A fully automatic high-precision probe station

Abstract

The present application relates to the technical field of semiconductor testing equipment, and particularly relates to a full-automatic high-precision probe station, which utilizes multi-physical field coupling sensing and compensation to significantly improve positioning accuracy: through comprehensive collection of multi-physical field data such as temperature field and stress field, combined with coupling algorithm optimization compensation strategy, multi-factor interference is effectively offset, positioning repeat accuracy can reach ±0.05 mu m, and the testing demand of advanced semiconductor technology is met; dynamic three-dimensional modeling technology enhances dynamic adaptability: a dynamic three-dimensional modeling system of multi-physical field data influence size can update the testing target morphology in real time, avoids testing deviation caused by target deviation or surface undulation, and improves the adaptive ability of the equipment to complex testing scenes. High degree of automation improves testing efficiency: the full process of data collection, modeling, compensation and testing can be completed without manual intervention, the calibration process is simplified, the testing period is shortened, and it is suitable for batch detection scenes.

Description

Technical Field

[0001] This invention relates to the field of semiconductor testing equipment technology, and more specifically to a fully automated high-precision probe station. Background Technology

[0002] Wafer probe stations are key equipment in the semiconductor manufacturing process, primarily used for electrical characteristic testing before chip dicing and packaging. They obtain electrical performance data by guiding probes to test points on the wafer, directly impacting chip production quality and efficiency. As semiconductor technology advances towards 3D packaging and heterogeneous integration, wafer sizes continue to increase and circuit density rises, placing higher demands on the positioning accuracy and testing stability of probe stations.

[0003] Although existing fully automated probe stations employ high-precision motion platforms, they still have several shortcomings in practical applications: First, changes in ambient temperature and heat generated during equipment operation can cause deformation of components such as the motion platform, optical components, and grating ruler base, resulting in multi-physical field coupling interference such as temperature and stress fields, affecting probe positioning accuracy; Second, existing calibration methods rely on laser interferometers, which are complex to set up, time-consuming, and unable to update calibration data in real time; Third, they lack dynamic perception of the test target's shape, making it difficult to cope with test deviations caused by surface undulations or positional shifts; Fourth, the single-physical-field data acquisition and compensation method cannot fully explore the inherent correlation between various physical-field data, easily leading to information loss and inaccurate compensation. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a fully automatic high-precision probe station, which is based on multi-physics field coupling intelligent compensation and dynamic three-dimensional modeling. This solves the technical problems of existing probe stations, such as easy interference with positioning accuracy, complex calibration, and poor dynamic adaptability, thereby improving the testing accuracy and stability of the probe station.

[0005] The technical solution is as follows: a fully automatic high-precision probe station, characterized in that it includes a multi-physics data acquisition module, a dynamic three-dimensional modeling system, an intelligent compensation unit, a high-precision motion control module, and a central control module;

[0006] The multiphysics data acquisition module includes a temperature sensor, a stress sensor, and a temperature sensor, which are used to acquire parameters of temperature field, stress field, and ambient temperature data, respectively.

[0007] The dynamic 3D modeling system acquires the original image data of the test target and reconstructs the size of the target by using a coupling algorithm model that influences the size of the multiphysics data acquisition module.

[0008] The intelligent compensation unit has a built-in multiphysics coupling algorithm. It receives real-time data from the multiphysics data acquisition module and model parameters from the dynamic 3D modeling system. It then uses weighted coefficients to calibrate and fuse key parameters to generate targeted compensation instructions.

[0009] The high-precision motion control module includes a linear motor four-axis motion platform and a probe drive unit. It receives compensation commands from the intelligent compensation unit and drives the probe to complete positioning and testing actions. The positioning repeatability is not less than ±0.05μm.

[0010] The central control module is used to coordinate the data interaction and working sequence of each module, store test data and modeling results, and support data export and visualization.

[0011] A further feature is that the length deformation of the part is denoted as .

[0012] ΔL=L0α L (T-T0)+ L0σ / E (1),

[0013] The equation for the influence of temperature field on size is ΔL. W =L0α L (T-T0) characterizes the effect of temperature change on the dimensions of critical components, where L0 is the length of the critical component, and α is the value of T. L Here, T is the coefficient of thermal expansion of the key component, T0 is the ambient temperature, and the equation for the influence of the stress field on the dimensions is ΔL. f =L0σ / E, characterizing the influence of vibration on the size of key parts, where σ is the stress of key parts and E is the elastic modulus; in the dynamic three-dimensional modeling system, such as the design size at room temperature superimposed with the thermal expansion and contraction size after temperature change, a dynamically updated three-dimensional model of the test target is constructed, and the model size is updated using formula (1). The dynamic three-dimensional modeling system integrates a high-definition industrial camera and a laser ranging unit, and the modeling error is less than 0.1μm;

[0014] The workflow of the dynamic 3D modeling system is as follows:

[0015] (1) Use a high-definition industrial camera and a laser ranging unit to synchronously acquire surface images and distance data of the test target;

[0016] (2) Update the original data using a coupled algorithm model that uses the influence of multiphysics data parameters on size, and use size-driven update model;

[0017] (3) Based on the contact state between the probe and the target during the test, the three-dimensional model is updated every 5ms to ensure the consistency between the model and the actual target;

[0018] The multiphysics coupling algorithm includes the following steps:

[0019] (1) Preprocess the collected multi-physics field data, remove outliers and standardize the data. If a sensor data shows a large and unreasonable sudden change, it is considered abnormal. The specific value can be determined based on the data during the prototype testing process.

[0020] (2) By using the probe station working state model, key influencing parameters are determined, and the optimal parameter values ​​and corresponding weighting coefficient combinations are found. The obtained sensor parameters directly drive the change in the size of the three-dimensional model.

[0021] (3) Adjust the weighting coefficients in real time according to the shape changes of the dynamic three-dimensional model to generate a dynamic compensation strategy: Based on the dynamic correction size and the reference dynamic size of the three-dimensional model, the compensation is driven by the algorithm, the high-precision motion control module is started, and fine-tuning and calibration are performed.

[0022] The intelligent compensation unit also includes a self-learning module, which optimizes the parameter mapping relationship of the multi-physics coupling algorithm and improves the compensation accuracy by accumulating test data under different working conditions.

[0023] The high-precision motion control module has a built-in grating ruler feedback unit that collects displacement data of the motion platform in real time and forms a closed-loop control with intelligent compensation commands.

[0024] By employing this invention, multi-physics field coupled sensing and compensation significantly improves positioning accuracy: Through comprehensive acquisition of multi-physics field data such as temperature and stress fields, combined with optimized compensation strategies using coupled algorithms, multi-factor interference is effectively offset, achieving a positioning repeatability accuracy of ±0.05μm, meeting the testing requirements of advanced semiconductor technologies. Dynamic 3D modeling technology enhances dynamic adaptability: The dynamic 3D modeling system, where multi-physics field data influences dimensions, can update the test target's shape in real time, avoiding test deviations caused by target offsets or surface undulations, and improving the equipment's adaptability to complex testing scenarios. High automation improves testing efficiency: The entire process of data acquisition, modeling, compensation, and testing can be completed without manual intervention, simplifying the calibration process, shortening the testing cycle, and making it suitable for batch testing scenarios.

[0025] Furthermore, it has strong self-learning capabilities and continuously optimized stability: the self-learning module of the intelligent compensation unit can accumulate working condition data and continuously optimize algorithm parameters to ensure that the equipment maintains stable testing accuracy during long-term use. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the modular structure of the fully automatic high-precision probe station of the present invention;

[0027] Figure 2 This is a schematic diagram illustrating the workflow of a multiphysics coupling algorithm.

[0028] Figure 3Schematic diagram of a state-based 3D modeling system;

[0029] Figure 4 This is a flowchart of the probe testing method of the present invention;

[0030] Figure 5 Probe station workflow Figure 1 ;

[0031] Figure 6 Probe station workflow Figure 2 . Detailed Implementation

[0032] See Figures 1 to 6 As shown, a fully automatic high-precision probe station includes a multi-physics data acquisition module, a dynamic three-dimensional modeling system, an intelligent compensation unit, a high-precision motion control module, and a central control module.

[0033] The multiphysics data acquisition module includes a temperature sensor, a stress sensor, and an ambient temperature sensor, used to collect parameters of the temperature field, stress field, and ambient temperature data, respectively. The temperature sensor has a measurement range of -20℃ to 80℃ and an accuracy of ±0.1℃; the stress sensor has a measurement range of 0 to 500N and an accuracy of ±0.1N, with a sampling frequency of no less than 100Hz to ensure the real-time performance and integrity of the data, providing data support for subsequent compensation calculations; assuming the length deformation of the part is...

[0034] ΔL=L0α L (T-T0)+ L0σ / E (1),

[0035] The equation for the influence of temperature field on size is ΔL. W =L0α L (T-T0) characterizes the effect of temperature change on the dimensions of critical components, where L0 is the length of the critical component, and α is the value of T. L Here, T is the coefficient of thermal expansion of the key component, T0 is the ambient temperature, and the equation for the influence of the stress field on the dimensions is ΔL. f =L0σ / E, characterizing the impact of vibration on the dimensions of critical components, where σ is the stress of the critical component and E is the elastic modulus.

[0036] The dynamic 3D modeling system combines a high-definition industrial camera (5000×5000 pixels resolution) and a laser ranging unit (laser ranging unit measurement range 0~100mm, accuracy ±0.05μm) to acquire the original image data of the test target. The system reconstructs the size of the multi-physics data by using a coupling algorithm model that influences the size of the multi-physics data acquisition module. The key area mesh size of the coupling algorithm model that influences the size of the multi-physics data parameters is 0.05μm, and the non-key area mesh size is 0.2μm. The modeling update cycle is 5ms. The model size is updated by using size-driven updates, such as the design size at room temperature superimposed with the thermal expansion and contraction size after temperature changes. A dynamically updated 3D model of the test target is constructed, and the model size is updated using formula (1). The dynamic 3D modeling system integrates a high-definition industrial camera and a laser ranging unit, and the modeling error is less than 0.1μm. The system updates the 3D model every 5ms, which can dynamically track the shape changes of the test target and avoid test deviations caused by target offset or surface undulations.

[0037] The intelligent compensation unit uses a microcontroller as its core processing chip, supports multi-threaded parallel processing, and incorporates a multiphysics coupling algorithm. Implemented in C language, it receives real-time data from the multiphysics data acquisition module and model parameters from the dynamic 3D modeling system. It then calibrates and fuses key parameters through a weighted coefficient combination (with a weighted coefficient adjustment step of 0.001) to generate targeted compensation instructions. The multiphysics coupling algorithm includes the following steps:

[0038] (1) Preprocess the collected multi-physics field data, remove outliers and standardize the data. If a sensor data shows a large and unreasonable sudden change, it is considered abnormal. The specific value can be determined based on the data during the prototype testing process.

[0039] (2) By using the probe station working state model, key influencing parameters are determined, and the optimal parameter values ​​and corresponding weighting coefficient combinations are found. The obtained sensor parameters directly drive the change in the size of the three-dimensional model.

[0040] (3) Adjust the weighting coefficients in real time according to the shape changes of the dynamic three-dimensional model to generate a dynamic compensation strategy: Based on the dynamic correction size and the reference dynamic size of the three-dimensional model, the algorithm drives the compensation, the high-precision motion control module is started, and fine-tuning and calibration are performed.

[0041] The high-precision motion control module includes a linear motor four-axis motion platform and a probe drive unit. It receives compensation commands from the intelligent compensation unit and drives the probe to complete positioning and testing actions. The linear motor four-axis motion platform has a stroke of 500mm×500mm×100mm×360° (X / Y / Z / θ axes), a maximum motion speed of 100mm / s, and a grating ruler resolution of 0.01μm. The probe drive unit uses a piezoelectric ceramic actuator with a displacement resolution of 0.001μm. The positioning repeatability is not less than ±0.05μm.

[0042] The central control module uses an industrial-grade computer, equipped with an Intel Core i7 processor, 16GB of RAM, and a 1TB solid-state drive. The visual user interface is developed based on Qt and supports test parameter settings, real-time data curve display, 3D model visualization, and test report export. It coordinates data interaction and workflow among modules, stores test data and modeling results, and supports data export and visualization.

[0043] The intelligent compensation unit also includes a self-learning module, which optimizes the parameter mapping relationship of the multi-physics coupling algorithm and improves the compensation accuracy by accumulating test data under different working conditions.

[0044] The high-precision motion control module has a built-in grating ruler feedback unit that collects displacement data of the motion platform in real time and forms a closed-loop control with intelligent compensation commands.

[0045] The specific testing method is given below, and the steps are as follows:

[0046] (1) Start the equipment and after the self-test of each module is completed, fix the wafer or chip on the stage. The multi-physics data acquisition module starts to collect ambient temperature, temperature of each component of the equipment, and stress data. The dynamic three-dimensional modeling system is initialized and moved above the wafer to collect the original image and distance data of the wafer surface.

[0047] (2) The dynamic 3D modeling system reconstructs the dimensions affected by multi-physics data through a coupling algorithm model of multi-physics data parameters based on the acquired raw data, and constructs the initial 3D model of the wafer. The intelligent compensation unit generates a benchmark compensation scheme based on the initially acquired multi-physics data and sends it to the high-precision motion control module.

[0048] (3) The high-precision motion control module drives the probe to move at a speed of 50 mm / s to 1 mm above the first test point on the wafer according to the reference compensation scheme. At this time, the multiphysics data acquisition module uploads the latest data in real time.

[0049] (4) The intelligent compensation unit receives real-time data and dynamically updated wafer 3D model, calculates that the platform deformation deviation caused by the current temperature field is 0.12μm and the probe offset caused by the stress field is 0.08μm, and then generates compensation command to control the X / Y / Z axes to adjust by +0.12μm, -0.08μm and +0.05μm respectively to correct the positioning deviation;

[0050] (5) The probe descends and contacts the test point to complete the electrical parameter test. The test data is stored by the central control module. Then the probe is lifted up, the dynamic three-dimensional modeling system updates the wafer three-dimensional model, and the intelligent compensation unit generates a new compensation instruction based on the position of the next test point and the real-time physical field data. Steps S3-S5 are repeated until all test points on the wafer are detected.

[0051] (6) After all tests are completed, the probe is reset and the central control module generates a test report, which includes the electrical parameters of each test point, the positioning deviation compensation value and the three-dimensional model data, and supports exporting to Excel or PDF format.

[0052] Through testing and verification on 10 silicon wafers with a diameter of 300mm, the fully automated high-precision probe station of this invention exhibits the following effects: Positioning accuracy: The actual deviation of the probe contacting the test point is less than ±0.05μm, meeting the requirements of high-precision testing; Testing stability: After continuous testing of 1000 test points, the deviation fluctuation range is less than ±0.02μm, demonstrating excellent stability; Testing efficiency: The testing time for a single wafer is 8 minutes, which is 30% shorter than that of traditional probe stations; Environmental adaptability: Under environmental conditions of 15℃~35℃ and 30%RH~70%RH, the testing accuracy does not change significantly, demonstrating strong anti-interference ability.

Claims

1. A fully automatic high-precision probe station, characterized in that, It includes a multiphysics data acquisition module, a dynamic 3D modeling system, an intelligent compensation unit, a high-precision motion control module, and a central control module; The multiphysics data acquisition module includes a temperature sensor and a stress sensor. The temperature sensor is used to collect ambient temperature and temperature parameters of various components of the equipment, and the stress sensor is used to collect stress data parameters. The dynamic 3D modeling system acquires the original image data of the test target, and reconstructs the size affected by multiphysics data through a coupling algorithm model of multiphysics data parameter influence on size, thereby constructing an initial 3D model of the test target. The intelligent compensation unit has a built-in multiphysics coupling algorithm. It receives real-time data from the multiphysics data acquisition module and model parameters from the dynamic 3D modeling system. It then uses weighted coefficients to calibrate and fuse key parameters to generate targeted compensation instructions. The high-precision motion control module includes a linear motor four-axis motion platform and a probe drive unit. It receives compensation commands from the intelligent compensation unit and drives the probe to complete positioning and testing actions. The positioning repeatability is not less than ±0.05μm. The central control module is used to coordinate the data interaction and working sequence of each module, store test data and modeling results, and support data export and visualization.

2. The fully automatic high-precision probe station according to claim 1, characterized in that, it is provided that... The deformation of the test target length is ΔL=L0α L (T-T0)+ L0σ / E (1), The equation for the influence of the temperature field on the size is ΔL. W =L0α L (T-T0) characterizes the effect of temperature change on the size of the test target, L0 is the length of the test target, and α L The test target is the coefficient of thermal expansion, T is the real-time temperature, T0 is the ambient temperature, and the equation for the influence of the stress field on the dimensions is ΔL. f =L0σ / E, characterizing the effect of vibration on the size of the test target, where σ is the stress of the test target and E is the elastic modulus; In the dynamic three-dimensional modeling system, the design size of the test target at room temperature is superimposed with the thermal expansion and contraction size after temperature change to construct a dynamically updated three-dimensional model of the test target. The model size is updated using formula (1). The dynamic three-dimensional modeling system integrates a high-definition industrial camera and a laser ranging unit, and the modeling error is less than 0.1μm.

3. The fully automatic high-precision probe station according to claim 2, characterized in that, The workflow of the dynamic 3D modeling system is as follows: (1) Use a high-definition industrial camera and a laser ranging unit to synchronously acquire surface images and distance data of the test target; (2) The original data is updated by using the coupling algorithm model of the influence of multi-physics data parameters on size. The size-driven update model is used: the design size of the test target at room temperature is superimposed with the thermal expansion and contraction size after temperature change, and a dynamically updated three-dimensional model of the test target is constructed. The model size is updated using formula (1). (3) Based on the contact state between the probe and the target during the test, the three-dimensional model is updated every 5ms to ensure the consistency between the model and the actual target.

4. The fully automatic high-precision probe station according to claim 1, characterized in that, The multiphysics coupling algorithm includes the following steps: (1) Preprocess the collected multi-physics field data, remove outliers and standardize the data. If a sensor data shows a large and unreasonable sudden change, it is considered to be abnormal. The specific value is determined based on the data during the prototype testing process. (2) By using the probe station working state model, key influencing parameters are determined, and the optimal parameter values ​​and corresponding weighting coefficient combinations are found. The obtained sensor parameters directly drive the change in the size of the three-dimensional model. (3) Adjust the weighting coefficients in real time according to the shape changes of the dynamic three-dimensional model to generate a dynamic compensation strategy: Based on the dynamic correction size and the reference dynamic size of the three-dimensional model, the compensation is driven by the algorithm, the high-precision motion control module is started, and fine-tuning and calibration are performed.

5. The fully automatic high-precision probe station according to claim 1, characterized in that, The intelligent compensation unit also includes a self-learning module, which optimizes the parameter mapping relationship of the multiphysics coupling algorithm and improves the compensation accuracy by accumulating test data under different working conditions.

6. The fully automatic high-precision probe station according to claim 1, characterized in that, The high-precision motion control module has a built-in grating ruler feedback unit that collects displacement data of the motion platform in real time and forms a closed-loop control with intelligent compensation commands.

Images