A method for processing small current data in integrated circuit testing

Abstract

The present invention provides a method for processing low-current data in integrated circuit testing, comprising: obtaining a number of low-current data from existing test equipment, fitting a function of current variation over time: y = a × ln (b × t + c) + d; obtaining low-current data to be processed; fitting the low-current data to be processed to obtain a linear function of current variation over time: y = e × t + f and a curve function: y = a1 × ln (b1 × t + c1) + d1; setting a new test equipment reference current value as y BL , according to the formula y=e×t+f, we can get t BL ; There are two intersection points between y=e×t+f and y=a1×ln(b1×t+c1)+d1, and the horizontal coordinate values ​​are t1 and t2. BL The N test time ranges are calculated based on the values ​​of t1 and t2. The N low current data points within the N test time ranges are averaged to obtain the standard low current data. This low current data processing method allows customers to compensate for the low current data of new test equipment when using it, matching the low current data of their existing equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of integrated circuit device testing, and in particular relates to a method for processing low current data in integrated circuit testing. Background Art

[0002] The integrated circuit testing link is an important link that can ensure that the integrated circuit meets the required performance, quality and other parameter requirements. It is an indispensable part of the design, production, manufacturing and application of the entire integrated circuit product, and has become one of the key technologies to ensure the high reliability of integrated circuit products.

[0003] Currently, there are no relevant national, state, or industry standards for testing equipment in the integrated circuit industry. Test equipment manufacturers all use their own standards to calibrate and test their equipment. This results in a certain offset in the data tested by manufacturers and designers when using different test equipment. Users can only process and analyze test data from the same device during use, and are unable to process and analyze test data from different test equipment. Small current data accounts for a very large proportion of the data tested by test equipment, and the accuracy of this data plays a very important role in integrated circuit design, production, manufacturing, and its application. In order to facilitate the use of test equipment by customers, it is necessary to find a simple and accurate method for processing small current data so that the tested small current data matches the test data of the customer's existing test equipment, maintain data consistency, and enable customers to process and analyze data when using different test equipment, thereby improving the test utilization rate of the equipment and optimizing the user's operating experience. Summary of the Invention

[0004] The present invention provides a method for processing low current data in integrated circuit testing, which can compensate for the low current data tested by new testing equipment and make the test data of the new testing equipment consistent with the test data of the customer's existing testing equipment.

[0005] To achieve the object of the above invention, a method for processing small current data in integrated circuit testing provided by a technical solution of the present invention includes: Step S1. Obtain a number of small current data obtained by existing tests of a test device, and fit to obtain an objective function of current varying with time: y = a×ln(b×t + c) + d, where a, b, c, and d are constants; Step S2. Obtain newly measured small current data to be processed by the test device; Step S3. Fit the small current data to be processed to obtain a linear function of current varying with time: y = e×t + f, where e and f are constants; Fit the small current data to be processed to obtain a curve function of current varying with time: y = a1×ln(b1×t + c1) + d1, where a1, b1, c1, and d1 are constants; Step S4. Let the reference current value measured by a reference device be y BL , BL , BL , BL , BL , BL , BL , BL , BL , BL , BL , BL , and inversely deduce t according to the formula y = e×t + f BL ; Step S5. The linear function y = e×t + f and the curve function y = a1×ln(b1×t + c1) + d1 have two intersection points, and record the abscissa values of the two intersection points as t1 and t2, where t1 < t2; If t BL < t1, calculate (t1 - t) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t BL ) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t BL ​​​​​​​​​​​​​​​​​​t2) / Δt, round the calculated value to get n, the number of tests N=2×n, the time range of N tests is (t2n×Δt, t2+n×Δt), the optimal test stabilization time=t2n×Δt, where Δt is the time interval between the collection of two adjacent small current data to be processed; Step S6. Take the average value of the N small current data within the test time range of the N times among the several small current data to be processed, and the average value is the obtained standard small current data. Through the above-mentioned method for processing small current data, when the customer uses the new test equipment, the small current data of the new test equipment can be compensated and matched with the small current data of the customer's existing equipment, that is, the benchmark equipment in this solution, so that the customer can use the data processing and analysis method of the existing equipment when using the new test equipment, and can effectively analyze and compare with the test data of the existing equipment; at the same time, the method for processing small current data can also determine the optimal test stabilization time of the new test equipment during the test process, so that the customer can know when to obtain stable optimal small current data, which greatly improves the test efficiency.

[0006] Furthermore, after obtaining the low current data to be processed in step S2, the method further includes removing unstable low current data at the start of the test, and fitting the retained low current data in step S3. The unstable low current data may be low current data whose data change rate exceeds a preset range. By removing unstable data at the start of the test, the accuracy of low current data processing can be improved.

[0007] After obtaining the low current data to be processed in step S2, the step further includes removing the stable low current data during a period of testing. The stable low current data may be the low current data with a data change rate within a preset range.

[0008] The step S2 further includes calculating the standard deviation of the retained low current data and removing a number of abnormal points. Preferably, the standard deviation of the low current data after removing the abnormal points is no greater than 3σ.

[0009] Furthermore, the method further includes presetting a quantity level of the plurality of low current data in step S1, and when obtaining the plurality of low current data obtained by the test equipment, the amount of the plurality of low current data is within the range of the quantity level. By obtaining a sufficient amount of low current data from the test equipment, the accuracy of fitting the objective function can be improved.

[0010] The objective function in step S1 and the curve function in step S3 are both fitted using the least squares method.

[0011] Preferably, after several fittings, the function with the minimum residual error is taken as the objective function in step S1 and the curve function in step S3.

[0012] Compared with the prior art, the main beneficial effects of the present invention include: according to the processing method of small current data in integrated circuit testing provided by the present invention, the data compensation and matching of the test equipment are carried out, so that customers can also realize data processing and analysis when using test equipment from different manufacturers, which is convenient for customers to use products and helps to reduce customers' production and R&D costs; at the same time, the optimal test stabilization time in the test process determined by this method can help users know when the measured data is the optimal small current data, greatly improving the test efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the specific embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 This is a flow chart of a method for processing low current data in integrated circuit testing according to a first specific embodiment of the present invention. DETAILED DESCRIPTION

[0015] The following is a clear and complete description of the technical solutions in the specific embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0016] A flowchart of a method for processing low current data in integrated circuit testing provided by the first embodiment of the present invention is shown in FIG. Figure 1 Please refer to Figure 1 , the method for processing the low current data specifically includes:

[0017] Step S1. Obtain a number of small current data obtained from existing tests of the test equipment, and fit the target function of the current changing with time: y=a×ln(b×t+c)+d, where a, b, c, and d are constants.

[0018] Specifically, the several small current data obtained by the test equipment refers to the data retained by the previous test of the test equipment or other test equipment of the same model as the test equipment. For example, it can be the data measured by the test equipment before leaving the factory, or it can be the data measured by the customer during the previous use. The amount of the several small current data in this step can be preset by its corresponding quantity level, such as ten thousand. Then, the several small current data obtained in the execution of step S1 must be within the quantity range of 10000-99999. For example, 50,000 data of small current data obtained by the previous test of the test equipment are extracted. In actual use, the customer can set the quantity level according to actual needs. After obtaining a sufficient amount of small current data, fitting is performed according to the corresponding relationship between current value and time to obtain the values ​​of coefficients a, b, c, and d in the objective function y=a×ln(b×t+c)+d, where y is the current value and t is the time.

[0019] Step S2: Obtain low current data to be processed that is newly measured by the test equipment.

[0020] Specifically, the small current data to be processed is the small current data newly obtained by the customer's test equipment, which is generally the small current data tested for a single product. Taking into account that the test equipment may be unstable at the beginning of the test, in this embodiment, when obtaining the small current data to be processed, the unstable small current data at the beginning of the test and the stable small current data during the test are removed. For example, the unstable small current data can be processed by setting a data change rate R. When the data change rate exceeds R, the corresponding small current data is recorded as unstable small current data; the stable small current data can also be processed by setting a data change rate M as mentioned above, and removing the small current data for a period of time when the data change rate is less than M. In other embodiments, the customer can set the values ​​of the data change rates R and M according to actual conditions, and can also choose whether to remove part of the small current data and which part of the small current data to remove.

[0021] To ensure more accurate low-current data, this step also includes calculating the standard deviation of the retained low-current data and removing several outliers, ensuring that the standard deviation of the retained low-current data is no greater than 3σ. In other embodiments, the customer may determine the standard deviation range based on actual circumstances, or may not perform this step.

[0022] In this embodiment, after removing unstable and partially stable low current data and several abnormal points, the operation of step S3 is continued on the remaining low current data.

[0023] Step S3. Fit the retained small current data to obtain a linear function of current varying with time: y = e×t + f, where the coefficients e and f are constant values determined by current and time;

[0024] Fit the retained small current data to obtain a curve function of current varying with time: y = a1×ln(b1×t + c1) + d1, where the coefficients a1, b1, c1, and d1 are also constant values determined by current and time.

[0025] In this embodiment, for the curve fitting of the objective function y = a×ln(b×t + c) + d in Step S1 and the curve function a1×ln(b1×t + c1) + d1 in Step S3, the least squares method is used for fitting. After several fittings, the function with the smallest sum of residuals is taken as the final function.

[0026] Step S4. Let the reference current value measured by the reference device be y BL , and t is obtained by back - calculating according to the formula y = e×t + f BL ; where the reference device is the old test device used by the customer before, such as the test device produced by other manufacturers; the reference current value y BL is the inherent data of the reference device.

[0027] The linear function y = e×t + f and the curve function y = a1×ln(b1×t + c1) + d1 have two intersection points. Denote the abscissa values of the two intersection points as t1 and t2, where t1 < t2;

[0028] If t BL < t1, calculate (t1 - t BL ) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t1 - n×Δt, t1 + n×Δt), and the best test stable time = t1 - n×Δt, where Δt is the time interval between the acquisitions of two adjacent small current data to be processed;

[0029] If t1 < t BL < t2, calculate (t BL - t1) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t BL - n×Δt, t BL + n×Δt), and the best test stable time = t BL - n×Δt;

[0030] If t BL > t2, calculate (t BL(t2) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t2n×Δt, t2 + n×Δt), the optimal test stable time = t2n×Δt, where Δt is the time interval between the acquisitions of two adjacent small current data to be processed.

[0031] It should be noted that in other embodiments, when t BL < t1, the time range of N tests can also be (t BL n×Δt, t BL + n×Δt), the optimal test stable time = t BL n×Δt. This solution saves some time compared to the solution of this embodiment. The reason is that in this embodiment, t BL < t1. If the time range of N tests is defined by t1, the time axis will be longer than that defined by t BL to define the time range of N measurements. The customer can determine which solution to use according to actual needs.

[0032] Step S6. Take the average of the N small current data within the test time range of N among several small current data to be processed, and this average value is the obtained standard small current data.

[0033] The above has introduced in detail a method for processing small current data in integrated circuit testing provided by the present invention. Specific examples are used in this article to elaborate on the structure and working principle of the present invention. The description of the above embodiments is only used to help understand the method and core idea of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for processing low current data in integrated circuit testing, characterized in that: include: Step S1. Obtain a number of small current data obtained from existing tests of the test equipment, and fit the target function of current variation over time: y = a × ln (b × t + c) + d, where a, b, c, and d are constants, y is the current value, and t is time; Step S2. Obtaining the low current data to be processed that is newly measured by the test equipment; Step S3. Fitting the low current data to be processed to obtain a linear function of the current changing with time: y=e×t+f, where e and f are constants; Fitting the low current data to be processed to obtain a curve function of current change over time: y = a1 × ln (b1 × t + c1) + d1, where a1, b1, c1, and d1 are constants; Step S4. Assume that the reference current value tested by the benchmark device is y BL , according to the formula y=e×t+f, we can get t BL ; Step S5. The linear function y=e×t+f and the curve function y=a1×ln(b1×t+c1)+d1 have two intersection points. The horizontal coordinates of the two intersection points are t1 and t2, where t1 <t2; If t BL <t1, calculate (t1 - t BL ) / Δt, round the calculated value to obtain n, the number of tests N = 2 × n, the time range of N tests is (t BL n × Δt, t BL + n × Δt), the optimal test stabilization time = t BL n × Δt, or, the time range of N tests is (t1 - n × Δt, t1 + n × Δt), the optimal test stabilization time = t1 - n × Δt; where, Δt is the time interval between the acquisitions of two adjacent small current data to be processed; If t1 < t BL < t2, calculate (t BL - t1) / Δt, round the calculated value to obtain n, the number of tests N = 2×n, the time range of N tests is (t BL + n×Δt, t BL + n×Δt), the optimal test stabilization time = t BL + n×Δt; If t BL >t2, calculate (t BL t2) / Δt, round the calculated value to get n, the number of tests N = 2×n, the time range of N tests is (t2n×Δt, t2+n×Δt), the optimal test stabilization time = t2n×Δt, where Δt is the time interval between two adjacent low current data to be processed; Step S6: averaging N small current data within the N test time ranges among the plurality of small current data to be processed, where the average value is the obtained standard small current data.

2. The method for processing low current data in integrated circuit testing according to claim 1, characterized in that: After obtaining the low current data to be processed in step S2, the unstable low current data at the beginning of the test is removed, and the retained low current data is fitted in step S3.

3. The method for processing low current data in integrated circuit testing according to claim 2, characterized in that: After obtaining the low current data to be processed in step S2, the step also includes removing the low current data that is stable for a period of time.

4. The method for processing low current data in integrated circuit testing according to claim 2, characterized in that: The unstable low current data is low current data whose data change rate exceeds a preset range.

5. The method for processing low current data in integrated circuit testing according to claim 3, characterized in that: The stable low current data is low current data whose data change rate is within a preset range.

6. The method for processing low current data in integrated circuit testing according to any one of claims 2 to 5, characterized in that: The step S2 also includes calculating the standard deviation of the retained low current data and removing several abnormal points.

7. The method for processing low current data in integrated circuit testing according to claim 6, characterized in that: The standard deviation of the low current data after removing several abnormal points is no greater than 3σ.

8. The method for processing low current data in integrated circuit testing according to claim 1, characterized in that: The quantity level of the plurality of small current data in step S1 is preset. When the plurality of small current data obtained by the test equipment is acquired, the data amount of the plurality of small current data is within the range of the quantity level.

9. The method for processing low current data in integrated circuit testing according to claim 1, characterized in that: The objective function in step S1 and the curve function in step S3 are both fitted using the least squares method.

10. The method for processing low current data in integrated circuit testing according to claim 9, characterized in that: After several fittings, the function with the minimum residual error is taken as the target function in step S1 and the curve function in step S3.

Images