A method of correcting errors of a flash memory comprises steps of modifying the data of a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein, checking for the presence or absence of an error of not properly modifying the data of the group of memory units and determining the completion of proper modification of the data of the group of memory units provided that an error is detected and the error can be corrected.
|
1. A method of controlling a flash memory system comprising the steps of:
modifying data of a group of flash memory cells adapted to erasing data therefrom and writing data therein;
checking for the presence or absence of an error of not properly modifying said data of said group of memory units cells; and
determining the completion of proper modification of said data of said group of memory units cells provided that an error is detected and said error can be corrected.
18. A flash memory device comprising:
a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; and
an error judgement section for determining the completion of proper data modification of data provided that the number of errors detected during data modification by an error detection/correction unit detecting/correcting errors in the data written in said group of memory units is not greater than a number of errors within a correctable range.
20. A computer-readable recording medium storing a program for realizing in a computer:
a function of ordering modification of data to an external memory device containing a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; and
a function of checking for the presence or absence of an error of not properly modifying said data of said group of memory units and determining the completion of proper modification of said data of said group of memory units provided that an error is detected and said error can be corrected.
19. A flash memory device comprising:
a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein;
an error detection unit for reading the data from said group of memory units and detecting errors during data modification; and
an error judgement section for counting the number of errors detected by said error detection unit and determining the completion of proper data modification of data provided that the number of errors detected by said error detection unit is not greater than a number which can be corrected in accordance with an error check code.
4. A flash memory system comprising:
a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein;
an error detection unit for reading the data written in said group of memory units and detecting errors during data modification; and
an error judgement section for counting the number of errors detected by said error detection unit and determining the completion of proper data modification of data provided that the number of errors detected by said error detection unit is not greater than a number which can be corrected in accordance with an error check code.
2. A method of controlling a flash memory system comprising steps of:
erasing data written in a group of flash memory cells adapted to erasing data therefrom and writing data therein;
reading the data written in said group of memory units cells having said data erased and checking the completion of proper erasure of said data;
counting the number of errors of not being properly erased provided that said data are determined not to be properly erased as a result of said checking step; and
determining completion of proper erasure of said data of said group of memory units cells provided that the counted number of errors is within a correctable range.
3. A method of controlling a flash memory system comprising steps of:
writing data in a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein;
reading the data written in said group of memory units and checking the completion of proper writing of said data;
counting the number of errors of not being properly written provided that said data are determined not to be properly written as a result of said checking step; and
determining the completion of proper writing of said data of said group of memory units provided that the counted number of errors is within a correctable range.
22. A computer-readable recording medium storing a program for realizing in a computer; :
a function of ordering writing of data written in a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; and
a function of reading the data written in said group of memory units and checking the completion of proper writing of said data counting the number of errors of not being properly written provided that said data are not properly written as a result of said checking step and determining the completion of proper writing of said data of said group of memory units provided that the counted number of errors is within a correctable range.
21. A computer-readable recording medium storing a program for realizing in a computer; :
a function of ordering erasure of the data written in a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; and
a function of reading the data written in said group of memory units having said data erased and checking the completion of proper erasure of said data, counting the number of errors of not being properly erased provided that said data are not properly erased as a result of said checking step and determining the completion of proper erasure of said data of said group of memory units provided that the counted number of errors is within a correctable range.
14. A flash memory device comprising:
a memory unit, including a plurality of flash memory cells, for storing data which can be electrically changed;
a control circuit for reading out the data stored in said memory unit and for controlling said memory unit in order to change the data stored in said memory unit;
an error judgement section, being included in said control circuit, for counting the number of failure data regarded as having not been successfully stored in the memory unit;
wherein, if the number of failure data is not larger than a predetermined number, said control circuit outputs status data which means that changing of data has been successfully done, wherein the counting is responsive to the changing of data during an erasing or a writing operation.
9. A flash memory system comprising:
a memory cluster comprising a plurality of external flash memory cells;
at least one or more than one memory sectors constituting said memory cluster;
a flash memory control unit for ordering writing of the data in said memory sectors;
an error detection/correction unit for detecting write errors in the data read from said memory cluster and correcting write errors up to n (n is a number of bits in a memory sector, that can be corrected in accordance with an error check code) bits attributable to the data; and
an error judgement section for counting the number of bits showing write errors and determining the completion of proper data writing provided that the number of bits showing write errors is not greater than m (1≦m≦n) bits in each and every memory sector.
8. A flash memory system comprising:
memory sectors having a plurality of flash memory cells;
a flash memory control unit for ordering writing of the data in each of said memory sectors;
an error detection/correction unit for detecting write errors in the data read from said memory cluster and correcting write errors up to n (n is a number of symbols in a memory sector, that can be corrected in accordance with an error check code (1 symbol=k bits, k≧2)) symbols (1 symbol=k bits, k≧2) attributable to the data; and
an error judgement section for counting the number of symbols defective in terms of writing as detected by said detection/correction unit aid and determining the completion of proper data modification provided that the number of symbols showing errors is not greater than m (1≦m≦n) in each and every memory sector.
7. A flash memory system comprising:
a memory cluster comprising a plurality of external flash memory cells;
at least one or more than one memory sector constituting said memory cluster;
a flash memory control unit for ordering writing of the data in said memory sectors;
an error detection/correction unit for detecting errors in the data read from said memory cluster and correcting erase errors up to n (n is the number of memory cells in a memory sector, that can be corrected in accordance with an error check code) attributable to flash memory cells; and
an error judgement section for counting the number of memory cells defective in terms of writing and contained in each memory sector and determining the completion of proper data writing provided that the number of memory cells defective in terms of writing is not greater than n in each and every memory sector.
5. A flash memory system comprising:
a memory cluster having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein;
at least one or more than one memory sectors constituting said memory cluster;
a flash memory control unit for ordering erasure of the data written in said memory cluster;
an error detection/correction unit for detecting erase errors in the data read from said memory cluster and correcting erase errors up to n (n is a number of memory cells in a memory sector, that can be corrected in accordance with an error check code) attribute attributable to memory cells; and
an error judgement section for counting the erase errors of each memory sector and determining the completion of proper data erasure provided that the completion of proper data erasure provided that is judged by the number of memory cells storing unerased data is not greater than m (1≦m≦n) in each and every memory sector.
6. A flash memory system comprising:
a memory cluster having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein;
at least one or more than one memory sectors constituting said memory cluster;
a flash memory control unit for ordering erasure of the data written in said memory cluster;
an error detection/correction unit for detecting erase errors in the data read from said memory cluster and correcting erase errors up to n (n is a number of symbols in a memory sector that can be corrected in accordance with an error check code (1 symbol=k bits, k≧2)) symbols attributable to the data;
a counter unit for counting the number of symbols showing erase errors and contained in each memory sector; and
an error judgement section for determining the completion of proper data modification provided that the number of symbols showing erase errors is not greater than m (1≦m≦n) in each and every memory sector.
10. A flash memory system comprising:
a memory unit, including a plurality of flash memory cells, for storing data which can be electrically changed;
a control circuit for reading out the data stored in said memory unit and for controlling said memory unit in order to change the data stored in said memory unit;
an error detection and correction unit for detecting whether an error has been occurred exists in read-out data read out by said control circuit and for correcting errors in the read-out data; and
an error judgement section, being implemented independent of said error detection and correction unit, for counting the number of failure data regarded as having not been successfully stored in the memory unit;
wherein said flash memory system settles that changing of data has been successfully done if the number of failure data is not larger than a predetermined number, said predetermined number satisfying the condition that the predetermined number of failure data can be corrected by said error detection and correction unit.
23. A flash memory system comprising:
a flash memory cell array;
a control circuit for erasing data from designated part of said memory cell array, for writing data into given part of said memory cell array, and for reading data from said designated or given part of said memory cell array, thereby it can be detected how many errors have been occurred within said designated or given part of said memory cell array after erasing or writing;
a counter for counting the number of the errors based on data read by said control circuit after the erasing or writing;
an error judgement section for outputting information that it is successfully erased or written if the number counted is smaller than a number which can be corrected in accordance with an error check code; and
an error detection/correction unit equipped beside said counter and said error judgement section for generating said error check code accompanying with data to be written into said given part of said memory cell array, said error check code being also written into said given part, and for correcting errors existing in data read from said designated or given part of said memory cell array.
27. A flash memory device comprising:
a flash memory cell array;
a control circuit for erasing data from a designated part of said memory cell array, for writing input data and check code incidental to said input data into a given part of said memory cell array, said input data being also input from external to the flash memory device, and for reading data from said designated or given part of said memory cell array, thereby whereby it can be detected how many errors have been occurred exist within said designated or given part of said memory cell array after erasing or writing;
a counter for counting the number of the errors based on data read by said control circuit after the erasing or writing; and
an error judgement section for outputting information that it a part of the said memory cell array is successfully erased or written if the number counted is smaller than a number that can be corrected in accordance with an error check code;
wherein if read instruction and address are input to the flash memory device, the input data and the check code stored in part of said memory cell array addressed by said address are read out by said control circuit and output to external to the flash memory device without any modification.
12. The flash memory system according to
13. The flash memory system according to
16. The flash memory device according to
17. The flash memory device according to
24. The flash memory system according to
25. The flash memory system according to
26. The flash memory system according to
28. The flash memory device according to
29. The flash memory device according to
30. The flash memory device according to
0. 31. The method of
0. 32. The method of
0. 33. The method of
0. 34. The method of
0. 35. The method of
0. 36. The method of
0. 37. The method of
0. 38. The method according to
0. 39. The method of
0. 40. The flash memory system of
0. 41. The flash memory system of
0. 42. The flash memory system according to
0. 43. The flash memory system of
0. 44. The system according to
0. 45. The system according to
0. 46. The system according to
0. 47. The system according to
0. 48. The system according to
0. 49. The system according to
0. 50. The system according to
0. 51. The system according to
0. 52. The system according to
0. 53. The system according to
0. 54. The system according to
0. 55. The system of
0. 56. The flash memory device according to
0. 57. The flash memory device according to
0. 58. The flash memory device according to
0. 59. The flash memory device according to
0. 60. The flash memory device according to
0. 61. The flash memory device according to
0. 62. The flash memory device according to
0. 63. The flash memory device according to
0. 64. The flash memory device according to
0. 65. The flash memory device according to
0. 66. The flash memory device according to
0. 67. The flash memory device according to
|
|||||||||||||||||||||
This invention relates to a method of controlling a flash memory. It also relates to a flash memory system and a flash memory device using such a control method.
Hard disk devices are popularly used as computer memory devices that are available at low cost relative to the memory capacity they provide. In many hard disks, file data are written and read out on a unit by unit basis and the unit is referred to as sector (e.g., 512 bytes).
Hard disks can give rise to errors due to a scar and/or particles of dirt that make it no longer possible to carry out the normal operation of writing data into or retrieving data from a sector. However, such errors can be detected and corrected by adding an annex data referred to as error check code to each file data. With the use of an error check code, the presence or absence of an error in a file data can be detected by using the file data and the error check code. If there is an error, then it can be corrected.
In many file management systems that are used for computers, a larger unit is formed by a plurality of sectors so that file data are mostly controlled on a cluster basis on the computer.
A hard disk consumes electric power at a relatively high rate for driving its motor. Additionally, vibrations generated by the motor can damage the disk and hence its reliability. Flash memories are introduced to bypass this problem. A flash memory consumes power at a lower rate and can withstand vibrations. Additionally, it does not require a power back up system comprising batteries. Thus, flash memories are popularly used in portable electronic equipment.
As far as this patent application is concerned, a memory device using a flash memory is referred to as flash memory system. In many instances, a flash memory is designed to operate just like a hard disk from the viewpoint of computer.
In other words, a flash memory system is operated in such a manner that a hard disk driver is driven for operation. This is because it is a great advantage for any computers to be able to access a hard disk and a flash memory without discrimination.
To do this, the computer is required to be able to write data into and read data from a flash memory on a sector basis. In other words, a flash memory is required to be provided with a data register that can temporarily store data by a volume to be collectively stored in a sector. Then, a volume of data to be written into and read from a sector will be collectively handled by the computer. This technique is well known and a system adapted to collectively write data into and read data from a sector can operate at high speed.
NAND type flash memories are known to be particularly adapted to the above technique. With an NAND type flash memory, for instance, data can be collectively written into and read from a sector by 528 bytes and four sectors may form a cluster. Then, data can be collectively erased on a cluster by cluster basis.
When the data stored in a cluster are erased in a flash memory system comprising an NAND type flash memory, the computer checks if all the data in the memory cells of the cluster have been erased or not (an operation including VERIFY/ERASE). If there is a memory cell in the cluster where a data has not been erased and remains there, the cluster is deemed as faulty and the cluster is disabled.
Similarly, when data are written into a sector, the computer checks if data are written correctly into all the memory cells or not (an operation including VERIFY/WRITE). If there is a memory cell in the sector where a data cannot be written, the sector is deemed as faulty and the cluster containing the sector is disabled.
Redundant clusters are provided in order to replace disabled clusters. When the number of disabled clusters exceeds a given level, it is no longer possible to replace a disabled cluster and the entire flash memory becomes faulty.
Multi-valued NAND type flash memories where a plurality of threshold voltages, e.g., four threshold voltages, are provided for each memory cell and a 2-bit data is written there are known. Such flash memories require a rigorous control of threshold voltage values far more than any ordinary NAND type flash memories. Thus, such flash memories can show a higher rate of appearance of faulty memory cells than any ordinary NAND type flash memories because their data retaining performance can often degrade.
In view of this problem, flash memory systems comprising a multi-valued NAND type flash memory are required to reliably detect and correct errors. When the data stored in a cluster are erased in such a flash memory system, the computer also checks if all the data in the memory cells of the cluster have been erased or not (an operation including VERIFY/ERASE). If there is a memory cell in the cluster where a data has not been erased and remains there, the cluster is deemed as faulty and the cluster is disabled.
Similarly, when data are written into a sector, the computer checks if data are written correctly into all the memory cells or not (an operation including VERIFY/WRITE). If there is a memory cell in the sector where a data cannot be written, the sector is deemed as faulty and the cluster containing the sector is disabled.
As described above, because of the provision of error check codes that are annexed to data in a system where a plurality of memory cells can be checked for error detection and error correction, only a single faulty memory cell disables the entire cluster for any operation of erasing and writing data before the system operates in an error detection/correction mode.
In other words, while the memory system may be relieved of a faulty condition, it can often give rise to a faulty sector or a faulty cluster and consequently a large disabled memory area that can leads to a system fault.
In view of the above identified circumstances, it is therefore the object of the present invention to provide a flash memory system that can correct errors in such a way that any unnecessary expansion of disabled memory areas is effectively prevented.
(1) A method of controlling a flash memory system according to the present invention comprises: modifying the data of a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; checking for the presence or absence of an error of not properly modifying the data of the group of memory units; and determining the completion of proper modification of the data of the group of memory units provided that an error is detected and the error can be corrected.
(2) A method of controlling a flash memory system according to the present invention comprises: erasing the data written in a group of memory units; each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; reading the data written in the group of memory units having the data erased and checking the completion of proper erasure of the data; counting the number of errors of not being properly erased provided that the data are not properly erased as a result of the checking step; and determining the completion of proper erasure of the data of the group of memory units provided that the counted number of errors is within a correctable range.
(3) A method of controlling a flash memory system according to the present invention comprises: writing data in a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; reading the data written in the group of memory units and checking the completion of proper writing of the data; counting the number of errors of not being properly written provided that the data are not properly written as a result of the checking step; and determining the completion of proper writing of the data of the group of memory units provided that the counted number of errors is within a correctable range.
(4) A flash memory system according to the present invention comprises: a group of memory units, each having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; an error detection/correction unit for reading the data written in the group of memory unit and detecting/correcting errors up to a predetermined number; an error judgement section for counting the number of errors detected by the error detection/correction unit and determining the completion of proper data modification of data provided that the number of errors detected by the error detection/correction unit is not greater than the predetermined number.
(5) A flash memory system according to the present invention comprises: a memory cluster having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; at least one or more than one memory sectors constituting the memory cluster; a flash memory control unit for ordering erasure of the data written in the memory cluster; an error detection/correction unit for detecting erase errors in the data read from the memory cluster and correcting erase errors up to n attributable to memory cells; an error judgement section for counting the erase errors of each memory sector and determining the completion of proper data erasure provided that the number of memory cells storing unerased data is not greater than m (1≦m≦n) in each and every memory sector.
(6) A flash memory system according to the present invention comprises: a memory cluster having a plurality of flash memory cells adapted to erasing data therefrom and writing data therein; at least one or more than one memory sectors constituting the memory cluster; a flash memory control unit for ordering erasure of the data written in the memory cluster; an error detection/correction unit for detecting erase errors in the data read from the memory cluster and correcting erase errors up to n symbols (1 symbol=k bits, k≧2) attributable to the data; a counter unit for counting the number of symbols showing erase errors and contained in each memory sector; an error judgement section for determining the completion of proper data modification provided that the number of symbols showing erase errors is not greater than m (1≦m≦n) in each and every memory sector.
(7) A flash memory system according to the present invention comprises: a memory cluster comprising a plurality of external flash memory cells; at least one or more than one memory sectors constituting the memory cluster; a flash memory control unit for ordering writing of the data in the memory sectors; an error detection/correction unit for detecting errors in the data read from the memory cluster and correcting erase errors up to n attributable to flash memory cells; an error judgement section for counting the number of memory cells defective in terms of writing and contained in each memory sector and determining the completion of proper data writing provided that the number of memory cells defective in terms of writing is not greater than n in each and every memory sector.
(8) A flash memory system according to the present invention comprises: memory sectors having a plurality of flash memory cells; a flash memory control unit for ordering writing of the data in each of the memory sectors; an error detection/correction unit for detecting write errors in the data read from the memory cluster and correcting write errors up to n symbols (1 symbol=k bits, k≧2) attributable to the data; an error judgement section for counting the number of symbols defective in terms of writing as detected by the detection unit aid and determining the completion of proper data modification provided that the number of symbols showing errors is not greater than m (1≦m≦n) in each and every memory sector.
(9) A flash memory system according to the present invention comprises: a memory cluster comprising a plurality of external flash memory cells; at least one or more than one memory sectors constituting the memory cluster; a flash memory control unit for ordering writing of the data in the memory sectors; an error detection/correction unit for detecting write errors in the data read from the memory cluster and correcting write errors up to n bits attributable to the data; an error judgement section for counting the number of bits showing write errors and determining the completion of proper data writing provided that the number of bits showing write errors is not greater than m (1≦m≦n) bits in each and every memory sector.
As well as the above-described methods, the functions of the error judgment section and the error detection/correction unit can be also realized, using a computer-readable recording medium that stores programs for enabling the computer to execute predetermined functions.
Further, the invention related to a method for controlling a flash memory system can be realized by a computer-readable recording medium, which stores programs for enabling the computer to execute a procedure corresponding to the method invention.
In the invention, data control unit 8 9, control circuit 11 receives a data and carries out an erase, write or read operation at the specified physical address in the memory cell array 10.
The control circuit 11 includes error judgement section 111. The error judgement section 111 counts, using a counter 112, errors which may occur in each of data erasing and writing modes. If the number of errors falls within a range in which the error detection/correction unit 7 can correct them in accordance with a predetermined algorithm, the error judgement section 111 permits the existence of the errors and judges that normal erasion or writing has been executed.
In this embodiment, a 4-valued memory system is used to store a 2 bits data in each memory cell of the flash memory. As shown in
Additionally, error check code C0-y and error check code C1-y (y=0, 1, . . . , 42) are paired as shown in the column indicated by ECC in FIG. 7 and stored in a memory cell (hereinafter error check code is referred to as check code). Check codes C0-y and C1-y are generated in error detection/correction unit 7 in FIG. 6. Check code C0-y is annexed to data D0-0 through D0-2047 so as to be able to correct errors for n bits by means of a given error correction algorithm. Similarly, check code C1-y is annexed to data D1-0 through D1-2047 so as to be able to correct errors of n bits by means of a given error correction algorithm.
Thus, if n memory cells become defective in the above embodiment, the n errors can successfully be corrected. The number of bits that can be remedied is limited by the check code length. While the check code length is equal to 43 bits in the above instance, the number of bits that can be remedied can be made greater than 43 by increasing the number of memory cells if necessary.
Logical addresses A0 through A5 in each of LA0 through LA6 are stored in three memory cells. In other words, 7 sets of same address data (LA0 through LA6) are provided. As a result, up to 3 bits can be corrected for an address data Az (z=0 through 5) on the basis of majority logic (the boarder line being defined by 4:3). Thus, it will be appreciated that up to 3 defective memory cells can be remedied. The number of correctable bits can be increased by increasing the number of address data sets.
Keeping the above discussed data arrangement in mind, now a principal part of the second embodiment will be described below by referring to
The control circuit 11 erases the cluster in question and checks the erased state of the cluster (an operation including VERIFY/ERASE).
After the erasing operation, if the control circuit 11 determines that there are one or more than one sectors that contain m (2≦m≦n+1) or more than m memory cells where the stored data have not been erased or there can be one or more than one sectors that contain m (2≦m≦n+1) or more than m memory cells where the stored data has not been erased, it concludes that the normal erase operation has not been carried out successfully for the cluster and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct errors of up to three memory cells for each sector. Assume here that each sector is allowed to make a memory cell error at the time of an erase operation so that the operation is carried out as normal erase operation. Then, if any of the sectors has a single error, the control circuit 11 sends status information of “pass” back to the flash memory control unit 8.
While errors of up to three memory cells are allowed in this embodiment, it is actually so arranged that only a memory cell error is allowed there in order to provide a degree of redundancy for accommodating data errors that may arise as the system degrades with time.
The control circuit 11 writes data in sectors and check the written state of each of the sectors (including an optical of VERIFY/WRITE). If the control circuit 11 determines that there are m (2≦m≦n+1) or more than m memory cells where a data has not been written or there can be m (2<m<n+1) or more than m memory cells where a data has not been written, it concludes that the normal write operation has not been carried out successfully for the sector and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct errors of up to three memory cells for each sector. Assume here that each sector is allowed to make a memory cell error at the time of an write operation so that the operation is carried out as normal write operation. Then, if any of the sectors has a single error, the control circuit 11 sends status information of “pass” back to the flash memory control unit 8.
Now, a third embodiment of the invention will be described also by referring to
In this embodiment, the error detection/correction unit 7 divides data into a plurality of groups (symbols) for so many memory cells and generates check codes C0-1 through C0-42 and C1-1 through C1-42 so that n symbols may be corrected. Popular Reed-Solomon codes may be used for them.
With this arrangement, errors can be corrected if up to the number of memory cells equal to that of n symbols become defective. The number of symbols that can be corrected is limited by the check code length. While the check code length is equal to 43 bits in the above instance, the number of correctable symbols can be made greater than 43 by increasing the number of memory cells if necessary.
Logical addresses A0 through A5 in each of LA0 through LA6 are stored in three memory cells. In other words, 7 sets of same address data (LA0 through LA6) are provided. As a result, up to 3 bits can be corrected for an address data Az (z=0 through 5) on the basis of majority logic. Thus, it will be appreciated that up to 3 defective memory cells can be remedied. The number of correctable bits can be increased by increasing the number of address data sets.
Keeping the above discussed data arrangement in mind, now a principal part of the third embodiment will be described below by referring to
The control circuit 11 erases the cluster in question and checks the erased state of the cluster (an operation including VERIFY/ERASE).
After the erasing operation, if the control circuit 11 determines that there are one or more than one sectors that contain errors equal to m (2≦m≦n+1) or more than m symbols or there can be one or more than one sectors that contain errors equal to m (2≦m≦n+1) or more than m symbols, it concludes that the normal erase operation has not been carried out successfully for the cluster and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct errors of up to three symbols errors for each sector. Assume here that each sector is allowed to make a symbol error at the time of an erase operation so that the operation is carried out as normal erase operation. Then, if any of the sectors has a single symbol error, the control circuit 11 sends status information.of “pass” back to the flash memory control unit 8.
The control circuit 11 writes data in sectors and check the written state of each of the sectors (including an optical of VERIFY/WRITE). If the control circuit 11 determines that there errors equal to m (2≦m≦n+1) symbols or more than m or there can be errors equal to m (2≦m≦n+1) or more than m symbols, it concludes that the normal write operation has not been carried out successfully for the sector and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct up to three symbol errors for each sector. Assume here that each sector is allowed to make a symbol error at the time of an write operation so that the operation is carried out as normal write operation. Then, if any of the sectors has a single symbol error, the control circuit 11 sends status information of “pass” back to the flash memory control unit 8.
Now, a fourth embodiment of the invention will be described also by referring to
In this embodiment, the error detection/correction unit 7 generates check codes C0-1 through C0-42 and C1-1 through C1-42 for data D0-0 through D0-2047 and D1-0 through D1-2047 so that n bits may be corrected.
With this arrangement, errors can be corrected if up to n memory cells become defective. The number of bits that can be corrected is limited by the check code length. While the check code length is equal to 86 bits in the above instance, the check code length can be made greater than 86 by increasing the number of memory cells depending on the value of n if necessary.
Logical addresses A0 through A5 in each of LA0 through LA6 are stored in six memory cells. In other words, 7 sets of same address data (LA0 through LA6) are provided. As a result, up to 3 bits can be corrected for an address data Az (z=0 through 5) on the basis of majority logic. Thus, it will be appreciated that up to 3 defective memory cells can be remedied. The number of correctable bits can be increased by increasing the number of address data sets.
Keeping the above discussed data arrangement in mind, now a principal part of the fourth embodiment will be described below by referring to
The control circuit 11 erases the cluster in question and checks the erased state of the cluster (an operation including VERIFY/ERASE).
After the erasing operation, if the control circuit 11 determines that there are one or more than one sectors that contain m (2≦m≦n+1) or more than m memory cells where the stored data have not been erased or there can be one or more than one sectors that contain m (2≦m≦n+1) or more than m memory cells where the stored data has not been erased, it concludes that the normal erase operation has not been carried out successfully for the cluster and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct errors of up to three memory cells for each sector. Assume here that each sector is allowed to make a memory cell error at the time of an erase operation so that the operation is carried out as normal erase operation. Then, if any of the sectors has a single error, the control circuit 11 sends status information of “pass” back to the flash memory control unit 8.
The control circuit 11 writes data in sectors and check the written state of each of the sectors (including an optical of VERIFY/WRITE). If the control circuit 11 determines that there are m (2≦m≦n+1) or more than m errors, it concludes that the normal write operation has not been carried out successfully for the sector and sends a “fail” status message back to the flash memory control unit 8.
For example, if n=2 and m=3, the system can correct up to three bit errors for each sector. Assume here that each sector is allowed to make a memory cell error at the time of an write operation so that the operation is carried out as normal write operation. Then, if any of the sectors has a single error, the control circuit 11 sends status information of “pass” back to the flash memory control unit 8.
Each sector comprises a total of 2,112 memory cells connected to a single word line and stores 4,224 bits data in a manner as illustrated in FIG. 7. Each cluster comprises fur four sectors sandwiched by a pair of selection transistors. A write or read operation is conducted on a sector by sector basis and an erase operation is conducted on a cluster by cluster basis.
A number of NAND type memory cells M are connected in series. If one of the memory cells connected in series gives rise to an error, all the memory cells M connected in series are deemed to be defective. For example, if four memory cells are connected in series and one of them becomes defective and no longer electrically conductive, all the remaining three memory cells are also deemed to be no longer electrically conductive.
With such an arrangement of memory cells, the control circuit 11 of
In view of this fact, it may be appreciated that a cluster (or a sector) does not necessarily have to be disabled if it gives rise to an error and may often operate properly as normal cluster (or sector). Therefore, excessive occurrences of defective sectors and those of defective clusters can be suppressed for NAND type cells by applying the method of the present invention. As a result, any unnecessary increase of disabled memory areas in a memory system of the type under consideration can be suppressed to consequently reduce the rate of occurrence of faulty systems.
Firstly, the algorithm of the operation of the memory system in a read mode will be described by referring to FIG. 9. Upon receiving a read request from the host computer, the system starts a read operation (Step S1). Then, the flash memory control unit 8 issues a read instruction to the flash memory 9 (Step S2). Subsequently, as the physical address of the sector from which a data is be read is specified (Step S3), the control circuit 11 outputs the data from the corresponding cell array 10 to the flash memory control unit 8 so that the flash memory control unit 8 acquires the data (Step S4).
The flash memory control unit 8 transfers the data to error detection/correction unit 7, which error detection/correction unit 7 by turn detects the number of errors in the sector (Step S5). The number of errors is expressed in terms of number of bits if the error correction method uses a bit as unit and in terms of number of symbols if the error correction method uses a symbol as unit.
The existence or non-existence of an error is detected (Step S6) and, if it is determined that there is no error (Step S7) as a result of the detection, the data read out from the flash memory 9 is directly output to the host computer except the check code to terminate the read operation (Step S11).
If it is determined that there is at least an error, then it is determined if the number of errors is equal to or less than the number of correctable errors (3 in this embodiment) (Step S8) and if the number of errors is equal to or less than the number of correctable errors, the errors are corrected (Step S9). As a result, it is determined that the read operation is carried out successfully (Step S7) and the corrected data is output to the host computer except the check code to terminate the read operation (Step S11).
If the number of errors is more than the number of correctable errors, it is determined that the read operation is unsuccessful (Step S10) and the host computer is notified of the unsuccessful result to terminate the read operation (Step S11).
Now, the algorithm of the operation of the memory system in an erase mode will be described by referring to FIG. 10. Upon receiving an erase instruction from the host computer or when an erase instruction is issued from the MPU 2 prior to a data writing operation of the host computer, an erase operation starts (Step S12). Then, the flash memory control unit 8 issues an erase command to the flash memory 9 (Step S13). Subsequently, as the physical address of the cluster from which the data is erased is specified (Step S14), the control circuit 11 erases the data of the cluster having the specified physical address from the cell array 10 (an operation including VERIFY/ERASE).
The control circuit 11 checks if the erase operation is carried out successfully or not and prepares for outputting status information for the erase operation. Then, the flash memory control unit 8 reads out the status information for the erase operation from the flash memory 9 (Step S15).
The control circuit 11 determines if the status information is that of “pass” or not (Step S16). If the status information is that of “pass”, it determines that the erase operation is carried out successfully (Step S17) and terminates the erase operation (Step S21). If the status information is not that of “pass”, the flash memory control unit 8 reads out the data of the memory cells from the flash memory 9 (Step S18).
Then, the flash memory control unit 8 counts the number of errors by means of the counter provided for this purpose, using the data read out in Step S18. Then, it determines if the counted number is greater than 1 or not (Step S19) and if the counted number is not greater than 1, it determines that the erase operation is successfully carried out (Step S17) and terminates the erase operation (Step S21).
The error judgement section 81 of
Now, the algorithm of the operation of the memory system in a write mode will be described by referring to FIG. 11. Upon receiving a write instruction from the host computer, a write operation starts (Step S22). Then, the error detection/correction unit 7 generates a check code from the data input by the host computer (Step S23). The flash memory control unit 8 issues a write command to the flash memory 9 (Step S24). Subsequently, the physical address of the sector where a data is written is specified (Step S25) and the data to be written, the check code and the physical address are input to the flash memory 9 (Step S26).
Then, the control circuit 11 writes the data onto the sector of the cell array 10 having the specified physical address. The control circuit 11 checks if the write operation is carried out successfully or not and prepares for outputting status information for the write operation. Then, the flash memory control unit 8 reads out the status information for the write operation from the flash memory 9 (Step S27).
The control circuit 11 determines if the status information is that of “pass” or not (Step S28). If the status information is that of “pass”, it determines that the write operation is carried out successfully (Step S29) and terminates the write operation (Step S33). If the status information is not that of “pass”, the flash memory control unit 8 reads out the data of the memory cells from the flash memory 9 (Step S30).
Then, the flash memory control unit 8 counts the number of errors by means of the counter provided for this purpose, using the data read out in Step S30. Then, it determines if the counted number is greater than 1 or not (Step S31) and if the counted number is not greater than 1, it determines that the write operation is successfully carried out (Step S29) and terminates the write operation (Step S33).
The error judgement section 81 of
In any of the above described embodiments of error correctable flash memory system according to the invention, thee are provided groups of memory units, each comprising a plurality of flash memory cells where data are erased or written, and errors can be detected and corrected out of the data read out from the groups of memory units if the number of errors is found below a predetermined level. When the data of the groups of memory units are modified and the data stored in the groups of memory units are checked for the state in which the data are stored, errors may be detected because the data are not stored in a desired state. However, if the number of errors is found below a predetermined level and hence can be detected and corrected, the system deems that the data are stored in the groups of memory units in a desired state.
Thus, in a flash memory system according to the invention, if the number of errors produced as a result of an erase or write operation is below a predetermined level and hence can be corrected, it deems that the erase or write operation is successful. As a result, any unnecessary increase of disabled memory areas in a flash memory system according to the invention can be suppressed to consequently reduce the rate of occurrence of faulty systems and improve the reliability.
In the embodiments of the invention, the functions of the error judgement section and the error detection/correction unit can be also realized by software. As well as the above-described methods, the functions of the error judgement section and the error detection/correction unit can be also realized, using a computer-readable recording medium that stores programs for enabling the computer to execute predetermined functions. The present invention is by no means limited to the above embodiments. The flash memory may not necessarily comprise NAND type memory cells. It may alternatively comprise NOR type memory cells, virtual grand memory cells or memory cells of some other type.
While each cluster comprises four sectors in the above description, it may alternative comprise eight, nine, sixteen or more sectors depending on the desired characteristics of the system. Still alternatively, a cluster may comprise only a sector.
When a memory cell is designed to store a 2k value (k≧2), it is possible to allow the presence of a total of k n-bit logical sectors in n memory cell groups. Thus, the present invention is feasible when memory sectors do not show a 1 to 1 correspondence relative to physical memory cell groups.
The present invention can be modified or altered in various different ways without departing from the scope of the invention.
Additional advantages and modifications will readily occur to those skilled in the art. Therefore, the invention in its broader aspects is not limited to the specific details and representative embodiments shown and described herein. Accordingly, various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and their equivalents.
| Patent | Priority | Assignee | Title |
| 7962838, | Sep 08 2009 | Kabushiki Kaisha Toshiba | Memory device with an ECC system |
| 8055957, | May 22 2007 | Renesas Electronics Corporation | Semiconductor integrated circuit device having fail-safe mode and memory control method |
| 8122320, | Jan 22 2008 | Polaris Innovations Limited | Integrated circuit including an ECC error counter |
| 8291303, | Aug 13 2007 | Kabushiki Kaisha Toshiba | Memory device with error correction system for detection and correction errors in read out data |
| 8327238, | Dec 22 2004 | SanDisk Technologies LLC | Erased sector detection mechanisms |
| 8468434, | Feb 12 2010 | TOSHIBA MEMORY CORPORATION | Error detection and correction system including a syndrome element calculating unit |
| Patent | Priority | Assignee | Title |
| 5361227, | Dec 19 1991 | Kabushiki Kaisha Toshiba | Non-volatile semiconductor memory device and memory system using the same |
| 5396468, | Mar 15 1991 | SanDisk Technologies LLC | Streamlined write operation for EEPROM system |
| 5504760, | Mar 15 1991 | SanDisk Technologies LLC | Mixed data encoding EEPROM system |
| 5611067, | Mar 31 1992 | Kabushiki Kaisha Toshiba | Nonvolatile semiconductor memory device having means for selective transfer of memory block contents and for chaining together unused memory blocks |
| 5822256, | Sep 06 1994 | Intel Corporation | Method and circuitry for usage of partially functional nonvolatile memory |
| JP10222995, | |||
| JP200030500, | |||
| JP6131884, | |||
| JP6131886, | |||
| JP676586, |
| Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
| Jan 20 2005 | Kabushiki Kaisha Toshiba | (assignment on the face of the patent) | / |
| Date | Maintenance Fee Events |
| Jul 22 2008 | M1552: Payment of Maintenance Fee, 8th Year, Large Entity. |
| Jul 11 2012 | M1553: Payment of Maintenance Fee, 12th Year, Large Entity. |
| Date | Maintenance Schedule |
| Apr 22 2011 | 4 years fee payment window open |
| Oct 22 2011 | 6 months grace period start (w surcharge) |
| Apr 22 2012 | patent expiry (for year 4) |
| Apr 22 2014 | 2 years to revive unintentionally abandoned end. (for year 4) |
| Apr 22 2015 | 8 years fee payment window open |
| Oct 22 2015 | 6 months grace period start (w surcharge) |
| Apr 22 2016 | patent expiry (for year 8) |
| Apr 22 2018 | 2 years to revive unintentionally abandoned end. (for year 8) |
| Apr 22 2019 | 12 years fee payment window open |
| Oct 22 2019 | 6 months grace period start (w surcharge) |
| Apr 22 2020 | patent expiry (for year 12) |
| Apr 22 2022 | 2 years to revive unintentionally abandoned end. (for year 12) |