JPS6048834B2 - Magnetic bubble storage system for both M/m loop and single loop - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a magnetic bubble storage device, and particularly to a loop method for transferring and storing magnetic bubble information, in which a magnetic bubble element with the same configuration is used in both a major-minor loop method and a single-loop method. The present invention provides a magnetic bubble storage method that can be used.

As is well known, the loop configuration in a magnetic bubble element includes a major-minor loop (hereinafter referred to as "M/minor loop") as shown in Figure 1.
There are two types of configurations: an "m-loop" configuration and a single-loop configuration as shown in FIG. 2, and each has its advantages and disadvantages.

That is, in the M/m loop configuration, a plurality of minor loops m... are arranged in parallel for one major loop M, and each minor loop m...
Transfer gate T between ... and major loop M
By operating ・・・・・・・・・ at the same time, a magnetic bubble (hereinafter referred to as “bubble”) on the major loop is created.
all at once into the minor loop.
Or conversely, transfer all the bubbles stored in each minor loop to the major loop (
(transfer). In the single-loop system shown in FIG. 2, bubbles are transferred in a line along one serially arranged transfer path northward. In addition, G is a bubble generator, S
is a replicator, D is a bubble detector, and A is a bubble extinguisher. As described above, the configuration and control method of the loop that transfers and stores bubble information is different, and the M/m loop method can shorten the access time and is suitable for large-capacity devices compared to the single loop method. , control is complex, and each has its own pros and cons. In this way, the M/m loop and the single loop have completely different configurations of bubble information transfer/storage loops, so patterns such as permalloy patterns and conductor patterns for bubble elements are created separately and cannot be manufactured independently. ing.

Therefore, when purchasing a magnetic bubble storage device, the user must specify either the M/m loop configuration or the single loop configuration. Since they are designed and manufactured separately, the cost is also high. Apart from this, in the case of the M/m loop type, a predetermined memory capacity can be maintained by creating extra minor loops in advance and allocating them to replenishment when a defective loop occurs. In the case of a single-loop type, such operations are difficult and have become a troublesome problem. The present invention eliminates these conventional problems. In other words, elements with the same configuration can be used as either the M/m loop method or the single loop method by simply changing the control method, and even when operating in the single loop method, the M/m loop method can be used. It is an object of the present invention to provide a magnetic bubble storage method that can avoid defective loops and perform predetermined operations in exactly the same way as in the case of . To achieve this object, the present invention provides a magnetic bubble device having a major loop and a minor loop, including a readout line connected to the major loop, and each minor loop connected to the readout line via a transfer gate. By performing the transfer from the major loop to the minor loop and the transfer out from the minor loop to the readout line at a 1/n period (n = number of minor loops x bit interval between minor loops), a single loop method is implemented. The configuration is such that it can be operated as a computer. Next, how the M/m loop and single loop magnetic bubble storage system according to the present invention is actually implemented will be explained using an example.

Figure 3 is. A diagram showing the configuration of the M/m loop and single loop magnetic bubble element according to the present invention, FIG. 4 is an enlarged diagram showing the relationship between each minor loop, major loop, and readout loop, and FIGS. 5 and 6 are control 3 is a time chart showing the timing of . In the case of the present invention, for one major loop M, multiple minor loops ml...
・・・・・・The one is Transfer Gate T...
It is the same as the conventional M/m loop configuration in that it is arranged in parallel via... However, the present invention further includes a readout line L connected to the major loop QM, and this readout line L is connected to the transfer gate TO for transfer at a position different from the transfer gate T. It is connected to each minor loop ml through . Like the major loop M and each minor loop, the configuration of the read line L is made up of an array of transfer patterns made of permalloy patterns or the like. In this configuration, when used as an M/m loop method, control is performed in exactly the same manner as in the conventional method.

That is, a series of information bubbles sequentially generated by the generator corresponding to the information "1 and o" are transmitted to each transfer gate T.
When it reaches the position of ・・・・・・・・・, the control conductor C
Apply a current pulse to 1 and create a minor loop M all at once.
......Moved to Akira and stored there. By repeating this series-to-parallel conversion of the bubble array, the bubbles transferred to the minor loop are processed into each minor loop M, . . . M. The information bubble is circulated inside by the action of the rotating magnetic field 7, and when the information bubble to be read reaches the position of the transfer gate T..., the transfer gate T... Run ... and transfer the bubbles in the minor loop ml to the major loop all at once, and transfer them to the replicator S side with the same series arrangement as when they were transferred. ,
Read with detector D. When using as a single loop method, the other transfer gate TO...
Use together with...

That is, in the same way as the transfer in the M/m loop method, bubbles are block transferred and stored in each minor loop, but after a certain time, another transfer gate TO
. . . The blocks are transferred to the readout line L and arranged in series, and then transferred to the major loop M along the readout line L and read by the detector D. Therefore, it is the same as the M/m loop method in that the bubbles circulate in the stored state in the minor loop, but if we focus on the bubble string of one block, as shown in the figure, when there are four minor loops, , the four series of bubbles sent from the generator are transferred to the transfer gate T...
・At the same time, the minor loop ml is transferred to the fish and circulated, and when it is the turn to transfer, the reading line L
The data is transferred to the major loop M in the same 4-piece series arrangement as before the transfer. Therefore, if bubbles are sequentially transferred to the minor loop in units of 1 block (4 bits) and then transferred and read out one block at a time in the same order as when they were transferred, the bubbles will move in exactly the same way as in the single-loop method. That means you are doing it. As is clear from the above explanation, when operating as a single loop,
Transfer gate T on the major loop M side...
. . . is for transfer only, and transfer gate TO on the read line L side is for transfer only. The operation of the single loop type will be specifically explained. Z Now, assuming that the number of single loops is 4 and the single loop interval in the major loop is 2 bits, in order to minimize the number of empty bits in the single loop, the number of bits in each minor loop is, for example, as shown in Figure 4. A possible 19 bits are considered. In this first configuration, the series array bubbles S,,S that have been transferred every other bit from the generator
2, S3, S4 are transfer gates T...
...When it comes to the top, transfer the head nok to the minor loop at once. The block of Babur transferred for the first time in this way becomes the 1 bit in each minor loop. Bubble 1 transferred during the minor loop
While the block is circulating in a loop with a rotating magnetic field,
The next block of bubbles transferred from the generator (bubbles S,...S2, S3, S, shown in a chain line circle) are transferred into the block, but when these are transferred into the block at the transfer gate, Since bubble 1, which was transferred first, has been moved by 8 bits in the minor loop, the bubble of the second block is transferred to position 2, which is 8 bits after address 1. Similarly below,
For every 8 rotations of the magnetic field, one block is converted from series to parallel during each minor loop, and 34
5... are stored in each minor loop in the order of... Then, after the lth block is transferred, the bubble of the 19th block is transferred and stored in the vacant position after the next 8 rotations of the magnetic field. After storing the first bubble, after the next 8 rotations of the magnetic field, the transferee is stored at address 1 in each minor loop.

To do this, address 1 must be transferred: T.
. . . It is necessary to transfer address 1 and make it an empty bit before reaching the position. for that,
The block at address 1 is transferred to the readout line L by the readout gate TO in preparation for the next block transfer to the second block. Address 1 is read all at once.Flash line L
When the block is transferred, the data becomes a series arrangement on the readout line, the same as the arrangement on the major loop M before the block transfer, and in that state, it is transferred to the major loop M along the readout line and read by the detector D.

Alternatively, it continues through the major loop and is stored again in each minor loop. In the same way, the bubble at address 2 in the minor loop is transferred and then 21
The 3rd block is transferred in, the bubble at address 3 is transferred out, and then the 2nd block is transferred in, and the operation is repeated in sequence. If we take a comprehensive look at the arrangement and movement of the bubbles from the bubble generator G to the detector D, we can see that the bubbles are generated in blocks, generator G → major loop M → transfer gate T... →Minor loop ml・・・・・・・・・Question→Transfer gate TO・・・・・・
...→Readout line L→Measure loop M→Replicator S→Detector D and annihilator A. And the bubble array is 1111, then 2222...
... [stock] [phase] gon [, O [phase] mouth [phase], and within each block, the order of generation from the generator, that is, S
,,S. ,S3,S. The order is as follows. These arrangements and operations are exactly the same as those of bubbles in a single loop configuration. In FIG. 4, at the time when the bubble at address 1 is transferred onto the readout line L, the [phase] address has not yet reached the position of the gate T.

Therefore, as shown in Fig. 5-2, the conductor C2 of the transfer gate is
The control pulse is applied to the bit ahead and transferred, and the vacant bit to be stored in the next bit is transferred to the transfer gate T...
When it reaches the position of ・, a current pulse is applied to the conductive wire C, and the wire is transferred. In the illustrated example, the timing of transfer and transfer is different, but in any case, the transferred blocks are transferred in the order of J, so that the transfer gate T... It is sufficient to perform control such that the bit becomes empty while the bit returns to , and prepares for the next transfer. The position of the transfer bit in the minor loop is the transfer gate TO.
When the position of ・・・・・・・・・ is reached, the empty bit in the minor loop to be transferred is exactly at the transfer gate T.
Change the number of bits of the minor loop and the gate T so that it is at the position of...
By configuring the arrangement of . . . , conductors C,,C. It is also possible to simultaneously apply a current pulse to both and move in and out at the same time. In FIG. 5, A indicates the timing of the clock signal, and the rotating magnetic field rotates in synchronization with this.

In Fig. 4, a minor loop M, ... ... ... is provided for every other bit with respect to the major loop M, so one bit is output from the generator G as shown in the diagram. Bubble information is generated every once in a while. And C, as shown in Figure 2, conductors C,,C for transfer-in/transfer-out. A current pulse is applied every 8 cycles of a clock signal, that is, a rotating magnetic field, and block transfer/transfer is performed one block (4 bits) at a time. That is, if n = (number of minor loops) × (bit interval between minor loops), then in the illustrated example, n = 4 × 2 = 8, and the two control conductors C, C2 each receive a clock signal. 11F
3. In other words, j current pulses are flowing at a period of 1/n.

In the illustrated example, a case has been described in which the number of minor loops is 4 and the minor loop interval is 2 bits, but these can be freely selected.

However, depending on the value of n and the number of bits in the minor loop, more empty bits than necessary may occur during the minor loop. In other words, although it is sufficient to have only one free bit before reaching the transfer position, there are cases where two or more bits are left vacant, which is wasted, so the optimal value is designed to minimize such waste. can. Also, close each minor loop to the major loop,
That is, it is also possible to arrange them at 1-bit intervals.

At this time, of course, as shown in Figure 6, the generator G continuously generates bubbles in synchronization with the clock signal, so that the bubbles are transferred in and out at regular intervals as shown in Figure 2. This will be done consecutively for 2 bits at a time. By adopting the above configuration and control method, it is possible to create extra minor loops in advance and not use them when a defective loop occurs, so it is possible to operate as a single loop method. Even if there is, unlike conventionally,
Defect loops can be easily dealt with in the same manner as the M/m loop method, and as a result, manufacturing yield is improved.

Moreover, in addition to the conventional M/m loop configuration, by providing a readout line connected to the major loop and connecting each minor loop to the readout line via a transfer gate, the M/m loop system can be used. When using the single-loop method, the bubbles in each minor loop are block-transferred in a fixed order, and then block-transferred, so that the same element can be It can be used as either a loop method or a single loop method. For this reason,
Only one type of bubble element is required, which significantly reduces costs. As a user, there is no need to consider and specify whether it is an M/m loop or a single loop each time, and the same bubble element can be used for either the M/m loop type or the single loop type each time, which is convenient.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing a conventional M/m loop configuration, Fig. 2 is a diagram showing a conventional single loop configuration, and Fig. 3 is a diagram illustrating a transfer circuit of an element for both M/m loop and single loop according to the present invention. , FIG. 4 is an enlarged view of the main part of FIG. 3, and FIGS. 5 and 6 are diagrams showing clock signals, bubble generation, and operation timings of two transfer gates.

Claims (1)

[Claims] 1. A magnetic bubble device having a major loop and a minor loop is provided with a readout line connected to the major loop, and each minor loop is connected to the readout line via a transfer gate, and a transfer from the major loop to the minor loop is provided. By performing transfer out from the fine and minor loops to the readout line at a 1/n period (n = number of minor loops x bit interval between minor loops), it is possible to operate as a single loop system. M/
Magnetic bubble storage system for both m-loop and single-loop.