An advanced process and wide-bandgap semiconductor probe testing method and system
The alignment results are obtained through an optical calibration system, and the probe pressure test, contact resistance test and visual inspection are performed to generate the first measurement result. This solves the problem of probe and pad alignment separation, realizes the pre-judgment of measurement trigger signal and the continuous link of measurement results, and meets the needs of advanced process and wide bandgap semiconductor probe testing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING GOLDEN ANT GUOCHUANG TECH CO LTD
- Filing Date
- 2026-05-22
- Publication Date
- 2026-07-03
AI Technical Summary
In existing semiconductor testing technologies, the alignment result of the probe and the pad is separated from the contact state, the measurement trigger signal lacks a pre-judgment, and the contact states of multiple probes and pads are difficult to correspond, resulting in an incomplete correlation link between the measurement result and the number of tests, which makes it difficult to meet the needs of advanced process and wide bandgap semiconductor probe testing.
The alignment results are obtained by moving the optical calibration system along the X, Y, Z and θ directions. The needle pressure test, contact resistance test and visual inspection of the external probe and the puncture area are performed to generate the first measurement result. Based on this result, the needle pressure test, contact resistance test and visual inspection of multiple probes and solder pads are performed to generate the second measurement result. Finally, the measurement trigger signal is generated by comparison and processing to realize the measurement action processing.
It achieves a unified process for probes and pads, and the correlation link between measurement results and test counts is continuously recorded. The measurement trigger signal is stable, ensuring the accuracy and consistency of measurement results.
Smart Images

Figure CN122330639A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor testing technology, and in particular to an advanced process and wide-bandgap semiconductor probe testing method and system. Background Technology
[0002] In the field of semiconductor testing technology, existing solutions typically use a vacuum tray to support the wafer, employ multiple probes on a probe card to contact the solder pads, and perform alignment by moving the probes along the X, Y, Z, and θ directions using an optical calibration system before executing the measurement. However, this approach suffers from limitations such as separation of alignment results from contact status, lack of pre-judgment criteria for measurement trigger signals, and an increase in the number of tests required. Existing methods often rely on directly performing pin pressure testing, contact resistance testing, or other measurements after alignment.
[0003] In scenarios where the measurement trigger signal is realized by comparing the first measurement result and the second measurement result, problems may arise such as the contact state formed by the external probe and the puncture area not participating in the determination of the measurement action of the chip under test, and the contact state formed by multiple probes and the pads being difficult to correspond to the previous contact state, making it difficult to meet the requirement of stable realization of the measurement trigger signal.
[0004] For the joint processing of the first measurement result, the second measurement result, the external probe, the puncture area, multiple probes, the solder pad, and the measurement trigger signal, existing technologies generally lack a comparison processing link that integrates needle pressure testing, contact resistance testing, and visual inspection into the same automated testing process. This makes it difficult to form a consistent process of acquisition, alignment, comparison processing, measurement action, and marking processing in advanced process and wide bandgap semiconductor probe testing application scenarios. As a result, the measurement trigger signal does not correspond adequately to the actual contact state, and the correlation link between the measurement result and the number of tests is incomplete. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides an advanced process and wide-bandgap semiconductor probe testing method, comprising:
[0006] S100: Acquire the wafer, chip under test, solder pads, pin piercing area, probe card, multiple probes, external probes and vacuum tray; perform optical calibration system and X, Y, Z and θ direction movement processing to obtain alignment results;
[0007] S200. Based on the alignment results, perform needle pressure test, contact resistance test and visual inspection processing between the external probe and the puncture area to obtain the first measurement result;
[0008] S300. Based on the first measurement result, perform multiple probe-pad pin pressure tests, contact resistance tests, and visual inspection processing to obtain the second measurement result;
[0009] S400. Based on the second measurement result, perform a comparison process between the first measurement result and the second measurement result to obtain a measurement trigger signal;
[0010] S500. Based on the measurement trigger signal, perform measurement action processing to obtain the measurement result.
[0011] Furthermore, the process of optical calibration system and movement processing in the X, Y, Z, and θ directions includes:
[0012] Based on the alignment marks on the wafer, the optical calibration system performs image acquisition and feature point extraction processing to generate a preliminary offset.
[0013] Based on the initial offset, a two-stage movement process of coarse and fine movement is performed in the X, Y, Z and θ directions. The two-stage movement process includes first performing a descent in the Z direction to a safe height close to the surface of the solder pad, then performing alignment in the X, Y and θ directions, and finally performing a contact descent in the Z direction.
[0014] During the Z-direction contact descent process, the approach distance between the external probe and the needle puncture area is monitored in real time. When the distance is less than a preset threshold, the descent is paused and the current Z coordinate is recorded as the alignment reference.
[0015] Based on the alignment reference and feature point extraction results, alignment results are generated.
[0016] Furthermore, the alignment result includes the center coordinates of the needle puncture area, the center coordinates of the solder pad, the Z-axis safety height, and the contact height.
[0017] Furthermore, the processes of needle pressure testing, contact resistance testing, and visual inspection of the external probe and the puncture area include:
[0018] Within the same probe descent stroke, the outer probe is controlled to descend from the safe height of the Z-axis in the alignment result to contact the needle puncture area, and needle pressure test, contact resistance test and visual inspection are performed in parallel.
[0019] The needle pressure test includes: acquiring the pressure waveform when the external probe contacts the puncture area using a piezoelectric sensor, extracting the peak pressure and steady-state pressure, calculating the deviation value from the preset ideal needle pressure value, and generating a needle pressure deviation result.
[0020] Furthermore, the processes of needle pressure testing, contact resistance testing, and visual inspection of the external probe and the puncture area also include:
[0021] The contact resistance test includes: measuring the contact resistance between the external probe and the puncture area using a four-wire Kelvin connection, calculating the deviation from the preset ideal resistance value, and generating a contact resistance deviation result.
[0022] Furthermore, the processes of needle pressure testing, contact resistance testing, and visual inspection of the external probe and the puncture area also include:
[0023] The visual inspection includes: acquiring contact images of the external probe tip on the surface of the puncture area using a high-resolution camera, extracting the offset between the center coordinates of the tip and the center coordinates of the puncture area, as well as the tip deformation features, and generating tip image deviation results.
[0024] The needle pressure deviation result, contact resistance deviation result, and needle tip image deviation result are organized into a structured field set containing field name, deviation direction, and deviation amplitude to obtain the first measurement result.
[0025] Furthermore, the process of multiple probes undergoing probe pressure testing, contact resistance testing, and visual inspection of the solder pads includes:
[0026] Based on the set of structured fields in the first measurement result, determine the same field names and the same deviation expression format that need to be collected in the second measurement result;
[0027] Within the same probe descent stroke as S200, multiple probes are controlled to descend from the Z-axis safety height in the alignment result to contact the pads of the chip under test, and probe pressure test, contact resistance test and visual inspection are performed in parallel.
[0028] The needle pressure test includes: acquiring the pressure waveform of each of the multiple probes when it contacts the corresponding solder pad, extracting the peak pressure and steady-state pressure, calculating the deviation value from the preset ideal needle pressure value, and generating multi-channel needle pressure deviation results;
[0029] The contact resistance test includes: measuring the contact resistance between each probe and the corresponding pad using a four-wire Kelvin connection, calculating the deviation from the preset ideal resistance value, and generating multi-channel contact resistance deviation results.
[0030] The visual inspection includes: acquiring contact images of each probe tip on the corresponding solder pad surface, extracting the offset between the tip center coordinates and the solder pad center coordinates, as well as the tip deformation features, and generating multi-channel tip image deviation results.
[0031] The multi-channel needle pressure deviation results, multi-channel contact resistance deviation results, and multi-channel needle tip image deviation results are organized into the same field structure as the first measurement results to obtain the second measurement results.
[0032] Furthermore, the process of comparing the first measurement result with the second measurement result includes:
[0033] The needle pressure deviation, contact resistance deviation, and needle tip image deviation results in the first measurement result are compared in parallel with the deviation results of the corresponding fields in the second measurement result.
[0034] The parallel comparison includes: simultaneously judging three fields: needle pressure deviation, contact resistance deviation, and needle tip image deviation. Each field has an independent allowable deviation range, which includes an upper limit threshold and a lower limit threshold.
[0035] When the needle pressure deviation value of a certain channel in the second measurement result falls within the ± allowable deviation range of the corresponding deviation value in the first measurement result, and the contact resistance deviation value falls within the ± allowable deviation range of the corresponding deviation value, and the needle tip image deviation value falls within the ± allowable deviation range of the corresponding deviation value, the contact state of that channel is determined to be qualified.
[0036] When the contact status of all channels is qualified, a measurement trigger signal containing a trigger enable flag and a diagnostic code is generated;
[0037] When the contact status of a channel is unqualified, a measurement trigger signal containing a trigger blocking flag and a diagnostic code is generated. The diagnostic code is used to identify the unqualified field type and the unqualified channel number.
[0038] After generating the trigger blocking flag, automatic rollback control is executed: based on the diagnostic code, the alignment, reference measurement and actual measurement are re-executed in reverse until the contact status of all channels is qualified or the preset retry limit is reached.
[0039] Furthermore, the process of measuring motion processing includes:
[0040] The measurement action processing includes the measurement action processing of pulse signal generator, drive unit, drive motherboard, drive daughterboard, drive IC, buffer capacitor and bus capacitor;
[0041] When the measurement trigger signal contains a trigger enable flag, the pulse signal generator is activated to generate a dynamic test pulse sequence that matches the type of the chip under test;
[0042] The driving unit receives the pulse sequence and selects the default driving voltage and current based on the diagnostic code in the measured trigger signal;
[0043] The driving motherboard distributes the driving signal to multiple driving daughterboards, and each driving daughterboard is connected to the corresponding probe through a driving IC.
[0044] During the measurement process, buffer capacitors and bus capacitors are used to stabilize the power supply voltage, while the output voltage and current waveforms of the driver IC are monitored in real time.
[0045] When an abnormal voltage or current is detected, the measurement action is automatically interrupted and an abnormality record is generated.
[0046] After the measurement operation is completed normally, the electrical response signals returned by each probe are collected, and after analog-to-digital conversion, they are organized into structured measurement results containing measurement timestamps, driving parameters and response waveforms according to the channel number.
[0047] Furthermore, an advanced process and wide-bandgap semiconductor probe testing system includes: an alignment result generation module, a first measurement result generation module, a second measurement result generation module, a comparison processing module, a measurement action processing module, and a marking processing module; the modules are connected in sequence to implement the method described in any of the above embodiments.
[0048] The key innovations of this invention include:
[0049] (1) After the alignment result is formed, the external probe and the puncture area are subjected to needle pressure test, contact resistance test and visual inspection to generate the first measurement result, and the first measurement result is used as the pre-processing input for the subsequent processing of the multiple probes and the solder pad.
[0050] (2) Based on the first measurement result, perform needle pressure test, contact resistance test and visual inspection on the plurality of probes and the solder pad to generate the second measurement result, so that the processing link of the plurality of probes and the solder pad is in correspondence with the processing link of the external probe and the puncture area.
[0051] (3) Based on the second measurement result, the first measurement result and the second measurement result are compared and processed to generate the measurement trigger signal, and the measurement trigger signal calls the pulse signal generator, drive unit, drive motherboard, drive daughterboard, drive integrated circuit, buffer capacitor and bus capacitor to perform measurement action processing.
[0052] The following are its main beneficial effects:
[0053] (1) In view of the problem that the alignment result and the contact state are separated in the existing solution, the first measurement result is formed by the external probe and the needle puncture area, so that the alignment result is first formed as a unified input content for needle pressure test, contact resistance test and visual inspection, and the measurement action processing no longer depends solely on the alignment completion state.
[0054] (2) In view of the problem that the contact state formed by multiple probes and solder pads in the existing solution is difficult to correspond to the previous contact state, the second measurement result is generated based on the first measurement result, so that the needle pressure test, contact resistance test and visual inspection processing of the multiple probes and solder pads and the corresponding processing of the external probe and the puncture area are connected in a sequential manner.
[0055] (3) In view of the problem that the measurement trigger signal lacks a prior judgment basis in the existing scheme, the measurement trigger signal is generated by comparing the first measurement result and the second measurement result, so that the measurement action processing is based on the correspondence between the needle pressure test result, the contact resistance test result and the needle tip image.
[0056] (4) By the sequential calling relationship between the measurement trigger signal and the pulse signal generator, the driving unit, the driving motherboard, the driving daughterboard, the driving integrated circuit, the buffer capacitor and the bus capacitor, the measurement result and the comparison processing form a continuous link, and the measurement result is consistent with the previous processing state.
[0057] (5) Through the continuous organization of the alignment result, the first measurement result, the second measurement result, the measurement trigger signal and the measurement result, the advanced process and wide bandgap semiconductor probe testing form a unified process of acquisition, alignment, comparison processing and measurement action processing, and the correlation link between the measurement result and the number of tests has a continuous recording basis. Attached Figure Description
[0058] Figure 1 A schematic flowchart illustrating an advanced process and wide-bandgap semiconductor probe testing method provided in this application embodiment;
[0059] Figure 2 This is a structural block diagram of an advanced process and wide bandgap semiconductor probe testing system provided in an embodiment of this application. Detailed Implementation
[0060] Example 1: Refer to Figure 1 This is a flowchart illustrating an advanced process and wide-bandgap semiconductor probe testing method provided by an embodiment of the present invention. The process may include at least steps S100-S500:
[0061] S100: Acquire the wafer, chip under test, solder pads, pin piercing area, probe card, multiple probes, external probes and vacuum tray; perform optical calibration system and X, Y, Z and θ direction movement processing to obtain alignment results;
[0062] S200. Based on the alignment results, perform needle pressure test, contact resistance test and visual inspection processing between the external probe and the puncture area to obtain the first measurement result;
[0063] S300. Based on the first measurement result, perform multiple probe-pad pin pressure tests, contact resistance tests, and visual inspection processing to obtain the second measurement result;
[0064] S400. Based on the second measurement result, perform a comparison process between the first measurement result and the second measurement result to obtain a measurement trigger signal;
[0065] S500. Based on the measurement trigger signal, perform measurement action processing to obtain the measurement result.
[0066] S100: Acquire the wafer, chip under test, solder pads, pin piercing area, probe card, multiple probes, external probes and vacuum tray; perform optical calibration system and X, Y, Z and θ direction movement processing to obtain alignment results;
[0067] Specifically, the wafer is a circular or square semiconductor substrate, and its surface is formed with multiple repeating chip unit regions through semiconductor processes. Each chip unit region is a chip under test (DUT). The entire wafer is held and supported by a vacuum tray. The DUT can be an advanced process logic chip, a memory chip, or a wide bandgap semiconductor device, including silicon carbide (SiC) or gallium nitride (GaN) devices. The surface of the DUT has multiple pads for electrical signal connections. These pads are square or circular conductive pads made of metal, and their surfaces are exposed in the openings of the chip passivation layer. The probe area is a dedicated metal area on the wafer that is physically isolated from the pads of the DUT. Its material is the same as or compatible with the pads. It is used to contact external probes to form an independent reference measurement loop. The position and size of the probe area are predefined during wafer manufacturing. The probe card is a ring-shaped or polygonal printed circuit board structure for mounting... On the head clamp of the probe station, multiple probes are arranged on its lower surface in a layout corresponding to the solder pads of the chip under test. Simultaneously, one or more external probes are positioned at its edge or specific locations. The multiple probes refer to a group of elastic conductive probes, matching the number of solder pads on the chip under test. Each probe is independently connected to a driver board or measurement circuit via wiring on a probe card, used to directly contact the solder pads and perform electrical measurements. The external probe is at least one dedicated probe independent of the multiple probes, with its tip facing the puncture area, used to establish contact with the puncture area before formal measurement to obtain a reference measurement result. The vacuum tray is set on the X, Y, and θ direction moving platform of the probe station. Its surface has multiple vacuum adsorption holes. A negative pressure generated by an external vacuum source is used to flatly adsorb and fix the wafer onto the tray, which moves with the tray in the horizontal and rotational directions. Simultaneously, the probe station controls the probe card to move up and down in the Z direction, achieving contact and separation between the probes and corresponding areas on the wafer. All of the above components together constitute the physical input basis of the test method of this invention, and each plays a specific functional role in subsequent alignment, measurement, and comparison processing.
[0068] Specifically, the wafer serves as the carrier of the chip under test (DUT), which is distributed on the wafer. The bonding pads are disposed on the surface of the DUT, and the pin-punch area is disposed on the wafer and spaced apart from the bonding pads. A probe card is mounted on a probe stage, multiple probes are disposed below the probe card, and an external probe is disposed on the probe card facing the pin-punch area. A vacuum tray is disposed on the movable end of the probe stage and carries the wafer. The wafer, the DUT, the bonding pads, the pin-punch area, the probe card, the multiple probes, the external probe, and the vacuum tray constitute the input for this step. This input is processed by the optical calibration system along the X, Y, Z, and θ directions.
[0069] Specifically, at the start of the automated testing process, the wafer is placed on the vacuum tray and fixed by a vacuum positioning mechanism. The distribution of the chip under test (DUT) on the wafer, the distribution of the solder pads on the DUT, and the position of the pin-piercing area on the wafer remain unchanged after the wafer is fixed. After the probe card is installed on the probe stage, the multiple probes and the external probe face the wafer on the vacuum tray. At this time, there is no contact between the multiple probes and the solder pads, and between the external probe and the pin-piercing area. The optical calibration system enters the working state, and the X, Y, Z, and θ direction movements enter the receiving state.
[0070] In this step, the optical calibration system acquires image information of the plurality of probes, the outer probe, the solder pads, and the pin-piercing area, and determines the position based on the image information. The X-direction and Y-direction movements change the position of the vacuum tray in the plane, the θ-direction movement changes the angular position of the wafer relative to the probe card, and the Z-direction movement changes the vertical spacing between the plurality of probes, the outer probe, and the wafer. The optical calibration system first acquires images of the outer probe and the pin-piercing area, then acquires images of the plurality of probes and the solder pads, and subsequently outputs adjustment commands for the X, Y, Z, and θ-direction movements. These adjustment commands are executed by the probe station, and after execution, the optical calibration system acquires images again and performs position determination, thus forming a process of acquisition, determination, movement, and re-acquisition.
[0071] In the X, Y, and θ direction movement processing, the optical calibration system first determines the positional correspondence between the external probe and the puncture area. If there is a positional offset between the external probe and the puncture area, the X-direction movement, Y-direction movement, and θ-direction movement are executed. After the positional offset decreases, the optical calibration system then determines the positional correspondence between the multiple probes and the solder pads. If there is a positional offset between the multiple probes and the solder pads, the X-direction movement, Y-direction movement, and θ-direction movement continue to be executed. After the external probe and the puncture area, and the multiple probes and the solder pads are all in a positional correspondence state, the Z-direction movement begins to perform contact preparation processing. The contact preparation processing does not initiate contact action; it only moves the external probe and the multiple probes to the starting distance for needle pressure testing, contact resistance testing, and visual inspection.
[0072] During the Z-direction movement process, the optical calibration system continuously acquires tip images of the external probe and the puncture area, as well as tip images of the multiple probes and the solder pads. If the tip images show a height deviation between the external probe or the multiple probes, the current Z-direction movement is stopped, and the process returns to the X, Y, and θ-direction movement processes. If the tip images show that the external probe and the puncture area, and the multiple probes and the solder pads are all in the same contact preparation state, the current X-direction movement position, Y-direction movement position, Z-direction movement position, and θ-direction movement position are recorded. The recorded information includes the positional correspondence between the external probe and the puncture area, the positional correspondence between the multiple probes and the solder pads, the current position of the vacuum tray, and the current installation status of the probe card.
[0073] Regarding boundary constraints and anomaly handling, if the optical calibration system detects in any round of image acquisition that the puncture area exceeds the current field of view, the solder pad exceeds the current field of view, the external probe fails to enter the current field of view, or multiple probes fail to enter the current field of view, or if it detects that the image state has not been updated after movement in the X, Y, Z, and θ directions, it terminates the current round of position determination, recalls the movement in the X, Y, Z, and θ directions to return to the previously recorded position, and then re-executes the current round of image acquisition. The previously recorded position is stored in the automated testing process record of the probe station. The automated testing process record is continuously updated in this step until the optical calibration system completes the final position determination. After the final position determination is completed, the positional correspondence between the external probe and the puncture area, the positional correspondence between the multiple probes and the solder pad, the termination status of the movement in the X direction, the termination status of the movement in the Y direction, the termination status of the movement in the Z direction, and the termination status of the movement in the θ direction are collectively recorded as the alignment result.
[0074] The alignment result is the output of this step. The alignment result includes the alignment of the external probe with the puncture area, the alignment of the multiple probes with the solder pad, the current state of the vacuum tray, and the final position determination corresponding to the optical calibration system. After this step, the alignment result is input into S200 and serves as the direct input for the needle pressure test, contact resistance test, and visual inspection processing of the external probe and the puncture area in S200. Based on this, S200 proceeds with the acquisition of the first measurement result.
[0075] The technical effect of this step is that the puncture area, the external probe, the multiple probes, and the solder pads enter the same alignment link, and the optical calibration system and the X, Y, Z, and θ direction movement processing no longer correspond to a single solder pad. The alignment result simultaneously covers the external probe and the puncture area, and the multiple probes and the solder pads, so the subsequent first measurement result and the second measurement result have the same positional basis. The automated testing process forms a unified starting point in this step, and the subsequent measurement trigger signals have direct preceding inputs.
[0076] S200. Based on the alignment results, perform needle pressure test, contact resistance test and visual inspection processing between the external probe and the puncture area to obtain the first measurement result;
[0077] Specifically, the alignment result output by S100 serves as the direct input for this step. The alignment result includes the positional correspondence between the external probe and the piercing area, the current position of the vacuum tray, and the current tip position status given by the optical calibration system. The external probe is a probe mounted on the probe card. The piercing area is a metal pad mounted on the wafer. The external probe does not engage with the measurement action of the chip under test. In this step, the external probe cooperates with the piercing area to first establish a current contact state, and then provides a first measurement result to subsequent steps. The first measurement result is generated in this step and input to S300 after this step, serving as a preliminary reference for the multiple probes and the pads to perform pin pressure testing, contact resistance testing, and visual inspection processing.
[0078] Specifically, after receiving the alignment result, the probe station first locks the current positions of the X-direction movement, the Y-direction movement, and the θ-direction movement, and then invokes the Z-direction movement to perform contact preparation. The contact preparation is jointly completed by the vacuum tray, the probe card, and the external probe. The vacuum tray maintains the wafer in a fixed state. The probe card remains in its installed state. The external probe moves along the Z-direction towards the puncture area. The optical calibration system continuously acquires tip images during the descent of the external probe and correlates these images with the current position in the alignment result. If the tip image shows that the external probe deviates from the puncture area, the probe station stops the current Z-direction movement, invokes the X-direction movement, the Y-direction movement, or the θ-direction movement for overlap processing, and then re-executes the Z-direction movement. After the overlap processing is completed, the optical calibration system re-acquires tip images until the external probe enters the contact preparation position of the puncture area.
[0079] Further, the needle pressure test is performed after the outer probe and the puncture area enter the contact preparation position. The input of the needle pressure test includes the current position of the outer probe, the current position of the puncture area, and the Z-direction movement state of the current wheel. The processing procedure of the needle pressure test is as follows: control the outer probe to continue descending in the Z-direction, so that the outer probe makes contact with the puncture area; after contact, record the contact state during the continued descent of the outer probe; and then form the needle pressure test result based on the contact state. The needle pressure test result reflects the degree of contact between the outer probe and the puncture area in the current wheel. If the needle tip image jumps, the outer probe deviates from the puncture area, or the Z-direction movement is interrupted during the needle pressure test, the needle pressure test of the current wheel is stopped, the abnormal record is retained, and the test is re-executed at the contact preparation position.
[0080] Furthermore, the contact resistance test and the needle pressure test are performed consecutively. The input for the contact resistance test is the contact state between the external probe and the puncture area after the needle pressure test is completed. The contact resistance test is conducted by the continuity state between the probe card and the puncture area. The contact resistance test acquires multiple contact resistance test values while the external probe maintains its current contact state. These multiple contact resistance test values are then evaluated for stability to form a contact resistance test result. If a sudden change occurs in the multiple contact resistance test values during continuous acquisition, the current round of contact resistance test is recorded as an abnormal record, and the Z-direction movement is re-initiated to perform upward and downward processing again, and the contact resistance test result is acquired again. The abnormal record and the needle pressure test result are saved accordingly for subsequent steps.
[0081] Furthermore, the visual inspection is performed both before and after the needle pressure test and the contact resistance test. The visual inspection uses the optical calibration system to acquire the tip image of the external probe and the image state of the puncture area. The visual inspection process includes pre-contact tip image acquisition, during-contact tip image acquisition, and post-contact tip image acquisition. The pre-contact tip image is used to confirm that the external probe and the puncture area are in the position corresponding to the alignment result. The during-contact tip image is used to confirm that the contact position between the external probe and the puncture area has not shifted. The post-contact tip image is used to confirm the state of the external probe and the puncture area after the current round of contact is completed. If the tip image at any stage shows that the external probe deviates from the puncture area, or shows that the image state of the external probe and the puncture area is discontinuous, the visual inspection output for the current round is an abnormal record, and a readjustment action is triggered in the X-direction movement, Y-direction movement, Z-direction movement, and θ-direction movement.
[0082] Understandably, in a complete and operable engineering embodiment, after the wafer is fixed to the vacuum tray, step S100 has already provided the alignment result between the external probe and the piercing area. At the beginning of this step, the probe station first keeps the vacuum tray stationary. Then, it controls the external probe to descend. The optical calibration system first acquires an image of the probe tip before contact. Next, it performs the needle pressure test and simultaneously acquires an image of the probe tip during contact. Then, it performs the contact resistance test, continuously outputting the contact resistance test value. Finally, it acquires an image of the probe tip after contact. The probe station combines the needle pressure test result, the contact resistance test result, and the probe tip image and records them as the first measurement result. If an abnormal record occurs at any time, the probe station stops the current round of processing and recalls the current position from the alignment result to re-perform contact preparation processing. The above process is automatically completed by the probe station without manual intervention.
[0083] At the end of this step, the first measurement result is generated as the output product. The first measurement result includes the pin pressure test result, the contact resistance test result, and the pin tip image. The first measurement result is written into the automated test process record and directly input into S300 after this step. After reading the first measurement result, S300 performs pin pressure testing, contact resistance testing, and visual inspection processing on the multiple probes and the solder pads. S400 then compares the first measurement result with the second measurement result.
[0084] The technical effect of this step can be summarized as follows: the external probe and the puncture area first form a first measurement result, and the subsequent processing of the multiple probes and the solder pads has a unified pre-processing input. The needle pressure test, the contact resistance test, and the visual inspection are executed continuously within the same contact link, and the field content of the first measurement result is complete. After the first measurement result enters S300 and S400, the formation of the measurement trigger signal has a direct pre-processing basis.
[0085] S300. Based on the first measurement result, perform multiple probe-pad pin pressure tests, contact resistance tests, and visual inspection processing to obtain the second measurement result;
[0086] Specifically, the first measurement result output by S200 serves as the direct input to this step. The first measurement result includes the pin pressure test result, the contact resistance test result, and the pin tip image. The plurality of probes are positioned on the underside of the probe card, and the solder pads are positioned on the surface of the chip under test, with each probe corresponding to one of the solder pads. The plurality of probes are specially customized for the solder pads and are arranged in a single plane. The first measurement result serves as a contact reference in this step. After entering this step, the first measurement result first participates in the contact preparation between the plurality of probes and the solder pads, and then participates in the continuous operation of the pin pressure test, the contact resistance test, and the visual inspection.
[0087] Specifically, after reading the first measurement result, the probe station retrieves the alignment result from S100 and maintains the current positions of the vacuum tray, the probe card, and the plurality of probes. The needle pressure test result in the first measurement result is used to provide a contact reference when the plurality of probes descend to make contact in this round. The contact resistance test result in the first measurement result is used to provide a contact reference for the conduction state in this round. The needle tip image in the first measurement result is used to provide an image reference for the needle tip position in this round. Based on the above references, the probe station controls the plurality of probes to move along the Z direction to approach the solder pad, and the optical calibration system synchronously acquires the needle tip images of the plurality of probes and the solder pad. If the needle tip image shows that the plurality of probes are completely deviated from the solder pad, the current round of contact processing is paused, and the X-direction movement, the Y-direction movement, and the θ-direction movement are called back to the position corresponding to the alignment result before the current round of contact processing is re-executed.
[0088] Further, after the multiple probes enter the contact preparation position, the pin pressure test begins. The pin pressure test is completed by the probe station controlling the multiple probes to continue descending. After the multiple probes contact the solder pad, the probe station records the contact state of each probe and forms a corresponding pin pressure test result. The pin pressure test result is not a single numerical record, but rather a record of the contact state of each of the multiple probes, which is then summarized into an overall pin pressure test result for this round of multiple probes and the solder pad. When acquiring the overall pin pressure test result, the probe station performs an internal check against the pin pressure test result in the first measurement result. If the overall pin pressure test result deviates significantly from the pin pressure test result in the first measurement result, the contact resistance test and visual inspection for this round are stopped, the current state is recorded as an abnormal record, and the contact processing between the multiple probes and the solder pad is repeated.
[0089] Further, after the needle pressure test is completed, the contact resistance test is performed. The contact resistance test is performed while the multiple probes maintain their current contact state with the solder pad. The probe holder forms a conductive link with the solder pad, and the probe station continuously acquires multiple contact resistance test values and performs a stability judgment on the multiple contact resistance test values. The stability judgment includes three operating actions: continuous acquisition, continuous recording, and continuous comparison. Continuous acquisition is used to obtain multiple contact resistance test values in the same round of contact. Continuous recording is used to form the contact resistance test trajectory of each of the multiple probes. Continuous comparison is used to compare the contact resistance test trajectory with the contact resistance test result in the first measurement result. If any of the multiple probes experiences a sudden change in contact resistance test value in the current round, or if the overall contact resistance test trajectory of the multiple probes deviates from the contact reference in the first measurement result, the subsequent processing of this round is terminated, and the probe station controls the multiple probes to rise and detach from the solder pad, returning to the contact preparation position.
[0090] Furthermore, the visual inspection is performed throughout the entire process of the pin pressure test and the contact resistance test. The visual inspection is executed by the optical calibration system. The optical calibration system acquires a pre-contact pin tip image before the multiple probes descend, a contact pin tip image during contact, and a post-contact pin tip image before the multiple probes rise. The pre-contact pin tip image is used to confirm the initial positional state of the multiple probes relative to the solder pad. The contact pin tip image is used to confirm the continuity of the positional state of the multiple probes during the contact process. The post-contact pin tip image is used to confirm the pin tip state after the current round of contact ends. The optical calibration system compares the above pin tip images with the pin tip images in the first measurement result. If the image comparison shows that any of the multiple probes exhibits a positional jump, pin tip deflection, or detachment from the solder pad, the current round corresponding to that probe is recorded as an abnormal record, and the processing of the contact resistance test results for that round is terminated.
[0091] Understandably, in one engineering embodiment, the chip under test (DUT) is an advanced process chip, the solder pads are distributed on the surface of the DUT, and the plurality of probes on the probe card correspond one-to-one with the solder pads. In another engineering embodiment, the DUT is a wide bandgap power semiconductor device, including silicon carbide (SiC) devices or gallium nitride (GaN) devices. The operating sequence is consistent in both engineering embodiments. First, the first measurement result is read, then the plurality of probes are controlled to descend, and then the pin pressure test, the contact resistance test, and the visual inspection are performed. In any engineering embodiment, the probe station uniformly writes the contact state, contact resistance test value, and pin tip image corresponding to each probe into the automated test process record.
[0092] At the end of this step, the probe station summarizes and records the overall pin pressure test results, overall contact resistance test results, and overall pin tip image of the multiple probes and the solder pad as the second measurement result. The second measurement result is generated as the output of this step. The second measurement result is directly input into S400 after this step. S400 compares the first measurement result with the second measurement result and generates the measurement trigger signal accordingly. If an anomaly is recorded in this step, the second measurement result is written to the update status, and the corresponding automated test process record is retained before the S400 call.
[0093] The technical effect of this step is that the first measurement result and the actual contact processing of the multiple probes enter the same operational chain, and the field content of the second measurement result is complete. The pin pressure test, the contact resistance test, and the visual inspection are carried out continuously between the multiple probes and the solder pads, and the S400 has direct input content when reading the second measurement result. The contact state of the multiple probes and the solder pads is formed into a unified record in this step, and the subsequent measurement trigger signal has a basis for prior comparison.
[0094] S400. Based on the second measurement result, perform a comparison process between the first measurement result and the second measurement result to obtain a measurement trigger signal;
[0095] Specifically, the first measurement result output by S200 and the second measurement result output by S300 are used together as inputs for this step. The first measurement result includes the needle pressure test result, the contact resistance test result, and the needle tip image. The second measurement result includes the needle pressure test result, the contact resistance test result, and the needle tip image. The comparison processing is performed within the probe station. After reading the first and second measurement results, the probe station first establishes a record of the same round of automated testing, and then reads the corresponding fields in the two measurement results. The corresponding fields include the needle pressure test result, the contact resistance test result, and the needle tip image. The record of the same round of automated testing is written at the beginning of this step and saved at the end of this step.
[0096] Specifically, the comparison process first performs a needle pressure test result comparison process. The probe station reads the needle pressure test result from the first measurement result, then reads the needle pressure test result from the second measurement result, and establishes a correspondence based on the contact state between the outer probe and the puncture area, and the contact state between the multiple probes and the solder pad. The needle pressure test result comparison process includes difference judgment, sequence judgment, and continuous state judgment. The difference judgment is used to determine whether the two needle pressure test results are at the same contact level. The sequence judgment is used to determine whether the sequential state of the multiple probes during the contact process is consistent with the contact state of the outer probe. The continuous state judgment is used to determine whether there is an interruption, jump, or rollback during the current round of contact. If an inconsistent state occurs in the needle pressure test result comparison process, the current round is recorded as an inactive state, and subsequent measurement actions are paused.
[0097] Further, the comparison process then performs a contact resistance test result comparison process. The probe station reads the contact resistance test results from the first measurement result and the second measurement result, and performs interval judgment and continuity judgment on the two contact resistance test results. The interval judgment is used to confirm whether the conduction state formed by the multiple probes and the solder pad matches the conduction state formed by the outer probe and the piercing area. The continuity judgment is used to confirm whether the contact resistance test result in the second measurement result remains continuous in the current round of automated testing process records. If there is a sudden change in the contact resistance test result in the second measurement result, or if there is a significant deviation between the contact resistance test result in the first measurement result and the contact resistance test result in the second measurement result, the probe station records the contact resistance test result comparison process as failed and writes the current round status into the exception record.
[0098] Furthermore, the comparison process also performs tip image comparison processing. The optical calibration system has already obtained the tip image in steps S200 and S300; this step calls upon the tip image to determine its position. The tip image comparison processing includes pre-contact image correspondence judgment, in-contact image continuity judgment, and post-contact image return judgment. The pre-contact image correspondence judgment confirms that the outer probe and the puncture area, and the multiple probes and the solder pad, are aligned on the same basis before the contact action. The in-contact image continuity judgment confirms that the tip position has not shifted during the contact process. The post-contact image return judgment confirms that the multiple probes do not exhibit abnormal dwelling after the current wheel contact ends. If there is a position jump, tip skew, or partial detachment during the tip image comparison processing, the probe station writes the current wheel state into an error record and keeps the measurement trigger signal disabled.
[0099] Understandably, the comparison processing does not judge a single field, but rather executes the needle pressure test result comparison processing, the contact resistance test result comparison processing, and the needle tip image comparison processing in a fixed order. If the previous processing outputs a failure status, the next processing continues to execute, and the probe station saves a complete record. When all three current processing steps output a pass status, the probe station generates an enabled status. If any of the three current processing steps outputs a failure status, the probe station generates a disabled status. The enabled and disabled statuses constitute the status field of the measurement trigger signal. The measurement trigger signal is generated in this step but does not involve performing any measurement action in this step.
[0100] In one engineering embodiment, the chip under test is an advanced process chip. The probe station first reads the pin pressure test result from the first measurement result, then reads the pin pressure test result from the second measurement result, completing the comparison processing of the current round pin pressure test results. Subsequently, it reads the contact resistance test results from the first measurement result and the contact resistance test results from the second measurement result, completing the comparison processing of the current round contact resistance test results. Then, it calls the pin tip image stored in the optical calibration system to complete the comparison processing of the current round pin tip image. If all three processes pass, the measurement trigger signal of the enabled state is output. If any process fails, the measurement trigger signal of the disabled state is output, and the current round state is marked in the automated test process record for reuse by S100, S200, and S300.
[0101] In another engineering embodiment, the chip under test (DUT) is a wide-bandgap power semiconductor device. The wide-bandgap power semiconductor device includes silicon carbide (SiC) devices or gallium nitride (GaN) devices. In this engineering embodiment, the execution sequence of the comparison processing is consistent with the aforementioned engineering embodiment. The probe station first aligns the fields of the first and second measurement results, then performs three comparison processes, and finally outputs the measurement trigger signal. When the current measurement trigger signal is in an inactive state, the automated testing process record retains the current round of anomaly records for reuse in the optical calibration system and the X, Y, Z, and θ direction movement processing.
[0102] At the end of this step, the probe station writes the comparison conclusion between the first measurement result and the second measurement result into the measurement trigger signal. The measurement trigger signal is generated as the output product of this step. After this step, the measurement trigger signal is directly input into S500. Based on the measurement trigger signal, S500 performs measurement operations on the pulse signal generator, drive unit, drive motherboard, drive daughterboard, drive IC, buffer capacitor, and bus capacitor, and obtains the measurement result.
[0103] The technical effect of this step is that the first measurement result and the second measurement result are recorded in the same round of automated testing process, and the measurement trigger signal has a clear input basis. The needle pressure test result comparison processing, the contact resistance test result comparison processing, and the needle tip image comparison processing run continuously in this step, and the status field of the measurement trigger signal is complete. When S500 reads the measurement trigger signal, the preceding judgment link is complete.
[0104] In one specific embodiment:
[0105] In S400, based on the second measurement result, a comparison process is performed between the first measurement result and the second measurement result to obtain a measurement trigger signal.
[0106] The first stage of the comparison process is the needle pressure test result comparison process. The input for this stage is the needle pressure test result field from the first measurement result output by S200. (Derived from the peak pressure and steady-state pressure deviation when the external probe contacts the puncture site), and the needle pressure test result field in the second measurement result output by S300. (Source: from multiple probes) The deviation between the peak pressure and steady-state pressure of the solder pad corresponding to the root probe. , (Total number of probes). After reading the above fields, the probe station first performs an alignment operation recorded in the same round of automated testing to establish the contact state vector between the outer probe and the puncture area. Contact state vector between multiple probes and solder pads The correspondence between them. To quantify the degree of difference between needle pressure test results, a weighted bias fusion method with tolerance intervals is adopted. Formula ① calculates the first... Channel needle pressure deviation score:
[0107]
[0108] in: : No. The needle pressure deviation score for the channel is calculated using this formula;
[0109] : Probe channel index, value range ,in This represents the total number of probes.
[0110] The needle pressure deviation value in the first measurement result comes from the deviation between the peak pressure and the steady-state pressure when the external probe contacts the puncture area in step S200.
[0111] The second measurement result The needle pressure deviation value of the channel originates from step S300. The deviation between the peak pressure and steady-state pressure of the solder pad corresponding to the root probe;
[0112] : To prevent the loss of a small amount, a fixed value is taken. g is used to avoid the denominator being zero;
[0113] : No. The needle pressure weighting coefficient of the channel is derived from the preset probe card configuration table, and its value range is... .
[0114] Simple numerical example: Let g, g, , ,but This score represents the relative magnitude of needle pressure deviation; if it exceeds a preset threshold... Then it is determined to be inconsistent.
[0115] Formula ① solves the quantification problem of "difference judgment" in needle pressure test results, converts absolute deviation into relative deviation, and allows different channels to use different weights.
[0116] After obtaining the needle pressure deviation scores for each channel, it is necessary to further determine the consistency and continuity of the contact process. Therefore, formula ② is used to construct a needle pressure timing consistency index:
[0117]
[0118] in: Needle pressure timing consistency index, value It is calculated using this formula;
[0119] Total number of probes, a positive integer, derived from the number of physical channels on the probe card;
[0120] : Probe channel index, same as formula ①;
[0121] : Indicator function, which takes the value when the condition within the parentheses is true. Otherwise ;
[0122] Same as formula ①, for the first... Channel needle pressure deviation score;
[0123] Needle pressure deviation scoring threshold, preset value is taken as follows: This is used to determine whether the needle pressure is consistent;
[0124] : No. The moment when the channel probe reaches steady-state pressure is provided by the timing record inside the probe station.
[0125] The moment when the external probe reaches steady-state pressure, source same as... ;
[0126] Maximum allowable time difference, preset value is... ms.
[0127] Simple numerical example: Let , (≤1.0) If the value is greater than 1.0, then the indicator function values are 1 and 0 respectively; ms, ms, ms, ms, then the first item is The second item is ; .
[0128] This indicator is below the preset consistency threshold. At that time, the probe station determines that the needle pressure test result comparison process is not passed and records the unenabled state.
[0129] The output field at this stage is the needle pressure consistency flag. (0 for failure, 1 for success) and the deviation score array for each channel. This output is then directly used in the next stage of contact resistance comparison.
[0130] The second stage of the comparison process is the comparison processing of contact resistance test results. This stage follows from the output of the first stage. and Additionally, the contact resistance deviation value from the first measurement result is read. (The deviation of the four-wire Kelvin resistance measured between the external probe and the puncture area from the ideal value) and the contact resistance deviation value in the second measurement result. (No. (Contact resistance deviation between the channel probe and the solder pad). First, determine the interval, and then calculate the contact resistance matching degree using formula ③:
[0131]
[0132] in: : No. Contact resistance matching degree of the channel, value ;
[0133] : Probe channel index, same as before;
[0134] : Natural exponential function;
[0135] The contact resistance deviation value in the first measurement result comes from the deviation between the four-wire Kelvin measured resistance between the outer probe and the puncture area and the ideal value in step S200.
[0136] The second measurement result The contact resistance deviation value of the channel originates from step S300. Contact resistance deviation between the root probe and the solder pad;
[0137] The resistance deviation scale parameter is derived from historical data statistics. mΩ;
[0138] Needle pressure penalty coefficient, taken as ;
[0139] Indicator function, same as formula ②;
[0140] : Same as formula ①;
[0141] : Same as formula ②.
[0142] Simple numerical example: Let mΩ, mΩ, ,but ;like If the value is less than or equal to 1.0, then the penalty term is 0. .like The penalty item is , The matching degree is below the threshold. The interval is judged as mismatched.
[0143] After determining the interval, it is also necessary to detect the continuous state of the contact resistance, that is, to determine whether the contact resistance in the second measurement result has a sudden change in the current automated testing process.
[0144] Therefore, formula ④ constructs the resistance continuity stability index:
[0145]
[0146] in: : Resistance continuity stability index;
[0147] Total number of probes, same as before;
[0148] : Probe channel index, same as before;
[0149] Same as formula ③, the second measurement result The contact resistance deviation value of the channel;
[0150] : No. The sliding average of the contact resistance deviation values of the channel in the first three tests in the current round is read from the probe station's internal buffer.
[0151] : To prevent the loss of a small amount, a fixed value is taken. mΩ.
[0152] Simple numerical example: Let The first three moving averages mΩ, in this measurement mΩ, ,but , A cutoff value of 0 indicates a severe mutation.
[0153] If this time mΩ, then , .when Less than the continuity threshold At that time, the probe station determined that the contact resistance comparison process failed.
[0154] Output resistance consistency indicator at this stage and the matching degree of each channel This output is used in the third stage of tip image comparison processing.
[0155] The third stage of the comparison processing is the needle tip image comparison processing. This stage follows the output of the first two stages. , and matching degree It also retrieves the tip image data stored in S200 and S300 by the optical calibration system.
[0156] Specifically, a reference image is obtained before the external probe makes contact with the needle puncture area. Images in contact and the repositioning image after contact Simultaneously acquire the first of multiple probes Image of the root probe before contact on the corresponding solder pad Images in contact Images after contact All images are grayscale, resolution... Pixels. First, perform image correspondence determination before contact, then calculate the position offset using formula ⑤:
[0157]
[0158]
[0159] in: : No. The sum of squared differences in image gradients of the channel in the X direction (unit: pixel²) is calculated using this formula;
[0160] : Probe channel index, same as before;
[0161] Image region Area (number of pixels) Near the needle tip The region of interest of a pixel, therefore pixel²;
[0162] Image pixel coordinates Horizontal index (unit: pixels). Vertical index (unit: pixels);
[0163] Region of interest, from near the tip of the needle. Pixel composition;
[0164] : No. Image in pixels before channel probe contact The gradient along the horizontal direction is calculated by the Sobel operator;
[0165] Image in pixels before external probe contact The gradient along the horizontal direction originates from the same source as above;
[0166] : No. The sum of squared differences in image gradients in the Y direction of the channel (in pixels²) is defined as follows: Similarly, simply replace the horizontal gradient with the vertical gradient. ;
[0167] “similarly”: indicates The calculation method and They are completely symmetrical, differing only in gradient direction.
[0168] Simple numerical example: Define the region of interest The accumulated sum of squared gradient differences is calculated to be... ,but Pixel². Similarly. Pixels². Total offset Pixels. Preset position offset threshold. If the value is 2 pixels, it is considered a pass.
[0169] Formula ⑤ solves the problem of image correspondence determination before contact, ensuring that the external probe and multiple probes are on the same alignment basis.
[0170] Next, the continuity of images during contact is determined using formula ⑥ to calculate the temporal consistency score of the images during the contact process:
[0171]
[0172] in: Image temporal consistency score, with values... It is calculated using this formula;
[0173] Total number of probes, same as before;
[0174] : Probe channel index, same as before;
[0175] The Structural Similarity (SSIM) function takes two grayscale images as input and outputs a range of values. The similarity value;
[0176] The images of the external probe during the contact process are derived from contact images stored in the optical calibration system;
[0177] : No. Images of the channel probe during the contact process, sourced from the same source as above;
[0178] : No. The acquisition time of the image during channel contact is recorded by the probe station timing.
[0179] The image acquisition time during external probe contact is the same as above;
[0180] : Time window width, fixed value ms;
[0181] : Same as formula ③, No. Contact resistance matching degree of the channel.
[0182] Simple numerical example: Let SSIM Time difference ms, Then the time term ;set up ,but .
[0183] If a preset image consistency threshold is set If the result is not met, the probe station will record an anomaly. The post-contact image repositioning determination uses a similar formula (⑤), but the comparison object is the gradient difference between the post-contact image and the pre-contact image; this will not be repeated here.
[0184] If all three image judgments pass, and the first two stages are successful... , If the probe station generates an enabled measurement trigger signal, the status field will be "Enabled" and accompanied by a diagnostic code "00"; otherwise, an disabled state will be generated, and the diagnostic code will indicate the failure stage (e.g., "01" indicates needle pressure failure, "10" indicates resistance failure, and "11" indicates image failure). This measurement trigger signal is the output of this step and is directly input into S500 for measurement processing.
[0185] At the end of this step, the probe station writes the comparison conclusion between the first and second measurement results into the measurement trigger signal and saves the record of the same round of automated testing.
[0186] This section summarizes the technical effects: Through three-stage continuous comparison processing, the needle pressure deviation and timing consistency, contact resistance matching degree and continuity, and needle tip image position and timing consistency are quantified respectively. The output of the previous stage is used as the weighted input of the next stage, realizing the joint determination of multi-modal contact state. Finally, a measurement trigger signal carrying a diagnostic code is output, providing a traceable and reversible gating basis for subsequent measurement actions.
[0187] S500. Based on the measurement trigger signal, perform measurement action processing to obtain the measurement result.
[0188] Specifically, the measurement trigger signal output by S400 serves as the direct input to this step. The measurement trigger signal includes an enabled state and an disabled state. The pulse signal generator, the drive unit, the drive motherboard, the drive daughterboard, the drive IC (Integrated Circuit), the buffer capacitor, and the bus capacitor together constitute the execution link for the measurement action processing. The drive unit includes the drive motherboard and the drive daughterboard. The drive daughterboard is plugged into and plugged into the drive motherboard. The drive daughterboard is equipped with the drive IC and the buffer capacitor. The drive motherboard is equipped with the bus capacitor. After the measurement trigger signal enters this step, it is first read by the drive unit, which then determines whether the current round of measurement action should enter the execution state.
[0189] Specifically, when the measurement trigger signal is enabled, the pulse signal generator starts outputting a pulse signal. The pulse signal enters the driving unit. The driving unit first receives the pulse signal from the driving motherboard, and then the driving motherboard sends the pulse signal to the driving daughterboard. The driving IC performs driving processing on the pulse signal entering the driving daughterboard. The buffer capacitor participates in the current wheel driving processing on the driving daughterboard. The bus capacitor participates in the current wheel driving processing on the driving motherboard. After the driving processing begins, the probe card maintains its current contact state with the chip under test, and the multiple probes maintain their current wheel contact positions with the solder pads, subsequently entering the measurement operation.
[0190] Furthermore, the measurement action is executed continuously after the pulse signal generator outputs a pulse signal. The input content of the measurement action includes the measurement trigger signal, the pulse signal, the drive processing status, and the contact status between the multiple probes of the current wheel and the solder pads. The operation process of the measurement action is as follows: first, the pulse signal generator outputs the pulse signal of the current wheel; then, the drive motherboard and the drive daughterboard complete the signal transmission of the current wheel; next, the drive IC completes the drive processing of the current wheel; simultaneously, the buffer capacitor and the bus capacitor maintain the drive link status of the current wheel; then, the probe card and the multiple probes perform the current wheel measurement action on the chip under test; and the measurement result is output after the measurement action is completed. The measurement result is the output field name of this step. The measurement result is written into the current wheel automated test process record and saved corresponding to the measurement trigger signal.
[0191] Furthermore, when the measurement trigger signal is in an inactive state, the pulse signal generator remains silent and does not output pulse signals. The drive motherboard, the drive daughterboard, the drive IC, the buffer capacitor, and the bus capacitor remain in the current wheel pending execution state. The probe card does not enter the measurement action. The probe station writes the current wheel state into the automated test process record and calls the corresponding steps S100, S200, and S300 to re-perform the optical calibration system and the X, Y, Z, and θ direction movement processing, the external probe contact processing with the puncture area, and the multiple probes contact processing with the solder pads. In this step, the inactive state does not output new pulse signals, only the current wheel blocking record.
[0192] Understandably, in one engineering embodiment, the chip under test (DUT) is an advanced process chip. After the S400 outputs the measurement trigger signal indicating an enabled state, the pulse signal generator outputs the current round pulse signal. The driver motherboard receives and transmits the pulse signal, the driver IC on the driver daughterboard completes the drive processing, and the buffer capacitor and the bus capacitor maintain the current round drive link state. Subsequently, the measurement action is performed on the DUT through the probe card and the multiple probes, and the measurement result of the current round is output. The measurement result is directly bound and saved to the current round automated test process record after output.
[0193] In another engineering embodiment, the chip under test (DUT) is a wide-bandgap power semiconductor device. The wide-bandgap power semiconductor device includes silicon carbide (SiC) or gallium nitride (GaN). In this engineering embodiment, the execution chain of the measurement action is consistent with the aforementioned engineering embodiment. When the measurement trigger signal is in the enabled state, the pulse signal generator output is executed first, then the drive unit processes the signal, and finally the current round measurement action on the probe card is executed. After the measurement result is output, it is written into the automated test process record and used for subsequent marking processing of the external probe on the puncture area based on the measurement result, thereby obtaining the number of tests.
[0194] At the end of this step, the measurement result is generated as the output of this step. The measurement result comes from the continuous measurement operation processing of the pulse signal generator, the drive unit, the drive motherboard, the drive daughterboard, the drive IC, the buffer capacitor, and the bus capacitor. After this step, the measurement result is written into the automated test process record and then used for subsequent marking of the needle puncture area by the external probe based on the measurement result.
[0195] The technical effect of this step is that the measurement trigger signal and the measurement action are seamlessly connected within the same operational link, and the measurement result has a direct preceding input. The pulse signal generator, the drive unit, the drive motherboard, the drive daughterboard, the drive IC, the buffer capacitor, and the bus capacitor operate continuously in a fixed order in this step, and the source of the measurement result is clear. When the measurement result is used for subsequent marking of the puncture area by the external probe, the current round of automated testing is fully recorded.
[0196] Example 2: Figure 2 A structural block diagram of an advanced process and wide-bandgap semiconductor probe testing system according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include:
[0197] Alignment result generation module 01 is used to acquire a wafer, a chip under test (DUT), solder pads, a pin piercing area, a probe card, multiple probes, an external probe, and a vacuum tray, and to perform optical calibration system and X, Y, Z, and θ direction movement processing to obtain alignment results. Specifically, the alignment result generation module receives the wafer, DUT, solder pads, pin piercing area, probe card, multiple probes, external probe, and vacuum tray. The wafer is placed on the vacuum tray. The DUT is distributed on the wafer. The solder pads are disposed on the surface of the DUT. The pin piercing area is disposed on the wafer and separated from the solder pads. The probe card is mounted on a probe stage, and the multiple probes and the external probe are disposed under the probe card. The optical calibration system acquires images of the external probe and the pin piercing area and images of the multiple probes and the solder pads, and drives the X, Y, Z, and θ direction movement processing. The X, Y, and θ direction movement processing is used to adjust the planar and angular positions of the vacuum tray and the probe card. The Z direction movement processing is used to adjust the vertical spacing between the multiple probes, the outer probe, and the wafer. The alignment result generation module performs a position determination after each round of image acquisition. If the image state is inconsistent with the current movement state, the X, Y, Z, and θ direction movement processing is repeated until the outer probe and the pin-piercing area form a positional correspondence, and the multiple probes and the solder pads form a positional correspondence. The alignment result generation module records the current position state as the alignment result and outputs the alignment result to the first measurement result generation module.
[0198] The first measurement result generation module 02, connected to the alignment result generation module, is used to perform needle pressure testing, contact resistance testing, and visual inspection processing on the external probe and the puncture area based on the alignment result, to obtain a first measurement result. Specifically, the first measurement result generation module receives the alignment result output by the alignment result generation module and calls the external probe and the puncture area into a contact preparation state. The first measurement result generation module first maintains the X, Y, and θ direction movement states according to the alignment result, and then calls the Z direction movement processing to bring the external probe closer to the puncture area. After the external probe contacts the puncture area, the first measurement result generation module continuously executes needle pressure testing, contact resistance testing, and visual inspection processing. The needle pressure testing process records the contact state between the external probe and the puncture area. The contact resistance testing process records the conduction state. The visual inspection processing records the needle tip image. If the needle tip image shows a position jump, or the contact resistance test shows a sudden change, the first measurement result generation module stops the current contact processing and re-calls the current position state in the alignment result. The first measurement result generation module summarizes the needle pressure test result, contact resistance test result and needle tip image to form the first measurement result, and outputs the first measurement result to the second measurement result generation module and the comparison processing module.
[0199] The second measurement result generation module 03, connected to the first measurement result generation module, is used to perform pin pressure testing, contact resistance testing, and visual inspection processing on the multiple probes and the solder pad based on the first measurement result, to obtain a second measurement result. Specifically, the second measurement result generation module receives the first measurement result output by the first measurement result generation module and reads the current position state in the alignment result. The second measurement result generation module controls the multiple probes to move along the Z direction to approach the solder pad. After the multiple probes and the solder pad enter a contact state, the second measurement result generation module sequentially performs pin pressure testing, contact resistance testing, and visual inspection processing. The pin pressure testing process establishes contact states for each of the multiple probes and then summarizes them into an overall pin pressure testing result. The contact resistance testing process continuously acquires the conduction state between the multiple probes and the solder pad and then forms an overall contact resistance testing result. The visual inspection processing acquires the pin tip images of the multiple probes before contact, during contact, and after contact. The second measurement result generation module internally references the current pin pressure testing result, contact resistance testing result, and pin tip image with the corresponding content in the first measurement result. If the current wheel state is inconsistent with the first measurement result, the second measurement result generation module stops the current contact processing and re-calls the alignment result and the first measurement result. The second measurement result generation module summarizes the needle pressure test results, contact resistance test results, and needle tip images formed by the multiple probes and the solder pads to form the second measurement result, and outputs the second measurement result to the comparison processing module.
[0200] The comparison processing module 04, connected to the first measurement result generation module and the second measurement result generation module, is used to compare the first measurement result with the second measurement result based on the second measurement result to obtain a measurement trigger signal. Specifically, the comparison processing module simultaneously receives the first measurement result output by the first measurement result generation module and the second measurement result output by the second measurement result generation module. The comparison processing module first compares the needle pressure test result in the first measurement result with the needle pressure test result in the second measurement result, then compares the contact resistance test result in the first measurement result with the contact resistance test result in the second measurement result, and finally compares the needle tip image in the first measurement result with the needle tip image in the second measurement result. If all three comparisons pass, the comparison processing module generates the measurement trigger signal in an enabled state. If any comparison fails, the comparison processing module generates the measurement trigger signal in an disabled state and calls the alignment result generation module, the first measurement result generation module, and the second measurement result generation module to re-execute the corresponding processing. The comparison processing module outputs the measurement trigger signal to the measurement action processing module.
[0201] The measurement action processing module 05, connected to the comparison processing module, is used to perform measurement action processing on the pulse signal generator, drive unit, drive motherboard, drive daughterboard, drive integrated circuit, buffer capacitor, and bus capacitor based on the measurement trigger signal, to obtain the measurement result. Specifically, the measurement action processing module receives the measurement trigger signal output by the comparison processing module. The measurement action processing module includes the pulse signal generator, the drive unit, the drive motherboard, the drive daughterboard, the drive integrated circuit, the buffer capacitor, and the bus capacitor. The drive unit includes the drive motherboard and the drive daughterboard. The drive daughterboard is plugged into and plugged into the drive motherboard. The drive integrated circuit and the buffer capacitor are disposed on the drive daughterboard. The bus capacitor is disposed on the drive motherboard. When the measurement trigger signal is in the enabled state, the measurement action processing module calls the pulse signal generator to output a pulse signal, which is transmitted through the drive motherboard and the drive daughterboard, driven by the drive integrated circuit, and maintained by the buffer capacitor and the bus capacitor. The probe card and the plurality of probes perform measurement actions on the chip under test in the current contact state. If the measurement trigger signal is in an inactive state, the measurement action processing module remains in a pending state and returns a blocking state to the comparison processing module. The measurement action processing module records the content formed by the current round of measurement action processing as the measurement result and outputs the measurement result to the marking processing module.
[0202] The marking processing module 06, connected to the measurement action processing module, is used to mark the puncture area by the external probe based on the measurement result, thereby obtaining the number of tests. Specifically, the marking processing module receives the measurement result output by the measurement action processing module. After the measurement result of the current round is generated, the marking processing module calls the external probe to perform marking processing on the puncture area. The marking processing module first reads the state corresponding to the measurement result of the current round, then controls the external probe to enter the corresponding position of the puncture area and forms the current round mark on the puncture area. The marking processing module counts the number of marks on the puncture area and records the count as the number of tests. If the measurement result of the current round is in a blocking state, the marking processing module does not perform the marking processing and maintains the state of the number of tests in the previous round. The marking processing module returns the number of tests to the alignment result generation module for the next round of wafer, puncture area, and external probe acquisition to read the current round state.
Claims
1. A method for probe testing advanced process and wide bandgap semiconductors, comprising: include: S100: Acquire the wafer, chip under test, solder pads, pin piercing area, probe card, multiple probes, external probes and vacuum tray; perform optical calibration system and X, Y, Z and θ direction movement processing to obtain alignment results; S200. Based on the alignment results, perform needle pressure test, contact resistance test and visual inspection processing between the external probe and the puncture area to obtain the first measurement result; S300. Based on the first measurement result, perform multiple probe-pad pin pressure tests, contact resistance tests, and visual inspection processing to obtain the second measurement result; S400. Based on the second measurement result, perform a comparison process between the first measurement result and the second measurement result to obtain a measurement trigger signal; S500. Based on the measurement trigger signal, perform measurement action processing to obtain the measurement result.
2. The method of claim 1, wherein, The process of optical calibration system and movement processing in the X, Y, Z and θ directions includes: Based on the alignment marks on the wafer, the optical calibration system performs image acquisition and feature point extraction processing to generate a preliminary offset. Based on the initial offset, a two-stage movement process of coarse and fine movement is performed in the X, Y, Z and θ directions. The two-stage movement process includes first performing a descent in the Z direction to a safe height close to the surface of the solder pad, then performing alignment in the X, Y and θ directions, and finally performing a contact descent in the Z direction. During the Z-direction contact descent process, the approach distance between the external probe and the needle puncture area is monitored in real time. When the distance is less than a preset threshold, the descent is paused and the current Z coordinate is recorded as the alignment reference. Based on the alignment reference and feature point extraction results, alignment results are generated.
3. The method of claim 2, wherein, The alignment results include the center coordinates of the needle puncture area, the center coordinates of the solder pad, and the Z-axis safety height and contact height.
4. The method of claim 3, wherein, The process of external probe and needle pressure testing, contact resistance testing, and visual inspection of the puncture area includes: Within the same probe descent stroke, the outer probe is controlled to descend from the safe height of the Z-axis in the alignment result to contact the needle puncture area, and needle pressure test, contact resistance test and visual inspection are performed in parallel. The needle pressure test includes: acquiring the pressure waveform when the external probe contacts the puncture area using a piezoelectric sensor, extracting the peak pressure and steady-state pressure, calculating the deviation value from the preset ideal needle pressure value, and generating a needle pressure deviation result.
5. The method of claim 4, wherein, The process of external probe and needle pressure testing, contact resistance testing, and visual inspection also includes: The contact resistance test includes: measuring the contact resistance between the external probe and the puncture area using a four-wire Kelvin connection, calculating the deviation from the preset ideal resistance value, and generating a contact resistance deviation result.
6. The method of claim 5, wherein, The process of external probe and needle pressure testing, contact resistance testing, and visual inspection also includes: The visual inspection includes: acquiring contact images of the external probe tip on the surface of the puncture area using a high-resolution camera, extracting the offset between the center coordinates of the tip and the center coordinates of the puncture area, as well as the tip deformation features, and generating tip image deviation results. The needle pressure deviation result, contact resistance deviation result, and needle tip image deviation result are organized into a structured field set containing field name, deviation direction, and deviation amplitude to obtain the first measurement result.
7. The method of claim 6, wherein, The process of multiple probes and solder pads undergoing pin pressure testing, contact resistance testing, and visual inspection includes: Based on the set of structured fields in the first measurement result, determine the same field names and the same deviation expression format that need to be collected in the second measurement result; Within the same probe descent stroke as S200, multiple probes are controlled to descend from the Z-axis safety height in the alignment result to contact the pads of the chip under test, and probe pressure test, contact resistance test and visual inspection are performed in parallel. The needle pressure test includes: acquiring the pressure waveform of each of the multiple probes when it contacts the corresponding solder pad, extracting the peak pressure and steady-state pressure, calculating the deviation value from the preset ideal needle pressure value, and generating multi-channel needle pressure deviation results; The contact resistance test includes: measuring the contact resistance between each probe and the corresponding pad using a four-wire Kelvin connection, calculating the deviation from the preset ideal resistance value, and generating multi-channel contact resistance deviation results. The visual inspection includes: acquiring contact images of each probe tip on the corresponding solder pad surface, extracting the offset between the tip center coordinates and the solder pad center coordinates, as well as the tip deformation features, and generating multi-channel tip image deviation results. The multi-channel needle pressure deviation results, multi-channel contact resistance deviation results, and multi-channel needle tip image deviation results are organized into the same field structure as the first measurement results to obtain the second measurement results.
8. The method of claim 7, wherein, The process of comparing the first measurement result with the second measurement result includes: The needle pressure deviation, contact resistance deviation, and needle tip image deviation results in the first measurement result are compared in parallel with the deviation results of the corresponding fields in the second measurement result. The parallel comparison includes: simultaneously judging three fields: needle pressure deviation, contact resistance deviation, and needle tip image deviation. Each field has an independent allowable deviation range, which includes an upper limit threshold and a lower limit threshold. When the needle pressure deviation value of a certain channel in the second measurement result falls within the ± allowable deviation range of the corresponding deviation value in the first measurement result, and the contact resistance deviation value falls within the ± allowable deviation range of the corresponding deviation value, and the needle tip image deviation value falls within the ± allowable deviation range of the corresponding deviation value, the contact state of that channel is determined to be qualified. When the contact status of all channels is qualified, a measurement trigger signal containing a trigger enable flag and a diagnostic code is generated; When the contact status of a channel is unqualified, a measurement trigger signal containing a trigger blocking flag and a diagnostic code is generated. The diagnostic code is used to identify the unqualified field type and the unqualified channel number. After generating the trigger blocking flag, automatic rollback control is executed: based on the diagnostic code, the alignment, reference measurement and actual measurement are re-executed in reverse until the contact status of all channels is qualified or the preset retry limit is reached.
9. The method according to claim 8, characterized in that, The process of measuring motion processing includes: The measurement action processing includes the measurement action processing of pulse signal generator, drive unit, drive motherboard, drive daughterboard, drive IC, buffer capacitor and bus capacitor; When the measurement trigger signal contains a trigger enable flag, the pulse signal generator is activated to generate a dynamic test pulse sequence that matches the type of the chip under test; The driving unit receives the pulse sequence and selects the default driving voltage and current based on the diagnostic code in the measured trigger signal; The driving motherboard distributes the driving signal to multiple driving daughterboards, and each driving daughterboard is connected to the corresponding probe through a driving IC. During the measurement process, buffer capacitors and bus capacitors are used to stabilize the power supply voltage, while the output voltage and current waveforms of the driver IC are monitored in real time. When an abnormal voltage or current is detected, the measurement action is automatically interrupted and an abnormality record is generated. After the measurement operation is completed normally, the electrical response signals returned by each probe are collected, and after analog-to-digital conversion, they are organized into structured measurement results containing measurement timestamps, driving parameters and response waveforms according to the channel number.
10. An advanced process and wide-bandgap semiconductor probe testing system, characterized in that, include: Alignment result generation module, first measurement result generation module, second measurement result generation module, comparison processing module, measurement action processing module, and mark processing module; The modules are connected in sequence to implement the method as described in any one of claims 1-9.