KR101234306B1 - Burn-in testing method using auto strobe - Google Patents

Abstract

PURPOSE: A burn-in test method using an auto strobe is provided to accurately output expectation data based on test data when performing a high speed test. CONSTITUTION: A test program is loaded, and delay time and a strobe control signal are set up(S10). When n number of strobe control signals are set up, the test program transmits and inputs test data to a DUT(Device Under Test) arranged in a burn-in board(S20). When the test data are inputted the test program outputs the test data which is stored in the DUT arranged in one row among the DUT, as delay data by delaying using one of the strobe control signals occurred in each delay time interval(S30). When expectation data is outputted, the test program confirms which outputted row of the expectation data is the last one(S40). If the row is the last one, the test program determines good or bad quality of the DUT using the expectation data(S50). [Reference numerals] (AA) Start; (BB) End; (S11) Loading a test program; (S12) D_T; (S13) Setting up S_cont; (S21) Selecting an input mode; (S22) Matrix input mode?; (S23,S25) Inputting T_data; (S24) Selecting a row; (S27) Finishing input?; (S31) Selecting a column; (S32) Outputting E_data; (S40,S26) Last?; (S50) Determining good or bad quality

Description

Burn-in testing method using auto strobe}

The present invention relates to a burn-in test method using an auto strobe, and more particularly, automatically outputs a strobe control signal for each n columns when outputting expected data from a device under test (DUT) arranged in an m × n matrix on a burn-in board. The present invention relates to a burn-in test method using an auto strobe that can accurately output expected data according to test data during high speed test.

Burn-in test equipment is one of the semiconductor test equipment used for monitoring burn in test (MBT). The burn-in test apparatus installs a packaged semiconductor component in a socket of a burn-in board, loads it into a rack in which a plurality of burn-in boards can be stacked, and burn-in the rack on which the burn-in board is loaded. It is placed in a chamber provided in the device to apply electrical signals and heat for a period of time to verify the operational reliability of the packaged semiconductor component in extreme environments.

Korean Patent Laid-Open Publication No. 2005-0047928 (Patent Document 1) relates to a test method for simultaneously performing a burn-in test and a mounting test, and includes loading a test program and performing a mounting test. Loading the test program loads the test program into each of a plurality of device under tests (DUTs) mounted on the test board in the burn-in chamber of the burn-in system, and performing the mounting test is performed by using a burn-in test apparatus. The burn-in test is performed on each of the DUTs, and a test for each of the plurality of DUTs is performed using the loaded test program.

As in Patent Literature 1, the conventional burn-in test method performed burn-in test using one strobe control signal. In the conventional burn-in test method, since the test rate of the burn-in test apparatus is 10 MHz and the test period is 100 ms, the test data may be input to a plurality of DUTs arranged in a matrix for 100 ms, and then outputted.

Expected data is output data for test data stored in semiconductor parts. When one strobe control signal is set to approximately 50 ms when outputting expected data, it can be output as multiple DUTs arranged in rows on a burn-in board. . On the other hand, in the conventional burn-in test method using one strobe control signal, the test period is extremely short when the test speed of the burn-in test apparatus is high, and the expected data is lost when one strobe control signal is used. There is a problem that cannot be correctly tested.

Korean Patent Publication No. 2005-0047928 (published: 2005.05.23)

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems, and to automatically apply a strobe control signal for each n columns when outputting expected data from a device under test (DUT) arranged in an m × n matrix on a burn-in board. It provides burn-in test method using auto strobe to accurately output expected data according to test data during test.

Another object of the present invention is to provide a burn-in test method using an auto strobe that can input test data by selecting a DUT arranged in an m × n matrix on a burn-in board in a row or column direction.

Another object of the present invention is to select the DUT arranged in the m × n matrix on the burn-in board to enter the test data by selecting in the column or column direction to select the DUT arranged in the m × n matrix on the burn-in board It is to provide burn-in test method using auto strobe.

The burn-in test method using the auto strobe of the present invention includes the steps of loading a test program and setting delay time and n strobe control signals; Transmitting and inputting test data to a device under test (DUT) arranged in an m × n matrix on a burn-in board by a test program when the n strobe control signals are set; When the test data is input, one of the strobe control signals of the n strobe control signals, which are generated at a delay time interval, from the test data stored in the m DUTs arranged in one column among the DUTs arranged in the m × n matrix by the test program. Delaying and outputting the expected data; When the expected data is output, confirming whether a column in which expected data is output by a test program is the last of n columns; If the column is the last, it comprises the step of determining the good or bad of the DUT by using the expected data by a test program, wherein m and n are each a natural number of two or more.

The burn-in test method using the auto strobe of the present invention automatically applies a strobe control signal for each n columns at the time of outputting expected data from a device under test (DUT) arranged in an m × n matrix on a burn-in board, and tests at a high speed test. It has the advantage of accurately outputting the expected data according to the data, and by selecting the DUT arranged in the m × n matrix on the burn-in board in the row or column direction, the test data can be input to the m × n matrix on the burn-in board. The advantage is that you can select and test a DUT arranged with.

1 is a flow chart showing a burn-in test method using the auto strobe of the present invention,
2 is a block diagram showing the configuration of a burn-in test apparatus to which the burn-in test method using the auto strobe of the present invention is applied;
3 is a timing chart showing an input / output state of the format controller shown in FIG. 2;

An embodiment of the burn-in test method using the auto strobe of the present invention will be described with reference to the accompanying drawings.

In the burn-in test method using the auto strobe of the present invention as shown in FIGS. 1 to 3, first, a delay time D_t and n strobe control signals S_cont1, S_cont2, ..., S_contn are set by loading a test program. (S10). When n strobe control signals (S_cont1, S_cont2, ..., S_contn) are set, DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in an m × n matrix on the burn-in board 20 by the test program ( Device Under Test) to transmit and input the test data (T_data) (S20). When the test data (T_data) is input, m DUTs (aDUT1, aDUT2, ....) arranged in one column of the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix by the test program. , the test data T_data stored in aDUTn is delayed by one strobe control signal S_cont1 among the n strobe control signals S_cont1, S_cont2, ..., S_contn which are generated at intervals of delay time D_t, respectively. Output as expected data (S30). When the expected data is output, the test program checks whether the output column of the expected data is the last of the n columns (S40). If the column is the last, the test program determines the good or bad of the DUTs (aDUT1, aDUT2, ...,, mDUTn) by using the expected data (S50), and n and n each use two or more natural numbers.

The burn-in test method using the auto strobe of the present invention having the above configuration will be described in detail with reference to the accompanying drawings.

First, the burn-in test apparatus to which the burn-in test method using the auto strobe of the present invention is applied is composed of a test board 10 and a burn-in board 20 as shown in FIG. 2, and the test board 10 and the burn-in board 20 A connection board (not shown) is provided for the connection of signal transmission therebetween.

The test board 10 includes a test controller 11 and m format controllers (aFC, bFC,..., MFC). The test controller 11 stores a test program for overall control of the burn-in test method using the auto strobe of the present invention, and after the loading of the test program is activated, the common control signal G_cont, the test data T_data, and the DUT Select signals CS_1, CS2, ..., CSn are respectively generated and transmitted to the burn-in board 20 as DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in an m × n matrix for input or test data. The expected data E_data according to (T_data) is received to determine whether the DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in the m × n matrix are good or bad.

The test controller 11 also includes an ALPG (Algorithmic Pattern Generator) (not shown) that generates a common control signal (G_cont) such as an address pattern, a data pattern of a test data, a control pattern and a timing control signal and an embedded clock generation control signal. A TG (Time Generator) (not shown) for generating a common control signal G_cont such as a timing clock signal in response to the timing control signal is provided.

m format controllers (aFC, bFC, ..., mFC) are connected to the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in m × n matrices, respectively, to input test data (T_data). Expected data (E_data1, E_data2, ..., E_datan: shown in FIG. 3) according to the test data are outputted from the DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in an m × n matrix.

The burn-in board 20 is mounted such that the DUTs (aDUT1, aDUT2, ...., mDUTn) are arranged in an m × n matrix as shown in FIG. 2, and each DUT (aDUT1, aDUT2, ...., mDUTn) Is connected to m format controllers (aFC, bFC, ..., MFC). The DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in m rows of the DUTs (aDUT1, aDUT2, ...., and mDUTn) arranged in m × n rows are daisy chained with each other. The DUTs (aDUT1, aDUT2, ...,, mDUTn) shown in FIG. 3 represent a state in which the semiconductor memory devices or the semiconductor memory modules in which the input / output pins are commonly used are daisy chained in the row direction.

 Referring to the burn-in test method using the auto strobe of the present invention using the burn-in test apparatus having the above configuration as follows.

 In order to set the delay time D_t and the n strobe control signals S_cont1, S_cont2, ..., S_contn by loading a test program as shown in FIGS. 1 to 3 (S10), first, in the test controller 11 The test program is loaded (S11). When the test program is loaded, the delay time D_t calculated by converting the length between the DUTs (aDUT1, aDUT2, ...,, mDUTn) into time is set in the test program (S12). Delay time (D_t) is expected to be the physical distance deviation (M) and distance deviation (M) between the DUT (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix on the burn-in board 20 It is calculated using the transmission speed passed by (E_data) and set in the test program.

When the delay time (D_t) is set, the test period, which is the test time of the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix, the number of columns in the m × n matrix and the rising time of the digital signal ( The n strobe control signals S_cont1, S_cont2, ..., S_contn calculated using T1) and the polling time T2 are set in the test program (S12). The setting of the n strobe control signals S_cont1, S_cont2, ..., S_contn is to calculate each period T of the n strobe control signals S_cont1, S_cont2, ..., S_contn. The period T of the n strobe control signals S_cont1, S_cont2, ..., S_contn calculates the output time of the expected data E_data when the test speed of the fast burn-in test method of the present invention is 200 MHz. The output time of the expected data E_data is 2.5 ms when the test rate is 200 MHz, but the rising time or polling time is 0.8 ms, respectively. Therefore, when the rising and falling times are subtracted from 2.5 ms, the n strobe control signals S_cont1 and S_cont2 are respectively. The period T of the ..., ..., S_contn becomes 0.9 ms, respectively, and the period T of the n strobe control signals S_cont1, S_cont2, ..., S_contn is set to 0.9 ms or less, respectively. Expected data E_data outputted from the DUTs (aDUT1, aDUT2, ...,, mDUTn) including strobe information (not shown) in the setting period T is converted into DUT (aDUT1, aDUT2, ...., mDUTn). In the case of a memory device, the success or failure of each physical address or cell is determined.

If the delay time D_t and n strobe control signals S_cont1, S_cont2, ..., S_contn are set, the DUTs (aDUT1, aDUT2, .. .., mDUTn) to transmit and input the test data (T_data) (S20). In order to transmit the test data (T_data) to the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in an m × n matrix, the step of transmitting and inputting the test data (T_data) (S20) includes n strobes. When the control signals S_cont1, S_cont2, ..., S_contn are set, an input mode for selecting one of the column input mode and the row input mode is selected (S21). The operator selects the input mode, and the operator inputs using a user interface (not shown) provided in the burn-in data device. ;

If the input mode selection is selected, it is checked whether the matrix input mode is selected by the test program (S22). As a result of confirming that the matrix input mode is selected, when the matrix input mode is selected, a matrix test for inputting test data (T_data) into DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in m × n columns by a test program. Data T_data is input (S23). On the contrary, if the row input mode is selected as a result of confirming that the column input mode is selected, one of the n rows is selected by the test program (S24).

When one of the n rows is selected, a row test data input is performed in which test data (T_data) is input to n DUTs (aDUT1, aDUT2, ...,, aDUTn) arranged in the row selected by the test program. (S25). When the test data T_data is input, the test program checks whether the row into which the test data T_data is input is the last row (S26). As a result of the check, if the row into which the test data (T_data) is input is not the last row, the method returns to step S24 of selecting one row out of n rows. Enter test data in the selected rows from aDUT1, aDUT2, ...., mDUTn). Conversely, as a result of the verification, the test data is either n DUTs (mDUT1, mDUT2, ...., mDUTn) arranged in the last row or DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in an m × n matrix. It is checked in the test program whether input of T_data is completed (S27).

As a result of checking in step S27, when the input of the test data T_data is not completed with the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix or the n DUTs arranged in the last row, The test data is input to the DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in the m × n matrix by returning to the column test data input step S23 or the row test data input step S25. .

By selecting the matrix or column direction in the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix on the burn-in board 20, the test data can be input to the burn-in board 20 by m. You can optionally test the DUTs (aDUT1, aDUT2, ...,, mDUTn) arranged in an xn matrix. That is, when the DUTs are arranged in the 16 × 4 row in the burn-in board 20 in a state where the DUTs arranged in the 16 × 20 row can be arranged in the burn-in board 20, the test data (T_data) is inputted to the arranged DUT. By outputting the expected data (E_data), only the desired DUT can be tested among the flexibility of the test and m × n DUTs (aDUT1, aDUT2, ...., mDUTn).

When the test data (T_data) is input, m DUTs (aDUT1, aDUT2, ....) arranged in one column of the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix by the test program. The test data T_data stored in the aDUTn is delayed by one strobe control signal S_cont1 among the n strobe control signals S_cont1, S_cont2, ..., S_contn, which are generated at intervals of the delay time D_t. Output as expected data (S30). To this end, first, when test data (T_data) is input to a DUT arranged in an m × n matrix, one column of n columns is selected (S31). When a column is selected, it is stored in m DUTs (aDUT1, bDUT1, ...., mDUT1) arranged in the selected column among the DUTs (aDUT1, aDUT2, ...., mDUTn) arranged in the m × n matrix by the test program. The test data T_data is delayed by one strobe control signal S_cont1 among n strobe control signals S_cont1, S_cont2, ..., S_contn, which are generated at intervals of the delay time D_t, respectively, and are expected data E_data1. Output to (S31).

When the expected data E_data1 is output, as shown in FIG. 3, it is checked whether the column in which the expected data E_data1 is output by the test program is the last of n columns (S40). That is, as shown in FIGS. 2 and 3, the first strobe control signal includes the expected data E_data1 output from m DUTs (aDUT1, bDUT1,..., MDUT1) arranged in the first column in the burn-in board 20. Delay (S_cont1) by the first delay time (D_t x 0.5 ms) and output to m format controllers (aFC, bFC, ...., mFC) to delay and output.

After the first expected data (E_data1) is output, the expected data (E_data2) output from the m DUTs (aDUT2, bDUT2, ...., mDUT2) arranged in the second column is converted into the second strobe control signal (S_cont2). The delay is delayed by the second delay time D_t x 1 ms and output to the m format controllers aFC, bFC, ..., mFC. It is checked whether the column for outputting such continuous operation expectation data is the last of n columns (S40).

As a result of checking whether it is the last column, if it is not the nth column, it selects a column in which expected data (E_data: shown in FIG. 2) is not output again, and the nth arrayed m DUTs (aDUTn, bDUTn, ...., mDUTn) is the expected data (E_datan) output from m DUTs (aDUTn, bDUTn, ...., mDUTn) and nth strobe control signal (S_contn) is nth delay time (D_t × Delay by n ms and output to m format controllers aFC, bFC,..., mFC to delay and output.

If the expected value (E_datan) output from the last m-th arrayed m DUTs (aDUTn, bDUTn, ...., mDUTn) is output, all expected values (E_data: E_data1, E_data2, ..., E_datan) are returned. Received by the test controller 11 determines the good or bad of the DUT (aDUT1, aDUT2, ...., mDUTn) (S50). In order to select the column and output the expected value E_data, the test controller 11 is a DUT (aDUT1, aDUT2, ...,, mDUTn) arranged in an m × n matrix, respectively, and the DUT activation signals CS1 and CS2. , ..., CSn) and the delay time (D_t) and strobe control signal (S_cont) are calculated and set in the test program.The test program automatically outputs the expected data (E_data) according to the test speed and burns in. You can run the test.

In the burn-in test method using the auto strobe of the present invention, the delay time D_t is a representative generic name of the first delay time D_t × 0.5 ms to the nth delay time D_t × n ms, and the strobe control signal S_cont. Is the representative generic name of the first strobe control signal S_cont1 to the n th strobe control signal S_contn, and the expected data E_data is used as the representative generic name of the first expected data E_data1 to the nth expected data E_datan. .

As described above, the burn-in test method using the auto strobe of the present invention automatically applies a strobe control signal for each n columns when the expected data is output from the DUT arranged in the m × n matrix on the burn-in board to test data during high-speed test. It is possible to output the expected data accordingly, and by selecting the DUT arranged in the m × n matrix on the burn-in board in the row or column direction, the test data can be input to the DUT arranged in the m × n matrix on the burn-in board. You can choose to test it.

The burn-in test method using the auto strobe of the present invention can be applied to the manufacturing field of the burn-in test apparatus.

10: Test board
11: test controller
20: burn-in board

Claims (5)

Loading a test program to set delay time and n strobe control signals;
Transmitting and inputting test data to a device under test (DUT) arranged in an m × n matrix on a burn-in board by a test program when the n strobe control signals are set;
When the test data is input, one of the strobe control signals of the n strobe control signals, which are generated at a delay time interval, from the test data stored in the m DUTs arranged in one column among the DUTs arranged in the m × n matrix by the test program. Delaying and outputting the expected data;
When the expected data is output, confirming whether a column in which expected data is output by a test program is the last of n columns;
If the column is the last, it comprises the step of determining the good or bad of the DUT by using the expected data by the test program,
M and n are each a natural number of two or more burn-in test method using an auto strobe. The method of claim 1, wherein setting the delay time and the n strobe control signals comprises: loading a test program;
Setting the calculated delay time in the test program by converting the length between the DUTs into time when the test program is loaded;
Setting the strobe control signals calculated using the test cycles and the number of columns, the rising time and the polling time of the digital signal, which are the test time periods of the DUTs arranged in the m × n rows when the delay time is set in the test program. Burn-in test method using an auto strobe, characterized in that consisting of. The method of claim 1, wherein the transmitting and inputting of the test data comprises: an input mode selection step of selecting one of a row input mode and a row input mode when the n strobe control signals are set;
When the input mode is selected, checking whether a matrix input mode is selected by a test program;
A matrix test data input step of inputting test data into a DUT arranged in an m × n matrix by a test program when a matrix input mode is selected as a result of confirming that the matrix input mode is selected;
Selecting one of the n rows by a test program when the row input mode is selected as a result of checking whether the row input mode is selected;
A row test data input step of inputting test data into n DUTs arranged in rows selected by a test program when one row of the n rows is selected;
Checking whether the row into which the test data is input is the last row by the test program when the test data is input;
Comprising the step of checking in the test program whether the input of the test data is completed in the DUT arranged in the m × n column or the n DUT arranged in the last row,
When the test data is not input to the DUT arranged in the m × n column or the n DUTs arranged in the last row, the test returns to the column test data input step or the row test data input step. Burn-in test method using strobes. 4. The auto strobe of claim 3, wherein in the checking whether the row is the last row, when the test data is not the last row, the auto strobe is returned. Burn-in test method used. The method of claim 1, wherein the outputting of the expected data comprises: selecting one column of n columns when test data is input to a DUT arranged in m × n rows;
When the column is selected, the test data stored in the m DUTs arranged in the selected column among the DUTs arranged in the m × n matrix by the test program is controlled by one strobe control signal among the n strobe control signals generated at delay intervals. A burn-in test method using an auto strobe, characterized by comprising the steps of delaying and outputting expected data.

Images