Wafer test-based probe suite multifunctional test method and system

Abstract

The invention relates to a probe suite multifunctional test method and system based on wafer test, and the method comprises the steps: obtaining the to-be-tested data of a wafer and the parameter information of a probe and a probe card; and through a preset intelligent algorithm, the contact state of the probe tip and the wafer is judged according to the working parameters, and the contact pressure and the probe impedance are dynamically adjusted. Meanwhile, environmental parameters including temperature, humidity and electromagnetic interference intensity in a test area are obtained, and when the environmental parameters exceed a preset range, the environment is automatically optimized. And detecting wafer characteristics based on the to-be-tested data, preprocessing the initial test data to obtain wafer test data, and performing real-time analysis and anomaly detection. And finally, according to an analysis detection result, whether the probe is adjusted or self-repaired is judged, and meanwhile, the test parameters and the working mode of the probe card are dynamically adjusted, so that high-precision and high-stability wafer test is realized, the performance of the probe and a test system is guaranteed, and the accuracy and the reliability of semiconductor test are improved.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor testing technology, and particularly to a multi-functional testing method and system for a probe kit based on wafer testing. Background Art

[0002] In the field of semiconductor testing technology, wafer testing is a key link to ensure the quality and performance of semiconductor devices. With the continuous development of semiconductor technology, the integration degree of wafers is getting higher and higher, which puts forward higher requirements for the accuracy, efficiency and multi-functionality of testing. Traditional testing methods for probe kits based on wafer testing have many limitations. On the one hand, during the contact process between the probe tip and the wafer, it is difficult to monitor and adjust key working parameters, as well as the pressure and temperature of the probe operation in real time and accurately. This may lead to problems such as poor contact and inaccurate test signals, thus affecting the reliability of test results. On the other hand, the monitoring and repair mechanism for probe loss is not perfect enough, and it is impossible to repair in time according to the actual loss situation of the probe, reducing the service life of the probe and the stability of testing, and it is difficult to meet the requirements of modern semiconductor testing for high precision and high reliability. Summary of the Invention

[0003] The main object of the present invention is to provide a multi-functional testing method and system for a probe kit based on wafer testing, so as to achieve the purpose of realizing high-precision and high-stability wafer testing, ensuring the performance of the probe and the testing system, and improving the accuracy and reliability of semiconductor testing.

[0004] To achieve the above object, the present invention provides a multi-functional testing method for a probe kit based on wafer testing, including the following steps:

[0005] Obtain the data to be tested for the wafer, and obtain the parameter information of the probe and the probe card carried in the data to be tested, where the parameter information includes the material, size and tip shape of the probe, as well as the substrate material, wiring method and impedance matching circuit of the probe card;

[0006] Continuously monitor the working parameters during the contact process between the probe tip and the wafer, where the working parameters include resistance parameters, contact status, and the frequency and amplitude of the test signal, and at the same time monitor the pressure and temperature of the current probe operation;

[0007] Based on the working parameters, judge whether the current contact status between the probe tip and the wafer meets the preset conditions through a preset intelligent algorithm, and dynamically adjust the contact pressure between the probe tip and the wafer and the probe impedance;

[0008] Obtain the environmental parameters of the test area corresponding to the probe card, where the environmental parameters include temperature, humidity and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, execute the probe card environmental automatic optimization program;

[0009] Based on the data to be tested, detect the characteristics of the wafer, obtain the initial test data including electrical performance, signal integrity, and functional testing, and preprocess the initial wafer test data to obtain wafer test data;

[0010] Perform real-time analysis on the wafer test data, and perform anomaly detection based on the above analysis to obtain an analysis and detection result;

[0011] Based on the analysis and detection result, determine whether it is necessary to adjust the probe or the self-repair state. If necessary, perform the corresponding operation. At the same time, according to the analysis and detection result, dynamically adjust the test parameters and working mode of the probe card. The test parameters include but are not limited to test voltage, current, and frequency, and the working mode includes scanning mode and test sequence.

[0012] Further, after the step of continuously monitoring the working parameters during the contact process between the probe tip and the wafer, the working parameters include resistance parameters, contact state, and frequency and amplitude of the test signal, and at the same time monitoring the current working pressure and temperature of the probe, the following steps are also included:

[0013] Obtain the resistance parameters during the contact process between the probe tip and the wafer, and the resistance parameters include resistance value and resistance change rate;

[0014] Based on the resistance parameters, evaluate the current loss value of the probe tip through a pre-established resistance and wear relationship model; the resistance and wear relationship model is obtained by analyzing the wear amount of the probe tip at each stage of a large number of probes and the corresponding resistance change data of the contact between the probe tip and the wafer, and obtaining the mapping relationship between the resistance and the release amount of the probe repair material;

[0015] According to the preset loss threshold of the probe and the loss value, trigger the probe self-repair mechanism, and determine the probe repair quantity through the resistance and wear relationship model;

[0016] Release the quantitative repair material to the probe loss area, and at the same time detect the temperature change data of the probe loss area and adjust the catalyst component in the repair material.

[0017] Further, the step of judging whether the current contact state between the probe tip and the wafer meets the preset conditions based on the working parameters and dynamically adjusting the contact pressure between the probe tip and the wafer and the probe impedance through a preset intelligent algorithm includes:

[0018] Based on the currently obtained contact state between the probe tip and the wafer, the frequency and amplitude of the test signal, and the current working pressure and temperature of the probe, perform analysis through a preset intelligent algorithm;

[0019] Based on the model analysis results, determine whether the contact state between the current probe tip and the wafer meets the preset conditions. When it is determined that the contact state between the tip and the wafer does not meet the preset conditions, adjust the contact pressure of the probe tip through the intelligent algorithm until the tip pressure reaches the optimal state;

[0020] Based on the model analysis results, determine whether the current probe impedance matches the test signal. When the probe impedance does not match the test signal, generate impedance matching parameters and adjust the probe impedance until the impedance of the probe matches the test signal.

[0021] Further, the intelligent algorithm is obtained by training based on the historical data collected on the contact state between the probe tip and the wafer, the frequency amplitude of the test signal, the working pressure, and the temperature, and by selecting a machine learning algorithm framework.

[0022] Further, obtain the environmental parameters of the corresponding test area of the probe card. The environmental parameters include temperature, humidity, and electromagnetic interference intensity. After performing the steps of the probe card environment automatic optimization program if it is detected that the environmental parameters exceed the set range, it further includes:

[0023] Obtain the environmental parameters of the corresponding test area of the probe card. The environmental parameters include temperature, humidity, and electromagnetic interference intensity data;

[0024] According to the preset environmental parameter threshold and the current environmental parameters, determine the deviation degree of the environmental parameters;

[0025] According to the deviation degree of the environmental parameters, start the corresponding optimization device. The optimization device includes a temperature control device, a humidity adjustment device, and / or an electromagnetic shielding device;

[0026] According to the deviation degree of the environmental parameters, obtain the control parameters of the corresponding optimization device;

[0027] Send the control parameters to the corresponding optimization device, start the temperature control, humidity adjustment, and / or electromagnetic shielding process, continuously obtain the real-time environmental parameters, and perform dynamic optimization and adjustment on the test environment until the environmental parameters of the probe card return to within the preset environmental parameter threshold.

[0028] Further, use a digital filtering algorithm to remove noise from the initial wafer test data, suppress interference signals through an adaptive interference cancellation technique, and adjust the algorithm parameters according to the spectral characteristics of the signal to obtain the wafer test data.

[0029] Further, the steps of performing real-time analysis on the wafer test data and performing anomaly detection based on the above analysis to obtain the analysis and detection results include:

[0030] Classify the wafer test data according to electrical performance, signal integrity, and functional testing to obtain various data characteristics, including electrical performance data, analyzing parameter changes of resistance, capacitance, and inductance, signal transmission delay and signal distortion rate of signal integrity data, and logical function state changes and functional response time of functional testing data;

[0031] Compare the obtained various data characteristics with a pre-constructed benchmark model. The benchmark model includes the standard value range, fluctuation range, and normal fluctuation law of each characteristic parameter. If the data characteristics deviate from the set threshold of the benchmark model, the corresponding data is determined as suspected abnormal data;

[0032] Train a convolutional neural network model with historical normal data, and analyze the abnormal data through the trained convolutional neural network model to generate an analysis and detection result. The analysis and detection result includes whether the abnormality is determined, the type of abnormality, and the degree of influence of the abnormality on the test result.

[0033] Furthermore, if the analysis and detection result shows that the deviation of the test signal exceeds the preset range and this deviation is associated with the wear or position change of the probe, it is determined that the probe needs to be adaptively adjusted or self-repaired.

[0034] Furthermore, when the analysis and detection result indicates that the electrical performance data of the wafer is abnormal, according to the preset electrical performance and parameter adjustment rules, gradually adjust the test voltage and current in a gradient descent manner. After each adjustment, re-obtain the wafer test data until the wafer test data meets the expected range.

[0035] The present invention also provides a multifunctional test system for a probe kit based on wafer testing, including:

[0036] A data acquisition unit for acquiring the data to be tested for the wafer and obtaining the parameter information of the probe and the probe card carried in the data to be tested;

[0037] A monitoring unit for continuously monitoring the working parameters during the contact process between the probe tip and the wafer. The working parameters include resistance parameters, contact status, and the frequency and amplitude of the test signal, and simultaneously monitoring the pressure and temperature at which the probe is currently working;

[0038] A probe adjustment unit for judging whether the contact state between the current probe tip and the wafer meets the preset conditions according to the working parameters through a preset intelligent algorithm, and dynamically adjusting the contact pressure between the probe tip and the wafer and the probe impedance;

[0039] An environment optimization unit for obtaining the environmental parameters of the test area corresponding to the probe card. The environmental parameters include temperature, humidity, and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, the probe card environment automatic optimization program is executed;

[0040] A data processing unit for detecting the wafer characteristics according to the data to be tested, obtaining the initial test data including electrical performance, signal integrity, and functional test aspects, and preprocessing the wafer initial test data to obtain the wafer test data;

[0041] A data analysis unit for performing real-time analysis on the wafer test data, performing anomaly detection based on the above analysis, and obtaining the analysis and detection results;

[0042] An adaptive adjustment unit for judging whether it is necessary to adjust the probe or self-repair the state based on the analysis and detection results. If necessary, the corresponding operations are executed. At the same time, according to the analysis and detection results, the test parameters and working modes of the probe card are dynamically adjusted.

[0043] The present invention also provides a computer device, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the steps of the above-mentioned multi-functional test method of the probe kit based on wafer testing are implemented.

[0044] The multi-functional test method and system of the probe kit based on wafer testing provided by the present invention have the following beneficial effects:

[0045] The working parameters of the contact between the probe and the wafer are monitored in real time, and the contact pressure and impedance are dynamically adjusted through intelligent algorithms to ensure the accurate transmission of test signals, greatly improving the test accuracy. The probe loss is evaluated based on the resistance and wear relationship model, and automatic triggering and accurate quantitative repair are performed to extend the service life of the probe, reduce costs, and ensure test stability. By real-time monitoring the environmental parameters of the test area, when the range is exceeded, the optimization program is automatically started to regulate the temperature, humidity, and shield electromagnetic interference, reducing the influence of the environment on the test. Description of the Drawings

[0046] Figure 1 It is a flow chart of the multi-functional test method of the probe kit based on wafer testing in an embodiment of the present invention;

[0047] Figure 2 It is a structural block diagram of the multi-functional test system of the probe kit based on wafer testing in an embodiment of the present invention;

[0048] Figure 3 It is a structural schematic block diagram of the computer device in an embodiment of the present invention.

[0049] The realization, functional features and advantages of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. Detailed implementation manners

[0050] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0051] Referring to Figure 1 , which is a schematic flow chart of a multi-functional test method for a probe kit based on wafer testing proposed by the present invention, including the following steps:

[0052] S1. Obtain the data to be tested for the wafer, and obtain the parameter information of the probes and the probe card carried in the data to be tested. The parameter information includes the material, size and tip shape of the probes, as well as the substrate material, wiring method and impedance matching circuit of the probe card.

[0053] S2. Continuously monitor the working parameters during the contact process between the probe tips and the wafer. The working parameters include resistance parameters, contact states, and the frequency and amplitude of the test signals. At the same time, monitor the pressure and temperature at which the probes are currently working.

[0054] S3. Based on the working parameters, use a preset intelligent algorithm to determine whether the current contact state between the probe tips and the wafer meets the preset conditions, and dynamically adjust the contact pressure between the probe tips and the wafer and the probe impedance.

[0055] S4. Obtain the environmental parameters of the corresponding test area of the probe card. The environmental parameters include temperature, humidity and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, execute the automatic environmental optimization program for the probe card.

[0056] S5. Based on the data to be tested, detect the wafer characteristics, obtain the initial test data including electrical performance, signal integrity and functional testing, and preprocess the wafer initial test data to obtain the wafer test data.

[0057] S6. Perform real-time analysis on the wafer test data, and perform anomaly detection based on the above analysis to obtain the analysis and detection results.

[0058] S7. Based on the analysis and detection results, determine whether it is necessary to adjust or self-repair the probes. If necessary, execute the corresponding operations. At the same time, dynamically adjust the test parameters and working modes of the probe card according to the analysis and detection results. The test parameters include but are not limited to test voltage, current and frequency, and the working modes include scanning methods and test sequences.

[0059] As described in the above step S1, obtain the data to be tested for the wafer, and obtain the parameter information of the probes and the probe card carried in the data to be tested. Among them, the parameter information includes the material, size, and tip shape of the probes, as well as the substrate material, wiring method, and impedance matching circuit of the probe card. The material of the probes directly affects their key performance such as conductivity, hardness, and wear resistance. For example, tungsten probes have high hardness and good conductivity, and are suitable for test scenarios that require high mechanical strength and stable electrical connections; while beryllium copper alloy probes have a better balance between elasticity and conductivity. Obtaining the probe material data helps to evaluate their performance and potential loss during the test. The size of the probes includes length, diameter, etc. Different wafer test site spacings and layouts require the probes to have corresponding appropriate sizes. Precise size control is crucial to ensure that the probes can accurately contact the test points on the wafer. Size deviation may cause problems such as poor contact or short circuit. The tip shape determines the contact method and contact area between the probes and the wafer surface. Common tip shapes include conical, spherical, planar, etc. Different shapes have differences in contact resistance, signal transmission efficiency, and the degree of damage to the wafer surface. For example, conical tips can provide a smaller contact area, which is beneficial to improving the test resolution, but may cause certain scratches on the wafer surface; spherical tips can disperse the pressure to a certain extent and reduce the damage to the wafer.

[0060] As the basic structure for supporting and connecting the probes, the material characteristics of the probe card substrate have an important impact on the performance of the entire test system. Commonly used substrate materials include ceramics, epoxy glass fiber (such as FR-4), etc. Ceramic substrates have good high-temperature resistance, insulation performance, and dimensional stability, and are suitable for high-precision and high-reliability tests; while FR-4 substrates have lower costs and are widely used in some occasions with relatively less demanding performance requirements. Understanding the substrate material helps to evaluate the overall performance and applicable scenarios of the probe card. The wiring on the probe card is responsible for connecting the probes to the test instruments to achieve signal transmission. Different wiring methods will affect the signal transmission quality, such as signal attenuation, crosstalk, etc. A reasonable wiring design can optimize the signal path, reduce signal interference, and improve the test accuracy. For example, adopting hierarchical wiring, reasonable line spacing design, and shielding measures, etc., can effectively reduce signal crosstalk and ensure the integrity of the test signal. In the process of high-speed signal transmission, impedance matching is crucial. Impedance mismatch will cause signal reflection, distort the signal, and affect the accuracy of the test results. The impedance matching circuit on the probe card is used to adjust the impedance of the signal transmission line to match the input and output impedances of the test instruments and the wafer, ensuring that the signal can be transmitted efficiently and accurately. For example, by adding appropriate resistors, capacitors, inductors, etc. in the wiring to form a matching network to achieve impedance matching.

[0061] As described in the above step S2, continuously monitor the working parameters during the process of the probe tip contacting the wafer. The working parameters include resistance parameters, contact status, and the frequency and amplitude of the test signal. At the same time, monitor the current working pressure and temperature of the probe. Among them, the resistance value and its change rate can reflect the good degree of contact between the probe and the wafer and the probe loss. For example, if the resistance suddenly increases, the contact may be poor; if it changes continuously, it may imply probe wear. Judging whether the probe and the wafer are in normal contact, partial contact or non-contact directly affects the test signal transmission and the accuracy of the result. The frequency and amplitude are key characteristics. Different test requirements have different signals. Monitoring ensures compliance with the test standards and avoids result deviation caused by abnormal signals. The pressure affects the contact reliability and stability. If the pressure is inappropriate, the contact may be poor or the wafer may be damaged. The temperature change affects the performance of the probe and the wafer. For example, the conductivity of the material changes. Monitoring the temperature and controlling the temperature ensure the stability of the test environment.

[0062] As described in the above step S3, based on the working parameters, through a pre-set intelligent algorithm, judge whether the current contact status between the probe tip and the wafer meets the preset conditions, and dynamically adjust the contact pressure between the probe tip and the wafer and the probe impedance. Use the pre-set intelligent algorithm, and this algorithm is trained using the neural network algorithm framework. During the training process, collect a large amount of historical data on the contact status between the probe tip and the wafer, the frequency and amplitude of the test signal, the working pressure and temperature as training samples. Divide these data into input features (such as contact status, signal frequency and amplitude, pressure and temperature, etc.) and the corresponding expected outputs (such as whether the contact preset conditions are met, whether the pressure and impedance need to be adjusted, etc.). Through multiple iterative trainings, adjust the weights and biases of the neural network so that the model can accurately predict reasonable output results according to the input features. After the training is completed, this algorithm can comprehensively analyze the current obtained working parameters such as resistance, contact status, signal frequency and amplitude, pressure and temperature, etc., and thereby judge whether the contact status between the probe tip and the wafer meets the pre-set ideal conditions.

[0063] If the algorithm judges that the contact status does not meet the preset conditions, it will automatically adjust the contact pressure of the probe tip through the intelligent algorithm until the optimal state is reached, ensuring good contact between the probe and the wafer and guaranteeing the stable transmission of the test signal. At the same time, the intelligent algorithm will also judge whether the probe impedance matches the test signal. If they do not match, the algorithm generates impedance matching parameters to adjust the probe impedance to achieve their matching, avoiding signal reflection or attenuation and improving the test accuracy.

[0064] As described in step S4 above, obtain the environmental parameters of the corresponding test area of the probe card. The environmental parameters include temperature, humidity, and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, execute the automatic probe card environment optimization program. Use high-precision temperature sensors, humidity sensors, and electromagnetic interference detection equipment to accurately measure the temperature, humidity, and electromagnetic interference intensity in the test area of the probe card. These parameters will significantly affect the test results, and the data obtained should be timely and accurate.

[0065] Pre-determine the normal range of environmental parameters according to the test standards and the characteristics of the wafer and the probe card. For example, for the wafer and probe card in this test, the set temperature range is 20°C - 30°C, the humidity range is 30% - 60% RH, and the electromagnetic interference intensity is less than 50 μT. Compare the obtained environmental parameters with this set range to determine whether the parameters exceed the limit and decide whether to start the optimization program.

[0066] Once it is detected that the parameters exceed the range, the system automatically triggers the probe card environment optimization program. Determine the deviation degree according to the specific parameters that exceed the range. For example, if the temperature exceeds the range by 2°C, the temperature deviation degree is defined as moderate. Different deviation degrees correspond to different optimization strategies. For temperature deviation, if it is a mild deviation, the cooling or heating power can be fine-tuned through the temperature control device; if it is a moderate deviation, the power is adjusted significantly; if it is a severe deviation, it may be necessary to start the standby temperature control equipment. The same applies to the humidity adjustment device and the electromagnetic shielding device. By calculating the deviation degree, obtain the control parameters of the corresponding optimization device, such as the power adjustment value of the temperature control device, the humidification or dehumidification amount of the humidity adjustment device, and the shielding intensity adjustment value of the electromagnetic shielding device. Send these control parameters to the corresponding optimization device, start the temperature control, humidity adjustment, and / or electromagnetic shielding process, continuously obtain the real-time environmental parameters, and dynamically optimize and adjust the test environment until the environmental parameters of the probe card are restored within the preset environmental parameter threshold.

[0067] As described in step S5 above, based on the data to be tested, detect the wafer characteristics, obtain the initial test data including electrical performance, signal integrity, and functional test, and preprocess the wafer initial test data to obtain the wafer test data. According to the data to be tested, conduct electrical performance, signal integrity, and functional tests on the wafer to obtain the initial test data. For example, detect the resistance and capacitance values in electrical performance; the signal transmission delay in signal integrity; whether the logic function is normal in functional test, etc., to comprehensively understand the wafer characteristics.

[0068] The digital filtering algorithm is used to remove noise from the acquired initial test data. In this embodiment, a Butterworth low-pass filter is adopted, and the cut-off frequency is set according to the frequency characteristics of the signal to filter out the noise signals higher than the cut-off frequency. At the same time, the adaptive interference cancellation technology is used to suppress the interference signals. This technology compares the reference signal (a signal related to but obtainable from the interference signal) with the test signal containing interference, and adjusts the cancellation coefficient through an adaptive algorithm to make the reference signal and the interference signal cancel each other as much as possible. And the algorithm parameters are adjusted according to the spectral characteristics of the signal. For example, if spectral analysis finds that the signal is mainly concentrated in a certain frequency band, the filtering and interference cancellation parameters in this frequency band are optimized accordingly. Finally, the wafer test data for subsequent analysis is obtained to ensure the reliability of the analysis results based on this data.

[0069] As described in step S6 above, the wafer test data is analyzed in real time, and anomaly detection is performed based on the above analysis to obtain the analysis and detection results. During the test process, the wafer test data is continuously processed to quickly identify the data characteristics. For example, analyze the changes in resistance and capacitance in the electrical performance data, and the signal transmission delay in signal integrity, etc.

[0070] A benchmark model is pre-constructed, which is constructed by collecting data of a large number of wafers of the same model under normal test conditions. Statistical analysis is performed on these data to determine the standard value range, fluctuation range, and normal fluctuation law of each characteristic parameter. For example, for the resistance in the electrical performance data, after testing 1000 normal wafer samples, the standard value range is determined to be 90Ω - 110Ω, the fluctuation range is ±5Ω, and the normal fluctuation law conforms to a certain statistical distribution (such as normal distribution).

[0071] The data characteristics obtained from the real-time analysis are compared with the pre-constructed benchmark model. If the data deviates from the threshold set by the benchmark model, it is marked as a suspected anomaly. Then, the trained convolutional neural network model is used for in-depth analysis. This convolutional neural network model is trained using a large number of labeled normal and abnormal data samples, and the network parameters are optimized through multiple iterations to enable it to accurately identify different types of abnormal data. The suspected abnormal data is analyzed through this model to determine whether it is truly abnormal, the type of anomaly, and the degree of influence on the test results, and finally the analysis and detection results are obtained.

[0072] As described in step S7 above, based on the analysis and detection results, it is determined whether the probe needs to be adjusted or self-repaired. If so, the corresponding operations are performed. At the same time, according to the analysis and detection results, the test parameters and working modes of the probe card are dynamically adjusted. The test parameters include but are not limited to test voltage, current, and frequency, and the working modes include scanning methods and test sequences. According to the analysis and detection results of step S6, it is determined whether the probe causes test anomalies due to factors such as wear and position changes. If the anomaly is related to the probe, the adjustment or self-repair operation is started. The preset loss threshold of the probe is determined in advance through a large number of experiments. For example, after 500 simulation tests on probes of the same model, the changes in the test performance of the probes under different wear degrees are statistically analyzed. It is determined that when the resistance change caused by probe wear exceeds 15%, it will have a significant impact on the test results. Therefore, 15% is used as the preset loss threshold.

[0073] When it is detected that the probe loss value reaches this threshold, the self-repair mechanism is triggered. The quantitative repair of the probe is determined through the resistance-wear relationship model. This model analyzes in detail the probes collected at multiple different wear stages, measures the wear amount of the probe and the corresponding resistance change at each stage, and establishes a functional relationship between the resistance and the wear amount of the probe. According to the currently detected resistance change, substituting it into this functional relationship calculates the amount of material that needs to be repaired, which is the quantitative repair of the probe.

[0074] The quantitative repair material is accurately released to the probe loss area through microelectromechanical system (MEMS) technology. At the same time, a micro temperature sensor is used to detect the temperature change data of the probe loss area. A temperature-catalyst component adjustment relationship table is established in advance. For example, for every 1°C increase in temperature, the proportion of a certain component in the catalyst increases by 0.5%. According to the detected temperature change, the catalyst component in the repair material is searched and adjusted from the relationship table to ensure that the repair material adheres to the probe loss area quickly and evenly, completing the repair.

[0075] At the same time, according to the analysis and detection results, the test parameters and working modes of the probe card are flexibly changed. Parameters such as test voltage, current, and frequency, as well as working modes such as scanning methods and test sequences, can all be adjusted as needed, so as to optimize the test process and ensure the accuracy of subsequent test data. For example, if the analysis and detection results show that the electrical performance data of the wafer is abnormal, according to the preset electrical performance and parameter adjustment rules, the test voltage and current are gradually adjusted in a gradient descent manner. The adjustment amplitude each time is determined according to the severity of the anomaly. For example, if the anomaly is less severe, the voltage is adjusted by 0.05V each time; if the anomaly is more severe, the voltage is adjusted by 0.1V each time. After each adjustment, the wafer test data is re-obtained until the wafer test data meets the expected range. For the working mode, if the current scanning method results in low test efficiency, it can be adjusted to a more efficient scanning method according to the analysis results, such as changing from line-by-line scanning to spiral scanning to improve the test efficiency.

[0076] In one embodiment, a wafer used for manufacturing high-end smartphone chips is tested. First, the data to be tested for the wafer is obtained. These data are provided by the front-end process of the wafer manufacturing production line and stored in the standard XML format. The parameter information of the probes and the probe card carried therein is as follows:

[0077] Probes: The material is tungsten, the length is 5 mm, the diameter is 0.1 mm, and the tip shape is conical.

[0078] Probe card: The substrate material is ceramic, the wiring method uses multi-layer PCB hierarchical wiring, and the impedance matching circuit is designed as a 50Ω matching network.

[0079] During the contact process between the probe tip and the wafer, the working parameters are continuously monitored by a high-precision sensor. For example, during a certain test time period, the real-time monitored resistance parameter is a resistance value of 0.5Ω, and the resistance change rate is 0.01Ω per 10 seconds; the contact state is good contact; the test signal frequency is 1 GHz, and the amplitude is 0.5V; at the same time, the current working pressure of the probe is monitored to be 50 mN, and the temperature is 25°C.

[0080] The environmental monitoring equipment is used to obtain the environmental parameters of the test area corresponding to the probe card. In this test scenario, the initially detected temperature is 23°C, the humidity is 40%RH, and the electromagnetic interference intensity is 10 μT. The preset environmental parameter range is a temperature of 20 - 30°C, a humidity of 30% - 60%RH, and an electromagnetic interference intensity less than 50 μT. The current environmental parameters are all within the set range.

[0081] Based on the currently obtained working parameters, a pre-set intelligent algorithm is used for analysis. This intelligent algorithm is trained based on a large amount of historical data using a machine learning algorithm framework (such as a neural network). By inputting data such as the current contact state, test signal frequency and amplitude, working pressure and temperature into the trained model, it is judged whether the current contact state between the probe tip and the wafer meets the preset conditions.

[0082] After analysis by the intelligent algorithm, it is judged that the contact pressure between the current tip and the wafer is slightly low, resulting in a slightly high contact resistance and not meeting the preset conditions. At this time, the intelligent algorithm adjusts the contact pressure of the probe tip according to the model analysis result.

[0083]

[0084] P opt represents the optimal contact pressure, P 0 represents the initial pressure, ΔP iLet \(\Delta P_i\) denote the \(i\)-th pressure adjustment amount, and \(n\) denote the number of adjustments. This formula describes the optimization process of the probe contact pressure. The pressure is gradually adjusted by increasing 5 mN each time. After 3 adjustments, the pressure reaches 65 mN, at which time the contact resistance drops to 0.4 \(\Omega\), meeting the preset conditions, and the tip pressure reaches the optimal state.

[0085] While adjusting the contact pressure, the intelligent algorithm determines whether the current probe impedance matches the test signal. After analysis, the current probe impedance is 60 \(\Omega\), which does not match the required 50 \(\Omega\) of the test signal. At this time, the intelligent algorithm generates impedance matching parameters by adjusting the impedance matching circuit elements (such as trim capacitors and inductors) on the probe card.

[0086]

[0087] Z probe represents the probe impedance, \(Z_0\) 0 represents the initial impedance, \(\Delta Z_j\) j denotes the \(j\)-th impedance adjustment amount, and \(m\) denotes the number of adjustments. This formula describes the adjustment process of the probe impedance. After 5 fine-tuning operations, the probe impedance is adjusted to 50 \(\Omega\), achieving impedance matching with the test signal.

[0088] Obtain the resistance parameters during the contact process between the probe tip and the wafer, and based on the pre-established relationship model between resistance and wear, evaluate the current loss value of the probe tip.

[0089]

[0090] Let \(W\) represent the probe loss value, \(R\) represent the contact resistance, \(dR / dt\) represent the resistance change rate, and \(f\) represent the loss evaluation function. This formula describes the relationship between the probe loss and the resistance parameters. This model is obtained by analyzing 1000 sets of data on the wear amount of the probe tip at different stages and the corresponding resistance changes during the contact between the probe tip and the wafer, and derives the mapping relationship between the resistance and the release amount of the probe repair material. Based on the current resistance value of 0.5 \(\Omega\) and the resistance change rate, the current probe tip loss value is evaluated to be 10% (expressed as the relative wear degree) in combination with the model.

[0091] The preset probe loss threshold is 15%. Although the current loss value of 10% has not reached the threshold, considering that long-term testing may lead to increased loss, the probe self-repair mechanism is triggered when the loss value reaches 10%.

[0092] M repair = g(W, \(W_0\)) threshold )

[0093] M repair represents the quantitative amount of the repair material, \(W\) represents the current loss value, \(W_0\) thresholdLet \(\theta\) represent the loss threshold, and \(g\) represent the quantitative calculation function of the repair material. This formula describes the relationship between the quantitative repair material and the loss value. According to the resistance-wear relationship model, the quantitative repair of the probe is determined to be 5 mg (assuming the amount of repair material is expressed in mass). The quantitative repair material (such as a repair slurry containing a specific metal component) is released to the probe loss area, and at the same time, the temperature change data of the probe loss area is detected by a micro temperature sensor. During the repair process, it is detected that the temperature rises by 2 °C. According to the preset relationship between temperature and catalyst components, the catalyst components in the repair material are adjusted to ensure that the repair material adheres to the probe loss area quickly and evenly, completing the repair.

[0094] Based on the data to be tested, the wafer is characterized to obtain initial test data. For example, in terms of electrical performance testing, the resistance of a certain area of the wafer is measured to be 100 Ω, the capacitance is 10 pF, and the inductance is 5 nH; in signal integrity testing, the signal transmission delay is 1 ns and the signal distortion rate is 2%; in terms of functional testing, the logical function state changes normally and the function response time is 50 ns.

[0095] The digital filtering algorithm is used to remove noise from the obtained initial test data of the wafer, and the interference signal is suppressed by the adaptive interference cancellation technology. For example, in the digital filtering algorithm, the cut-off frequency is set to 500 MHz to remove high-frequency noise; the adaptive interference cancellation technology automatically adjusts the cancellation parameters according to the detected characteristics of the interference signal to suppress the interference signal. At the same time, the algorithm parameters are adjusted according to the spectral characteristics of the signal, and the wafer test data is obtained after processing.

[0096] The wafer test data is classified according to electrical performance, signal integrity, and functional testing to obtain the characteristics of each type of data. For example, the parameter changes of resistance, capacitance, and inductance are analyzed from the electrical performance data; the signal transmission delay and signal distortion rate are obtained from the signal integrity data; the changes in the logical function state and the function response time are concerned in the functional test data.

[0097] The characteristics of each type of data obtained are compared with a pre-constructed benchmark model. The benchmark model includes the standard value range, fluctuation range, and normal fluctuation law of each characteristic parameter. For example, in the electrical performance data, the standard value range of resistance is 90 - 110 Ω, the standard value range of capacitance is 8 - 12 pF, and the standard value range of inductance is 4 - 6 nH; in the signal integrity data, the standard value range of signal transmission delay is 0.8 - 1.2 ns, and the standard value range of signal distortion rate is less than 3%; in the functional test data, the change of the logical function state should conform to specific logical rules, and the standard value range of the function response time is 40 - 60 ns. If the data characteristics deviate from the threshold set by the benchmark model, the corresponding data is determined to be suspected abnormal data.

[0098] The convolutional neural network model is trained with historical normal data, and the amount of training data is 5,000 groups. The trained convolutional neural network model is used to analyze the suspected abnormal data to generate analysis and detection results. For example, the analysis and detection results show that the signal transmission delay in this test data is 1 ns, which is within the standard range, so it is determined that there is no abnormality; the signal distortion rate is 2%, which is within the standard range, so it is determined that there is no abnormality; all indicators of the electrical performance and function test data are within the normal range, and the overall test result is normal.

[0099] Since the analysis and detection results of this test show that the test is normal, there is no need to adjust the test parameters and working mode of the probe card for the time being. However, assuming in other test scenarios, if the analysis and detection results show that the deviation of the test signal exceeds the preset range and this deviation is associated with the wear or position change of the probe, it is determined that the probe needs to be adaptively adjusted or self-repaired, and at the same time, according to the analysis and detection results, the test parameters and working mode of the probe card are dynamically adjusted.

[0100]

[0101] V test(i+1) represents the test voltage for the next iteration, V test(i) represents the current test voltage, α represents the learning rate, represents the gradient of the objective function with respect to the current test voltage. This formula describes the dynamic adjustment process of the test parameters. For example, if the electrical performance data is abnormal, according to the preset rules for electrical performance and parameter adjustment, the test voltage and current are gradually adjusted in a gradient descent manner. Assume that the current test voltage is 1 V, and the adjustment amplitude each time is 0.05 V. After each adjustment, the wafer test data is re-obtained until the wafer test data meets the expected range. In terms of the working mode, if it is found that the current scanning method results in low test efficiency, it is adjusted to a more efficient scanning method according to the analysis results, such as changing from row-by-row scanning to spiral scanning.

[0102] Through the above specific embodiments, the technical implementation process of the multi-functional test method and system of the probe kit based on wafer test is fully demonstrated. Through real-time monitoring, intelligent judgment and adjustment, the accuracy and reliability of wafer test are ensured.

[0103] Refer to Figure 2 , which is the structural block diagram of the multi-functional test system of the probe kit based on wafer test in an embodiment of the present invention, including:

[0104] A data acquisition unit, configured to acquire the data to be tested for the wafer, and acquire the parameter information of the probe and the probe card carried in the data to be tested;

[0105] The monitoring unit is used to continuously monitor the working parameters during the process of the probe tip contacting the wafer. The working parameters include resistance parameters, contact status, and the frequency and amplitude of the test signal. At the same time, it monitors the pressure and temperature at which the probe is currently working.

[0106] The probe adjustment unit is used to judge whether the current contact status between the probe tip and the wafer meets the preset conditions according to the working parameters through a preset intelligent algorithm, and dynamically adjust the contact pressure between the probe tip and the wafer and the probe impedance.

[0107] The environment optimization unit is used to obtain the environmental parameters of the test area corresponding to the probe card. The environmental parameters include temperature, humidity, and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, it executes the automatic probe card environment optimization program.

[0108] The data processing unit is used to detect the wafer characteristics according to the data to be tested, obtain the initial test data including electrical performance, signal integrity, and functional testing, and preprocess the wafer initial test data to obtain the wafer test data.

[0109] The data analysis unit is used to perform real-time analysis on the wafer test data, perform anomaly detection based on the above analysis, and obtain the analysis and detection results.

[0110] The adaptive adjustment unit is used to judge whether it is necessary to adjust the probe or self-repair the state based on the analysis and detection results. If necessary, it executes the corresponding operations. At the same time, according to the analysis and detection results, it dynamically adjusts the test parameters and working mode of the probe card.

[0111] For the specific implementation of each module in the above device example, please refer to that described in the above method embodiment, and details will not be elaborated here.

[0112] Refer to Figure 3 , in the embodiment of the present invention, a computer device is further provided. The computer device can be a server, and its internal structure can be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0113] Those skilled in the art can understand that Figure 3 The structure shown in Figure 3 is only a block diagram of some structures related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0114] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0115] In summary, obtain the data to be tested for the wafer, and obtain the parameter information of the probes and the probe card carried in the data to be tested; based on the working parameters, judge whether the contact state between the current probe tip and the wafer meets the preset conditions through a preset intelligent algorithm, and dynamically adjust the contact pressure between the probe tip and the wafer and the probe impedance; obtain the environmental parameters of the test area corresponding to the probe card, and the environmental parameters include temperature, humidity, and electromagnetic interference intensity. If it is detected that the environmental parameters exceed the set range, execute the automatic optimization program for the probe card environment; based on the data to be tested, detect the wafer characteristics, obtain the initial test data including electrical performance, signal integrity, and functional test, preprocess the wafer initial test data to obtain the wafer test data; perform real-time analysis on the wafer test data, perform anomaly detection based on the above analysis to obtain the analysis and detection results; based on the analysis and detection results, judge whether it is necessary to adjust or self-repair the state of the probe. If necessary, execute the corresponding operation. At the same time, according to the analysis and detection results, dynamically adjust the test parameters and working mode of the probe card to achieve high-precision and high-stability wafer testing, ensure the performance of the probe and the test system, and improve the accuracy and reliability of semiconductor testing.

[0116] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0117] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, apparatus, article, or method including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, apparatus, article, or method including the element.

[0118] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A multifunctional test method for a probe kit based on wafer testing, characterized in that: The following steps are involved: Acquire the test data of the wafer, and acquire the parameter information of the carried probe and probe card from the test data, wherein the parameter information includes the material, size and needle tip shape of the probe, and the substrate material, wiring method and impedance matching circuit of the probe card; Continuously monitor the working parameters of the probe tip and the wafer during contact, the working parameters including resistance parameters, contact state, and frequency and amplitude of the test signal, and monitor the current working pressure and temperature of the probe; Based on the working parameters, a pre-set intelligent algorithm is used to determine whether the current contact state between the probe tip and the wafer meets the preset conditions, and the contact pressure between the probe tip and the wafer and the probe impedance are dynamically adjusted; Acquire the environmental parameters of the test area corresponding to the probe card, wherein the environmental parameters include temperature, humidity and electromagnetic interference intensity. If it is detected that the environmental parameters are out of the set range, execute the automatic optimization program of the probe card environment; Based on the data to be tested, detecting wafer characteristics, obtaining initial test data including electrical performance, signal integrity and functional testing, and preprocessing the initial wafer test data to obtain wafer test data; Performing real-time analysis on the wafer test data, performing abnormality detection based on the analysis, and obtaining analysis and detection results; Based on the analysis and detection results, determine whether the probe needs to be adjusted or self-repaired. If necessary, perform corresponding operations. At the same time, according to the analysis and detection results, dynamically adjust the test parameters and working mode of the probe card. The test parameters include but are not limited to test voltage, current and frequency. The working mode includes scanning mode and test sequence.

2. The multifunctional test method of the probe kit based on wafer test according to claim 1, characterized in that: After the step of continuously monitoring the working parameters of the probe tip in the contact process with the wafer, wherein the working parameters include resistance parameters, contact state, and frequency and amplitude of the test signal, and simultaneously monitoring the pressure and temperature of the current working of the probe, the method further includes: Obtaining resistance parameters during the contact between the probe tip and the wafer, wherein the resistance parameters include resistance value and resistance change rate; Based on the resistance parameter, the loss value of the current probe tip is evaluated through a pre-established resistance and wear relationship model; the resistance and wear relationship model obtains a mapping relationship between resistance and the amount of probe repair material released after analyzing a large amount of probe tip wear at each stage and corresponding resistance change data of the probe tip in contact with the wafer; According to the preset loss threshold of the probe and the loss value, the probe self-repair mechanism is triggered, and the probe repair quantity is determined by the resistance and wear relationship model; A quantitative repair material is released into the probe loss area, and temperature change data of the probe loss area is detected at the same time to adjust the catalyst component in the repair material.

3. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: The step of judging whether the current contact state between the probe tip and the wafer meets the preset conditions based on the working parameters and dynamically adjusting the contact pressure between the probe tip and the wafer and the probe impedance includes: Based on the currently acquired contact status between the probe tip and the wafer, the frequency and amplitude of the test signal, and the current working pressure and temperature of the probe, analysis is performed using a pre-set intelligent algorithm; Based on the model analysis results, it is determined whether the current contact state between the probe tip and the wafer meets the preset conditions. When it is determined that the contact state between the probe tip and the wafer does not meet the preset conditions, the contact pressure of the probe tip is adjusted by the intelligent algorithm until the tip pressure reaches the optimal state; Based on the model analysis result, it is determined whether the current probe impedance matches the test signal. When the probe impedance does not match the test signal, an impedance matching parameter is generated to adjust the probe impedance until the impedance of the probe matches the test signal.

4. The multifunctional test method of the probe kit based on wafer test according to claim 3, characterized in that: The intelligent algorithm is based on the collection of historical data on the contact status between the probe tip and the wafer, the test signal frequency amplitude, the working pressure and the temperature, and is trained using a machine learning algorithm framework.

5. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: The step of obtaining the environmental parameters of the test area corresponding to the probe card, wherein the environmental parameters include temperature, humidity and electromagnetic interference intensity, and if it is detected that the environmental parameters exceed the set range, executing the automatic optimization program of the probe card environment, further includes: Acquire environmental parameters of a test area corresponding to the probe card, wherein the environmental parameters include temperature, humidity, and electromagnetic interference intensity data; Determine the degree of deviation of the environmental parameters according to the preset environmental parameter threshold and the current environmental parameters; According to the degree of deviation of the environmental parameter, a corresponding optimization device is started, wherein the optimization device includes a temperature control device, a humidity adjustment device and / or an electromagnetic shielding device; According to the degree of deviation of the environmental parameters, a control parameter of a corresponding optimization device is obtained; The control parameters are sent to the corresponding optimization device to start the temperature control, humidity adjustment and / or electromagnetic shielding process, continuously obtain real-time environmental parameters, and dynamically optimize and adjust the test environment until the environmental parameters of the probe card are restored to within the preset environmental parameter thresholds.

6. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: A digital filtering algorithm is used to remove noise from the initial wafer test data, interference signals are suppressed through adaptive interference cancellation technology, and algorithm parameters are adjusted according to the frequency spectrum characteristics of the signal to obtain wafer test data.

7. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: The step of performing real-time analysis on the wafer test data, performing abnormality detection based on the analysis, and obtaining analysis and detection results includes: Classify the wafer test data according to electrical performance, signal integrity, and functional test to obtain various data features, including electrical performance data, analytical resistance, capacitance, and inductance parameter changes, signal integrity data signal transmission delay and signal distortion rate, and functional test data logic function state changes and function response time; Compare the acquired data features with a pre-built benchmark model, which contains the standard value range, fluctuation range, and normal fluctuation law of each feature parameter. If the data feature deviates from the threshold set by the benchmark model, the corresponding data is determined to be suspected abnormal data; The convolutional neural network model is trained with historical normal data, and the abnormal data is analyzed by the trained convolutional neural network model to generate analysis and detection results. The analysis and detection results include whether the abnormality is determined, the type of abnormality, and the degree of influence of the abnormality on the test results.

8. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: If the analysis and detection result shows that the deviation of the test signal exceeds a preset range and the deviation is associated with the wear or position change of the probe, it is determined that the probe needs to be adaptively adjusted or self-repaired.

9. The multifunctional test method of a probe kit based on wafer test according to claim 1, characterized in that: When the analysis and detection results indicate that the electrical performance data of the wafer is abnormal, the test voltage and current are gradually adjusted in a gradient descent manner according to the preset electrical performance and parameter adjustment rules, and the wafer test data is reacquired after each adjustment until the wafer test data is within the expected range.

10. A probe kit multifunctional test system based on wafer test, characterized in that: include: A data acquisition unit, used to acquire the test data for the wafer, and acquire parameter information of the carried probe and probe card from the test data; A monitoring unit, used to continuously monitor the working parameters during the contact between the probe tip and the wafer, wherein the working parameters include resistance parameters, contact state, and frequency and amplitude of the test signal, and also monitor the current working pressure and temperature of the probe; A probe adjustment unit, for determining whether the current contact state between the probe tip and the wafer meets a preset condition according to the working parameters and through a preset intelligent algorithm, and dynamically adjusting the contact pressure between the probe tip and the wafer and the probe impedance; An environmental optimization unit is used to obtain environmental parameters of the test area corresponding to the probe card, wherein the environmental parameters include temperature, humidity and electromagnetic interference intensity. If it is detected that the environmental parameters are out of the set range, an automatic optimization program of the probe card environment is executed; A data processing unit is used to detect wafer characteristics according to the data to be tested, obtain initial test data including electrical performance, signal integrity and functional test, and pre-process the initial wafer test data to obtain wafer test data; A data analysis unit, used to perform real-time analysis on the wafer test data, perform abnormality detection based on the analysis, and obtain analysis and detection results; The adaptive adjustment unit is used to determine whether the probe needs to be adjusted or self-repaired based on the analysis and detection results, and if necessary, perform corresponding operations. At the same time, according to the analysis and detection results, dynamically adjust the test parameters and working mode of the probe card.

Images