JPS5841592B2 - Test method for magnetic bubble memory - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for testing a magnetic bubble memory having a plurality of storage minor loops.

Magnetic bubble memories, which store bits of data in magnetic bubbles moving around in a magnetic thin film, generally consist of a major loop for reading or writing and a minor loop for storing.

In such a type of magnetic bubble memory, a spare loop is generally provided in order to ensure the necessary number of storage loops and increase yield even if the minor storage loop is defective.

That is, if a certain minor loop has a defect (hereinafter, a minor loop having this defect will be referred to as a defective loop),
An external circuit will mask this defective loop so that it is not used, and a spare loop will be used instead.

On the other hand, in order to test this type of magnetic bubble memory, a test device called a magnetic bubble memory tester or simply a tester is used, but the test method using this tester is conventionally used to test all minor loops. When a specific loop is found to be a defective loop, the loop number of the defective loop is set in the tester, and then when writing to the magnetic bubble memory, writing to the defective loop is prohibited.

Normally, information "1" corresponds to the existence state of a magnetic bubble (hereinafter abbreviated as bubble), and information rOJ corresponds to the absence state of a bubble. is the case where a bubble exists, that is, the information is "1".

Therefore, prohibiting writing of information to a defective loop is equivalent to not allowing a bubble to exist, that is, writing information "0".

On the other hand, when reading, the information written in the defective loop is meaningless, so the defective loop is read and discarded.

In this way, the magnetic bubble memory tester is written.

When writing, information is prohibited from being written to the defective loop, and when reading, information from the defective loop is discarded.
It is configured to provide a so-called defective loop skipping function.

The configuration and functions of a conventional magnetic bubble memory tester will be described below with reference to FIG.

1 is a loop counter indicating loop information, which counts the number of loops to be written or read to determine a loop number.

2 is a defective loop storage circuit for storing the contents of defective loops.

The output of the loop counter 1 is connected to the address of the defective loop memory circuit 2, and whether or not the test loop specified by the loop counter 1 is a defective loop is input to the defective loop memory circuit 2. Information on the defective loop is output to the signal line 3 labeled 5KIP (skip bar) of the defective loop storage circuit 2.

4 is a data generator that generates data for testing the magnetic bubble memory.

5 is an AND gate that inputs the data and the 5KIP signal; if the loop is not a defective loop, the data from the data generator 4 becomes WRITE DATA 6; if the loop is a defective loop, the data is WRITE DATA 6; , the signal output to the 5KIP signal line 3 is rLow
Therefore, the data obtained as the output of the AND gate 5 becomes i-LowJ, that is, 1-01.

On the other hand, at the time of reading, R read out from the bubble
EAD DATA (read data) 1 is compared with the output of data generator 4 by comparison gate 8, and the output of comparison gate 8 is READ DATA.
7 and the output of the data generator 4 do not match, that is, if there is an error, the result is "HighJ".

One output of the comparison gate 8 is connected to the input of the AND gate 9, and the other input is connected to the 5KIP signal line 3, and if the read loop is a defective loop,
Since the 5KIP signal line 3 is I-LowJ and the output of the AND gate 9 is I-LowJ, the data is not checked and is discarded.

Can you test magnetic bubble memory with a conventional tester that has this configuration and functionality? As a result of testing with J, it became clear that the conventional method is insufficient as a tester for magnetic bubble memory.

This point will be explained below.

The defective loop includes armpit defects where “0” becomes “1” and “1”.
There is a chip defect in which "" is "0".

In armpit defects, bubbles are naturally generated due to crystal defects in the magnetic thin film, and bubbles are always read out.

A defective chip occurs when a pattern such as a T-bar used as a transfer pattern has a defect, and this defect causes a transfer error when a bubble is transferred over the pattern, causing the bubble to disappear.

In the case of such a defective loop, proper testing can be performed by providing a skip function like the conventional tester described above.

By the way, until now we have thought that defective loops are a phenomenon caused only by defective loops, and that the influence from other loops can be completely ignored, but after conducting numerous tests on magnetic bubble memory, we have found that defective loops can be skipped. It has become clear that just doing so is not enough, and that loops adjacent to the defective loop affect the defective loop and cause malfunctions.

In other words, in such a case, even if the defective loop is skipped, when information is written to the loop adjacent to the defective loop, an unnecessary bubble may be generated in a specific location of the defective loop due to the influence of the bubble in the adjacent loop. Depending on the shape of the defect in the transfer pattern, the bubble that transfers the adjacent loop may split into two, and one may enter the defective loop.

When a bubble is generated in the defective loop in this way, the bubble in the defective loop cannot be controlled, so the bubble strays into the normal loop again and causes the normal loop to malfunction.

In this way, when a defective loop generates bubbles due to the influence of adjacent loops and causes malfunctions, the bubbles are not detected when they occur in the defective loop, and are detected only after the bubbles in the defective loop stray into the normal loop. This has the disadvantage that it takes a long time to cause malfunctions and to detect them, requiring a very long test time.

The present invention eliminates the above-mentioned drawbacks of the prior art, quickly detects defective loops that are affected by adjacent loops and generates unnecessary bubbles, and causes malfunctions, and tests can be performed excluding these defective loops. The purpose of this study is to provide a test method for magnetic bubble memory that can be used.

In order to achieve the above object, the present invention writes all bits 0 to a specific minor loop, writes an arbitrary pattern to other minor loops, then reads the stored data of the specific minor loop, and checks whether all bits are 0 or not. The present invention is characterized in that it checks whether the data is different from the other minor loops, reads out arbitrary data written in other minor loops, and compares the read data with the written data to see if there is a difference between the two.

The present invention will be explained below with reference to illustrated embodiments.

In FIG. 2, numerals 11, 12, and 14 are a loop counter, a defective loop storage circuit, and a data generator, respectively, which have the same configuration and function as those in FIG. 1.

15 is the 5KIP signal 13 from the defective loop storage circuit 12
WRITE DATA 16 is output by an AND gate which receives the output from the data generator 14 and the output from the data generator 14 as inputs.

WRITE DA which is the output of this AND gate 15
READ sent from TA16 and a detector (not shown)
DATA17 is connected to the input of comparison gate 18.

The input of this comparison gate 18, that is, the READ DATA Iγ and WRITE which are compared by the comparison gate 18
Among DATA16, WRITEDATAIγ is always "0" due to the 5KIP signal 13 in the case of a defective loop, so READDATA17 is compared with "0".

Therefore, for the present invention, the defective loop is typically always r
It is checked as OJ, but if a bubble occurs in the defective loop due to the influence of the neighboring loop, the data of the defective loop will be read as "1", and as a result, this loop will be affected by the neighboring loop and become 1. -1'' is output to the output of the comparison gate '-N8, so such chips can be removed.

As is clear from the above description, the method of the present invention writes all bits 0 and an arbitrary pattern to a specific inner loop and other minor loops, and then reads out these stored contents and compares them with the data at the time of writing. By checking whether there is a difference between the two, all defective loops, including defects that are affected by adjacent loops and cause unnecessary bubbles, can be detected in a short time, and this defective loop can be removed. This makes it possible to significantly improve work efficiency and reliability.

[Brief explanation of the drawing]

Figure 1 is a block diagram of a conventional magnetic bubble memory tester.
FIG. 2 is a block diagram of a magnetic bubble memory tester used in the testing method of the present invention. 11... Loop counter, 12... Defective loop storage circuit, 14... Data generator, 1
5...AND gate, 1B...Comparison gate.

Claims (1)

[Claims] 1. In a test method for a magnetic bubble memory having a plurality of storage minor loops, all bits O are written to a specific minor loop, arbitrary patterns are written to other minor loops, and then the memory data of the specific minor loop is read, and this The feature is that it checks whether the bit is O or not, reads arbitrary data written to this from another minor loop, and compares the read data with the written data to check whether there is a difference between the two. Test method for magnetic bubble memory.