Static information storage and retrieval – Read/write circuit – Bad bit
Reexamination Certificate
2002-07-03
2004-04-27
Lam, David (Department: 2818)
Static information storage and retrieval
Read/write circuit
Bad bit
C365S189011, C365S189120
Reexamination Certificate
active
06728146
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to digital memory systems and, more particularly, to a memory device, a method for automatically repairing defective memory cells, and a method for accessing a memory device.
2. Description of the Related Art
Memory devices typically include an array of memory cells. If one memory cell fails (i.e., fails to store the data written into the memory cell), then the complete array may have to be scrapped. Typical memory devices include several sectors or arrays. Each of the arrays includes several working cells and at least one redundancy cell. If one of the working cells fails, then the redundancy cell can be substituted for the failed working cell and the data can be stored in the redundancy cell in the same array.
One of the shortfalls of conventional memory devices is that if the number of failed working cells in a first array is greater than the number of redundant cells in the same array, then all of the failed working cells cannot be replaced. Therefore, the entire memory device has failed and must be replaced. In addition, even if a second array has unused redundant cells, the first array cannot use the unused redundant cells from the second array to replace failed working cells in the first array. Therefore, the unused redundant cells in the second array may go unused.
Memory devices are typically repaired in the fabrication facility. For example, in the case in which a redundant cell is used to replace a failed working cell, a fuse is set with, e.g., a laser, so that the redundant cell is activated or the failed working cell is deactivated. In the field, however, if the working cell fails (e.g., when the memory device is being used), then the failed working cell cannot be replaced and entire memory device must be replaced.
Therefore, in view of the foregoing, what is needed is a technique for automatically repairing a failed memory device that is capable of repairing the memory device either in the fabrication facility or in the field.
SUMMARY OF THE INVENTION
Broadly speaking, the present invention fills this need by automatically replacing a failed working memory cell with a redundancy memory cell. By way of example, the present invention may be implemented in the form of a system, an apparatus, a method, a device, or a computer readable media.
In accordance with one aspect of the present invention, a memory device that is capable of automatically repairing itself is provided. The memory device includes a working cell array, a redundancy cell array, a control logic, and a lookup table. The working cell array includes a plurality of working memory cells. The redundancy cell array includes a plurality of redundancy memory cells. The control logic controls reading and writing of data to the working cell array and the redundancy cell array. The lookup table stores addresses of the data stored in the redundancy cell array.
In one embodiment, the memory device includes a power down detector to detect a power off voltage. The memory device also may include a low voltage detector for detecting a power on voltage. In one embodiment, the control logic includes at least one control register and the lookup table includes at least one latch register.
In one embodiment, the redundancy cell array is distributed between at least two sectors. In another embodiment, the redundancy cell array is included in one sector. In one embodiment, the redundant cell array includes one data byte per sector and three address bytes per sector. In one embodiment, the redundancy cell array includes a data portion in a first sector and an address portion in a second sector. In another embodiment, the redundancy cell array includes a status portion in a third sector.
In accordance with another aspect of the present invention, a method for automatically repairing a defective memory cell in a memory device is provided. In this method, data is written to a working memory cell. The data written to the working memory cell is then read. The data written to the working memory cell and the data read from the working memory cell are compared. If the data written to the working memory cell and the data read from the working memory cell are not equal, i.e., the same, then the data is written to a redundancy memory cell and an address of the redundancy memory cell and an address of the working memory cell are updated in a lookup table.
In one embodiment, the data is read from the redundancy memory cell. In this embodiment, the data written to the redundancy memory cell is compared to the data read from the redundancy memory cell. If the data written to the redundancy memory cell and the data read from the redundancy memory cell are not equal, then the data is written to another redundancy memory cell.
In one embodiment, the method further includes detecting a power off condition in the memory device and storing the lookup table. In one embodiment, the lookup table is stored in a redundancy cell array that includes the redundancy memory cell.
In one embodiment, the method further includes detecting a power on condition in the memory device and initiating the lookup table. In this embodiment, the operation of initiating the lookup table includes reading the lookup table from a redundancy cell array that includes the redundancy memory cell.
In accordance with yet another aspect of the present invention, a method for accessing a memory device is provided. In this method, an address for data is received. It is then determined whether the address is included in a lookup table. If the address is not included in the lookup table, then the data is accessed in the working memory cell. If the address is included in the lookup table, then the data is accessed in the redundancy memory cell at an address corresponding to the address in the lookup table.
The present invention advantageously enables a failed working memory cell in a memory device to be automatically repaired at any time, e.g., in the fabrication facility or in the field. This significantly improves the reliability of memory devices. Such improved reliability reduces both the number of memory devices that must be scrapped during the manufacturing process and also the number of memory devices that must be replaced after use in the field and thereby reduces the costs associated with the manufacture and sale of memory devices.
REFERENCES:
patent: 6430096 (2002-08-01), Grass
patent: 6442083 (2002-08-01), Hotaka
patent: 6577534 (2003-06-01), Tsuruda
patent: 2002/0039311 (2002-04-01), Takeuchi et al.
Li Minfu
Liou Ken
Yen Min Nan
LandOfFree
Memory device and method for automatically repairing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Memory device and method for automatically repairing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Memory device and method for automatically repairing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3269860