TWI823913B - Semiconductor memory devices, memory systems, and methods of replacing semiconductor memory devices - Google Patents

Abstract

A method includes replacing an address of a first normal memory cell in a first column of a first memory block with a destination address that is an address of a second normal memory cell in a second column of the first memory block, and reassigning the address of the second normal memory cell in the second column of the first memory block to an address of a first redundancy memory cell in a redundancy block of the memory device.

Description

Semiconductor memory components, memory systems and replacement semiconductor memories component method [Cross-reference to related applications]

This application claims Korean Patent Application No. 10-2018-0036291 filed with the Korean Intellectual Property Office on March 29, 2018, and Korean Patent Application No. 10-2018-0036291 filed with the Korean Intellectual Property Office on July 9, 2018. No. 10-2018-0079345, filed with the Korean Intellectual Property Office on October 5, 2018. Korean Patent Application No. 10-2018-0119317, filed with the Korean Intellectual Property Office on January 30, 2019. Priority rights to Korean Patent Application No. 10-2019-0011563 and U.S. Serial Number 16/283,650 filed with the United States Patent and Trademark Office (USPTO) on February 22, 2019 Priority rights of patent applications, the disclosure contents of the Korean patent application and the U.S. patent application are fully incorporated into this case for reference.

The present disclosure relates to memories, and more specifically to semiconductor memory devices, methods of operating semiconductor memory devices, and memory systems.

Semiconductor wafers are manufactured through semiconductor manufacturing processes and then tested by test elements in a wafer state, a die state, or a package state. The defective portions of the defective wafer are selected through testing, and if some of the memory cells are defective, a repair operation is performed to save the semiconductor wafer. Currently, the size of semiconductor chips (eg, dynamic random access memory (DRAM)) continues to decrease through sophisticated manufacturing processes, and therefore, the likelihood of errors occurring during the manufacturing process increases. Additionally, if defects are not detected through the initial testing process, errors may occur during wafer operation.

One way to correct errors is to use redundant memory blocks that include arrays of memory cells that serve as backup cells for faulty cells in normal memory blocks. However, adding redundant memory blocks typically increases the overall size of the memory cell array and memory chip. Therefore, it would be advantageous to perform redundancy in such a way that the size of the memory chip increases less.

Therefore, the repair control circuit in the semiconductor memory element can use at least one normal cell in at least one memory block in the memory block to repair the faulty cell in the same memory block at least once, and can use the redundant area The redundant cells in the block replace the normal cells. Therefore, the semiconductor memory device can use redundant resources in the redundant block more efficiently.

According to exemplary embodiments (which may be described elsewhere in this summary) As part of one or more of the illustrated embodiments), a method replaces a memory cell in a first row of a memory block in a memory device, the memory device including a plurality of memory blocks and at least a first redundant block. The method includes replacing the address of the first normal memory cell in the first row of the first memory block with a destination address, the destination address being the address of the first normal memory cell in the second row of the first memory block. the address of a second normal memory cell; and reallocating the address of the second normal memory cell in the second row of the first memory block to the redundancy of the memory element The address of the first redundant memory cell in the block.

According to an exemplary embodiment, which may be part of one or more of the embodiments set forth elsewhere in this summary, a memory element includes a plurality of memory blocks, including a first memory block, each A memory block includes multiple rows of normal memory cells; at least one first redundant block includes multiple rows of redundant memory cells; and a repair control circuit. The repair control circuit is configured such that the second row of normal memory cells of the first memory block serves as the destination row of the first row of normal memory cells of the first memory block and is It is configured so that the first row of redundant memory cells of the redundant memory block stores data destined for the second row of normal memory cells of the first memory block.

According to an exemplary embodiment, which may be part of one or more of the embodiments set forth elsewhere in this summary, a memory element includes a plurality of memory blocks, including a first memory block, each A memory block includes a plurality of rows of normal memory cells; at least a first redundant block including a plurality of rows of redundant memory cells; and a repair control circuit. The repair control circuit is configured to replacing a first normal memory cell in a first row of the first memory block with a second normal memory cell in a second row of the first memory block and being configured to use the first The first redundant memory cells in the first row of redundant memory cells of the redundant block replace the second normal memory cells in the second row of the first memory block. Using the first redundant memory cell to replace the second normal memory cell may include causing the first redundant memory cell to store data destined for the second normal memory cell.

According to an exemplary embodiment, which may be part of one or more of the embodiments set forth elsewhere in this summary, a memory element includes a plurality of memory blocks, including a first memory block, each A memory block includes multiple rows of normal memory cells; at least one first redundant block includes multiple rows of redundant memory cells; and a repair control circuit. the repair control circuit is configured to replace the first normal memory cell in the first row of the first memory block with a second normal memory cell in the second row of the first memory block, and configured to replace the first redundant memory cell in the first row of redundant memory cells of the first redundant block with the first redundant memory cell in the second row of the first redundant block. Two normal memory cells. Replacing the second normal memory cell with the first redundant memory cell may include replacing the first normal memory cell with a third normal memory cell in a first row of a second memory block of the memory element. Two normal memory cells; and using the first redundant memory cell to replace the second normal memory cell and other normal memory cells of the second memory block or other memory blocks of the memory element Normal memory cells in .

According to an exemplary embodiment, which may be part of one or more of the embodiments set forth elsewhere in this summary, a memory element includes a plurality of memory blocks, including a first memory block, each A memory block includes multiple rows of normal memory cells; at least one first redundant block includes multiple rows of redundant memory cells; and a repair control circuit. the repair control circuit is configured to replace the first normal memory cell in the first row of the first memory block with a second normal memory cell in the second row of the first memory block, and configured to replace the first redundant memory cell in the first row of redundant memory cells of the first redundant block with the first redundant memory cell in the second row of the first memory block. Two normal memory cells. Replacing the second normal memory cell with the first redundant memory cell may include replacing the first normal memory cell with a third normal memory cell in a first row of a second memory block of the memory element. Two normal memory cells; and using the first redundant memory cell to replace the second normal memory cell and other normal memory cells of the second memory block or other memory blocks of the memory element Normal memory cells in .

According to an exemplary embodiment, which may be part of one or more of the embodiments set forth elsewhere in this summary, a memory element includes: a plurality of normal memory blocks, including a first memory block, Each normal memory block includes a plurality of rows of normal memory cells; at least one first redundant block, the first redundant block includes a plurality of rows of redundant memory cells; a plurality of row selection lines, For selecting the plurality of rows of normal memory cells and the plurality of rows of redundant memory cells, each row selection line is associated with a row address; and a repair control circuit. The repair control circuit is configured to: replace a first source address with a first destination address, wherein the first source address is the address of a first row select line connected to the first memory The first row of memory cells of the block, and the first destination address is the address of the second row selection line, the second row selection line is connected to the memory cell of the first memory block a second row of cells; and replacing the address of the second row of selection lines with the address of the first row of selection lines, the first row of selection lines being connected to the first row of the redundant block.

20:Memory system

100:Memory controller

200, 200a, 200b, 600: semiconductor memory components

205: Access control circuit

210, 210a: Control logic circuit

211:Command decoder

212: Mode register

220: Address register

230: Memory bank control logic

240: Column address multiplexer

245: Renew the counter

250: Row address latch

251:Address buffer

260, 261: column decoder

260a~260h: The first memory bank decoder ~ the eighth memory bank decoder

270: Line decoder

270a: Row decoder/first memory bank row decoder

270b~270h: Second memory bank row decoder ~ Eighth memory bank row decoder

280: Error Correction Code (ECC) Engine

285: Sense amplifier unit

285a~285h: First memory bank sense amplifier~Eighth memory bank sense amplifier

290: Input/output gate control circuit/gate control circuit

290a: Input/output (I/O) gate control circuit

291a, 291b, 291c, 291d, 291e, 292a, 292b, 292c, 292d: input/output circuit

293a, 293b, 293c, 293d, 293e, 296a, 296b, 296c, 296d: row select circuit

294a, 294b, 294c, 294d, 297a~297h: row selection transistor

295, 296: Data input/output buffer

300, 301: Memory cell array

310~380: The first memory bank array~the eighth memory bank array

310b, 310c: first memory bank array

311, 312, 313, MB, MB4, MB5, MB6, MB7~MB15: memory block

314: Redundant block/memory block

400, 400a, 400b, 614: Repair control circuit

401a: Unit Repair Controller/First Unit Repair Controller

401b, 401c, 401d, 402a, 402b, 402c, 501, 502, 503, 504: unit repair controller

401e: Redundant Repair Controller/Unit Repair Controller

402d: Redundant Repair Controller

405, 460: Table pointer

410: Fault address storage circuit/fault address storage table

411:Antifuse Array

412:Control unit

413, 425, 425b: sensing unit

414: Temporary register unit

415: Column address comparator

420, 420b: Address storage table

421, 421b: first storage unit

422: Antifuse

423, 423b: Second storage unit

426:NMOS transistor

430, 510, 515, 516, 518: Row address comparator

435:And gate

440: Select circuit

450, 550, 555, 556, 557: row select line driver

451, 452, 453, 454: drive transistor

455, 456, 517, 519: inverter

457: Anti-and gate

470: Redundant row select line driver

480, 480b, 480c: fuse circuit

500: Timing control circuit

511, 511a, 512, 512a, 513, 513a, 514, 514a, 515, 515a, 521, 531, 532, 533, 534: Reference number

520:Normal decoder

521a: Repair status/reference number

522:CSL0 operation

523:CSL1 operation

524:CSL2 operation

525:CSL3 operation/normal decoder

526:CSL0 operation/decoder

527:CSL1 operation/decoder

528:CSL3 operation

530: Destination decoder

540, 545, 546, 548: multiplexer

541:Repair status

542:CSL0 operation

543:CSL3 operation

544:CSL7 operation

610: First set of die/buffer die

612: Type 2 ECC engine

620: The second group of grains

620-1, 620-2~620-r-1, 620-p, 620-r, D11, D12, D13, D14: memory die

622: Type 1 Error Correction Code (ECC) Engine

623: Error injected into scratchpad group

632:TSV line group/data TSV line group

634: Co-located TSV line group

700: Three-dimensional wafer structure

710:Printed circuit board

720:Host die

2911, 2912, 2913, 2914, 2915, 2916, 2917, 2918, 2919: selection circuit

4251: First sub-sensing unit

4252: Second sub-sensing unit

AB1, AB2: additional bits

ADDR: address/access address

ABL1, ABL2~ABLq, BTL, BTL1, BTL2, BTL3, BTL4~BTLn-1, BTLn~BTLv: bit lines

AF: Antifuse

AWL1, AWL2~AWLp, WL, WL1, WL2, WL3~WLm-1, WLm~WLu: character lines

B10: Data bus

BANK_ADDR: memory bank address

BL0, BL1, BL2, BL3: clump length

CA: target row address

CA7, CA8, CA9: bits

CADDR: row address/access row address

CADDR1, CADDR4, CADDR8: row address

CADDR_LSB: lower bit

CADDR_MSB: higher bits

CLK: clock signal

CMD: command

CSL, CSL0, CSL1, CSL2, CSL3, CSL4, CSL5, CSL6, CSL7, CSLa, CSLa1, CSLb, CSLa2, CSLc, CSLd, CSLf, CSLg: row select line signal

CSLe: row select line signal/redundant row select line signal

CTL1: first control signal

CTL2: second control signal

DCA, DCA’: decoded target row address

DCADDR: first decoded row address

DDSCA: Second decoded row address

DQ:data

DS: Destination information

DSCA: destination line address

FB: flip chip bump

FCAI: bank address information

FFI, FI_MB0, FI_MB1, FI_MB2, FI_RMB3, MB0_L, MB1_L, MB2_L: Fuse information

FRAI: column address information

GIO: data line

HIT1, HIT21, HIT22, HIT31, HIT32: hit signal

L1~Lp:TSV line

L10~Lq: TSV line/co-located TSV line

MB0: Memory block/first memory block

MB1: Memory block/second memory block

MB2: Memory block/third memory block

MB3: Memory block/fourth memory block

MC: memory cell

MCB: microbump

MFB: Main fuse information

MFB&DS: consolidated information

MTH1: first matching signal/matching signal

MTH2: second matching signal

NCA: normal cell array

NO1: The first node

NO2: The second node

PCSLD: De-energize the main signal

PCSLE: Master Signal of Empowerment

PCSLM: row select main signal

RA, RADDR: column address

RBTL, RBTL1, RBTL2, RBTL3, RBTL4~RBTLt~RBTLv: redundant bit lines

RCA: redundant cell array

RCADDR4, RCADDR8: redundant row address

REF_ADDR: Renew column address

RM: row matching signal

RMB, RMB2: redundant block

RMB3: redundant block/redundant memory block

RMC: redundant cell

S[9]S[8]S[7]: Pattern

S100, S200, S310, S330, S350: Operation

SR: source information

SRA: spare column address

SRCA: source row address

SRFI: Source Fuse Information

TPS: Table pointing signal/pointer signal

VDD: power supply voltage

VSS: ground voltage

Exemplary embodiments will be explained in more detail below with reference to the accompanying drawings.

FIG. 1 is a block diagram illustrating a memory system according to an exemplary embodiment.

2A is a block diagram illustrating an example of the semiconductor memory element shown in FIG. 1 according to an exemplary embodiment.

2B illustrates a portion of the semiconductor memory element shown in FIG. 2A according to an exemplary embodiment.

3 is a block diagram illustrating a portion of the semiconductor memory element shown in FIG. 2A according to an exemplary embodiment.

4A is a block diagram illustrating an example of a first cell repair controller in the semiconductor memory element shown in FIG. 3 according to an exemplary embodiment.

4B is a circuit diagram illustrating an example of a row selection line driver in the first cell repair controller shown in FIG. 4A.

FIG. 5 is a block diagram illustrating an example of a redundancy repair controller in the semiconductor memory element shown in FIG. 3 according to an exemplary embodiment.

FIG. 6A illustrates a repair operation performed in the semiconductor memory element shown in FIG. 3 .

FIG. 6B illustrates a repair operation performed in the semiconductor memory element shown in FIG. 2B.

FIG. 6C shows data input/output when the repair operation shown in FIG. 6A is performed.

FIG. 6D shows data input/output when the repair operation shown in FIG. 6B is performed.

FIG. 7 shows an example of the address storage table in the first unit redundancy repair controller shown in FIG. 4A.

FIG. 8 is a diagram showing an example of the address storage table shown in FIG. 7 .

9A to 9C are diagrams illustrating methods of replacing faulty cells with normal cells in the same memory block and replacing normal cells with redundant cells.

10 is a flowchart illustrating a method of operating a semiconductor memory element according to an exemplary embodiment.

11 is a block diagram illustrating another example of a semiconductor memory element in the memory system shown in FIG. 1 according to an exemplary embodiment.

FIG. 12 shows an example of a first bank array in the semiconductor memory device shown in FIG. 11 .

13 is a block diagram illustrating a repair control circuit included in each of the memory bank row decoders in the semiconductor memory element shown in FIG. 12 according to an exemplary embodiment.

FIG. 14 shows an example of a fault address storage table in the repair control circuit shown in FIG. 13 according to an exemplary embodiment.

FIG. 15 illustrates a portion of the semiconductor memory element shown in FIG. 11 according to an exemplary embodiment.

FIG. 16A illustrates a repair operation performed in the semiconductor memory element shown in FIG. 15 according to an exemplary embodiment.

FIG. 16B shows another example of the first memory bank array in the semiconductor memory device shown in FIG. 15 .

FIG. 16C shows another example of the first memory bank array in the semiconductor memory device shown in FIG. 15 .

FIG. 17A shows an example of an address storage table in the repair control circuit shown in FIG. 13 according to an exemplary embodiment.

Figure 17B illustrates an example of reducing the number of fuses associated with a repair operation, according to an exemplary embodiment.

Figure 17C illustrates another example of reducing the number of fuses associated with a repair operation, according to an exemplary embodiment.

Figure 17D illustrates in detail an example of implementing the example shown in Figure 17C.

Figure 17E illustrates an example of a unit repair controller according to an exemplary embodiment.

Figure 17F illustrates another example of a unit repair controller according to an exemplary embodiment.

Figure 17G illustrates another example of a unit repair controller according to an exemplary embodiment.

Figure 17H illustrates another example of a unit repair controller according to an exemplary embodiment.

18 is a flowchart illustrating a method of operating a semiconductor memory element according to an exemplary embodiment.

19 is a block diagram showing a semiconductor memory element according to an exemplary embodiment.

20 is a cross-sectional view of a three-dimensional (3D) wafer structure employing the semiconductor memory element shown in FIG. 19 according to an exemplary embodiment.

Various exemplary embodiments will be explained more fully hereinafter with reference to the accompanying drawings, in which exemplary embodiments are shown.

As seen in the various claims and in the specification, the specific items set forth herein are set forth using the naming convention of "first," "second," "third," etc. . Unless context dictates otherwise, these terms are only used to distinguish different items from each other and do not necessarily indicate the physical location or order of operations of the items. Therefore, in the description or in different claims, different naming terms (for example, "first", "second", etc.) may be used to refer to specific items depending on the context of the discussion. or objects.

FIG. 1 is a block diagram illustrating a memory system according to an exemplary embodiment.

Referring to FIG. 1 , the memory system 20 may include a memory controller 100 and a semiconductor memory device 200 .

Memory controller 100 may control the overall operation of memory system 20 . The memory controller 100 can control the overall data exchange between the external host and the semiconductor memory device 200 . For example, the memory controller 100 may write data into the semiconductor memory device 200 or read data from the semiconductor memory device 200 in response to a request from the host. In addition, the memory controller 100 can issue operation commands to the semiconductor memory device 200 to control the semiconductor memory device 200 .

In some embodiments, the semiconductor memory device 200 includes, for example, the following Memory components of dynamic memory cells: dynamic random access memory (DRAM), 4th generation double data rate 4 (DDR4) synchronous dynamic random access memory (synchronous DRAM, SDRAM), low-power third generation Fourth generation double data rate (LPDDR4, low power DDR4) synchronous dynamic random access memory or low power fifth generation double data rate synchronous dynamic random access memory (LPDDR5 SDRAM).

The memory controller 100 transmits the clock signal CLK, the command CMD and the address (signal) ADDR to the semiconductor memory device 200, and exchanges data DQ with the semiconductor memory device 200.

The semiconductor memory device 200 includes a memory cell array (MCA) 300 for storing data DQ, a control logic circuit 210 and a repair control circuit 400. The memory cell array 300 may include a plurality of memory blocks and at least one redundancy block.

The control logic circuit 210 controls access to the memory cell array 300 based on the command CMD and the address ADDR, and the repair control circuit 400 can use normal cells in the first memory block among the plurality of memory blocks to repair Therefore, the faulty cell in the first memory block is replaced, and the first normal cell can be replaced with the redundant cell in the redundant block. Therefore, the repair control circuit 400 can use redundant resources in the redundant blocks more efficiently.

For example, the repair control circuit 400 may replace memory cells in a memory block with normal cells in the same memory block at least once and then may replace the normal cells with redundant cells. Therefore, the first memory cell to be replaced can be Faulty cells can be repaired using normal cells, and the normal cells can be replaced with redundant cells, or can be replaced with other normal cells, thus resulting in a final normal cell in a series of replacement normal cells. Cells are replaced with redundant cells instead of simply repairing faulty cells using only redundant cells. Therefore, the repair control circuit 400 can use redundant resources in the redundant blocks more efficiently.

2A is a block diagram illustrating an example of the semiconductor memory element shown in FIG. 1 according to an exemplary embodiment.

Referring to FIG. 2A , the semiconductor memory device 200a may include a control logic circuit 210a, an address buffer 251, a repair control circuit 400a, a column decoder 261, an input/output (I/O) gate control circuit 290a, and a data input. /Output buffer 296 and memory cell array 301.

The control logic circuit 210a receives the command CMD and the access address ADDR. The control logic circuit 210a can control the operation of the semiconductor memory device 200a based on the command CMD and the access address ADDR. The control logic circuit 210a may control the column decoder 261, the input/output gating circuit 290a and the repair control circuit 400a based on the command CMD and the address ADDR.

The address buffer 251 receives the access address ADDR, provides the column address RADDR of the access address ADDR to the column decoder 261, and provides the row address CADDR of the access address ADDR to the repair control circuit 400a. The repair control circuit 400a can use normal cells in a memory block to repair faulty cells in the same memory block at least once, and can be based on a comparison of the row address CADDR and the faulty row address stored in the row address CADDR. to replace the normal cells with redundant cells.

Column decoder 261 is coupled to memory cell array 301 via word line WL and input/output gating circuit 290a is coupled to memory cell array 301 via bit line BTL. The data input/output buffer 296 utilizes the memory controller 100 to receive/transmit data DQ via the input/output gating circuit 290a.

2B illustrates a portion of the semiconductor memory element shown in FIG. 2A according to an exemplary embodiment.

In FIG. 2B, a memory cell array 301, an input/output gate control circuit 290a, a data input/output buffer 296 and a repair control circuit 400a are shown.

Referring to FIG. 2B , the memory cell array 301 includes a normal cell array NCA and a redundant cell array RCA. The normal cell array NCA includes a plurality of memory blocks MB0, MB1, MB2 and MB3, and the redundant cell array RCA includes at least A redundant block RMB. The normal cell array NCA includes memory cells MC coupled to the word line WL and the bit line BTL, and the redundant block RMB includes the redundant cell RMC coupled to the word line WL and the redundant bit line RBTL.

The input/output gating circuit 290a includes a plurality of input/output circuits 291a, 291b, 291c, 291d and 291e and a plurality of row selection circuits 293a to 293e, and the row selection circuits 293a to 293e can select the input/output circuits 291a to 291e. One is connected to a corresponding one of the memory blocks MB0 to MB3 and the redundant block RMB. Each of the row selection circuits 293a to 293e may include a plurality of row selection transistors 294a to 294d, and the plurality of row selection transistors 294a to 294d respectively switch corresponding memories in response to the row selection line signals CSLa to CSLe. Multiple bit lines or one bit line in the block or redundant block RMB are connected to the corresponding input/output circuit. The plurality of input/output circuits 291a to 291e may be coupled to the data input/output buffer 296 through the data line GIO in response to the first control signal CTL1 from the control logic circuit 210a.

Although not shown in the figure, the row selection line signal CSLb may be applied to the row selection circuit 293b, the row selection line signal CSLc may be applied to the row selection circuit 293c, and the row selection line signal CSLd may be applied to the row selection circuit 293d. And the row selection line signal CSLe may be applied to the row selection circuit 293e.

The repair control circuit 400a may provide a corresponding one of the row selection line signals CSLa to CSLe to a corresponding one of the row selection circuits 293a to 293e in response to accessing the row address CADDR. The access row address CADDR is provided from the address buffer 251 shown in FIG. 2A, and the access row address CADDR specifies a bit without considering at least one faulty cell in the memory blocks MB0 to MB3. The address of the yuan line. Each of the row select line signals CSLa to CSLd is a signal that simultaneously selects a corresponding bit line in the memory blocks MB0 to MB3 based on accessing the row address CADDR. The semiconductor memory device 200a can simultaneously input and output data having a size corresponding to a burst length in response to each of the row select line signals CSLa to CSLd.

3 is a block diagram illustrating a portion of the semiconductor memory element shown in FIG. 2A according to an exemplary embodiment.

In FIG. 3, a memory cell array 301, an input/output gate control circuit 290a, a repair control circuit 400a and a data input/output buffer 296 are shown. When comparing Figure 3 with Figure 2B, the repair control circuit 400a is shown in detail, and will mainly The complex control circuit 400a will be described below.

Referring to FIG. 3 , the memory cell array 301 includes a normal cell array NCA and a redundant cell array RCA. The normal cell array NCA includes the plurality of memory blocks MBO, MB1, MB2 and MB3, and the redundant cell array RCA includes at least one redundant block RMB. The normal cell array NCA includes memory cells coupled to word lines WL and bit lines, and the redundant block RMB includes redundant cells coupled to word lines and redundant bit lines.

The repair control circuit 400a may include a plurality of unit repair controllers 401a to 401d and a redundancy repair controller 401e, and the unit repair controllers 401a to 401d and the redundancy repair controller 401e correspond to the memory blocks MB0, MB1, MB2 and MB3. And redundant block RMB. The repair control circuit 400a may be included, for example, in a row decoder in the semiconductor memory device 200a.

The plurality of input/output circuits 291a, 291b, 291c, 291d and 291e can control the memory blocks MB0, MB1, MB2 and MB3, the redundant block RMB and the data input/output buffer in response to the first control signal CTL1 296 connections. The row selection circuits 293a to 293e may connect one of the input/output circuits 291a to 291e to a corresponding one of the memory blocks MB0 to MB3 and the redundancy block RMB.

The unit repair controllers 401a to 401d and the redundancy repair controller 401e can jointly receive the access row address (ie, row address) CADDR, and can provide the corresponding row selection circuits 293a to 293e with information applied to the memory blocks MBO, MB0, The row select line signals CSLa to CSLe of MB1, MB2 and MB3 and the redundant block RMB. For example In other words, the same access row address may be transmitted to each of the unit repair controllers 401a to 401d and the redundancy repair controller 401e at a given time. Based on the status of each unit repair controller 401a to 401d and redundancy repair controller 401e and the information stored in each unit repair controller 401a to 401d and redundancy repair controller 401e, output is sent to select a row Each corresponding row selection circuit 293a to 293e performs memory access.

When the memory blocks MB0, MB1, MB2 and MB3 do not include faulty cells, the input/output circuit 291e cuts off the connection between the redundant block RMB and the data input/output buffer 296 in response to the first control signal CTL1 , and the input/output circuits 291a to 291d can transmit the data DQ from the memory blocks MB0, MB1, MB2 and MB3 to the data input/output buffer 296 through the row selection transistors 294a to 294d in response to the first control signal CTL1 Alternatively, data DQ may be transferred from data input/output buffer 296 to memory blocks MB0, MB1, MB2 and MB3. In this case, in each of the memory blocks MB0, MB1, MB2, and MB3, one or more bit lines located at the same location (eg, the same relative position within each memory block) The selection can be made in response to a corresponding one of the row selection line signals CSLa to CSLd, and the semiconductor memory element 200a can simultaneously input and output data having a size corresponding to the burst length of the semiconductor memory element 200a.

When at least one of the memory blocks MB0, MB1, MB2 and MB3 includes at least one faulty cell, the input/output circuit 291e is connected to the redundant block RMB in response to the first control signal CTL1, and can At least one faulty cell performs a repair operation.

For example, when each of the memory blocks MB0, MB2 and MB3 includes at least one faulty cell on the first cell line that would normally be selected by the row select line signal CSL0, by as reference Number 511 indicates that the row selection line signal CSL3 is enabled to replace the row selection line signal CSL0, and the faulty cells in the memory block MB0 are repaired by the normal cells in the memory block MB0. For example, by energizing the row selection line signal CSL3 in place of the row selection line signal CSL0, the row selection transistor 294c is connected to the memory block MB0 and the input/output circuit 291a instead of the row selection transistor 294a. As further explained below, the corresponding cell repair controller may enable the row select line signal CSL3 to replace the row select line signal CSL0. Additionally, normal cells in memory block MB0 are replaced by redundant cells in redundant block RMB by selecting redundant bit lines to replace bit lines in memory block MB0 as indicated by reference number 512 (For example, non-redundant cells that are not faulty cells in this example). For example, instead of energizing row select line signal CSL0 to select a second bit line coupled to a first faulty cell, a second bit line coupled to a first normal cell is selected by energizing row select line signal CSL3. With the first cell line, the repair control circuit 400a can use the first normal cell to repair the first faulty cell.

It is assumed that memory block MB1 does not contain faulty cells in any of the memory blocks, and therefore none of the rows of memory block MB1 need to be reallocated to or replaced by other rows.

By energizing the row select line signal CSL2 in place of the row select line signal CSL0 as indicated by reference number 513, the faulty cells in the memory block MB2 are repaired from the normal cells in the memory block MB2. Additionally, by reference numbers such as Instruction 514 selects a redundant bit line to replace the bit line in the memory block MB2, and replaces the normal cells in the memory block MB2 with the redundant cells in the redundant block RMB. By selecting redundant bit lines to replace the bit lines in memory block MB3 as indicated by reference number 515, the faulty cells in memory block MB3 are repaired by redundant cells in redundant block RMB.

4A is a block diagram illustrating an example of a first cell repair controller in the semiconductor memory element shown in FIG. 3 according to an exemplary embodiment.

Referring to FIG. 4A, the first unit repair controller 401a may include a table pointer (table pointer) 405, an address storage table 420, a row address comparator 430, a selection circuit (MUX) 440 and a column selection line (column selection). line, CSL) driver 450.

The table pointer 405 may generate a table pointing signal TPS that switches in response to the sequentially changing row address CADDR. The address storage table 420 may store at least one source row address SRCA and at least one destination row address DSCA corresponding to the at least one source row address SRCA as fuse information.

The row address comparator 430 compares the access row address CADDR with the source row address SRCA from the address storage table 420, and outputs a first match signal MTH1 indicating the comparison result. The selection circuit 440 may select one of the destination row address DSCA and the access row address CADDR from the address storage table 420 in response to the first matching signal MTH1 to output the selected one as the target row address CA. . The row select line driver 450 may output the row select line signal CSLa to select (enable) the bit line corresponding to the target row address CA.

When the access row address CADDR does not match the source row address SRCA, the selection circuit 440 may output the access row address CADDR as the target row address CA in response to the first matching signal MTH1 (eg, having a first value). . When the access row address CADDR matches the source row address SRCA, the selection circuit 440 may output the destination row address DSCA as the target row address CA in response to the first matching signal MTH1 (eg, having a second value). Therefore, when the row address of the bit line coupled to the at least one faulty cell in the memory block MBO is stored in the address storage table 420 as the source row address SRCA and as fuse information, and coupled to the pair When the row address of the bit line of the normal cell to be replaced by the faulty cell in the memory block MB0 is stored in the address storage table 420 as the destination row address DSCA and as fuse information, the row address in the memory block MB0 The faulty cells of are repaired using normal cells in the memory block MB0, and are therefore replaced with normal cells in the memory block MB0. In addition, the normal cells may be replaced by redundant cells in the redundant block RMB or may be reallocated to redundant cells in the redundant block RMB.

In an exemplary embodiment, the selection circuit 440 may be configured to perform an exclusive OR (XOR) operation on some higher bits of the access row address CADDR and the bits of the destination row address DSCA in response to the match signal MTH1 address translation circuit. For example, when the access row address CADDR includes six bits and the destination row address DSCA includes three bits, the address conversion circuit responds to the match signal MTH1 having a logic high level. The upper three bits of CADDR and the three bits of the destination row address DSCA perform a mutually exclusive OR operation to output the target row address.

Each of the unit repair controllers 401b, 401c and 401d shown in Figure 3 The configuration may be substantially the same as the configuration of the first unit repair controller 401a shown in FIG. 4A.

4B is a circuit diagram illustrating an example of a row selection line driver in the first cell repair controller shown in FIG. 4A.

Referring to FIG. 4B , the row select line driver 450 may include driving transistors 451 , 452 , 453 and 454 , inverters 455 and 456 and a NAND gate 457 .

The NAND gate 457 performs an NAND operation on the target row address CA and the enabling main signal PCSLE. The drive transistor 451 has a source coupled to the power supply voltage VDD, a gate receiving the output of the NAND gate 457 , and a drain coupled to the first node NO1 . The driving transistor 452 includes a drain coupled to the first node NO1 , a gate receiving the de-energizing main signal PCSLD, and a source coupled to the driving transistor 453 . The drive transistor 453 includes a drain coupled to the drive transistor 452 , a gate receiving the output of the NAND gate 457 , and a source coupled to the ground voltage VSS.

Inverter 455 inverts the logic level at the first node NO1 to provide its output at the second node NO2, and inverter 456 inverts the logic level at the second node NO2 to output the row select Line signal CSLa. The driving transistor 454 includes a drain coupled to the first node NO1, a gate coupled to the second node NO2, and a source coupled to the ground voltage VSS.

When a logic high level is applied to the target row address CA and a logic high level is applied to the enable master signal PCSLE, the output of the NAND gate 457 becomes a logic low level. Therefore, the driving transistor 451 is turned on, and the driving transistor 453 is turned off. Therefore, the first A node NO1 becomes a high level, the driving transistor 454 is turned off, and the inverter 456 outputs a row select line signal CSLa with a high level.

When a logic low level is applied to the target row address CA and a logic high level is applied to the enable master signal PCSLE, the output of the NAND gate 457 becomes a logic high level. Therefore, the driving transistor 451 is turned off, and the driving transistors 452 and 453 are turned on. Therefore, the inverter 456 outputs the row select line signal CSLa with a low level. The enable main signal PCSLE and the disable main signal PCSLD can be provided by a predecoder in the repair control circuit 400a or a row decoder including the repair control circuit 400a. The predecoder may control enablement by referencing fuse information associated with the failed cell and fuse information associated with row select line information stored in a fuse circuit included in redundancy repair controller 401e The logic levels of the main signal PCSLE and the disable main signal PCSLD. Therefore, by controlling the enabling main signal PCSLE and the deactivating main signal PCSLD, the repair control circuit 400a can select the first normal cell to replace the first faulty cell and can select the first redundant cell to replace the first normal cell. .

FIG. 5 is a block diagram illustrating an example of a redundancy repair controller in the semiconductor memory element shown in FIG. 3 according to an exemplary embodiment.

Referring to FIG. 5 , the redundancy repair controller 401 e includes a table pointer 460 , a fuse circuit 480 and a redundancy column selection line (SCL) driver 470 .

The table pointer 460 generates the table pointing signal TPS that switches in response to the sequentially changing row address CADDR. Fuse circuit 480 stores row select line information associated with each of the redundant bit lines in the redundant block RMB. Fuse circuit 480 storage When the row selection line signals CSL0, CSL2 and CSL3 are respectively enabled, the information of the repaired memory blocks MB3, MB2 and MB0 is obtained.

The redundant row select line driver 470 may output the redundant row select line signal CSLe in response to the table pointing signal TPS and select some of the redundant bit lines by referring to the row select line information in the fuse circuit 480 bit line.

Therefore, referring to FIGS. 3 to 5 , the repair control circuit 400a uses the first normal cell in the same memory block to repair the first faulty cell, and uses the first redundant cell in the redundant block to replace it. The first normal cell.

Figure 6A illustrates a repair operation performed in the semiconductor memory element shown in Figure 3, according to one embodiment.

In FIG. 6A , it is assumed that the repair status of the memory cell array 301 corresponds to the status indicated by reference number 521 . The repair status of the memory cell array 301 may be determined by taking into account the location of the faulty cell in each of the memory blocks MB0 to MB3. The repair status can be determined so that redundant resources for repairing faulty cells in memory blocks MB0, MB2, and MB3 or replacing normal cells in memory blocks MB0, MB2, and MB3 do not overlap, and can support semiconductor memory devices. 200a burst operation.

Referring to FIGS. 3 to 6A , when the CSL0 operation 522 in each of the memory blocks MB0 to MB3 is specified by the access row address CADDR, the first normal cell corresponding to CSL3 is selected in the memory block MB0 Instead of the faulty cell (for example, the row selection transistor 294d in the memory block MB0 is turned on instead of the row selection transistor 294a), the normal cell corresponding to CSL0 is selected in the memory block MB1. In block MB2, the normal cell corresponding to CSL2 is selected instead of the faulty cell, and in the memory block MB3, the redundant cell corresponding to CSL0 is selected instead of the faulty cell. During the CSL0 operation, a repair operation of repairing the faulty cell using the corresponding normal cell is performed in each of the memory blocks MB0 and MB2.

When the CSL1 operation 523 in each of the memory blocks MB0 to MB3 is specified by the access row address CADDR, the normal cell corresponding to CSL1 is selected in each of the memory blocks MB0 to MB3. When the CSL2 operation 524 in each of the memory blocks MB0 to MB3 is specified by the access row address CADDR, the normal cell corresponding to CSL2 is selected in each of the memory blocks MB0, MB1 and MB3. , and select the redundant cell corresponding to CSL2 in the redundant block RMB to replace the normal cell corresponding to CSL2 selected in the memory block MB2.

When the CSL3 operation 525 in each of the memory blocks MB0 to MB3 is specified by the access row address CADDR, the normal cell corresponding to CSL3 is selected in each of the memory blocks MB1, MB2 and MB3. , and select the redundant cell corresponding to CSL3 in the redundant block RMB to replace the normal cell corresponding to CSL3 selected in the memory block MB0. In this way and in conjunction with FIGS. 1 to 5 above, the repair control circuit is configured so that the second row of normal memory cells of the first memory block (for example, the row numbered 3 of MB0) is used as the first memory. The destination row of the first row of normal memory cells of the block (e.g., row number 0 of MB0), and makes the first row of redundant memory cells of the redundant memory block (e.g., row number 0 of RMB The row of 3) stores the data of the second row of the normal memory cell whose destination is the first memory block. The first row may be a row of a first memory block having at least one faulty memory cell. for For a read operation, the repair control circuit may be configured such that the second row of normal memory cells of the first memory block serves as the destination read row of the first row of normal memory cells of the first memory block. , and the first row of the redundant memory cells of the redundant memory block is used as the destination read row of the second row of the normal memory cells of the first memory block. For write operations, the repair control circuit reallocates data destined for the first failed row to the second row, and reallocates data destined for the second row to the redundant block. A similar process can occur in memory block MB2. In this way, the first memory cell in the first row of the first memory block can be repaired using the normal memory cell in the second row of the first memory block, and is therefore repaired by the first memory cell of the first memory block. Normal memory cell replacement in the second row.

As discussed above, the fuse circuit (480) may be part of the repair control circuit and may store corrections between redundant blocks and row select lines of memory blocks in the plurality of memory blocks. In addition, the address storage table can store the source address and the corresponding destination address for the repair control circuit, so that the second row of the normal memory cells of the first memory block is used as the normal memory cell of the first memory block. Used when the first row of elements is the destination row. When the first row of the redundant memory cells of the redundant memory block is stored as the data of the second row of the normal memory cells of the first memory block, the repair control circuit further uses the source address and the corresponding Destination address. As shown in the example shown in FIG. 6A and other examples, the first row of the redundant memory cells of the first redundant block (for example, the row numbered 3 in the RMB) may have the following values in the first redundant block: The second row of the first memory block (for example, the row numbered 3 of MB0) has the same relative position within the first memory block. Similarly, the redundancy of the first redundant block The second row of the remaining memory cells (e.g., the row numbered 2 of RMB) may have the same value as the second row of the second memory block (e.g., the row numbered 2 of MB2) within the first redundant block. The faulty cell (used to repair the faulty cell in row 0 of MB2) has the same relative position in the second memory block. The second row of the first memory block (e.g., row 3 of MB0) may have the same value as the second row of the second memory block (e.g., row 2 of MB2) within the first memory block. ) different relative positions.

FIG. 6B illustrates a repair operation performed in the semiconductor memory element shown in FIG. 2B.

Referring to FIG. 2B and FIG. 6B , the repair status 521a of the memory cell array 301 is as follows. Memory block MB0 includes faulty cells associated with row select line signal CSL1, and thus with the first row of memory block MB0, by pairing with the second row of memory block MB0 as indicated by reference number 511a. The associated row selection line signal CSL3 is enabled to replace the row selection line signal CSL1, and the first normal cell in the memory block MB0 is used to repair the faulty cell. As indicated by the reference number 512a, the first normal cell in the memory block MB0 is replaced with a second normal cell in the memory block MB1 that is different from MBO, and as indicated by the reference numbers 513a, 514a and 515a, using The first redundant cell in the redundant block RMB replaces the second normal cell in the memory block MB1. The repair control circuit 400a uses the first normal cell in the first memory block MB0 to repair the first faulty cell in the first memory block MB0, and uses the second memory block MB1 that is different from the first memory block MB0. to replace the first normal cell with the second normal cell in the second memory block MB1, to replace the second normal cell with the third normal cell in the third memory block MB2 that is different from the second memory block MB1, and with The fourth normal cell in the third memory block MB2 is different from the fourth memory block MB3 to replace the third normal cell, and the first redundant cell in the redundant block RMB is used to replace the fourth normal cell. Yuan. The first normal cell, the second normal cell, the third normal cell, the fourth normal cell and the first redundant cell may have the same row selection line address. In this way, the first normal memory cell (e.g., located in the row numbered 3 in the first memory block MB0) uses the first redundant memory cell (e.g., located in the row numbered 3 in the redundant block RMB). in the row of 3) is replaced by a shift operation. For example, the shift operation may include replacing the second normal memory cell with the third normal memory cell in the first row of the second memory block of the memory element (eg, located in the row numbered 3 of the memory block MB1). normal cells (e.g., located in row number 3 of memory block MB0), and then replace the final memory area with a row of memory cells from an adjacent memory block in the plurality of memory blocks A row of memory cells of each memory block outside the block, and a row in the final memory block of the plurality of memory blocks is replaced with a first row of redundant memory cells of the redundant block.

When the CSL0 operation 526 in each of the memory blocks MB0 to MB3 and the redundant block RMB is specified by the access row address CADDR, the selection corresponding to CSL0 in each of the memory blocks MB0 to MB3 is of normal cells. When the CSL1 operation 527 in each of the memory blocks MB0 to MB3 and the redundant block RMB is specified by the access row address CADDR, the first normal cell corresponding to CSL3 is selected in the memory block MB0 to The first faulty cell corresponding to CSL1 is replaced, and a normal cell corresponding to CSL1 is selected in each of the memory blocks MB1 to MB3.

When each of the memory blocks MB0 to MB3 and the redundant block RMB When the CSL3 operation 528 in is specified by the access row address CADDR, select the second normal cell corresponding to CSL3 in the memory block MB1 to replace the first normal cell corresponding to CSL3 in the memory block MB0, and select the memory The third normal cell corresponding to CSL3 in block MB2 is selected to replace the second normal cell corresponding to CSL3 in memory block MB1, and the fourth normal cell corresponding to CSL3 in memory block MB3 is selected to replace the memory block. The third normal cell corresponding to CSL3 in MB2 is selected, and the first redundant cell corresponding to CSL3 is selected in the redundant block RMB to replace the fourth normal cell corresponding to CSL3 in the memory block MB3. Therefore, the semiconductor memory device 200a can use the redundant resources in the redundant block RMB to support data input/output having a size corresponding to the burst length.

FIG. 6C shows data input/output when the repair operation shown in FIG. 6A is performed.

Referring to FIG. 6C, when the repair status of the memory cell array 301 corresponds to the status indicated by the reference number 521, the row selection circuits 293a to 293e and the selection circuits 2911 to 2915 in the input/output gating circuit 290a operate in the memory area. The data selected in the blocks MB0 to MB3 and the redundant block RMB are provided to the data input/output buffer 296 in units of burst lengths BL0 to BL3. That is, the selection circuit 2911 can select the output of one of the row selection circuits 293a and 293e, the selection circuit 2912 can select the output of one of the row selection circuits 293b and 293e, and the selection circuit 2913 can select the row selection circuits 293c and 293e. The output of one of the row selection circuits 293d and 293e may be selected by the selection circuit 2914.

In CSL0 operation, in response to the signal {0,0, 0,1}, the selection circuit 2911 can select the data output from the memory block MB0, the selection circuit 2912 can select the data output from the memory block MB1, the selection circuit 2913 can select the data output from the memory block MB2, and the selection circuit 2914 can select the data output from the redundant block RMB.

FIG. 6D shows data input/output when the repair operation shown in FIG. 6B is performed.

6D, when the repair status of the memory cell array 301 corresponds to the status indicated by the reference number 521a, the row selection circuits 293a to 293e and the selection circuits 2916 to 2919 and 2915 in the input/output gating circuit 290a are configured. The data selected in the memory blocks MB0 to MB3 and the redundant block RMB are provided to the data input/output buffer 296 in units of burst lengths BL0 to BL3. That is, the selection circuit 2916 can select the output of one of the adjacent row selection circuits 293a and 293b, the selection circuit 2917 can select the output of one of the adjacent row selection circuits 293b and 293c, and the selection circuit 2918 can select The output of one of adjacent row selection circuits 293c and 293d, and selection circuit 2919 may select the output of one of adjacent row selection circuits 293d and 293e.

In CSL4 operation, in response to the signal {1,1,1,1} from the selection circuit 2915, the selection circuit 2916 can select the data output from the memory block MB1, and the selection circuit 2917 can select the data output from the memory block MB2. , the selection circuit 2918 can select the data output from the memory block MB3, and the selection circuit 2919 can select the data output from the redundant block RMB.

Figure 7 shows addresses in the first unit redundancy repair controller shown in Figure 4A Instance of storage table.

Referring to FIG. 7 , the address storage table 420 includes a first storage unit 421 , a second storage unit 423 and a sensing unit 425 . The first storage unit 421 (which may be a circuit) stores the source row address SRCA to be repaired, and the second storage unit 423 (which may be a circuit) stores the destination row address DSCA that replaces the source row address SRCA. The address storage table 420 may be implemented as an antifuse array or a content addressable memory (CAM). The sensing unit 425 outputs the source row address SRCA and the destination row address DSCA stored in the positions (indicated by the pointer signal TPS) in the first storage unit 421 and the second storage unit 423 in response to the pointer signal TPS. In FIG. 7 , the address storage table 420 stores the row address CADDR1 associated with CSL0 and the row address CADDR4 associated with CSL3 as the source row address SRCA, and stores the row address CADDR4 that replaces the row address CADDR1 and The redundant row address RCADDR4 replaces the row address CADDR4 as the destination row address DSCA. In this way, the address of the first normal memory cell (CADDR1) in the first row of the first memory block is replaced with the address of the second normal memory cell in the second row of the first memory block. The destination address of the address (CADDR4), and the address (CADDR4) of the second normal memory cell in the second row of the first memory block is replaced and reallocated to the redundant block of the memory element. The address of the first redundant memory cell (RCADDR4). It should be noted that in some embodiments, particularly those in which the redundant block RMB has the same number of memory cell rows as the memory blocks MB0 to MB3, the address storage table does not need to store the second normal memory cell to Reallocation of the first redundant memory cell. This reassignment may be accomplished using a fuse circuit (such as fuse circuit 480 shown in Figure 5).

FIG. 8 is a diagram illustrating an example of the address storage table shown in FIG. 7 according to an exemplary embodiment.

Referring to FIG. 8 , the address storage table 420 may be implemented by an antifuse array including a plurality of antifuses 422 . The electrical characteristics of antifuse 422 are opposite to those of the fuse assembly. Antifuse 422 is a resistive fuse component that has a relatively high resistance value when it is not programmed and a relatively low resistance value when it is programmed. The address storage table 420 may store the source row address SRCA and the destination row address DSCA by selectively programming the antifuse 422.

The sensing unit 425 includes a first sub-sensing unit 4251 and a second sub-sensing unit 4252 coupled to the first storage unit 421 and the second storage unit 423 respectively. Each of the first sub-sensing unit 4251 and the second sub-sensing unit 4252 may be implemented using an n-type metal oxide semiconductor (n-metal oxide semiconductor, NMOS) transistor 426. Therefore, the sensing unit 425 provides the source row address SRCA to the row address comparator 430 and provides the destination row address DSCA to the selection circuit 440 in response to the pointer signal TPS.

9A to 9C are diagrams for explaining a method of replacing faulty cells with normal cells in the same memory block and using redundant cells to replace normal cells.

In FIGS. 9A to 9C , the memory block MBO includes memory cells coupled to the word lines WL1 to WLu and the bit lines BTL1 to BTLv, and the redundant block RMB includes memory cells coupled to the word lines WL1 to WLu and redundant lines. Redundant cells of the remaining bit lines RBTL1 to RBTLv. In some embodiments, the redundant bit lines in the redundant block RMB The number is the same as the number of bit lines in each of the normal memory blocks (eg MB0). However, in other embodiments, the number of redundant bit lines in the redundant block RMB may be smaller or larger than the number of bit lines in each of the normal memory blocks.

FIG. 9A is a diagram illustrating replacement between bit lines. For example, when a fault occurs in a memory cell coupled to word line WL1 and bit line BTL1, bit line BTL1 is replaced with bit line BTL4 and bit line BTL4 can be replaced with redundancy. Bit line RBTL4.

9B is a diagram illustrating substitution between portions of a bit line (eg, segments of a bit line). A single bit line may be divided into two or more segments, with each segment connected to at least one memory cell. For example, when a fault occurs in a memory cell coupled to word line WL1 and bit line BTL1, the segment in bit line BTL1 is replaced with a segment in bit line BTL4 and the segment in bit line BTL4 The sectors of are replaced with sectors in redundant bit line RBTL4.

FIG. 9C is a diagram illustrating replacement between memory cells. For example, when a fault occurs in a memory cell coupled to word line WL1 and bit line BTL1, the memory cell with the fault is replaced with a memory cell coupled to bit line BTL4 and coupled to the bit line The memory cells of BTL4 are replaced with redundant cells coupled to redundant bit line RBTL4.

10 is a flowchart illustrating a method of operating a semiconductor memory element according to an exemplary embodiment.

Referring to FIGS. 2A to 10 , in a method of operating a semiconductor memory device including a memory cell array including a plurality of memory blocks and at least one redundant block, the modification The complex control circuit 400a uses the first normal cell in the first memory block among the plurality of memory blocks to repair the first faulty cell in the first memory block (S100). Before using the first normal cell in the first memory block to repair the first faulty cell in the first memory block, the repair control circuit 400a may determine the access row address and the specified number coupled to the first faulty cell. Whether the address of the first row of the first element line matches. When the access row address matches the first row address (source row address), the repair control circuit 400a performs a repair operation.

The first faulty cell in the first memory block and the first normal cell in the first memory block may have different row selection line addresses. For example, the first faulty cell and the first normal cell in the first memory block are coupled to different bit lines selected by different column selection line (CSL) signals. The first faulty cell and the first normal cell in the first memory block may be connected to the same input/output circuit. The repair control circuit 400a replaces the first normal cell in the first memory block with the first redundant cell in the redundant block (S200). The first normal cell and the first redundant cell may have the same row selection line address. The first normal cell and the first redundant cell may be respectively connected to different input/output circuits.

11 is a block diagram illustrating another example of a semiconductor memory element in the memory system shown in FIG. 1 according to an exemplary embodiment.

Referring to Figure 11, the semiconductor memory device 200b includes a control logic circuit 210, an address register 220, a memory bank control logic (bank control logic) 230, a refresh counter 245, a row address (RA) multiplexer ( multiplexer (MUX) 240, row address (column address (CA) latch 250, column decoding 260, row decoder 270, memory cell array 300, sense amplifier unit 285, input/output gate control circuit 290 and data input/output buffer 295.

In an exemplary embodiment, the semiconductor memory device 200b may include an error correction code (ECC) engine 280.

The control logic circuit 210, the repair control circuit 400 and the timing control circuit 500 may constitute the access control circuit 205.

The memory cell array 300 includes a first memory bank array 310 to an eighth memory bank array 380 . The column decoder 260 includes first to eighth memory bank column decoders 260a to 260h respectively coupled to the first to eighth memory bank arrays 310 to 380, and the row decoder 270 includes a first to eighth memory bank column decoder 260h respectively coupled to the first to eighth memory bank arrays 310 to 380. The first memory bank row decoder 270a to the eighth memory bank row decoder 270h of the memory bank array 310 to the eighth memory bank array 380, and the sense amplifier unit 285 includes a memory bank row decoder 270a to an eighth memory bank row decoder 270h respectively coupled to the first memory bank array 310 to the eighth memory bank array 380. The first to eighth bank sense amplifiers 285a to 285h of the bank array 380. The first to eighth memory bank arrays 310 to 380, the first to eighth memory bank column decoders 260a to 260h, the first to eighth memory bank row decoders 270a to 270h, and The first to eighth memory bank sense amplifiers 285a to 285h may form first to eighth memory banks. Each of the first to eighth memory bank arrays 310 to 380 includes a plurality of memory cells MC formed at intersections of a plurality of word lines WL and a plurality of bit lines BTL.

The address register 220 receives the address including the memory bank address BANK_ADDR, the column address RADDR and the row address CADDR from the memory controller 100 ADDR. The address register 220 provides the received memory bank address BANK_ADDR to the memory bank control logic 230 , provides the received column address RADDR to the column address multiplexer 240 , and provides the received row address CADDR Provided to row address latch 250.

The memory bank control logic 230 generates a memory bank control signal in response to the memory bank address BANK_ADDR. One of the first bank decoders 260a to the eighth bank decoder 260h corresponding to the bank address BANK_ADDR is enabled in response to the bank control signal, and the first bank row decoder One of the bank row decoders 270a to 270h corresponding to the bank address BANK_ADDR is enabled in response to the bank control signal.

The column address multiplexer 240 receives the column address RADDR from the address register 220 and receives the refreshed column address REF_ADDR from the refresh counter 245 . The column address multiplexer 240 selectively outputs the column address RADDR or the new column address REF_ADDR as the column address RA. The column address RA output from the column address multiplexer 240 is applied to the first to eighth bank column decoders 260a to 260h.

One of the first bank row decoders 260a to the eighth bank row decoder 260h that is enabled by the bank control logic 230 decodes the column address RA output from the column address multiplexer 240, And enable the word line corresponding to column address RA. For example, the enabled memory bank column decoder applies the word line driving voltage to the word line corresponding to the column address RA. Additionally, when enabling the RA with column address At the same time as the corresponding word line, the enabled memory bank column decoder activates the spare word line corresponding to the spare column address SRA output by the self-healing control circuit 400 .

The row address latch 250 receives the row address CADDR from the address register 220 and temporarily stores the received row address CADDR. In some embodiments, in burst mode, the row address latch 250 generates a row address that is incremented from the received row address CADDR. The row address latch 250 applies the temporarily stored or temporarily generated row addresses to the first to eighth memory bank row decoders 270a to 270h.

An activated bank row decoder among the first to eighth bank row decoders 270a to 270h enables the bank row decoder corresponding to the bank address BANK_ADDR and the row address CADDR through the input/output gating circuit 290 Sense amplifier. Each of the first to eighth memory bank row decoders 270a to 270h may include a repair control circuit, and one of the first to eighth memory bank row decoders 270a to 270h is enabled The repair control circuit included in a memory bank row decoder can use the first normal cell in at least one memory block of the corresponding memory bank array to repair the faulty cell in the same memory block, and can use the corresponding memory The first redundant cell in the redundant block in the library array is used to replace the first normal cell.

The input/output gating circuit 290 includes circuitry for gating input/output data, and further includes a read data lock for storing data output from the first memory bank array 310 to the eighth memory bank array 380 registers and write drivers for writing data to the first to eighth memory bank arrays 310 to 380 .

Data read from one of the first to eighth memory bank arrays 310 to 380 is sensed by a sense amplifier coupled to the one of the memory bank arrays to be read and stored. in the read data latch. The data stored in the read data latches may be provided to the memory controller 100 via the data input/output buffer 295 . Data to be written into one of the first memory bank arrays 310 to the eighth memory bank array 380 may be written into one memory bank array by a write driver.

When the semiconductor memory device 200b includes the ECC engine 280, the ECC engine 280 can perform ECC encoding on the data to be written to provide codewords to the input/output gate control circuit 290, and can read the codewords. ECC decoding is performed to provide corrected data to data input/output buffer 295.

The data input/output buffer 295 may provide data DQ from the memory controller 100 to the ECC engine 280 during a write operation of the semiconductor memory device 200b based on the clock signal CLK, and may provide data DQ during a read operation of the semiconductor memory device 200b. The data DQ is provided from the ECC engine 280 to the memory controller 100 .

The control logic circuit 210 may control the operation of the semiconductor memory device 200b. For example, the control logic circuit 210 can generate a control signal for the semiconductor memory device 200b to perform a write operation or a read operation. The control logic circuit 210 includes a command decoder 211 and a mode register 212. The command decoder 211 decodes the command CMD received from the memory controller 100. The mode register 212 sets the operating mode of the semiconductor memory device 200b.

For example, the command decoder 211 can write the enable signal, the column The address strobe signal, row address strobe signal, chip select signal, etc. are decoded to generate a control signal corresponding to the command CMD. The control logic circuit 210 provides the first control signal CTL1 to the input/output gating circuit 290 and the second control signal CTL2 to the ECC engine 280 .

FIG. 12 shows an example of a first memory bank array in the semiconductor memory device shown in FIG. 11 .

Referring to FIG. 12 , the first memory bank array 310 includes a normal cell array NCA and a redundant cell array RCA. The normal cell array NCA includes a plurality of word lines WL1 to WLm (m is a natural number greater than two), a plurality of bit lines BTL1 to BTLn (n is a natural number greater than two), and word lines WL1 to WLm. and multiple memory cells MC at intersections between bit lines BTL1 to BTLn. The redundant cell array RCA includes a plurality of redundant cells RMC disposed at intersections between the word lines WL1 to WLm and the plurality of redundant bit lines RBTL1 to RBTLt.

13 is a block diagram illustrating a repair control circuit included in each of the memory bank row decoders in the semiconductor memory element shown in FIG. 12 according to an exemplary embodiment.

Referring to FIG. 13 , the repair control circuit 400b may include a fault address storage circuit 410, a column address comparator 415 and a cell repair controller 402a.

Although repair control circuit 400b is shown as including unit repair controller 402a, repair control circuit 400b may include multiple unit repair controllers and redundant repair controllers as shown in FIG. 15 .

The fault address storage circuit 410 stores the normal address in the memory cell array 300 Column address information FRAI and row address information FCAI of at least one defective cell (ie, faulty cell) appearing in the cell array. The fault address storage circuit 410 includes a non-volatile memory element to store the location information of the at least one defective cell. For example, the fault address storage circuit 410 may include an antifuse to store the location information of the at least one defective cell. The location information of the at least one defective cell stored in the fault address storage circuit 410 can be updated.

For example, the location information of defective cells that appear in the normal cell array when the semiconductor memory device 200b is continuously used can be updated in the fault address storage circuit 410. In addition, the location information of additional defective cells that appear after the semiconductor memory device 200b is packaged can be updated in the fault address storage circuit 410. Such location information of defective cells can be obtained by testing whether a faulty bit occurs in the semiconductor memory device 200b. The testing may be performed before the semiconductor memory device 200b is packaged (ie, at the wafer level), or may be performed after the semiconductor memory device 200b is packaged. Post-package repair (PPR) may be performed using the repair control circuit 400 according to an exemplary embodiment.

The location information of the at least one defective cell may be column address information FRAI and row address information FCAI of the at least one defective cell.

The column address comparator 415 stores the column address information FRAI received from the fault address storage circuit 410 . The column address comparator 415 may receive the column address information FRAI from the fault address storage circuit 410 while the semiconductor memory element 200b is being driven or at a desired time period after the semiconductor memory element 200b is driven. The column address comparator 415 receives the column address RADDR of the access address ADDR, and compares the column address RADDR with the column address The column address information FRAI is compared, and a column matching signal RM is output when the column address RADDR matches the column address information FRAI.

The cell repair controller 402a may include a table pointer 405, an address storage table 420b, a row position comparator 430, an AND gate 435, a selection circuit 440, and a row selection line driver 450.

The address storage table 420b can sequentially store the row address information FCAI of the faulty cell and the row address information of the first normal cell that repairs the faulty cell as the source row address SRCA, and can sequentially store the first normal cell. The row address information of the first normal cell and the row address information of the second normal cell that repairs the first normal cell are used as the destination row address DSCA. The table pointer 405 may generate to the address storage table 420b a table pointing signal TPS that is switched in response to the sequentially changed access row address CADDR. The address storage table 420b may output the source row address SRCA and the destination row address DSCA corresponding to the source row address SRCA in response to the table pointing signal TPS.

The row address comparator 430 compares the access row address CADDR with the source row address SRCA from the address storage table 420b, and outputs a first match signal MTH1 indicating the comparison result. The AND gate 435 performs an AND operation on the column matching signal RM and the first matching signal MTH1 to output the second matching signal MTH2. The selection circuit 440 may select one of the destination row address DSCA and the access row address CADDR from the address storage table 420b in response to the second matching signal MTH2 to output the selected one as the target row address CA. . The row select line driver 450 may output a row select line signal CSLa for selecting (enabling) the bit line corresponding to the target row address CA.

For example, when the column match signal RM is low or when the row address is accessed When CADDR does not match the source row address SRCA, the selection circuit 440 may output the access row address CADDR as the target row address CA in response to the second matching signal MTH2. For example, when the column match signal RM is at a high level and when the access row address CADDR matches the source row address SRCA, the selection circuit 440 may output the destination row address DSCA as the target in response to the second match signal MTH2 Row address CA.

FIG. 14 shows an example of the fault address storage table in the repair control circuit shown in FIG. 13 .

Referring to FIG. 14 , the fault address storage table 410 includes an antifuse array 411 , a control unit 412 , a sensing unit 413 and a register unit 414 .

The antifuse array 411 includes p*q anti-fuses (AF) respectively connected to the intersections of p columns and q rows. The antifuse array 411 includes p word lines AWL1 to AWLp for accessing antifuses (AF) disposed at p columns and q bit lines ABL1 to ABLq disposed corresponding to q rows to transmit Information read by the reflexive fuse (AF).

The control unit 412 programs the location information of the faulty cell in the antifuse array 411 or reads the location information of the faulty cell from the antifuse array 411 . The sensing unit 413 can sense and amplify the position information of the faulty cell received from the anti-fuse array 411 and output the amplified result. The register unit 414 may temporarily store the location information of the faulty cell received from the sensing unit 413 . The register unit 414 outputs the column address information FRAI and row address information FCAI of the faulty cell to the column address comparator 415 and the address storage table 420b respectively.

FIG. 15 shows a part of the semiconductor memory element shown in FIG. 11 .

In Figure 15, a first memory bank array 310, an input/output gating circuit 290, a row decoder 270a and a data input/output buffer are shown.

Referring to FIG. 15 , the first memory bank array 310 includes a normal cell array NCA and a redundant cell array RCA. The normal cell array NCA includes a plurality of memory blocks MB0 to MB15, ie, 311 to 313, and the redundant cell array RCA includes at least one redundant block 314. The memory blocks 311 to 313 are memory blocks that determine the memory capacity of the semiconductor memory element 200b. Redundancy block 314 is used for redundancy repair.

In each of the memory blocks 311 to 313, a plurality of memory cells are arranged in columns and rows. In the redundant block 314, a plurality of redundant cells are arranged in columns and rows.

The gate control circuit 290 includes a plurality of input/output circuits 292a to 292d and a plurality of row selection circuits 296a to 296d, and the row selection circuits 296a to 296d can connect one of the input/output circuits 292a to 292d to the memory block 311, 312 and 313 and the corresponding one of the redundant block 314. Each of the row selection circuits 296a to 296d may include a plurality of row selection transistors 297a to 297h, and the plurality of row selection transistors 297a to 297h respectively switch the corresponding memory in response to the row selection line signals CSLa to CSLg. A plurality of bit lines or a bit line in the block or redundant block 314 are connected to corresponding input/output circuits. The plurality of input/output circuits 292a to 292d may be coupled to the data input/output buffer 296 through data lines (not shown) in response to the first control signal CTL1 from the control logic circuit 210. For example, when a row select line signal is applied to the row select transistor 297a, one or more of each of the memory blocks 311 to 313 and the redundant block coupled to the row select transistor 297a may be simultaneously selected. Multiple bit lines. When the row select line signal is applied to the row select transistor At 297h, one or more bit lines in each of the memory blocks 311-313 and the redundancy block coupled to the row select transistor 297h may be selected simultaneously.

Row decoder 270a may include a predecoder (not shown), a plurality of cell repair controllers 402a-402c, and a redundancy repair controller 402d. The predecoder may decode the access row address CADDR to collectively provide the decoded row address to the plurality of cell repair controllers 402a-402c and redundancy repair controller 402d.

The cell repair controllers 402a to 402c and the redundancy repair controller 402d can jointly receive the access row address CADDR or the decoded row address, and can provide the corresponding row selection circuits 296a to 296d with applied to the memory blocks 311 to 313 and the row select line signals CSLa to CSLg of the redundant block 314.

The repair control circuit 400b uses the first normal cell in at least one of the memory blocks 311 to 313 to repair at least one faulty cell in the same memory block, and uses the second normal cell in the same memory block to repair The first normal cell is replaced, and the second normal cell is replaced with the first redundant cell in the redundant block 314 . Therefore, the repair control circuit 400b can use the redundant resources in the redundant block 314 with substantially maximum efficiency.

FIG. 16A shows a repair operation performed in the semiconductor memory element shown in FIG. 15.

Referring to FIG. 16A , the repair status 541 of the first memory bank array 310 is as follows. Each of the memory blocks MB0 and MB15 includes faulty cells on the bit lines selected by the row select line signal CSL0. By enabling row select line signal CSL3 in place of row select line signal CSL0 as indicated by reference number 531, the memory block The first normal cell in MB0 repairs the faulty cell in the memory block MB0 by enabling the row select line signal CSL7 to replace the row select line signal CSL3 as indicated by reference number 532 to replace the memory block MB0 and as indicated by reference number 533, the second normal cell in the memory block MB0 is replaced with the corresponding redundant cell in the redundant block 314. In addition, as indicated by reference number 534, the faulty cell associated with CSL0 in the memory block MB15 is repaired using the corresponding redundant cell in the redundant block 314.

When the CSL0 operation 542 in each of the memory blocks 311 to 314 is specified by the access row address CADDR, the first normal cell corresponding to CSL3 is selected in the memory block MB0 to replace the faulty cell, in The normal cells corresponding to CSL0 are selected in the memory block MB1, and the redundant cells corresponding to CSL0 are selected to replace the faulty cells in the memory block MB15.

When the CSL3 operation 543 in each of the memory blocks 311 to 314 is specified by the access row address CADDR, the second normal cell corresponding to CSL7 is selected to replace the first normal cell in the memory block, The normal cell corresponding to CSL3 is selected in the memory block MB1, and the normal cell corresponding to CSL3 is selected in the memory block MB15.

When the CSL7 operation 544 in each of the memory blocks 311 to 314 is specified by the access row address CADDR, the corresponding redundant cell corresponding to CSL7 is selected to replace the normal cell corresponding to CSL7 in the memory block MB0. element, select the normal cell corresponding to CSL7 in the memory block MB1, and select the normal cell corresponding to CSL7 in the memory block MB15.

As can be seen in Figure 16A, similar to Figure 6A, the repair control circuit is configured to utilize the second normal memory cells in the second row of the first memory block (eg, in the row numbered 7 in the memory block MB0 cell) to replace the first normal memory cell in the first row of the first memory block (for example, the cell in the row numbered 3 in the memory block MB0), and is configured to use the first redundancy The first redundant memory cell in the first row of the redundant memory cell of the remaining block (for example, the row numbered 7 in the redundant block RMB) is used to replace the first redundant memory cell in the second row of the first memory block. Two normal memory cells. In this example, using the first redundant memory cell to replace the second normal memory cell includes causing the first redundant memory cell to store data destined for the second normal memory cell. As further shown in FIG. 16A , the repair circuit is further configured to use the first normal memory cell of the first memory block (eg, the cell in the row numbered 3 in the memory block MB0 ) to repair another row (eg, , the faulty memory cell in the row numbered 0 in the memory block MB0).

FIG. 16B shows another example of the first memory bank array in the semiconductor memory device shown in FIG. 15 .

In FIG. 16B , the size of each memory block MB0 to MB3 is larger than the size of the redundant block RMB2 in the first memory bank array 310b , and the size of each memory block MB0 to MB15 is the same as that of the first memory block RMB2 shown in FIG. 16A The size of the redundant block RMB in the library array 310. Each size of the memory blocks MB0 to MB3 is twice larger than the size of the redundant block RMB2 in the first memory bank array 310b shown in FIG. 16B.

In FIG. 16B, each of the memory blocks MB0 to MB3 may be based on The most significant bit (MSB) of the access row address is divided into an upper block corresponding to CSL0 to CSL3 and a lower block corresponding to CSL4 to CSL7. When the faulty cells are distributed as shown in Figure 16B, the repair status can be determined by processing without considering the MSB of the accessed row address, so that the faulty cells in the memory blocks MB0, MB1 and MB2 are The redundant resources in the redundant block RMB2 do not overlap to support the burst operation of the semiconductor memory device 200b.

That is, the fuse information FI_MB1 associated with the memory block MB1 is set to select the normal cell corresponding to CSL5 to replace the faulty cell corresponding to CSL4, and the fuse information associated with the memory block MB2 is set. FI_MB2 is set to select the normal cell corresponding to CSL2 to replace the faulty cell corresponding to CSL0. In addition, when CSL0 and CSL4 are specified in each of the memory blocks MB0 to MB3, the redundant cell corresponding to CSL0 is selected in the redundant block RMB2. The fuse circuit 480b may store fuse information MB0_L, MB1_L and MB2_L regarding the operations of CSL0, CSL1 and CSL2 respectively.

FIG. 16C shows another example of the first memory bank array in the semiconductor memory device shown in FIG. 15 .

In FIG. 16C , the size of each of the memory blocks MB0 to MB7 is the same as the size of the redundant block RMB3 in the first memory bank array 310c.

Referring to FIG. 16C , the first memory bank array 310c may include a plurality of memory blocks MB0 to MB7 and a redundant block RMB3.

The memory block MB0 includes a first fault cell and a second fault cell on the bit lines selected by the row select line signals CSL0 and CSL1, and the redundant memory block RMB3 includes a first fault redundant cell and a second fault redundant cell on the bit line selected by the row select line signals CSL0 and CSL1. By energizing the row selection line signal CSL3 in place of the row selection line signal CSL0, the first faulty cell in the memory block MB0 is repaired from the first normal cell in the memory block MB0. The first normal cell in the memory block MB0 is sequentially replaced by the second normal cell in each of the memory blocks MB1 to MB7 and the third redundant cell in the redundant block RMB3. By energizing the row selection line signal CSL3 in place of the row selection line signal CSL0, the first faulty cell in the memory block MB0 is repaired from the first normal cell in the memory block MB0. The second faulty cell in the memory block MB0 is sequentially replaced by the second normal cell in each of the memory blocks MB1 to MB7 and the second redundant cell in the redundant block RMB3, and Since the second redundant cell in the redundant block RMB3 is a faulty redundant cell, the second redundant cell in the redundant block RMB3 is replaced by the fourth redundant cell in the redundant block RMB3 Yuan.

In FIG. 16C , the burst operation of the semiconductor memory device 200 b can be supported by determining the repair status for repairing the faulty cells in the memory block MBO so that the redundant resources in RMB3 do not overlap.

That is, the fuse information FI_MB0 is set to select the normal cell corresponding to CSL2 to replace the faulty cell corresponding to CSL0 associated with the memory block MB0, and the fuse information FI_RMB3 is set to select the normal cell corresponding to CSL3 The redundant cell replaces the failed redundant cell corresponding to CSL1 associated with the redundant block RMB3. The fuse circuit 480c can store fuse information about CSL1 and CSL2 respectively.

FIG. 17A shows the address storage table in the repair control circuit shown in FIG. 13. Example.

Referring to FIG. 17A, the address storage table 420b includes a first storage unit 421b, a second storage unit 423b and a sensing unit 425b. The first storage unit 421b stores the source row address SRCA to be repaired, and the second storage unit 423b stores the destination row address DSCA that replaces the source row address SRCA. The address storage table 420b may be implemented as an antifuse array or a content addressable memory (CAM). The sensing unit 425b outputs the source row address SRCA and the destination row address DSCA stored in the positions (indicated by the pointer signal TPS) in the first storage unit 421b and the second storage unit 423b in response to the pointer signal TPS. In Figure 17, the address storage table 420b stores the row address CADDR1 associated with CSL0, the row address CADDR4 associated with CSL3, and the row address CADDR8 associated with CSL7 as the source row address SRCA, and stores the replacement The row address CADDR4 of the row address CADDR1, the row address CADDR8 that replaces the row address CADDR4, and the redundant row address RCADDR8 that replaces the row address CADDR8 serve as the destination row address DSCA.

Figure 17B illustrates an example of reducing the number of fuses associated with a repair operation, according to an exemplary embodiment.

Referring to FIG. 17B , the main fuse information MFB may include two bits, and the main fuse information MFB may have one of “00”, “01”, “10” and “11”. In FIG. 17B , the main fuse information MFB is merged with the fuse information FFI of the faulty cell.

In Figure 17B, six bits of fuse information can specify forty-eight addresses. For example, the main fuse information MFB with "00" can indicate that the repair operation is not applied, and the main fuse information MFB with "01", "10" and "11" can be used to specify the fault cell. Yuan information. In addition, the fuse information FFI of the faulty cell with four bits can specify sixteen addresses. That is, each of the main fuse information MFBs with "01", "10" and "11" merged with the fault cell's fuse information FFI can specify sixteen addresses, and six bits Fuse information can specify forty-eight addresses. When thirty-four of the forty-eight addresses are used, one of the six bits of fuse information can be reduced.

Figure 17C illustrates another example of reducing the number of fuses associated with a repair operation in accordance with an exemplary embodiment.

Referring to FIG. 17C, one of the source information SR and the destination information DS may be merged with the main fuse information MFB. FIG. 17C shows that the destination information DS is merged with the main fuse information MFB and generates the merged information MFB&DS. The combined information MFB&DS includes three bits, and the source fuse information SRFI includes six bits. In Figure 17C, one bit associated with memory block MB is shared by two adjacent row blocks and the number of fuses can be reduced.

Figure 17D illustrates in detail an example of implementing the example shown in Figure 17C.

Referring to FIG. 17D , the main fuse information MFB merged with the destination information corresponding to the merged information MFB&DS shown in FIG. 17C may include three bits of pattern S[9]S[8]S[7]. When the pattern S[9]S[8]S[7] has the value "000", it indicates that the repair operation is not used. When the pattern S[9]S[8]S[7] has a value other than "000", the master fuse information MFB may include bits of destination information or part of the bits specifying the address of a normal cell. As shown in Figure 17D, the row address CADDR of the normal cell as the destination information can be obtained by selectively modifying the upper three bits CA9, CA8 and CA7 is obtained by flipping.

Figure 17E illustrates an example of a unit repair controller according to an exemplary embodiment.

Referring to FIG. 17E, the cell repair controller 501 may include a row address comparator 510, a normal decoder 520, a destination decoder 530, a multiplexer 540, and a row select line driver 550.

The row address comparator 510 compares the access row address CADDR with the source row address SRCA output from the address storage table 420b shown in FIG. 13, and outputs a value indicating the access row address CADDR and the source row address SRCA. The hit signal HIT1 of the comparison result. The normal decoder 520 decodes the access row address CADDR in response to the main row select signal PCSLM to output the first decoded row address DCADDR.

The destination decoder 530 decodes the destination row address DSCA output from the address storage table 420b in response to the row select main signal PCSLM to output a second decoded row address DDSCA.

The multiplexer 540 selects one of the first decoded row address DCADDR and the second decoded row address DDSCA in response to the hit signal HIT1 to output the selected one as the decoded target row address DCA. . The row select line driver 550 receives the decoded target row address DCA and outputs a row select line signal to select (enable) the bit line corresponding to the decoded target row address DCA.

Figure 17F illustrates another example of a unit repair controller according to an exemplary embodiment.

Referring to Figure 17F, the cell repair controller 502 may include a row address comparator 515, inverter 517, multiplexer 545, normal decoder 525 and row select line driver 555.

The row address comparator 515 compares the access row address CADDR with the source row address SRCA and the additional bit AB1, and outputs a comparison result indicating the access row address CADDR with the source row address SRCA and the additional bit AB1. The hit signal HIT21. The additional bit AB1 may include four bits.

Three bits in the additional bits AB1 may correspond to the main fuse information MFB merged with the destination information DS explained with reference to FIG. 17C , and one bit in the additional bits AB1 may be two adjacent bits. Bits shared by row blocks.

The access row address CADDR may include higher bits CADDR_MSB and lower bits CADDR_LSB. The inverter 517 inverts the higher bits CADDR_MSB of the accessed row address CADDR. The multiplexer 545 outputs one of the output of the inverter 517 and the lower bit CADDR_LSB of the access row address CADDR in response to the hit signal HIT21.

The normal decoder 525 decodes the output of the multiplexer 545 and the lower bit CADDR_LSB of the access row address CADDR in response to the main row select signal PCSLM to output the decoded target row address DCA. The row select line driver 555 receives the decoded target row address DCA and outputs a row select line signal CSL to select (enable) the bit line corresponding to the decoded target row address DCA.

The higher bits CADDR_MSB of the access row address CADDR may be encoded, as shown in Figure 17D.

When using normal cells in the memory block to compare old cells in the same memory block When repairing defective cells, the unit repair controller 501 shown in FIG. 17E and the unit repair controller 502 shown in FIG. 17F can be used.

Figure 17G illustrates another example of a unit repair controller according to an exemplary embodiment.

Referring to FIG. 17G , the cell repair controller 503 may include a row address comparator 516, an inverter 517, a multiplexer 545, a normal decoder 525, and a row select line driver 555.

The row address comparator 516 compares the access row address CADDR with the source row address SRCA and the additional bit AB2, and outputs a hit indicating the comparison result of the access row address CADDR with the source row address SRCA and the additional bit AB2. Signal HIT22. The additional bit AB2 may include three bits.

The access row address CADDR may include higher bits CADDR_MSB and lower bits CADDR_LSB. The inverter 517 inverts the higher bits CADDR_MSB of the accessed row address CADDR. The multiplexer 545 outputs one of the output of the inverter 517 and the lower bit CADDR_LSB of the access row address CADDR in response to the hit signal HIT22.

The normal decoder 525 decodes the output of the multiplexer 545 and the lower bit CADDR_LSB of the access row address CADDR in response to the main row select signal PCSLM to output the decoded target row address DCA. The row select line driver 555 receives the decoded target row address DCA and outputs a row select line signal CSL to select (enable) the bit line corresponding to the decoded target row address DCA.

The upper bit CADDR_MSB of the access row address CADDR can be The encoding is shown in Figure 17D.

Since two adjacent memory blocks share a fuse, the additional bit AB2 includes three bits. In addition, when normal cells in the memory block are used to repair faulty cells in the same memory block, the unit repair controller 503 shown in FIG. 17G can be used.

Figure 17H illustrates another example of a unit repair controller according to an exemplary embodiment.

17H, the cell repair controller 504 may include a row address comparator 515, an inverter 517, a multiplexer 546, a decoder 526, a row selection line driver 556, a row address comparator 518, an inverter 519, Multiplexer 548, decoder 527 and row select line driver 557.

The row address comparator 515 compares the access row address CADDR with the source row address SRCA and the additional bit AB1 including four bits, and outputs an instruction to access the row address CADDR with the source row address SRCA and the additional bit. The hit signal HIT31 of the comparison result of AB1.

The access row address CADDR may include higher bits CADDR_MSB and lower bits CADDR_LSB. The inverter 517 inverts the higher bits CADDR_MSB of the accessed row address CADDR.

The row address comparator 518 compares the access row address CADDR with the source row address SRCA and the additional bit AB1, and outputs a hit indicating the comparison result of the access row address CADDR with the source row address SRCA and the additional bit AB1. Signal HIT32.

The inverter 519 accesses the higher bits of the row address CADDR. CADDR_MSB is inverted.

Multiplexer 546 outputs one of the output of inverter 517 and the lower bit CADDR_LSB of the accessed row address CADDR. The decoder 526 decodes the output of the multiplexer 546 and the lower bit CADDR_LSB of the access row address CADDR to output the decoded target row address DCA. The row select line driver 556 receives the decoded target row address DCA and outputs the row select line signal CSLa1 to select (enable) the bit line corresponding to the decoded target row address DCA.

Multiplexer 548 outputs one of the output of inverter 519 and the lower bit CADDR_LSB of the access row address CADDR. The decoder 527 decodes the output of the multiplexer 548 and the lower bit CADDR_LSB of the access row address CADDR to output the decoded target row address DCA'. The row select line driver 557 receives the decoded target row address DCA' and outputs the row select line signal CSLa2 to select (enable) the bit line corresponding to the decoded target row address DCA'.

When normal cells in two adjacent memory blocks are used to repair a faulty cell, the cell repair controller 504 shown in FIG. 17H can be used.

18 is a flowchart illustrating a method of operating a semiconductor memory element according to an exemplary embodiment.

Referring to FIGS. 11 to 18 , in a method of operating a semiconductor memory device 200 b including a memory cell array 300 including a plurality of memory blocks and at least one redundant block, the repair control circuit 400 b utilizes the plurality of memory blocks. The first normal cell in the first memory block is used to repair the first faulty cell in the first memory block (S310). The repair control circuit 400b is coupled to the first normal cell by using a designated Swapping the second row address of the second bit line specifies a first row address of the first bit line coupled to the first faulty cell to repair the first faulty cell using the first normal cell. The repair control circuit 400b replaces the first normal cell in the first memory block with the second normal cell in the first memory block among the plurality of memory blocks (S330). The first fault cell, the first normal cell and the second normal cell in the first memory block may have different row selection line addresses. For example, a first fault cell, a first normal cell and a second normal cell in the first memory block are coupled to different bit lines selected by different row select line (CSL) signals. The first fault cell, the first normal cell and the second normal cell in the first memory block may be connected to the same input/output circuit.

The repair control circuit 400a replaces the second normal cell in the first memory block with the first redundant cell in the redundant block (S350). The second normal cell and the first redundant cell may have the same row selection line address, and may have the same relative position within their respective memory cell arrays. The second normal cell and the first redundant cell may be connected to different input/output circuits respectively.

19 is a block diagram showing a semiconductor memory element according to an exemplary embodiment.

Referring to FIG. 19 , the semiconductor memory device 600 may include a first group of dies 610 and a second group of dies 620 in a stacked chip structure. The first group of dies 610 and the second group of dies 620 provide soft errors. Analysis and correction functions.

The first set of dies 610 may include at least one buffer die. The second group of dies 620 may include a plurality of memory dies 620-1 to 620-r stacked on the first group of dies 610 and via a plurality of strips. Silicon through hole (through silicon via, TSV) line transfers data.

At least one of the memory dies 620-1 to 620-r may include a first type error correction code (ECC) engine 622 and an error injection register set 623. The first type error correction code The engine 622 generates transmit parity bits based on the transmit data to be sent to the first group of dies 610 . The first type ECC engine 622 may be referred to as a "cellular core ECC engine."

The buffer die 610 may include a second type ECC engine 612 that uses transmission parity bits to correct transmission errors when a transmission error is detected from transmission data received over the TSV line and generates an error corrected information. The second type ECC engine 612 may be referred to as a "through-hole ECC engine." The buffer die 610 may include a repair control circuit (RCC) 614, and the repair control circuit 614 may be the repair control circuit 400b shown in FIG. 13.

The semiconductor memory device 600 may be a stacked chip type memory device or a stacked memory device that transmits data and control signals through TSV lines. TSV lines may also be called "through electrodes".

The first type ECC engine 622 may perform error correction on the data output from the memory die 620-p before sending the transmission data.

Through the above description, the TSV line group 632 formed at one memory die 620-r may include a plurality of TSV lines L1 to Lp, and the co-located TSV line group 634 may include a plurality of TSV lines L10 to Lq. The TSV lines L1 to Lp of the data TSV line group 632 and the co-located TSV lines L10 to Lq of the co-located TSV line group 634 may be connected to correspondingly formed micro-bumps MCB in the memory dies 620-1 to 620-r.

At least one of memory dies 620-1 to 620-r may include DRAM cells, each DRAM cell including at least one access transistor and one storage capacitor.

The semiconductor memory device 600 may have a three-dimensional (3D) chip structure or a 2.5-dimensional chip structure to communicate with the memory controller via the data bus B10. Buffer die 610 may be connected to the memory controller via data bus B10.

The first type ECC engine 622, represented as a cellular ECC engine, may output transmit parity bits and transmit data via the parity TSV line group 634 and the data TSV line group 632, respectively. The output transmission data may be error-corrected data by the first type ECC engine 622 .

The second type ECC engine 612 , represented as a through-hole ECC engine, may determine whether a transmission error occurs in the transmission data received via the data TSV line group 632 based on the transmission parity bits received via the parity TSV line group 634 . When a transmission error is detected, the second type ECC engine 612 may use transmission parity bits to correct the transmission error on the transmission data. When the transmission error is uncorrectable, the second type ECC engine 612 may output information indicating that an uncorrectable data error has occurred.

20 is a cross-sectional view of a three-dimensional wafer structure employing the semiconductor memory element shown in FIG. 19 according to an exemplary embodiment.

FIG. 20 shows a three-dimensional chip structure 700 in which a host is directly connected to a high-bandwidth memory (HBM) without an interposer layer.

Referring to Figure 20, host die 720 (e.g., system die (system-on-chip, SoC), central processing unit (CPU) or graphics processing unit (GPU)) can be disposed on a printed circuit board (printed circuit board) using flip-chip bumps FB. PCB)710. The memory dies D11 to D14 may be stacked on the host die 720 to implement the HBM structure. In Figure 20, the buffer die 610 or logic die shown in Figure 19 is omitted. However, the buffer die 610 or the logic die may be disposed between the memory die D11 and the host die 720 . To implement the HBM (620) structure, TSV lines can be formed at memory dies D11 and D14. The TSV lines can be electrically connected to the micro-bumps MCB placed between each memory die.

Aspects of the inventive concept may be applied to systems using semiconductor memory devices.

The foregoing is illustrative of exemplary embodiments and should not be construed as limiting the exemplary embodiments. Although several exemplary embodiments have been described, those skilled in the art will readily appreciate that many modifications may be made in the exemplary embodiments without materially departing from the novel teachings and advantages of the inventive concept.

290a: Input/output (I/O) gate control circuit

291a, 291b, 291c, 291d, 291e: input/output circuit

293a, 293b, 293c, 293d, 293e: row selection circuit

296: Data input/output buffer

301: Memory cell array

400a: Repair control circuit

401a: Unit Repair Controller/First Unit Repair Controller

401b, 401c, 401d: unit repair controller

401e: Redundant Repair Controller/Unit Repair Controller

511, 512, 513, 514, 515: Reference number

CADDR: row address/access row address

CSL, CSLa, CSLb, CSLC, CSLd: row select line signals

CSLe: row select line signal/redundant row select line signal

CTL1: first control signal

DQ:data

MB0: Memory block/first memory block

MB1: Memory block/second memory block

MB2: Memory block/third memory block

MB3: Memory block/fourth memory block

NCA: normal cell array

RCA: redundant cell array

RMB: redundant block

Claims (9)

A method of replacing memory cells in a first row of a memory block in a memory element, the memory element including a plurality of memory blocks and at least one first redundant block, the method comprising: using a destination address Replace the address of the first normal memory cell in the first row of the first memory block with the destination address being the address of the second normal memory cell in the second row of the first memory block. ;Reassign the address of the second normal memory cell in the second row of the first memory block to the first redundant memory cell in the redundant block of the memory element and replacing the bit of the second normal memory cell in the first row of the second memory block with the address of the second destination memory cell in the second row of the second memory block. address, wherein the second destination memory cell is a normal memory cell, and wherein: the first row of the first memory block and the first row of the second memory block are between The respective memory blocks have the same relative position, and the second row of the first memory block and the second row of the second memory block have different values in their respective memory blocks. relative position. The method as described in item 1 of the patent application, wherein: the first normal memory cell is a faulty cell, and using the destination address to replace the address of the first normal memory cell includes: The faulty cells are repaired. The method described in item 1 of the patent scope further includes: The first normal memory cells are used to repair the faulty cells in the third row of the first memory block. The method as described in item 1 of the patent application, wherein a first row selection line is used to access the first normal memory cell, and the address of the first normal memory cell includes the first The address of a row of selection lines, and further includes: replacing the address of the first row of selection lines with the address of a second row of selection lines, the second row of selection lines pairing the The second normal memory cell performs access. The method described in item 1 of the patent application further includes: matching the first normal memory cell in the first row of the first memory block with an address storage table based on storage source address matching. The addresses are replaced, and the source address pair corresponds to a source memory cell having a destination address, and the destination address corresponds to the destination memory cell. The method as described in item 5 of the patent application, wherein: the first normal memory cell is a faulty cell, and using the destination address to replace the address of the first normal memory cell includes: The faulty cell is repaired, and the source address corresponds to the faulty cell. The method described in item 5 of the patent application further includes: receiving the first row address of the first row; and comparing the first row address with the first row address stored in the address storage table. Row source addresses are compared; and When the first row address matches the first row source address stored in the address storage table, a target address is output to the row select line driver, and the target address is the first row Destination address, the first row of destination addresses corresponds to the first row of source addresses from the address storage table. The method described in item 5 of the patent application, wherein: the address storage table further stores destination address pairs, and the destination address pairs correspond to destination memory cells with redundant addresses, so The redundant address corresponds to a redundant memory cell, and further includes: reassigning the address of the second normal memory cell in the second row of the first memory block to the corresponding redundant memory cell. The redundant address of the address pair of the second normal memory cell. The method described in claim 1, wherein: the reassignment is based on information related to row selection lines and memory blocks stored in fuse circuits.

Images