxon in the second stop portion 92 remains in the second stop portion.

Therefore, when the signal current I82b indicating that it is the output timing of the calculation result is supplied from the other current source 82b related to the second stop section 92 at time t3, if the logic "1" is input, the corresponding current source 82b is supplied. Since there is no fluxon in the second stop section 92, naturally the output current Io does not flow through the output circuit 17, and therefore, while outputting logic "0" is achieved,
When the logic "0" is input, the fluxon stopped at the second stop section 92 is driven toward the output end, and eventually the logic "1" is input to the output circuit 17.
An output current Io representing .

When a logic "0" is output, the fluxon has already stopped again in the first stop section 91 at a stage before that, and even when a logic "1" is output,
By the mechanism already mentioned many times, fluxon is generated again in the Josefson track 1 via the feedback circuit, and it travels to the first stop section 91 and stops there, so from time t3 onward, we get:
It can be seen that this circuit can be reset to its initial state without applying any particular reset pulse or the like.

Next, in the embodiment shown in FIG. 6, the current source 81 associated with the first stop portion 91 may be composed of a plurality of sources, for example, for simplicity, two current sources 81a and 81b as shown in FIG. 8A. , each current source 81
By associating currents IA and IB from a and 81b with logic "1" when they are flowing and logic "0" when they are not flowing, a logical sum operation function is realized in the currents. can do.

Figure 8B is a time chart explaining the logical sum "1+1=1", and Figure 8C is a time chart explaining the logical sum "1+1=1".
1=1". In this embodiment as well, the current from the current source 81a
The application of IA is defined as the application of logic "1" of the first logic signal, and similarly the application of current IB from the current source 81b is defined as the application of logic "1" of the second logic signal,
Therefore, when each current IA, IB is not flowing, each corresponding logic signal is logic "0". The same applies to the output current Io, and it is assumed that the logic of the output logic signal is "1" when this is flowing.

The logical sum operation of this embodiment is obvious; if the device 10 is in the initial condition described above and one fluxon is stopped at the first stop section 91, then at time t1
When the first logic signal is converted into a current IA and applied to the first stop section 91, as shown in FIG. 8B, if this logic is "1", the first stop section 91 The fluxon, which had been stopped at , moves and reaches the second stop portion 92, where it stops. When the first logic signal is logic "0" as shown in FIG. 8C, the fluxon stopped at the first stop section 91 shows no behavior at this time t1.

Next, when the second logic signal is applied at time t2, since the fluxon has already moved to the second stop section 92 in FIG. Even if current IB of By applying IB, the fluxon that was still stopped at the first stop part 91 starts moving, reaches the second stop part 92, and stops there. After all, from time t2 until the next time t3, the conditions in both cases B and C in FIG. 8 are the same, that is, the fluxon reaches the second stop portion 92 and is stopped. .

Therefore, when the drive current I82 is supplied as a read current from the current source 82 associated with the second stop section 92 at the next calculation result read timing t3, in any case, the second stop section 92 stops. The fluxon that had been in use starts to move and reaches the line output end, and a part of the accompanying eddy current is output to the output circuit 17 as an output current Io representing logic "1", while the eddy current flows through the feedback circuit to the line input end. By feeding back the remainder of the current, one new fluxon is created by superimposing it on the DC bias current Ieo from the DC bias source 18, and it runs to the first stop part 91 and stops there, creating the circuit initial conditions, Wait for the next operation.

As is clear from the above, "1+0=1" is of course satisfied. In FIG. 8B, when the second logic signal is applied at time t2, even if the current IB does not flow in to represent logic "0", the fluxon moves to the second stop section 92 at a stage before that. It's on,
This is because the application of current to the first stop portion 91 at this time t1 has no effect on the operation. Furthermore,
It can be seen that the logical sum "0+0=0" is clearly satisfied since it does not have any effect on the fluxon under the initial condition stopped at the first stop portion 91.

Furthermore, the embodiment shown in FIG. 6 can also exhibit a frequency dividing function by devising the manner in which the drive current is applied to the first and second stop portions 91 and 92. That is, as shown in FIG. 9A, one driving current source 8 is used, and it is considered as an input frequency signal source with a period Ts, and the current I8 given for each period Ts may be equal in value. two resistors 19
What is necessary is to divide the flow at a and 19b and apply it to both stop portions 91 and 92 together.

As before, this circuit is placed under the initial condition where one fluxon is stopped at the first stop section 91, and the current I is set as the input frequency signal as shown in FIG. 9B.
8 is supplied at a certain period Ts, the fluxon in the first stop part 91 moves toward the second stop part 92 due to the shunt flow through the first resistor 19a accompanying the first input pulse at time t1, It will eventually stop there. Of course, as in each of the previous embodiments, the input pulse width is selected so that the fluxon that has started moving from the first stop part 91 falls before reaching the second stop part 92.

When the input pulse I82 is applied again at the next time t2 in this state, the fluxon that has reached the second stop portion 92 starts moving again due to the shunt flow through the second resistor 19b, and from the line output end to the output circuit. A part of the eddy current component is output to 17, and at the same time, the initial state of the circuit system is created again via the feedback circuit.

Therefore, for the third input pulse at time t3, the same operation occurs as for the first input pulse, and for the fourth input pulse at time t4, the same operation occurs as for the second input pulse. A similar operation occurs and is repeated thereafter, so that the train of output pulses Io has a period of 2Ts, which means that the frequency has been divided in half.

According to the configuration of the embodiment shown in FIG. 9, if the number of stop portions is further increased to n and the input signal current is allocated through each shunt resistor, the frequency division function of 1/n is satisfied. I understand that.

However, in that case, a current pulse width shorter than the fluxon travel time from one stop to the next stop is required.

However, current pulses with a fairly short pulse width can be obtained using a Josefson pulse oscillator, which has already been studied, and furthermore, as will be described later, by providing a current source for fluxon transit time control in the middle of the line, It is possible to intentionally slow down the transit time of the fluxon (while maintaining a sufficiently high speed range compared to conventional semiconductor devices, etc.), as long as the shunt of the input signal current can be kept at a significant value. It is possible to obtain a considerably high frequency division ratio.

Next, FIG. 10 shows an embodiment in which a large number of stop portions 91 are provided in the Josefson line loop 5 and yet another function is performed.

As shown in FIG. 10A, in this case, eight stop parts are illustrated, and each stop part 91 to 98
Current sources 81 to 88 are respectively provided in connection with the current sources 81 to 88. This circuit device system can be used, for example, as an 8-bit serial write/serial read type non-destructive read memory. The operation of this application example is explained in the 10th section.
This will be explained with reference to Figure B.

Information is written as follows.

First, a DC bias current Ieo as shown in equation (1) is supplied from the current source 18 connected to the input end of the Josefson line 1, and from time t1 to t8 with a predetermined time interval. At each write timing, currents I81 to I88 are supplied all at once from each current source 81 to 88 associated with each stop unit 91 to 98.

In synchronization with this, the fluxon generation current source 18
is used as the writing logic signal source, and depending on the bit pattern to be written superimposed on the above DC bias component,
A current pulse Ie for fluxon generation is selectively generated at the write timing of each bit. If the state in which fluxons exist is defined as logic "1", for example in the area shown in FIG. A write pulse Ie is generated from the current source 18 at t7, and is not generated at times t3, t5, and t8.

In this way, the fluxon generated at the line input end at time t1 reaches the first stop portion 91 and stops during the suspension period of the drive current I81 until the next time t2. Then, at time t2, which is the second bit write timing, when the drive current I81 is applied again to the first stop section 91 as in the other stop sections, the fluxon in the first stop section is transferred to the second stop section. 92, and the next time
During the drive current suspension period up to t3, the second stop section 9
2 and stops. On the other hand, the logic "1" bit signal fluxon generated at the second time t2 similarly reaches the first stop section 91 and stops by time t3.

By repeating these actions,
After time t8, a bit pattern that is exactly the same as the input bit pattern moves from the eighth stop section 98 to the first stop section 91 and is stored from the most significant bit to the least significant bit.

To read the information, the current sources 81 to 88 connected to each stop section are driven simultaneously, in the same way as above, with the DC bias current Ieo applied from the current source 18, just as in the writing shown in FIG. 10B. By applying current, the logic signal information that was stopped in the eighth stop section 98 at time t1 is sent to the output end of the Josephson line 1 and output to the output circuit (in this case,
Since this most significant bit is logic "1", the output circuit obtains an output current pulse Io as shown by the virtual line in Figure 10B), and at the same time sends the logic information from the next bit onwards to the next stop section. , and by feeding back the output logic information to the input end of the Josefson line 1 via the feedback circuit, the DC bias current is reduced.
It superimposes it on Ieo to generate the corresponding logic information again, and similarly, at the next time t2, the second bit logic information from the most significant bit, which has already been sent to the eighth stop section 98, is output to the output circuit 17. This can be repeated by returning the signal to the line input end via the feedback circuit, and then transferring the subsequent bit logic information to the next stop.

That is, since such transfer, output, and feedback operations are repeated in the same way at each subsequent readout time timing t3 to t8, after time t8 has elapsed, the output circuit 17 receives the virtual signal shown in FIG. 10B. As shown by the line, a bit string with the exact same pattern as the input bit pattern is obtained, and even after all bits are read out, the exact same bit pattern remains in the Josephson line loop 5 in a non-destructive manner. .

Information is erased in the read operation described above.
This can be achieved by not supplying the DC bias Ieo from the current source 18, so that fluxons cannot be generated at the line input end only by the feedback eddy current.

In addition, as already mentioned earlier, when only one fluxon is placed in the loop 5 in the configuration shown in FIG. 10A, and the drive current sources 81 to 88 are considered as input frequency signal sources, the structure shown in FIG. 10B is By applying this current pulse to each of the stop sections 91 to 98 at the same time at the timing shown, a frequency output that is divided by 1/8 with respect to this frequency can be obtained in the output circuit 17.
It can also function as an 8 frequency divider.

Of course, the memory and frequency divider described above can generally be considered to be expanded to n bits.

By the way, in the embodiments described so far, the output circuit 17 takes out the output current only from the output end of the Josephson line 1 of the Josephson line loop 5, that is, from the input end of the feedback circuit.
However, it is also possible to provide another output circuit in the middle of the line, and the device 1 of the embodiment shown in FIG.
0 has such a configuration.

A feedback circuit composed of first and second conductors 13 and 14 is provided between the input ends 6 and 7 and the output ends 11 and 12 of the Josephson line 1 constituting the Josephson line loop 5, and the input end carries a current. The first or main output circuit 17 is provided at the output end of the source 18, and one fluxon stop section 90 and a current source 8 connected to this are provided in the middle of the Josefson line 1. This is exactly the same as the basic embodiment shown in FIG. The difference is that the Josephson line 1 is divided in the middle, and there is an input side Josephson line section 1a and an output side Josephson line section 1.
b, and their opposing ends 20 and 21 are connected to each other by first and second conductors 13a and 14a, which may be similar to the first and second conductors described above.

A second output circuit 17a is connected to the output end 20 of the input Josephson line section 1a, while a current source 18a for selective generation of fluxons is connected to the input end 21 of the output Josephson line section 1b.
is connected. The output circuit 17a may have the same structure as the main output circuit 17, and the second current source 18a may have the same structure as the first or main current source 18.
Furthermore, as before, the first and second conductors connecting the Josefson line portions are represented by the resistor 13a in the figure.

Considering the basic operation of this device 10, fluxons generated by a mechanism similar to that described in connection with the second illustrated embodiment are
0, and in the process of being sent out to the main output circuit 17 by the supply current of the selective drive source 8, a part of the eddy current accompanying the fluxon is transferred to the output of the input Josephson line section 1a. edge 2
0 to the second or auxiliary output circuit 17a.

However, at the same time, the remaining part If of the eddy current is sent to the input end 21 of the Josephson line section 1b on the output side by the first and second conductors 13a and 14a forming a feed-forward circuit, and thus DC bias current from two current sources that satisfies the above equation (1)
If Ieo is given, fluxon can be generated again at the input terminal 21.

Therefore, when this fluxon reaches the output end of the Josefson line 1 as a whole and disappears,
Because the eddy current accompanying that time can supply an output current to the main output circuit 17,
As far as the output circuit 17 is concerned, this device
It is possible to perform exactly the same operation as the illustrated embodiment.

In other words, this device can be said to be a device that is subjected to various stresses exactly like the device of the second illustrated embodiment, but that has two independent output circuits.

Similarly, the device 10 shown in FIG.
This is a modification of the illustrated embodiment, and is a device that can provide two additional outputs in addition to the same frequency division function as the embodiment shown in FIGS. 6 and 9. In terms of configuration, the basic part is the same as the embodiment shown in FIGS. 6 to 9, and the Josephson line 1 is divided into an input line part 1a and an output line part 1b, and the connection part between them is shown in FIG. 11A. It is made up of the configuration shown in.

As can be imagined from the embodiments shown in FIG. 11, if the Josefson line 1 is divided into many more parts and each output circuit and DC bias power supply are connected to the division points as described above, even more output types can be connected. It can be a circuit. In particular, in a device such as the embodiment shown in Figure 10, if the dividing point configuration shown in Figure 11 is adopted between adjacent stop parts, the fluxons in each stop part will start to move at the same time and move to the next stop part. The presence or absence of each fluxon can also be monitored in parallel during the process leading to this.

By combining multiple devices of each of the above embodiments and devising the current supply sequence,
Many more significant functions can be fulfilled.
As an example of such an applied circuit, a configuration example of a binary counter 100 is shown in FIG. 12 and will be explained.

In the case of this application example, three devices 10-1, 10-2, and 10-3, each corresponding to the device 10 having the configuration of the embodiment shown in FIGS. 6 to 9, are connected in cascade. The count signal is sent to the first stage Josefson line loop device 10-
The pulse current I8 is sent from the current source 8 to the first and second stop portions 91-1 and 92-1 of No. 1 through the shunt resistors 19a-1 and 19b-1. That is, this binary counter 100 counts the number of pulse currents I8 as count data over time.

Josephson line loop device 10-1,
10-2, 10-3 Josephson line 1-
Current sources 18-1, 18-2, and 18-3 are connected to the input terminals of 1, 1-2, and 1-3, respectively, and from these current sources, a DC bias current Ieo expressed by equation (1) is generated.
is applied.

The circuit initial conditions of this counter 100 are as follows:
In this current source, a fluxon generated by superimposing a pulse current Ie of a significant magnitude on the DC bias current Ieo is transmitted to each loop device 10-1, 10-2, 1, as in the previous embodiments.
0-3 each Josephson line first stop 91-
1, 91-2, and 91-3.

In addition, the output current Io-1 to the output circuit 17-1 of the first-stage loop device 10-1 is applied to the first and second stop portions 91-2 of the second-stage loop device 10-2, Shunt resistance 19a-2, 1 also to 92-2
9b-2, and the output current Io-2 to the output circuit 17-2 of the second-stage loop device 10-2 is applied to the third-stage loop device 1.
0-3 to the first and second stop portions 91-3, 92-3 via shunt resistors 19a-3, 19b-3, while each loop device 10-1, The second stop portions 92-1, 92-2, 92-3 of 10-2, 10-3 further include selective fluxon drive current sources 82-1, 82-2, 8.
2-3 is connected, and this current source is connected to the output circuits 17-1 and 17- after completing data input as described later.
It is used to output the count results to groups 2 and 17-3.

The operation of such counter 100 can be explained as shown in FIG. 12B.

In Figure 12B, the leftmost number is the input data content,
That is, the pulse current I8 input during the data input period
The schematic figure in the center shows the number of fluxons in each loop according to the data content after input data is taken in each loop device 10-1, 10-2, 10-3. represents the location. That is, when a small circle mark is written on the top of an arc-shaped figure, each loop device 10-1,
10-2 and 10-3, each fluxon is shown in each corresponding first stop part 91-1 to 91-3, and when shown below, each fluxon is in each corresponding second stop part 92. -1 to 92-3.

In FIG. 12B, the group of numbers on the right end indicates each of the output circuits 17-1 to 17-3 by the count result output operation after input data is taken in, as will be described later.
The output circuit 17-1 indicates the least significant digit, and the output circuit 17-3 indicates the most significant digit.

Considering the transient state from when the input data starts to input to when the input ends, the first input pulse current I8 in the input data reaches both stop portions of the first loop device 10-1. 91-1,9
2-1, the fluxon in the first stop part 91-1 moves to the second stop part 92-1 and stops there.

When the input data is a number greater than or equal to "1" in decimal notation, the second input current pulse I8 is applied, and the second stop part 92-1 is moved in the first device 10-1. As shown by the circular arc-shaped imaginary line arrow in FIG. One new fluxon is generated at the input end of the Josefson line 1-1 through the feedback circuit with the help of the DC bias current of the current source 18-1, and the first stop section of the Josephson line 1-1 is generated. It reaches 91-1 and stops. At the same time, as shown by the imaginary straight arrow c in FIG. 2 both stop parts 91
-2, 92-2, the fluxon present in the first stop part 91-2 of the second loop device travels toward the second stop part 92-2 and stops.

In this way, the least significant digit loop device 10
-1 is transferred from the first stop part 91-1 to the second stop part 92-1 and from the second stop part 92-1 to the first stop part 91-1 every time the input current pulse I8 is applied. , and so on, repeating the operation of alternately running and stopping, and in the loop devices 10-1, 10-2 of the lowest digit and the second digit, the second stop portions 92-1, 92 When the fluxon at -2 returns to the first stop via the feedback circuit, the fluxon in the loop device one digit above it is moved to another stop, as shown by arrow c.

Therefore, the position occupied by the fluxon in each loop device after counting the input data is as shown in FIG. 12B.

Once the input data has been taken in, the next step is to output the count results. This corresponds to the current source 82-1 associated with each second stop section 92-1 to 92-3.
This is done by supplying current all at once from 82-3 to 82-3.

Therefore, when the count is finished, the current sources 82-1 to 82-1 are applied only to the output circuit 17-i (i=1, 2, or 3) of the loop device whose fluxon is stopped at the second stop portion. With the supply of count content read current from 82-3, the output current Io-
i flows.

In this way, if the flow of the output current is made to correspond to logic "1", the count result output will look like the group of numbers shown at the right end in FIG. 12B,
It can be seen that the expected binary count result is satisfied with the output circuit 17-1 side as the least significant digit.

However, once this reading is completed, it is necessary to reset the counter 100 as a whole to the above-mentioned initial state in order to prepare for the next counting operation. However, this is because the current source 82-
This can be easily done by supplying a current in the direction opposite to that of reading from 1 to 82-3. That is, after reading,
Even if there is a loop device in which the fluxon has moved to the second stop, if a negative current is supplied to the second stop, the fluxon in the second stop will shift to the left in the figure. This is because it can travel around and return to the first stop without affecting the fluxon in the loop device of the next digit.

Furthermore, as is obvious from the above mechanism, this counter 100 rotates the fluxon position once every two input current pulses, and changes the output current pulse or carry pulse to the next digit every two input current pulses. In order to send input current pulses to the loop device, if the loop device is connected continuously over n stages, it is possible to count up to 2 ⁿ -1 input current pulses.

The embodiment or applied example shown in FIG. 13 has substantially the same configuration as the applied example shown in 12th figure, but realizes the binary full adder 101 by changing the manner in which input is applied. In this case, the augend is "AB" from the upper digit, and the number to be added is "CD" from the upper digit, both of which are binary two-digit numbers, and the results of these additions are output to the output circuits 17-1 to 17-1. 17-3. As previously stated, the least significant digit of the addition result is represented by the output circuit 17-1 at the left end in the figure. In addition, when the number of digits A, B, C, and D in both the augend and the addition number are "1" in binary, currents IA, IB, IC, ID are generated from the corresponding current sources.
I promise that it will flow.

For the sake of simplicity, the existence of each shunt resistor will be omitted in the explanation. The currents IA and IB corresponding to each digit of the augend "AB" are as follows: The current IA of the upper digit is the loop device 10 of the second digit. -2 both stop parts 91-1, 92
−2, the current IB in the lower digit is the loop in the lowest digit.
Both stop parts 91-1, 92-1 of device 10-1
are given to each.

Similarly, each current IC, ID corresponding to each digit of the addition number "CD" is determined by the loop device 1 of the second digit when the current IC of the upper digit is different from the above current IA.
The current ID of the lower digit is applied to both the stop parts 91-2, 92-2 of the loop device 10-1 of the lowest digit at a timing different from the current IB. -1 respectively.

In the present adder 101 as well, in the initial state of the calculation, fluxons generated by the superimposed pulse currents from the current sources 18-1 to 18-3, which always flow a DC bias current, are addressed to each loop. This is a state in which the device is stopped in the first stop portions 91-1 to 91-3.

To explain with reference to the time chart shown in FIG. 13B, first, each binary value of the upper and lower digits of the augend is divided into two digits corresponding to each other as described above at time t1. Both stop parts 91-2, 92 of the loop device 10-2, 10-1 of the first digit
-2; Provided as currents IA and IB to 91-1 and 92-1. Of course, if the value of the corresponding digit is binary “0”, the corresponding current IA or IB
won't let this pass.

On the other hand, the binary value of the lower digit of the addition number is expressed as whether the current ID is flowing or not flowing at time t2, and the lower digit of the loop device 10-1 has both stop portions 91.
-1, 92-1, and the binary value of the upper digit is similarly transformed into whether or not the current IC flows at time t3, and the second digit loop device 10-
2, both stop portions 91-2 and 92-2.

As an example, if the augend is "10" and the addend is "11", first, at time t1, each binary information of the upper and lower digits of the augend corresponds to a loop. Both stop parts 91-2, 92-2 of devices 10-2, 10-1; 91-1, 92-1
given to. In this case, since the lower digit of the augend is "0", no significant current IB is passed through both stop portions 91-1 and 92-1 of the loop device 10-1 for the lowest digit. , a significant current is generated only in both stop portions 91-2, 92-2 of the loop device 10-2 in the second digit, corresponding to the upper digit value “1”.
IA is washed away.

As a result, the fluxon in each loop device, which was initially in the state shown in the top row of FIG. 13C, has its stopping position changed only in the second-digit loop device 10-2. It now moves to the second stop portion 92-2. This state is at time t1<
This is shown in FIG. 13C as a state where t<t2.

At the next time t2, the binary value "D=1" in the lower digit of the addition number is a significant current ID, and both stop portions 91-1 of the loop device 10-1 in the first digit,
92-1, but as a result, fluxon movement occurs in the loop device 10-1 of the first digit, and in FIG. 3C, time t2<t<t3
As shown in the state shown in FIG. - A state occurs in which the fluxon in the device still remains at the first stop portion 91-3.

Furthermore, at the next time t3, the binary value "C=1" of the upper digit of the addition number is changed to loop device 1 of the second digit.
Current to both stop parts 91-2, 92-2 of 0-2
When given as IC, the second digit of the loop
In the device 10-2, fluxon movement occurs,
As it disappears as it reaches the output end of the corresponding Josefson line 1-2, a part of the output current Io-2 flows into the third to most significant digit loop, as shown by the straight arrow c. - The third digit fluxon, which is given to both the stop parts 91-3 and 92-3 of the device 10-3 and which was in the first stop part 91-3, is sent to the second stop part 92-3.

This is shown as a state at time t3<t<t4 in FIG.
has completed the addition and contains the result.

To read the contents of this adder, each loop
Second stop portion 92- of devices 10-1 to 10-3
Current sources 82-1 to 8 connected to 1 to 92-3
From 2-3, current I82-1 as read current
~I82-3 are applied all at once at time t4.

Then, only from the loop devices 10-1 and 10-3 where the fluxon of the second stop part existed in the state of time t3<t<t4, a significant signal is sent to the corresponding output circuits 17-1 and 17-3. Output current Io
-1, Io-3 are output, and no output current flows from the second digit loop device 10-2.

Therefore, the result of reading the adder contents is "101" from the most significant digit, which is "10" as the result of the intended binary addition operation of the augend "10" and the addend "11".
It can be seen that “+11=101” is satisfied.

The time when the contents of the adder are read out by converting them into the output current is shown as the state at time t4<t<t5, but at the same time, due to the movement of the fluxon at this time, the second A loop device may result where fluxons remain at the stop. For example, in this case it is the second digit loop device 10
-2. Therefore, in order to prepare for the next addition operation, it is necessary to return the fluxon of such a loop device to the first stop section and return the present adder 101 as a whole to its initial state.

However, this can be easily done as in the case of the counter described above, and as shown at time t5, a current in the opposite direction from the current source 82-1 to 82-3 used for reading is applied to each second stop section. If fluxons are supplied all at once and are stopped at the second stop, they may be returned to the first stop in a counterclockwise direction in the figure. Therefore, in FIG. 3C, the state at time t<t1 representing the initial state of the circuit can also be said to be the state at time t5<t.

Although only the above example of addition is shown here, it can be easily verified that the present adder operates effectively in other cases as well. Furthermore,
By cascading loop devices with the same configuration over n+1 stages and creating a signal application sequence similar to the above, addition operations of n bits, that is, n digits of augends and addends, can be performed with a carry. It can be carried out.

Each of the embodiments described above can be combined with each other appropriately, and various other types of stress can be created by appropriately arranging the sequence of signal applied currents. I will now describe some slightly more detailed examples of modifications that can be applied.

As briefly mentioned above, it is also conceivable to connect the auxiliary current source 22 to a portion of the Josephson line other than where the stop portion 90 is provided, for example as shown in FIG. 14A. In this way, for example, the running speed of the fluxon running on the Josephson track can be adjusted, and the degree of freedom in circuit design can be increased.

Further, as shown in FIG. 14B, an auxiliary output circuit 23 can be provided to draw the output from the middle of the Josefson line. However, in this case, the output circuit 2 must be
It is desirable that the resistance component of the impedance No. 3 be at least two to three times the characteristic impedance (L/C) ^1/2 of the Josefson line. Conversely, the output current that can be taken out to such an output circuit 23 is
Since it is necessarily smaller than the output circuit 17 provided at the output end of the Josefson line, this must be taken into consideration.

Furthermore, in each of the above-described embodiments, the stop section 90 for stopping the fluxon was configured by connecting a pair of superconductors with a resistance member, but the invention is not limited to this, and the point is to stop the fluxon. It is sufficient if the possible portion is formed at a predetermined location on the Josephson track. Even if the same resistance member is used, the stop portion 90 can also be made with the configuration shown in FIG. 15, for example.

In other words, at least one of the superconductors (in this case, the upper superconductor 2) is provided with a resistance member 9 in at least one portion of the Josephson line 1 in the length direction.

In the distributed constant equivalent circuit of such a structure, a resistance component ro is placed in parallel with the inductance component in the length direction of the line connecting the multiple Josephson junctions between the upper and lower superconductors 2 and 3. The resistance member 9 can be represented by the shape.

Even in such a configuration, the fluxon is the resistance member 9.
When entering the resistor portion 90 marked with , the voltage applied to the inductance across the resistor portion 90 is also applied to the resistance component ro described above. Therefore, energy is consumed by this resistance component, and the fluxon in question reduces its kinetic energy, becomes unable to advance beyond this resistor portion 90, and eventually stops this resistor portion 90. It can be used as part 90.

Furthermore, in order to efficiently stop the fluxon in such a configuration, it is necessary to apply a voltage effectively across both ends of the resistance member 9 when viewed in the length direction of the line. It is desirable that the film thickness of the superconductor 2 in the resistor portion 90 be limited to at most twice the London magnetic field penetration distance λ _L .

However, if you do not want to be subject to such limitations, for example, as shown by the imaginary line 9' in FIG.
It is conceivable to embed this resistance member inside the superconductor. Conversely, even if there is no particular limit to the film thickness, if the resistance member is intentionally moved closer to the joint 4 in this way, the function of stopping fluxon can be increased. .

Further, the resistance member may be provided along both superconductors 2 and 3 as shown by the imaginary line 9''.In this case as well, the above-mentioned buried configuration may be adopted for one or both of the resistance members. It's good.

In any case, the above-mentioned current sources 8, 81 to 81 which provide a fluxon drive current to the resistor portion 90 are used.
88 does not necessarily have to be connected to the resistance member 9, but may be directly connected to the superconductor as long as it is in the stop portion 90 where the resistance member 9 is provided.

In the Josephson line device of the present invention described in detail above, one superconductor is a Pb-Au alloy, the other superconductor is a Pb-In-Au alloy, and the joint is a Pb-In
- Each layer is formed with a discharge oxide film of Au alloy, and the maximum Josephson current density of the Josephson line is determined by e.g.
1KA/ ^cm2 , the spatial spread of fluxons in the Josephson line can be reduced to about 20μm, and the upper limit of the fluxon movement speed can be set to about 10μm/ps, and the pulse current taken out to the output circuit can be reduced to about half that time. The price range can be sharp, with a range of 2 to 3 ps. Therefore, if the device dimensions are appropriately selected, the speed can be increased to the extent that each pulse interval in the operation description of each of the above embodiments is less than 10 ps, and an extremely high speed device can be expected as an applied device.

Furthermore, if the Josephson line is constructed of a superconductor with a large London penetration depth λ _L such as niobium nitride, the Josephson penetration depth λ _J can be made small, as is clear from the following equation (5). Penetration length λ _L is 0.5 to 0.6 μm).

λ _J = {〓/[2e・Io・μo・(d+2λ _L )]} ^0.5 d: Thickness of joint, μo: Vacuum permeability, e:
Elementary charge, 〓: (blank constant)/2π, Io: maximum Josephson current density, ...(5) Therefore, if such a Josephson line device is used, each device as shown in each of the above application examples can be When manufacturing, the entire size can be kept within a minute area, and high integration can be achieved.

Incidentally, although the above embodiments are all applications as digital functional devices, analog applications are also highly conceivable.

In any case, according to the present invention, it is possible to provide an extremely high-speed device with lower power consumption than conventional semiconductor devices as well as conventional Josefson devices, and the scope of its application is as described above. As can be understood from the fact that even this is only a small part, it is extremely wide-ranging, and the contribution of the present invention to this industry in the future is enormous.

[Brief explanation of drawings]

Figure 1 is an explanatory diagram of a general Josephson line.
FIG. 2 is a schematic configuration diagram of a basic embodiment of the Josephson line device according to the present invention, FIGS. 3 and 4,
and FIG. 5 are explanatory diagrams of each operation example of the device shown in the second diagram, and each of the diagrams from FIG. 6 to FIG. FIG. 14 is an explanatory diagram of how to attach an auxiliary current source and an auxiliary output circuit, and FIG. 15 is an explanatory diagram of another method of forming a fluxon stop portion. In the figure, 1, 1-1,..., 1-3 are Josephson lines, 2, 3 are superconductor layers, 4 is a junction, 5 is a loop containing the Josephson lines, 8, 81,..., 8
8 is a fluxon driving current source, 10, 10-1,
..., 10-3 is the Josephson line device of the embodiment of the present invention as a whole, 13, 13-1, ..., 1
3-3 is the first conductor or resistor that constitutes the feedback circuit;
14 is a second conductor constituting a feedback circuit, 17, 17
-1,...,17-3 are output circuits, 18,18-
1,..., 18-3 are current sources for fluxon generation, 9
0, . . . , 98 are fluxon stop units, 100 is a binary counter, and 101 is a binary two-digit full adder.

Claims (1)

[Claims] 1. A Josephson line loop formed by electrically connecting both lengthwise ends of a Josephson line consisting of a pair of superconductors and a junction between them to form a feedback circuit; a fluxon stop provided at at least a portion of the Josephson line forming the loop; a current source connected to the Josephson line forming the loop to control the behavior of the fluxon; the Josephson line also forming the loop; and an output circuit connected to and capable of receiving an output current from the fluxon.