| 6256243 | Test circuit for testing a digital semiconductor circuit configuration | July, 2001 | Savignac et al. | 365/201 |
| 6408412 | Method and structure for testing embedded analog/mixed-signal cores in system-on-a-chip | June, 2002 | Rajsuman | 714/724 |
| 6636998 | Semiconductor memory device and parallel bit test method thereof | October, 2003 | Lee et al. | 714/735 |
| 20040153910 | Integrated memory and method for testing the memory | August, 2004 | Thalmann et al. |
| JP05012900 | January, 1993 | SEMICONDUCTOR STORAGE CONTAINING TEST FUNCTION AND ITS TEST METHOD | ||
| KR2001-0040999 | May, 2001 | |||
| KR1020040036523 | April, 2004 |
A claim of priority is made to Korean Patent Application No. 10-2005-0011442, filed on Feb. 7, 2005, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein in its entirety by reference.
1. Field of the Invention
Example embodiments of the present invention generally relate to a semiconductor memory device. More particularly, example embodiment of the present invention relate to a semiconductor memory device in which memory cells are tested using several different test data patterns and method thereof.
2. Description of the Related Art
As the capacity of a semiconductor memory device has increased, so has the time required to test the semiconductor memory device. Accordingly, in order to reduce the test time, several semiconductor memory devices may be tested at the same time using a single test apparatus.
The number of input/output (I/O) pins capable of being tested by a single test apparatus is fixed, therefore, to test additional semiconductor memory devices at the same time, it may be necessary to reduce the number of I/O pins allocated to a memory device.
To test a semiconductor memory device using fewer number of I/O pins, the semiconductor memory device may be adapted to operate in a test mode in addition to a normal mode. In the test mode, the semiconductor memory device may process and output several internal parallel data through the I/O pins, and may expand and write data received through the I/O pins as parallel data.
However, when data received through the fewer number of I/O pins are expanded and several internal parallel data are generated, data patterns capable of being written in memory cells may not be varied.
FIG. 1 illustrates a conventional semiconductor memory device 100 capable of expanding input data and generating parallel data.
Referring to FIG. 1, the memory device 100 may include a plurality of transmission gates TG 11 to TG 116 and TG 21 to TG 216 . In a normal mode, the transmission gates TG 11 to TG 116 in response to a normal mode signal MODE_N may transfer received data DIN 1 to DIN 16 to corresponding memory cells (not shown). The normal mode signal MODE_N may be applied to NMOS transistors of the transmission gates TG 11 to TG 116 , and an inverted normal signal MODE_N, inverted by an inverter 11 , may be applied to PMOS transistors of the transmission gates TG 11 to TG 116 .
During a test mode, the transmission gates TG 21 to TG 216 may apply received data DIN 1 , DIN 5 , DIN 9 , and DIN 13 in response to a test mode signal MODE_T to corresponding memory cells (not shown).
The test mode signal MODE_T may be applied to NMOS transistors of the transmission gates TG 21 to TG 216 , and an inverted test mode signal MODE_T, inverted by an inverter 12 , may be applied to PMOS transistors of the transmission gates TG 21 to TG 216 .
As described above, in order for a single test apparatus to test several semiconductor memory devices at one time, the number of I/O pins allocated to a memory device must be reduced. The memory device 100 illustrated in FIG. 1 has an “×16” organization in which 16 data bits may be input/output at the same time, and the number of I/O pins used during the testing may be four.
The memory device 100 takes the four data DIN 1 , DIN 5 , DIN 9 , and DIN 13 in a test mode and generates (expands) 16 data DATA 1 to DATA 16 to be applied to the memory cells. The data DATA 1 to DATA 16 may be respectively written to the corresponding memory cells.
Since the memory device 100 uses only four input data DIN 1 , DIN 5 , DIN 9 , and DIN 13 during a test mode, the same value as the input data DIN 1 may be written to the memory cells to which data DATA 1 , DATA 2 , DATA 3 , and DATA 4 may be applied.
Likewise, the same value as the input data DIN 5 may be written to the memory cells to which data DATA 5 , DATA 6 , DATA 7 , and DATA 8 may be applied. This process also applies to DIN 9 and DIN 13 .
In general, in order to detect a defective semiconductor memory device, several different test data patterns may be required. However, as described above, it may not possible for the semiconductor memory test circuit 100 with the “×16” organization to have a value data DATA 1 that is different than a value of data DATA 2 . In the memory device 100 , the number of test data patterns capable of being used in a test mode may be limited.
FIG. 2 illustrates a conventional test data pattern generator 200 .
As described above with respect to the memory device 100 of FIG. 1, the number of test data patterns may be limited.
In order to solve the problem of the limited test data patterns available during a test mode, an address key method of combining input data DIN 1 to DINn with dummy addresses ADD 1 to ADDm to generate test data patterns P_DATA 1 to P_DATAn has been introduced in the conventional art.
However, because the number of available test data patterns depends on the number of available addresses and input data, and the addresses must be varied according to the test data patterns, there may be compatibility issues of using the address key method with existing package test patterns.
FIG. 3 illustrates another conventional memory device 300 which may include a selector 310 to select either data DOI output from a memory cell (not shown) in a normal mode or select a test result signal TRS in a test mode, in response to a mode control signal MODE. The memory device 300 may further include an output buffer 320 , and logic devices ENR 1 to ENR 16 , and N 1 to generate the test result signal TRS.
With reference to FIG. 3, a “merged DQ” method to reduce the number of I/O pins through which test results are output will be described. In the merged DQ method, test data values T_DATA 1 to T_DATA 16 stored in internal registers (not shown) may be respectively compared with data DOI 1 to DOI 16 read from memory cells, a pass/fail (P/F) may be determined from the comparison, and the determined result may be output through the output buffer 320 . By using the merged DQ method, it may be possible to reduce the number of I/O pins and output buffers 320 during a test mode.
By respectively comparing test data values T_DATA 1 to T_DATA 16 stored in internal registers (not shown) with data DOI 1 to DOI 16 read from memory cells using logic devices ENR 1 to ENR 16 , and N 1 , if at least one memory cell is defective, a test result signal TRS indicating the defect may be generated.
However, in the merged DQ method, problems still exist in that the method requires a predetermined command to change test data values T_DATA 1 to T_DATA 16 stored in internal registers (not shown) to be respectively compared with read data DOI 1 to DOI 16 . In addition, it may be difficult to change the test data values T_DATA 1 to T_DATA 16 . Accordingly, compatibility issues with test data patterns used with current test method still exists.
In an example embodiment of the present invention, a semiconductor memory device having a test circuit includes a switch unit adapted to transfer data or transfer test data to memory cells in response to a mode control signal, a plurality of storage units respectively adapted to store the test data in response to an input control signal, and a selection unit adapted to apply one of the test data to the switch unit in response to the input control signal.
In another example embodiment, a semiconductor memory device having a test circuit includes a plurality of selectors adapted to transfer internal data to corresponding output buffers or transfer test result signals in response to a mode control signal, and a plurality of test units adapted to determine whether the internal data are the same as the corresponding test data and generate the test result signals thereof, wherein the test data corresponding to the plurality of test units is different from each other.
And in another example embodiment, a method of testing memory cells includes transferring internal data or transferring test result signals in response to a mode control signal, and determining whether the internal data are the same as corresponding test data and generating the test result signals thereof, wherein the test data corresponding to the plurality of test units is different from each other.
The present invention will become more apparent with the description of the detail exemplary embodiments thereof with reference to the attached drawings in which:
FIG. 1 illustrates a circuit diagram of a conventional semiconductor memory device.
FIG. 2 illustrates a conventional test data pattern generator;
FIG. 3 illustrates a conventional memory device adapted to perform a merged DQ method for outputting test results for memory cells;
FIG. 4 is a circuit diagram of a semiconductor memory device according to an example embodiment of the present invention; and
FIG. 5 is a circuit diagram of a memory device in which memory cells may be tested using various test data patterns, according to another example embodiment of the present invention.
Hereinafter, example embodiments of the present invention will be described in detail with reference to the appended drawings. Like reference numbers refer to like components throughout the drawings.
FIG. 4 is a circuit diagram of a memory device 400 , in which memory cells may be tested using several different test data patterns, according to an example embodiment of the present invention.
Referring to FIG. 4, the memory device 400 may include a switch unit 410 , a plurality of storage units REG 1 to REG 16 , and a selector 420 .
The switch unit 410 may transfer m bits of data DIN 1 to DIN 16 or m bits of test data T_DATA received in response to a mode control signal MODE to memory cells (not shown), wherein m is a natural number. For the convenience of description but not limited thereto, in this example embodiment, m is 16.
The switch unit 410 may include a plurality of first transmission gates TG 11 to TG 116 , which may transmit 16 bits of data DIN 1 to DIN 16 in response to a first mode control signal MODE, and a plurality of second transmission gates TG 21 through TG 216 , which may transmit 16 bits of test data T_DATA in response to a second mode control signal MODE.
When the semiconductor memory device 400 is in a normal mode, the first transmission gates TG 11 to TG 116 may be turned on and transmit data DIN 1 to DIN 16 . In the present embodiment, the first mode control signal MODE may be high when the memory device 400 is in a normal mode. However, in another example embodiment of the present invention, in a normal mode, the first mode control signal may be low. The second transmission gates TG 21 to TG 216 may be turned on when the memory device 400 is in a test mode and transmit test data T_DATA. In the present embodiment, the second mode control signal MODE may be low when the memory device 400 is in the test mode. However, in another example embodiment of the present invention, in a test mode, the second mode control signal may be low. If the first transmission gates TG 11 to TG 116 may be turned on, the second transmission gates TG 21 to TG 216 may be turned off. If the second transmission gates TG 21 to TG 216 are turned on, the first transmission gates TG 11 to TG 16 are turned off.
The plurality of storage units REG 1 to REG 16 may each respectively store 16 bits of test data T_DATA in response to n bits of an input control signal DIN_CTRL. The input control signal DIN_CTRL may be a partial bit(s) of the data DIN 1 to DIN 16 . The number of the storage units REG 1 to REG 16 may be 2 n or less, wherein 2 n =m. For example, the number of the storage units REG 1 to REG 16 may be 16 or less and the input control signal DIN_CTRL may be 4 bits of data DIN 1 to DIN 16 . Each of the storage units REG 1 to REG 16 may be a register.
The selector 420 applies test data T_DATA stored in one of the plurality of storage units REG 1 to REG 16 to the switch unit 410 in response to an input control signal DIN_CTRL. The test data T_DATA may be applied to the second transmission gates TG 21 to TG 216 and written to a corresponding memory cell (not shown).
Test data T_DATA stored in the plurality of storage units REG 1 to REG 16 may be different from each other. The selector 420 may be a multiplexer.
First, before performing a write operation, test data T_DATA may be stored in the storage units REG 1 to REG 16 in a test mode. Since an input control signal DIN_CTRL may have 4 bits and each of the storage units REG 1 to REG 16 may store 16 bits (the same number of bits of data DIN 1 to DIN 16 applied to a memory cell) of test data T_DATA, the 16 bits of the test data T_DATA may be divided and written four times to a storage unit.
In order to store test data T_DATA in all the storage units REG 1 to REG 16 , 4×16 write operations may be performed. Through this initialization procedure, test data T_DATA to be used in the test mode may be stored in advance in the storage units REG 1 to REG 16 . The test data T_DATA stored in the storage unit REG 1 to REG 16 may be different from each other.
The number of the storage units REG 1 to REG 16 may be a maximum of 16. The selector 420 may select one of outputs of the storage units REG 1 to REG 16 using a combination of bits of an input control signal DIN_CTRL when test data T_DATA is actually written. The selected output may be written to a corresponding memory cell via a corresponding one of the second transmission gates TG 21 through TG 216 .
As necessary, the test data T_DATA stored in the storage units REG 1 to REG 16 may be readjusted during a test. The memory device 400 shown in FIG. 4 may diversify test data patterns for the testing memory cells using a simple circuit configuration.
FIG. 5 is a circuit diagram of a memory device 500 of another example embodiment of the present invention.
Referring to FIG. 5, the memory test circuit 500 may include a plurality of selectors MUX 1 and MUX 2 and plurality of test units T 1 to T 2 .
The selectors MUX 1 and MUX 2 may transfer m bits of internal data DOI to corresponding output buffers DQB 1 and DQB 2 or may transfer test result signals TRS 1 and TRS 2 , which contain information of whether memory cells (not shown) have defects, to the corresponding output buffers DQB 1 and DQB 2 , in response to a mode control signal MODE.
Here, m is a natural number. In the present example embodiment and for the convenience of description, m is 16. Also for convenience and purposes of explanation, FIG. 5 depicts a memory device having 2 selectors MUX 1 and MUX 2 and 2 test units T 1 and T 2 , but the number of selectors and test units are not limited thereto.
If the memory device 500 is in a normal mode, the selectors MUX 1 and MUX 2 may output internal data DOI received from memory cells in response to a mode control signal MODE.
If the semiconductor memory device 500 is in a test mode, the selectors MUX 1 and MUX 2 may output test result signals TRS 1 and TRS 2 received from the test units T 1 and T 2 in response to the mode control signal MODE.
The test units T 1 and T 2 may determine whether 16 bits of internal data DOI 1 to DOI 16 are the same as corresponding 16 bits of test data TA_DATA 1 to TA_DATA 16 , TB_DATA 1 to TB_DATA 16 , and may generate test result signals TRS 1 to TRS 2 .
That is, the test unit T 1 may determine whether the internal data DOI 1 to DOI 16 are the same as the test data TA_DATA 1 to TA_DATA 16 , and may generate the test result signal TRS 1 based on the determination. If the internal data DOI 1 to DOI 16 are the same as the test data TA_DATA 1 to TA_DATA 16 , the test result signal TRS 1 may indicate whether the corresponding memory cells are normal.
Also, the test unit T 2 may determine whether the internal data DIO 1 to DOI 16 are the same as the test data TB_DATA 1 to TB_DATA 16 , and may generate the test result signal TRS 2 . If the internal data DOI 1 to DOI 16 are the same as the test data TB_DATA 1 to TB_DATA 16 , the test result signal TRS 2 may also indicate whether the corresponding memory cells are normal.
Here, the test data TA_DATA 1 to TA_DATA 16 corresponding to the test unit T 1 are different from the test data TB_DATA 1 to TB_DATA 16 corresponding to the test unit T 2 . In other words, by respectively comparing the same internal data D 101 to DOI 16 with different test data TA_DATA 1 to TA_DATA 16 , TB_DATA 1 to TB_DATA 16 , it may be possible to diversify test data patterns. Such test operations may be simultaneously performed in a single clock cycle.
Each of the test units T 1 and T 2 may include first through sixteenth Exclusive OR devices, and an AND device.
For example, the first to sixteenth Exclusive OR devices ENR 11 through ENR 1 _ 16 of the first test unit T 1 respectively may perform Exclusive OR operation of 16 bits of internal data DOI 1 to DOI 16 with corresponding 16 bits of test data TA_DATA 1 to TA_DATA 16 . If the internal data DOI 1 to DOI 16 are the same as the test data TA_DATA 1 to TA_DATA 16 , the Exclusive OR devices ENR 11 to ENR 1 _ 16 may generate output high level signals. If the internal data DOI 1 to DOI 16 are different from the test data TA_DATA 1 to TA_DATA 16 , the Exclusive OR devices ENR 11 to ENR 1 _ 16 may generate output low level signals, and the AND device N 1 may generate a test result signal TRS 1 having a low level. The low level test result signal TRS 1 may indicate that a defect exists in a test memory cell.
Accordingly, if at least one bit of the internal data DOI 1 to DOI 16 is different from a corresponding bit of the test data TA_DATA 1 to TA_DATA 16 , the Exclusive OR devices ENR 11 to ENR 1 _ 16 may generate output signals with a low level, and the AND device N 1 may generate a test result signal TRS 1 with a low level. The test result signal TRS 1 with the low level may indicate that a defect exist in the tested memory cells.
The configuration and operation of T 2 may be similar to the operation of T 1 , therefore, a detail configuration and operation descriptions of T 2 are omitted.
The semiconductor memory device 500 may include storage units REG 1 to REG-n storing test data TA_DATA 1 to TA_DATA 16 , TB_DATA 1 to TB_DATA 16 . The test data TA_DATA to TA_DATA 16 and TB_DATA 1 to TB_DATA 16 respectively stored in the storage units REG 1 and REG 2 may be different from each other.
The selectors MUX 1 and MUX 2 may simultaneously output test result signals TRS 1 and TRS 2 in synchronization to a clock signal. Accordingly, two different test data TA_DATA 1 to TA_DATA 16 and TB_DATA 1 to TB_DATA 16 may be stored in advance in the storage units REG 1 and REG 2 , and internal data DOI 1 to DOI 16 may be compared with the different test data TA_DATA 1 to TA_DATA 16 and TB_DATA 1 to TB_DATA 16 in the same clock cycle.
By simultaneously outputting test result signals TRS 1 and TRS 2 corresponding to comparison results through corresponding output buffers DQB 1 and DQB 2 , the semiconductor memory device 500 may determine whether two different test data TA_DATA 1 to TA_DATA 16 and TB_DATA 1 to TB_DATA 16 are equal to the internal data DOI 1 to DOI 16 in a clock cycle. This method may be expansively applied to perform a plurality of pass/fail (P/F) determinations in a clock cycle.
As described above, in a semiconductor memory device according to example embodiments of the present invention, it may be possible to test memory cells using various test data patterns and simultaneously receive test results using the various test data.
While example embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the scope of the example embodiments of the present invention.