TWI798696B - Auto-calibration method used in chip test equipment and test element - Google Patents

Abstract

An auto-calibration method used in chip test equipment and test element are provided. The auto-calibration method includes the steps outlined below. Calibration data of the chip test equipment is retrieved to generate a distribution diagram. A slope of each two neighboring data points in the distribution diagram is calculated. The data points in the distribution diagram are categorized into K1 part and K2 part according to an X-axis or a Y-axis. Standard deviations of first slope data, first 2 slope data... and first predetermined number of slope data corresponding to the K1 part and the K2 part are calculated to obtain standard deviation groups, wherein the predetermined number corresponds to a total number of the corresponding slope data and is larger than 3. A maximum standard deviation of each two of the standard deviation groups is selected to use the corresponding data points as sectional transition points, to separate the data points into three sections and perform curve fitting to obtain a fitting curve. The method of the present invention uses a method of calculating standard deviations of the slopes to select the sectional transition points, to provide an operation standard of calibration that can be standardized and automatically performed.

Description

An automatic calibration method and test element for wafer testing machine

The invention relates to the technical field of semiconductors, and relates to an automatic calibration method and a test element for a wafer tester, in particular to an automatic calibration method applied to an integrated circuit tester and a test element applying the method.

Integrated circuit (chip) automatic testing machine (Automatic Test Equipment) in the semiconductor industry means integrated circuit (IC) automatic testing machine, which is used to test the integrity of integrated circuit functions and is the final process of integrated circuit manufacturing. To ensure the quality of integrated circuit manufacturing.

As a kind of measuring instrument, the accuracy of the wafer tester itself largely determines the accuracy of the wafer tester, that is, in practical applications, it is often necessary to calibrate the wafer tester at regular intervals, usually through an external high-level Precision digital multimeter is used as a calibration tool, such as using keysight 3458A to measure PE, PMU, and DPS resources on the automatic test machine. For the measured data, the two-point method is generally used to calculate the calibration parameters, although this method can be realized to a certain extent. Calibration of parameters, but because the calibration parameter curve of PE/PMU/DPS is non-linear, this leads to a large actual error of this calibration method, poor matching with the real situation, and it is becoming more and more difficult to meet the requirements of higher and higher precision. Wafer tester test requirements.

Therefore, it is of great practical significance to develop an automatic calibration method applied to integrated circuit testing machines with a small error and a high degree of matching with the real situation.

The purpose of the present invention is to overcome the defect that the actual error of the two-point calibration line in the prior art is relatively large and the matching degree with the real situation is poor, and to provide an automatic calibration applied to an integrated circuit testing machine with a small error and a high degree of matching with the real situation method.

In order to achieve the above object, the present invention provides the following technical solutions: an automatic calibration method for a wafer testing machine, applied to electronic equipment, comprising the following steps: (1) obtaining calibration data of a wafer testing machine, and drawing Scatter diagram; (2) Calculate the slope of every two adjacent points in the scatter diagram respectively, and the slope of every two adjacent points is the slope data; (3) Divide the scatter points in the scatter diagram according to the abscissa or ordinate into Part K1 and part K2; (4) For part K1 and part K2, respectively calculate the standard deviation of the first slope data, the first two slope data, ... the previous preset number of slope data, and obtain the standard deviation of the K1 part The standard deviation data set D2 of the data sets D1 and K2, wherein the preset number is the total number of corresponding slope data and is greater than 3; (5) Separately screen the maximum value of the standard deviation in the standard deviation data group D1 and the standard deviation data group D2, and use the point corresponding to the maximum value of the standard deviation as the complex segmentation point to divide all the scatter points of the scatter diagram into three segments, and for each segment Carry out curve fitting on the scattered points respectively to obtain a fitting curve; (6) extract the calibration parameters corresponding to the fitting curve obtained in step (5), and input the calibration parameters into the wafer testing machine to complete the automatic calibration of the wafer testing machine.

The automatic calibration method for the wafer testing machine of the present invention optimizes the calculation method of the commonly used two-point calibration parameters at present. The essence of the two-point calculation is to collect two points for linear fitting, which is suitable for PE/PMU/DPS As far as the calibration parameter curve is concerned, it performs well in the middle of the curve with better linearity, but there is a large deviation from the actual situation at both ends, and because the two-point sampling area is mainly the middle linear part, it is impossible to correct Compensation for the nonlinear region at both ends further leads to an increase in the numerical deviation at both ends, and the calibration method (segmented fitting) of the present invention overcomes the difficulty of applying the two-point calibration parameter curve to the nonlinear PE/PMU/DPS (Its form of expression is that the linearity of the middle section is better, and the nonlinearity at both ends is more obvious). For the problem of segment point selection, the present invention adopts the slope standard deviation method (calculate the slope of every two points and sequentially compare these slopes) Carrying out the standard deviation operation) select the standard deviation maximum point to be the subsection point, provide a kind of standardized and fully automatic operation standard for the calibration of the wafer testing machine test data, while the method of the present invention maintains the two-point calculation in the middle section of the curve While collecting and calculating areas with good linearity, additional scattered points can be collected for areas with poor linearity at both ends, and a third-order polynomial curve fitting can be performed to reduce the excessive deviation from the actual value when calibrating parameters in different sections of the curve. Very promising application.

As preferred technical scheme: a kind of automatic calibration method that is used for wafer testing machine as above, the specific steps of described step (4) are as follows: (4.1) For the K1 part, calculate the standard deviation of the first slope data, the first 2 slope data, ... the first N slope data respectively, and obtain the standard deviation data group D1 of the K1 part, and N is the total number of the slope data of the K1 part , and N is greater than 3; (4.2) For the K2 part, calculate the standard deviation of the first slope data, the first 2 slope data, ... the first M slope data, and obtain the standard deviation data group D2 of the K2 part, and M is K2 The total number of partial slope data, and M is greater than 3.

A kind of automatic calibration method that is used for wafer testing machine as above, described step (5) is specifically as follows: respectively screen the standard deviation maximum value X in the standard deviation data set D1 and the standard deviation maximum value in the standard deviation data set D2 Y, the maximum value of the standard deviation X is the standard deviation of the previous a slope data in the K1 part, and the maximum standard deviation Y is the standard deviation of the first b slope data in the K2 part, that is, the a+1th point in the K1 part and The b+1th point in part K2 is used as a segmentation point to divide all the scatter points in the scatter diagram into three segments, and perform curve fitting for each segment of the scatter points to obtain a fitted curve. The protection scope of the present invention is not limited thereto, and only a feasible technical solution is given here, and those skilled in the art may make some adaptive adjustments in practical applications.

As mentioned above in an automatic calibration method for a wafer testing machine, the fitting formula of the curve fitting is y = E + Fx or y =A+B x +Cx ² +D x ³ .

As mentioned above in an automatic calibration method for a wafer testing machine, the fitting formula of the middle section is y = E + Fx ; the fitting formula of the other two sections is y = A+B x +Cx ² +D x ³ .

An automatic calibration method for a wafer testing machine as described above, further comprising the following steps: (7) After completing the calibration parameter input operation, repeat the above operations of steps (1) to (6), and perform multiple calibrations on the wafer testing machine to improve the calibration accuracy.

The present invention also provides a test element applying the above-mentioned automatic calibration method for a wafer testing machine, including one or more processors, one or more memories, one or more programs and a calibration data acquisition device; The calibration data obtaining device is used to obtain the calibration data of the wafer testing machine, the one or more programs are stored in the memory, and when the one or more programs are executed by the processor, the The test element implements the auto-calibration method described above for the wafer tester.

As a preferred technical solution: the test element as described above further includes a wafer testing machine calibration parameter setting module connected to the processor; the wafer testing machine calibration parameter setting module is used for setting calibration parameters.

Beneficial effects: (1) The automatic calibration method for wafer testing machine of the present invention adopts segmented fitting, which can be used for both ends of the linear calibration while maintaining the two-point calculation in the middle section of the curve with good linearity. Additional scattered points are collected in areas with poor precision, and third-order polynomial curve fitting is carried out, so as to reduce the situation that the deviation from the actual value is too large when calibrating parameters in different sections of the curve; (2) the automatic calibration for wafer testing machine of the present invention Method, select this problem for subsection point, the present invention adopts slope standard deviation method (calculate the slope of every two points and carry out standard deviation operation to these slopes successively) select standard deviation maximum point to be subsection point, for chip testing machine test The calibration of the data provides a standardized and fully automatic operation standard, which has great application prospects; (3) The test element of the present invention has simple structure, low cost and high degree of automation, and can realize automatic calibration of a wafer testing machine with high calibration accuracy.

About the feature, implementation and effect of this case, hereby cooperate with drawing as preferred embodiment and describe in detail as follows.

500: test components

510: Processor

520: Memory

530: program

540: Calibration data acquisition device

550: Wafer testing machine calibration parameter setting module

[Fig. 1] is a flow chart of an automatic calibration method for a wafer testing machine of the present invention; [Fig. 2A] is a schematic diagram of a fitting curve obtained by linear fitting; [Fig. 2B] is a three-section curve of the present invention Schematic diagram of the fitting curve obtained by fitting; [Fig. 3] is a schematic diagram of comparison between a linear fitting and a three-segment curve fitting segment and variance; [Fig. 4] is a segmental linear fitting and a three-segment curve fitting segment The comparative schematic diagram of root mean square; and (Fig. 5) is the structural schematic diagram of the test element of the present invention.

The specific implementation manners of the present invention will be further elaborated below in conjunction with the accompanying drawings.

Example 1

Please refer to Figure 1. FIG. 1 is a flowchart of an automatic calibration method 100 for a wafer testing machine according to the present invention. The automatic calibration method 100 is applied to electronic equipment, and includes the following steps: (1) Step S110: Obtain the calibration data of the wafer testing machine, and draw a scatter diagram according to the calibration data; (2) Step S120: Calculate each The slope of two points, the slope of every two adjacent points is the slope data; (3) Step S130: divide the scattered points in the scatter diagram into K1 part and K2 part according to the abscissa or ordinate; (4) Step S140: For the K1 part and the K2 part, calculate the standard deviation of the first slope data, the first 2 slope data, ... the previous preset number of slope data respectively, and obtain the standard deviation of the K1 part standard deviation data group D1 and the standard deviation of the K2 part Data group D2, wherein the preset number is the total number of corresponding slope data and is greater than 3, specifically: (4.1) For the K1 part, calculate the first slope data, the first 2 slope data, ... the first N slope data The standard deviation of the standard deviation data set D1 of the K1 part is obtained, and N is the total number of slope data of the K1 part, and N is greater than 3; (4.2) For the K2 part, calculate the first slope data, the first two slope data, ... ...the standard deviation of the first M slope data to obtain the standard deviation data group D2 of the K2 part, M is the total number of the slope data of the K2 part, and M is greater than 3; (5) Step S150: Screen the standard deviation data group D1 respectively The maximum value of the difference X and the maximum value of the standard deviation Y in the data group D2, the maximum value of the standard deviation X is the standard deviation of the previous a slope data in the K1 part, and the maximum value of the standard deviation Y is the first b slope data in the K2 part The standard deviation of the scatter plot is divided into three segments by using the a+1 scatter point in K1 and the b+1 scatter point in K2 as segment points, and curve fitting is performed for each scatter point. (wherein the fitting formula of the middle section is y = E + Fx , and the fitting formula of the left and right sections is y =A+B x +Cx ² +D x ³ ) to obtain the fitting curve; (6) step S160: extract The calibration parameter corresponding to the fitting curve that step (5) (being step S150) obtains, said calibration parameter input wafer testing machine promptly completes the automatic calibration to wafer testing machine; (7) after finishing calibration parameter input operation, repeat many The above operations in the sub-steps (1)-(6) (ie step S110-step S160).

For the calibration data of the wafer testing machine shown in Table 1 below,

The segmentation points selected in step (5) are respectively X=6 and X=17, and the fitting curve is obtained as shown in Fig. 2B.

The fitting curve obtained by using the calculation method of two-point calibration parameters, ie, a segment of linear fitting, is shown in FIG. 2A .

For each point in Figure 2A and Figure 2B, the sum variance and root mean square were calculated in sections, and the results are shown in Figures 3 and 4.

Analyzing Figures 2A, 2B, 3 and 4, it can be found that no matter it is the better middle segment of linearity or the other two segments of non-linearity, the average error of each segment of the three-segment fitting of the present invention is less than that of the one-segment formula, that is, it is consistent with the real data The degree of deviation is smaller than the one-step fitting and closer to the actual measured value.

It has been verified that the automatic calibration method for a wafer testing machine of the present invention adopts segmented fitting, which maintains the two-point calculation in the middle section of the curve and collects and calculates in the area with good linearity. It is also possible to collect additional scattered points for areas with poor linearity at both ends, and perform third-order polynomial curve fitting to reduce the excessive deviation from the actual value when calibrating parameters in different segments of the curve; for the problem of segment point selection , the present invention adopts the slope standard deviation method (calculate the slope of every two points and carry out the standard deviation operation to these slopes successively) and select the point with the maximum standard deviation as the subsection point, which provides a standardized and comprehensive method for the calibration of the wafer testing machine test data The operation standard of automatic completion has great application prospect.

Example 2

A test element 500, as shown in FIG. 5 , includes one or more processors 510, one or more memories 520, one or more programs 530, a calibration data acquisition device 540 and a wafer testing machine connected to the processor 510 Calibration parameter setting module 550; the wafer testing machine calibration parameter setting module is used to set the calibration parameters, the calibration data acquisition device is used to obtain the calibration data of the wafer testing machine, one or more programs are stored in the memory, when one or When a plurality of programs are executed by the processor, the testing device is made to perform the same automatic calibration method for a wafer testing machine as in Embodiment 1.

It has been verified that the test element of the present invention has simple structure, low cost and high degree of automation, and can realize automatic calibration of a wafer testing machine with high calibration precision.

Although the specific implementations of the present invention have been described above, those skilled in the art should understand that these are only examples, and various changes or modifications can be made to these implementations without departing from the principle and essence of the present invention. .

100: Automatic calibration method

S110~S160: steps

Claims (8)

An automatic calibration method for a wafer testing machine, applied to an electronic device, comprising the following steps: (1) obtaining a calibration data of the wafer testing machine, and drawing a scatter diagram (scatter diagram) according to the calibration data; ( 2) Calculate a slope of every two adjacent points of the complex scatter points in the scatter diagram respectively, and the slope of every two adjacent points is a slope data; (3) the scatter points in the scatter diagram According to an abscissa or a ordinate, it is divided into a K1 part and a K2 part; (4) for the K1 part and the K2 part, respectively calculate a first slope data, a first 2 slope data, ... a One standard deviation of the previous preset number of slope data to obtain a standard deviation data set D1 of the K1 part and a standard deviation data set D2 of the K2 part, wherein the preset number is the total number of the corresponding slope data and is greater than 3; (5) Screen the maximum value of a standard deviation in the standard deviation data set D1 and the standard deviation data set D2 respectively, and use the scatter points corresponding to the maximum standard deviation value as complex segment points to make all the scatter diagrams The scatter points are divided into three sections, and a curve fitting is performed on each scatter point respectively to obtain a fitted curve; and (6) extracting a calibration parameter corresponding to the fitted curve obtained in step (5), the The automatic calibration of the wafer testing machine is completed when the calibration parameters are input into the wafer testing machine. The automatic calibration method for wafer testing machine as described in claim 1, wherein, step (4) further comprises: (4.1) For the K1 part, calculate the standard deviation of the first slope data, the first 2 slope data, ... the first N slope data respectively, and obtain the standard deviation data group D1, N of the K1 part is the total number of the slope data of the K1 part, and N is greater than 3; and (4.2) for the K2 part, calculate the first slope data, the first 2 slope data, ... the first M slope data The standard deviation is to obtain the standard deviation data set D2 of the K2 part, M is the total number of the slope data of the K2 part, and M is greater than 3. The automatic calibration method for wafer testing machine as described in claim 2, wherein, step (5) further comprises: respectively screening a standard deviation maximum value X in the standard deviation data set D1 and the standard deviation data set D2 The maximum value of a standard deviation Y, the maximum value of the standard deviation X is the standard deviation of the previous a slope data in the K1 part, the maximum value of the standard deviation Y is the standard deviation of the previous b slope data in the K2 part, That is, the a+1th scatter point in the K1 part and the b+1 th scatter point in the K2 part are used as the segment points to divide all the scatter points in the scatter diagram into three segments, and for each segment the scatter points Points were fitted to the curve to obtain the fitted curve. The automatic calibration method for a wafer testing machine as described in Claim 3, wherein a fitting formula of the curve fitting is y = E + Fx or y = A + B x + Cx ² +D x ³ . The automatic calibration method for wafer testing machine as described in claim item 4, wherein, the fitting formula of a middle section is y = E + Fx ; The fitting formula of other two sections is y =A+B x + Cx ² +D x ³ . The automatic calibration method for wafer testing machine as described in claim 1, wherein, also comprises the following steps: (7) After completing the calibration parameter input operation, repeat the above operations of steps (1)~(6). A test element applying the automatic calibration method for the wafer testing machine according to any one of claims 1 to 6, comprising one or more processors, one or more memories, one or more programs and calibration A data acquisition device; the calibration data acquisition device is used to obtain the calibration data of the wafer testing machine, one or more programs are stored in the memory, and when the program is executed by the processor, the test element executes as The automatic calibration method for the wafer testing machine according to any one of claims 1-6. The test component as described in claim 7 further includes a wafer testing machine calibration parameter setting module connected to the processor; the wafer testing machine calibration parameter setting module is used to set the calibration parameters.

Images