Temperature control device, temperature control method, program, prober, and method for generating a learning model

Abstract

A temperature control device, a temperature control method, a program, a probe, and a method for generating a learning model are provided, capable of performing automatic adjustment of a chuck temperature control parameter. The temperature control device (21) comprises a chuck temperature acquisition unit (100) for acquiring a chuck temperature, a classification unit (102) for applying a learning completed model that learns the corresponding relationship between the characteristics of the change in chuck temperature and the temperature change pattern and outputting a temperature change pattern when a chuck temperature is input, and a temperature control parameter setting unit (104) for deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit and setting the temperature control parameter, and controls the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter.

Description

Technology Field

[0001] The present invention relates to a temperature control device, a temperature control method, a program, a prober, and a method for generating a learning model, which are applied to the inspection of electrical characteristics of a semiconductor chip formed on a wafer. Background Technology

[0002] A plurality of semiconductor chips having identical electrical component circuits are formed on the surface of a wafer. The electrical characteristics of each semiconductor chip are inspected using a wafer test system equipped with a probe and a tester.

[0003] The prober, while holding the wafer on the wafer chuck, moves the probe card equipped with a probe needle relative to the wafer chuck to electrically connect the probe needle to the electrode pad of the semiconductor chip. The tester supplies various test signals to the semiconductor chip from the terminal electrically connected to the probe needle, receives signals output from the semiconductor chip, and interprets the received signals to test whether the semiconductor chip is operating normally.

[0004] Since semiconductor chips are used for a wide range of applications and across a wide temperature range, testing of semiconductor chips is performed at temperatures corresponding to the environments in which the chips are assumed to be used, such as room temperature, high temperature, and low temperature. Furthermore, the term "room temperature" as used here may include the concept referred to as ambient temperature. Low temperature refers to a temperature environment that is relatively lower than room temperature. Similarly, high temperature refers to a temperature environment that is relatively higher than room temperature.

[0005] The wafer chuck of the prober is equipped with a temperature control device including, for example, a heater mechanism, a chiller mechanism, and a heat pump mechanism, and heating or cooling is performed on a wafer maintained on the wafer chuck using the temperature control device.

[0006] Patent Document 1 describes an IC test handler equipped with a temperature sensor for a wafer and directly measuring the temperature of the wafer, which is a disturbance factor in chuck temperature control. The device described in said document measures the surface temperature of an IC package in which an IC is housed using a non-contact thermometer.

[0007] Patent Document 2 describes an inspection device that measures the temperature of a wafer using an infrared sensor provided on a probe card. When measuring an electronic device, the device described in the document measures the temperature of the electronic device using an infrared sensor provided on a probe card.

[0008] Patent Document 3 describes an inspection device that estimates the amount of heat generated by a wafer based on power output from a tester and performs temperature correction of a chuck based on the amount of heat generated by the wafer. In the device described in the document, power supplied to an electronic device to be inspected from a power supply unit is input, the amount of heat generated by the electronic device is estimated based on the power supplied to the electronic device, the temperature difference between the electronic device and the chuck is estimated from the amount of heat generated by the electronic device, and the temperature of the chuck is controlled using the estimated temperature difference value.

[0009] The devices described in each of Patent Documents 1 to 3 measure or estimate the temperature of a wafer, i.e., a disturbance, and perform temperature control using the measurement result or the estimate result, thereby improving the performance of temperature control in wafer inspection. In addition, the IC described in Patent Document 1 and the electronic devices described in Patent Documents 2 and 3 correspond to semiconductor chips formed on the said wafer. Prior art literature

[0010] Japanese Patent Publication No. 2018-80919, Japanese Patent Publication No. 2021-128965, Japanese Patent Publication No. 2022-90538 The problem to be solved

[0011] However, in the embodiment in which a temperature sensor is provided on the wafer described in Patent Document 1 and the embodiment in which a temperature sensor is provided on the probe card described in Patent Document 2, a temperature sensor provided outside the system of the probe is required. If a temperature sensor cannot be provided on the wafer or the probe card due to design issues or the like, there may be cases where precise temperature control cannot be performed.

[0012] In addition, in the embodiment of estimating the amount of heat generated by a wafer based on the power output from a tester described in Patent Document 3, the power output from the tester depends on the specifications of the tester, making it difficult to perform universal measurements for multiple wafers of different types.

[0013] In other words, to realize universal measurement for multiple wafers of different types, it is necessary to perform temperature control using information such as the chuck temperature and coefficients that the probe can directly acquire.

[0014] Furthermore, the amount of heat generated and the heating pattern of the wafer differ due to differences in wafer type, such as variations in the type of semiconductor chips formed on the wafer. Consequently, in wafer temperature control using conventional temperature control parameters, if temperature changes occur due to disturbances such as differences in wafer type, there is a concern that delays in reaching the target temperature and hunting of the control amount may occur.

[0015] Delays in reaching the target temperature cause a decrease in measurement throughput. Hunting of the control quantity leads to unstable temperature control, raising concerns about reduced temperature control precision. Reduced temperature control precision results in improper wafer inspection.

[0016] If there is a delay in reaching the target temperature, it becomes necessary to adjust the temperature control parameters for each wafer being measured. On the other hand, for example, since the temperature control of a chuck performs heating and cooling controls collectively, determining the temperature control parameters is not easy. Furthermore, because wafers are the property of the user of the inspection equipment, it is difficult to adjust the temperature control parameters using the wafers to be measured in advance of inspection; consequently, the parameters are adjusted manually during measurement. This leads to problems such as the over-reliance on individual temperature control parameter adjustments and a decrease in productivity.

[0017] The present invention has been made in consideration of these circumstances and aims to provide a temperature control device, a temperature control method, a program, a probe, and a method for generating a learning model, capable of performing automatic adjustment of control parameters of chuck temperature. means of solving the problem

[0018] A temperature control device according to a first aspect of the present disclosure comprises: a chuck temperature acquisition unit for acquiring a chuck temperature representing the temperature of a wafer chuck that holds a wafer having a plurality of semiconductor chips formed thereon; a classification unit for applying a learning completed model that learns the characteristics of a change in chuck temperature and the corresponding relationship of a number of temperature change patterns classified by the change in chuck temperature, and outputting a temperature change pattern according to the change in chuck temperature when the chuck temperature is input; a temperature control parameter setting unit for deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter; and a temperature control device for controlling the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter set using the temperature control parameter setting unit.

[0019] According to the temperature control device of the present disclosure, a classification unit to which a learned model is applied that has learned the corresponding relationship between the characteristics of a change in chuck temperature and the pattern of a change in chuck temperature is used, and when an acquired chuck temperature is input to the classification unit, a temperature control parameter output from the classification unit is applied to a temperature adjustment unit that adjusts the chuck temperature. By doing so, automatic adjustment of the temperature control parameter applied to the chuck temperature adjustment unit that adjusts the chuck temperature based on the chuck temperature can be performed.

[0020] The temperature control parameter may include a heating control parameter applied when heating the wafer chuck and a cooling control parameter applied when cooling the wafer chuck.

[0021] The temperature control parameter may include multiple components.

[0022] The temperature control device according to the second embodiment may include a chuck temperature change derivation unit that derives a change in chuck temperature from a plurality of chuck temperatures acquired at different timings, in the temperature control device according to the first embodiment.

[0023] According to this embodiment, temperature control parameters can be derived based on the characteristics of the change in chuck temperature.

[0024] A temperature control device according to a third embodiment may, in the temperature control device according to a first embodiment, have a learning completion model that learns a correspondence relationship between a feature quantity representing a characteristic of a change in chuck temperature and a temperature change pattern classified as a change in chuck temperature, and a classification unit may acquire a feature quantity as a characteristic of a change in chuck temperature and output a temperature change pattern corresponding to the feature quantity.

[0025] According to this embodiment, the change in chuck temperature can be classified based on quantified characteristic quantities of the change in chuck temperature.

[0026] In the temperature control device according to the fourth embodiment, in the temperature control device according to any one of the first to third embodiments, the temperature control parameter setting unit may derive a temperature control parameter in which an adjustment coefficient corresponding to a temperature change pattern is used.

[0027] In this embodiment, PID control parameters can be applied as temperature control parameters.

[0028] A temperature control device according to the fifth embodiment, in any one of the first to fourth embodiments, comprises a wafer information acquisition unit for acquiring wafer information including the type of wafer to be measured, and a plurality of learned models generated by learning for each type of wafer, and a classification unit may select one learned model from the plurality of learned models according to the type of wafer included in the wafer information acquired using the wafer information acquisition unit.

[0029] According to this embodiment, changes in chuck temperature can be classified according to the type of wafer.

[0030] A temperature control method according to the sixth aspect of the present disclosure is a temperature control method in which a computer performs the steps of: acquiring a chuck temperature representing the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon; using a classification unit to which a learned model is applied that has learned the characteristics of a change in chuck temperature and the corresponding relationship of a number of temperature change patterns classified by the change in chuck temperature, outputting a temperature change pattern according to the change in chuck temperature when the chuck temperature is input; deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter; and controlling the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter.

[0031] According to the temperature control method of the present disclosure, it is possible to obtain the same operational effect as the temperature control device according to the present disclosure.

[0032] In the temperature control method according to the present disclosure, specific details and similar details in any one of the second to fifth embodiments may be appropriately combined. In that case, a component responsible for a specific processing or function in a temperature control device may be identified as a component of a temperature control method responsible for a corresponding processing or function.

[0033] A program according to the seventh aspect of the present disclosure is a program that implements the following functions in a computer: a function to acquire a chuck temperature representing the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon; a function to output a temperature change pattern corresponding to the input chuck temperature change when the chuck temperature is input, using a classification unit to which a learned model is applied that has learned the characteristics of the chuck temperature change and the corresponding relationship of the chuck temperature change to a number of classified temperature change patterns; a function to derive a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and to set the temperature control parameter; and a function to control the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter.

[0034] According to the program of the present disclosure, it is possible to obtain the same operational effect as the temperature control device according to the present disclosure.

[0035] In the program according to the present disclosure, specific matters and similar matters in any one of the second to fifth embodiments may be appropriately combined. In that case, a component responsible for a specific processing or function in a temperature control device may be identified as a component of a program responsible for a corresponding processing or function.

[0036] A probe according to the eighth aspect of the present disclosure comprises a wafer chuck for holding a wafer having a plurality of semiconductor chips formed thereon, a probe card having a probe needle, a relative movement unit for moving the wafer chuck relative to the probe needle, a chuck temperature adjustment device for adjusting the temperature of the wafer chuck, and a temperature control device for controlling the operation of the chuck temperature adjustment device by applying a temperature control parameter. The temperature control device comprises a chuck temperature acquisition unit for acquiring a chuck temperature indicating the temperature of the wafer chuck, a learning completion model applied to the characteristics of the change in chuck temperature and the corresponding relationship of a number of temperature change patterns classified by the change in chuck temperature, a classification unit for outputting a temperature change pattern according to the change in the input chuck temperature when the chuck temperature is input, and a temperature control parameter setting unit for deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter, and a probe that controls the operation of the chuck temperature adjustment unit for adjusting the chuck temperature by applying the temperature control parameter set using the temperature control parameter setting unit.

[0037] According to the probe according to the present disclosure, it is possible to obtain the same operational effect as the temperature control device according to the present disclosure.

[0038] In the probe according to the present disclosure, a combination of specific details and similar details in any one of the second to fifth embodiments may be appropriately combined. In that case, a component responsible for a specific processing or function in a temperature control device may be identified as a component of a probe responsible for a corresponding processing or function.

[0039] A method for generating a learning model according to the ninth aspect of the present disclosure is a method for generating a learning model that generates a learning completed model by learning the characteristics of a change in chuck temperature, which is the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon, and the corresponding relationship between the number of temperature change patterns classified by the change in chuck temperature.

[0040] According to the method for generating a learning model according to the present disclosure, it is possible to provide a completed learning model applied to a temperature control device according to the present disclosure. Effects of the invention

[0041] According to the present invention, a classification unit to which a learned model is applied that has learned the corresponding relationship between the characteristics of a change in chuck temperature and the pattern of a change in chuck temperature is used, and when an acquired chuck temperature is input to the classification unit, a temperature control parameter output from the classification unit is applied to a temperature adjustment unit that adjusts the chuck temperature. By doing so, automatic adjustment of the temperature control parameter applied to the chuck temperature adjustment unit that adjusts the chuck temperature based on the chuck temperature can be performed. Brief explanation of the drawing

[0042] FIG. 1 is a schematic diagram of a probe according to an embodiment. Figure 2 is an external perspective view of the probe shown in Figure 1. Figure 3 is a top view of a wafer. Figure 4 is a functional block diagram showing the electrical configuration of the probe shown in Figure 1. Figure 5 is a functional block diagram showing an example of the configuration of the chuck temperature control unit shown in Figure 4. FIG. 6 is a functional block diagram showing a modified example of the chuck temperature control unit shown in FIG. 5. Figure 7 is a schematic diagram of the optimization of temperature control parameters during wafer inspection. Figure 8 is a schematic diagram of the adjustment of chuck temperature control parameters according to the prior art. Figure 9 is an explanatory diagram of the first pattern of change in chuck temperature. Figure 10 is an explanatory diagram of the second pattern of change in chuck temperature. Figure 11 is an explanatory diagram of the third pattern of change in chuck temperature. Figure 12 is an explanatory diagram of the fourth pattern of change in chuck temperature. Figure 13 is a schematic diagram of the creation of a completed learning model applied to the classification of patterns of change in chuck temperature. Figure 14 is a schematic diagram of the classification of patterns of change in chuck temperature to which the learning completed model shown in Figure 13 is applied. Figure 15 is a schematic diagram showing a specific example of chuck temperature control. Specific details for implementing the invention

[0043] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this specification, the same reference numerals are assigned to identical components, and redundant descriptions are appropriately omitted.

[0044] [Example of Prober Composition According to Embodiment]

[0045] FIG. 1 is a schematic diagram of a probe according to an embodiment. FIG. 2 is an external perspective view of the probe shown in FIG. 1. FIG. 3 is a top view of a wafer. The probe (10) shown in FIG. 1 and FIG. 2 is used in a wafer test system for inspecting the electrical characteristics of a plurality of semiconductor chips formed on a wafer. A wafer (W) to be inspected using the probe (10) is shown in FIG. 3.

[0046] In the drawing, the upper surface of a wafer (W) supported by a wafer chuck (20) is shown. A plurality of semiconductor chips (9) are formed on the wafer (W). A plurality of electrode pads (9a) are formed on each semiconductor chip (9).

[0047] The prober (10) shown in FIGS. 1 and 2 comprises a base (12), a Y stage (13), a Y moving part (14), an X stage (15), an X moving part (16), a Zθ stage (17), a Zθ moving part (18), and a wafer chuck (20) as shown in FIGS. 1. Additionally, the prober (10) comprises a support (23), a head stage (24), a card holder (25), and a probe card (26) as shown in FIGS. 2. Additionally, the prober (10) comprises a wafer position alignment camera (29), an upper / lower stage (30), a needle position alignment camera (31), a cleaning plate (32), and a temperature sensor (34) as shown in FIGS. 1. Furthermore, the configuration of the prober (10) is not limited to the examples shown in FIGS. 1 and 2 and can be appropriately modified.

[0048] On the upper surface of the base (12), the Y stage (13) is supported so as to be movable in the Y-axis direction using the Y moving part (14). The Y moving part (14) moves the Y stage (13) in the Y-axis direction on the upper surface of the base (12).

[0049] The Y moving part (14) is, for example, a guide rail placed on the upper surface of the base (12), a guide rail parallel to the Y-axis, a slider placed on the lower surface of the Y stage (13), a slider that hooks onto the guide rail, and an actuator such as a motor that moves the Y stage (13) in the Y-axis direction.

[0050] On the upper surface of the Y stage (13), the X stage (15) is supported so as to be movable in the X-axis direction using the X moving part (16). The X moving part (16) moves the X stage (15) in the X-axis direction on the upper surface of the Y stage (13).

[0051] The X moving part (16) is, for example, a guide rail placed on the upper surface of the Y stage (13), a guide rail parallel to the X-axis, a slider placed on the lower surface of the X stage (15), a slider that hooks onto the guide rail, and an actuator such as a motor that moves the X stage (15) in the X-axis direction.

[0052] On the upper surface of the X stage (15), a Zθ stage (17) and an upper / lower stage (30) are arranged. The Zθ stage (17) is equipped with a Zθ moving part (18). A wafer chuck (20) is supported on the upper surface of the Zθ stage (17).

[0053] The Zθ moving unit (18) is provided with, for example, a lifting mechanism for raising and lowering the Zθ stage (17) and a rotating mechanism for rotating the Zθ stage (17) around a rotation axis parallel to the Z axis. The Zθ moving unit (18) moves a wafer chuck (20) supported on the upper surface of the Zθ stage (17) in the Z-axis direction and also rotates the wafer chuck (20) around a rotation axis parallel to the Z axis.

[0054] A wafer (W) is supported on the upper surface of the wafer chuck (20) by applying various support methods, such as vacuum suction. Additionally, the wafer chuck (20) is equipped with a chuck temperature adjustment unit (20a). The chuck temperature adjustment unit (20a) adjusts the temperature of the wafer chuck (20) to adjust the temperature of the wafer (W) supported on the wafer chuck (20).

[0055] The chuck temperature adjustment unit (20a) is applied using known mechanisms, such as a heater mechanism, a chiller mechanism, and a heat pump mechanism. The operation of the chuck temperature adjustment unit (20a) is controlled based on a command signal transmitted from the chuck temperature control unit. The chuck temperature control unit is illustrated in FIG. 4 by assigning the reference numeral 21. Furthermore, the chuck temperature adjustment unit (20a) described in the embodiment is an example of a chuck temperature adjustment device for adjusting the temperature of a wafer chuck.

[0056] The wafer chuck (20) is supported so as to be movable in the XYZ axis direction using the above-described Zθ stage (17), etc., and is also supported so as to be rotatably around a rotation axis in a direction parallel to the Z axis. The above-described Zθ stage (17), etc. functions as a relative moving part that moves the wafer (W) and the probe needle (35) supported on the wafer chuck (20) relative to each other in the X direction, Y direction, Z direction, and rotation direction.

[0057] The support (23) shown in FIG. 2 is provided on the upper surface of the base (12) and supports the head stage (24) at a position above the Y stage (13), X stage (15) and Zθ stage (17). That is, the head stage (24) is fixed to the upper surface of the base (12) using the support (23).

[0058] A card holder (25) is provided in the central part of the head stage (24). The card holder (25) has a retaining hole (25a) formed therein to hold the outer circumference of the probe card (26). The probe card (26) is inserted into the retaining hole (25a) of the probe card (26) and is supported in a position facing the wafer (W) using the head stage (24) and the card holder (25). Additionally, the retaining hole (25a) is illustrated in FIG. 1.

[0059] The probe card (26) shown in FIG. 1 is equipped with a probe needle (35) that is positioned according to the arrangement of the electrode pad (9a) of the semiconductor chip (9) to be tested. The card holder (25) and the probe card (26) are exchanged according to the type of semiconductor chip (9).

[0060] The probe card (26) is provided with a connection terminal that is electrically connected to a probe needle (35). A tester is connected to the connection terminal. The tester supplies various test signals to the electrode pad (9a) of the semiconductor chip (9) through the connection terminal of the probe card (26) and the probe needle (35), and also receives signals output from the electrode pad (9a). The tester analyzes the signals output from the electrode pad (9a) to test whether the semiconductor chip (9) is operating normally. The configuration of the tester and the test method may utilize known technologies. Here, a detailed description of the configuration of the tester, etc., is omitted. Additionally, the illustration of the connection terminal and the tester is omitted.

[0061] A wafer positioning camera (29) photographs a semiconductor chip (9) on a wafer (W) supported by a wafer chuck (20). The image of the semiconductor chip (9) captured using the wafer positioning camera (29) is used to detect the position of the electrode pad (9a) of the semiconductor chip (9) to be inspected. The position where the wafer positioning camera (29) is placed and the structure of the wafer positioning camera (29) are not particularly limited.

[0062] The upper and lower stage (30) is equipped with a needle position alignment camera (31) and a cleaning plate (32). The needle position alignment camera (31) and the cleaning plate (32) are positioned opposite to the probe card (26), etc. Additionally, the upper and lower stage (30) is equipped with a lifting mechanism that supports the needle position alignment camera (31) and the cleaning plate (32) so as to be movable in the Z-axis direction. The upper and lower stage (30) can adjust the position of the needle position alignment camera (31) and the cleaning plate (32) in the Z-axis direction by operating the lifting mechanism. Additionally, the illustration of the lifting mechanism movable in the Z-axis direction is omitted.

[0063] The needle position alignment camera (31) and the cleaning plate (32) are configured to be movable in the Y-axis direction using the Y stage (13) and the Y moving part (14) via the upper and lower stage (30), and are also supported to be movable in the X-axis direction using the X stage (15) and the X moving part (16). That is, the needle position alignment camera (31) and the cleaning plate (32) and the probe needle (35) are configured to be movable relative to each other in the X-axis direction, the Y-axis direction, and the Z-axis direction.

[0064] The needle position alignment camera (31) photographs the probe needle (35). The image of the probe needle (35) captured using the needle position alignment camera (31) is used to detect the tip position of the probe needle (35). Specifically, the XY coordinates of the tip position of the probe needle (35) are detected based on the position coordinates of the needle position alignment camera (31), and the Z coordinates of the tip position of the probe needle (35) are detected based on the focal position of the needle position alignment camera (31).

[0065] When the semiconductor chip (9) of the wafer (W) is inspected, the tip position of the probe needle (35) is detected whenever the probe card (26) is replaced. The tip position of the probe needle (35) may also be detected whenever a prescribed number of semiconductor chips (9) are inspected.

[0066] In detecting the tip position of the probe needle (35), the needle position matching camera (31) is moved to the shooting position of the tip position of the probe needle (35), and the tip position of the probe needle (35) is photographed using the needle position matching camera (31), and the tip position of the probe needle (35) is detected based on the photographed image of the tip position of the probe needle (35).

[0067] Additionally, the position of the electrode pad (9a) of the semiconductor chip (9) is detected in the wafer (W) to be inspected, which is supported using the wafer chuck (20). Specifically, the wafer positioning camera (29) is moved to the position of the electrode pad (9a) of the semiconductor chip (9) to be inspected in the wafer (W), the electrode pad (9a) is photographed using the wafer positioning camera (29), and the position of the electrode pad (9a) is detected based on the photographed image of the electrode pad (9a).

[0068] Then, a probe needle (35) is electrically contacted to the electrode pad (9a) of the first semiconductor chip (9) to be tested, and the first semiconductor chip (9) to be tested is tested using a tester.

[0069] Below, the wafer (W) to be inspected is moved, and the probe needle (35) is electrically contacted with the electrode pad (9a) of the next semiconductor chip (9) to be inspected, and the semiconductor chip (9) to be inspected is inspected. By repeating this procedure, inspections of a plurality of semiconductor chips (9) to be inspected are performed in sequence. The tip of the probe needle (35) is properly cleaned and polished using a cleaning plate (32).

[0070] In addition, a specific inspection method for the semiconductor chip (9) may be applied using known inspection methods, such as the inspection method described in Japanese Patent Publication No. 2018-117095. Here, a detailed description of the inspection method for the semiconductor chip (9) is omitted.

[0071] A temperature sensor (34) is provided at a position opposite to the lower surface of the card holder (25) and the lower surface of the probe card (26). The lower surface of the card holder (25) is the side opposite to the side of the card holder (25) where the probe card (26) is supported. The lower surface of the probe card (26) is the side of the probe card (26) that is supported by the card holder (25).

[0072] Examples of placement of the temperature sensor (34) include the side of the Zθ stage (17) and the side of the upper / lower stage (30). The temperature sensor (34) is supported so as to be able to move relative to the card holder (25) and the probe card (26) using the Y stage (13), X stage (15), Zθ stage (17), and upper / lower stage (30).

[0073] For example, the temperature sensor (34) may be a non-contact temperature sensor using a radiation energy detection method, thereby enabling non-contact measurement of the temperature of the card holder (25) and the probe card (26). The card holder (25) and the probe card (26) may undergo thermal deformation due to the influence of the temperature of the wafer chuck (20). As a result of the thermal deformation of the card holder (25), the tip position of the probe needle (35) is displaced relative to the specified position.

[0074] The probe (10) may use a temperature sensor (34) to measure the temperature of the card holder (25) and the probe card (26) and predict the displacement of the tip position of the probe needle (35) accompanying thermal deformation of the card holder (25), etc. For example, the probe (10) may predict the amount of displacement and the direction of displacement of the tip position of the probe needle (35) as the displacement of the tip position of the probe needle (35).

[0075] [Example of electrical configuration of a probe according to an embodiment]

[0076] FIG. 4 is a functional block diagram showing the electrical configuration of the probe shown in FIG. 1. FIG. 4 mainly illustrates functions related to temperature control of the wafer chuck (20) and functions related to contact control between the probe needle (35) and the electrode pad (9a) of the semiconductor chip (9) on the wafer (W), and other functions are appropriately omitted.

[0077] The control device (40) controls each part of the probe (10) collectively. The control device (40) is a computer. The control device (40) executes various programs corresponding to the functions of each part of the probe (10) to realize the functions of each part of the probe (10). The control device (40) may be placed in the main body of the probe (10) or placed outside the probe (10).

[0078] The form of the computer may be a server, a personal computer, a workstation, or a tablet terminal, etc. The form of the computer may be a virtual machine.

[0079] Various programs may be stored in a memory device provided in the control device (40), or may be stored in a memory device provided inside the prober (10) and outside the control device (40). The control device (40) may acquire various programs from a memory device outside the prober (10).

[0080] The control unit (40) is equipped with an arithmetic circuit composed of various processors and memory. Examples of various processors include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), and a programmable logic device.

[0081] Examples of programmable logic devices include SPLDs (Simple Programmable Logic Devices), CPLDs (Complex Programmable Logic Devices), and FPGAs (Field Programmable Gate Arrays). Various functions of the control device (40) may be realized using a single processor or using multiple processors. The multiple processors may be multiple processors of the same type or multiple processors of different types.

[0082] The control device (40) is equipped with various communication interfaces. The control device (40) is connected to communicate with peripheral devices, such as a wafer positioning camera (29), a needle positioning camera (31), and a temperature sensor (34), through various communication interfaces. Various standards, such as USB (Universal Serial Bus), can be applied to the communication interfaces. The communication type of the communication interface may be either wired communication or wireless communication.

[0083] The control device (40) is provided with a chuck temperature control unit (21). The chuck temperature control unit (21) controls the operation of a chuck temperature adjustment unit (20a) provided in the wafer chuck (20). The chuck temperature control unit (21) obtains the chuck temperature from a chuck temperature sensor (20b) provided in the wafer chuck (20). Details of the chuck temperature control unit (21) will be described later.

[0084] The control device (40) is provided with a card temperature acquisition unit (42). The card temperature acquisition unit (42) acquires the temperature of the card holder (25) and the temperature of the probe card (26) output from the temperature sensor (34). The card temperature acquisition unit (42) is provided with a communication interface corresponding to the output signal output from the temperature sensor (34).

[0085] The control device (40) is provided with a needle position acquisition unit (44). The needle position acquisition unit (44) acquires a captured image of the tip position of the probe needle (35) output from the needle position matching camera (31). The needle position acquisition unit (44) acquires information of the tip position of the probe needle (35) from the captured image of the tip position of the probe needle (35). The acquisition of information may include the concept of processing original information to generate desired information.

[0086] The control device (40) is equipped with a movement control unit (52). The movement control unit (52) controls the operation of the X movement unit (16) to drive the X stage (15). The movement control unit (52) controls the operation of the Y movement unit (14) to drive the Y stage (13). The movement control unit (52) controls the operation of the Zθ movement unit (18) to drive the Zθ stage (17).

[0087] The movement control unit (52) obtains the position of the semiconductor chip (9) to be inspected from a captured image of the semiconductor chip (9) to be inspected transmitted from the wafer position alignment camera (29). Additionally, the movement control unit (52) obtains information on the tip position of the probe needle (35) from the needle position acquisition unit (44).

[0088] When performing an inspection of a wafer (W), the movement control unit (52) drives the X stage (15), Y stage (13), and Zθ stage (17) to move the wafer (W) relative to the probe needle (35) and sequentially contact a plurality of semiconductor chips (9) of the inspection target with respect to the probe needle (35).

[0089] The movement control unit (52) may use the card temperature acquisition unit (42) to acquire the temperature of the card holder (25) and the temperature of the probe card (26), and correct for the variation in the tip position of the probe needle (35) caused by temperature changes such as the temperature of the card holder (25).

[0090] The control device (40) is equipped with a wafer information acquisition unit (54). The wafer information acquisition unit (54) acquires wafer information that specifies the type of wafer (W) to be inspected. The wafer information includes information on the type of semiconductor chip (9) formed on the wafer (W). The chuck temperature control unit (21) controls the chuck temperature based on the wafer information acquired using the wafer information acquisition unit (54). The wafer information acquisition unit (54) may be a component of the chuck temperature control unit (21).

[0091] [Example of configuration for chuck temperature control unit]

[0092] FIG. 5 is a functional block diagram showing an example of the configuration of the chuck temperature control unit shown in FIG. 4. The chuck temperature control unit (21) may be configured as a temperature control device to which a computer is applied. The computer functioning as the temperature control device may be the same computer as the computer functioning as the control device (40) shown in FIG. 4.

[0093] The chuck temperature control unit (21) shown in FIG. 5 uses the chuck temperature to perform automatic adjustment of the chuck temperature control parameter applied to the chuck temperature adjustment unit (20a). The chuck temperature control unit (21) is equipped with a learning model that has completed learning.

[0094] The chuck temperature control unit (21) comprises a chuck temperature acquisition unit (100), a classification unit (102), a learning completion model memory unit (103), and a chuck temperature control parameter setting unit (104). The chuck temperature acquisition unit (100) acquires the chuck temperature by applying a predetermined sampling period.

[0095] The classification unit (102) classifies the change in chuck temperature into one temperature change pattern included in a plurality of predetermined temperature change patterns based on the chuck temperature acquired using the chuck temperature acquisition unit (100). The classification unit (102) applies a completed learning model obtained by performing supervised learning using the relationship between the characteristics of the change in chuck temperature and the temperature change pattern as learning data.

[0096] The learning completion model memory unit (103) stores one or more learning completion models applied to the classification unit (102). The classification unit (102) reads out a learning completion model from the learning completion model memory unit (103) and performs classification of changes in chuck temperature.

[0097] The learning completion model memory unit (103) may store multiple learning completion models corresponding to each of the multiple types of wafers (W). The classification unit (102) may classify the change in chuck temperature by selecting one learning completion model corresponding to the wafer information of the inspection target acquired using the wafer information acquisition unit (54).

[0098] The completed learning model stored in the completed learning model memory unit (103) can be generated using a computer. The computer used to generate the completed learning model may be a device outside the chuck temperature control unit (21) or a device provided in the chuck temperature control unit (21).

[0099] The chuck temperature control parameter setting unit (104) acquires a temperature change pattern output from the classification unit (102) according to the input of the chuck temperature and derives a chuck temperature control parameter corresponding to the temperature change pattern.

[0100] The chuck temperature control parameter setting unit (104) comprises a heating control parameter derivation unit (110) for deriving heating control parameters and a cooling control parameter derivation unit (112) for deriving cooling control parameters.

[0101] The chuck temperature control parameter setting unit (104) sets heating control parameters and cooling control parameters as chuck temperature control parameters for the chuck temperature adjustment unit (20a).

[0102] In this embodiment, PID control parameters are exemplified as examples of chuck temperature control parameters. The heating control PID shown in FIG. 5 is a heating control parameter derivation unit (110), and means that PID control parameters are derived as heating control parameters applied to heating control. Also, the cooling control PID shown in FIG. 5 is a cooling control parameter derivation unit (112), and means that PID control parameters are derived as cooling control parameters applied to cooling control. Additionally, P in PID stands for Proportional, I stands for Integral, and D stands for Derivative.

[0103] FIG. 6 is a functional block diagram showing a modified example of the chuck temperature control unit shown in FIG. 5. The chuck temperature control unit (21a) shown in FIG. 6 is equipped with a chuck temperature control parameter setting unit (104a) instead of the chuck temperature control parameter setting unit (104) shown in FIG. 5.

[0104] The chuck temperature control parameter setting unit (104a) comprises a heating control parameter derivation unit (110a) and a cooling control parameter derivation unit (112a). The heating control parameter derivation unit (110a) acquires a temperature change pattern output from the classification unit (102) and derives a heating control parameter corresponding to the temperature change pattern. Additionally, the heating control parameter derivation unit (110a) derives a cooling control amount based on the heating control parameter.

[0105] The cooling control parameter derivation unit (112a) obtains a cooling control amount from the heating control parameter derivation unit (110a) and derives a cooling control parameter based on the cooling control amount. The chuck temperature control parameter setting unit (104a) sets a heating control parameter and a cooling control parameter as chuck temperature control parameters for the chuck temperature adjustment unit (20a).

[0106] [Specific example of chuck temperature control during wafer inspection]

[0107] FIG. 7 is a schematic diagram of the optimization of temperature control parameters during wafer inspection. FIG. 7 illustrates the change in chuck temperature by applying a graph format in which the horizontal axis represents time and the vertical axis represents chuck temperature. During the inspection time of a single semiconductor chip (9), the chuck temperature is acquired multiple times, and the change in chuck temperature as shown in graph G10, etc. is acquired. The temperature range parallel to the time axis shown in graph G10, etc. represents the allowable temperature range of the chuck temperature during wafer (W) inspection.

[0108] The first contact shown in FIG. 7 represents the inspection of a semiconductor chip (91). For the inspection of the semiconductor chip (91), the initial value of the specified chuck temperature control parameter is applied, and the temperature control of the wafer chuck (20) is performed. In addition, the curve of graph G10, etc. shown in FIG. 8 is a hypothetical curve representing the temperature history.

[0109] Graph G10 represents the change in chuck temperature acquired during the inspection of the semiconductor chip (91). The inspection of the semiconductor chip (91) is the period from the time when the probe needle (35) contacts the semiconductor chip (91) until the time when the tester finishes the measurement and the contact of the probe needle (35) with the semiconductor chip (91) is separated.

[0110] The measurement of the chuck temperature during the inspection of the semiconductor chip (91) is performed as background processing for the inspection of the semiconductor chip (91). The measurement of the chuck temperature does not have to be performed by contact between the semiconductor chip (9) and the probe needle (35). Additionally, for the inspection of multiple semiconductor chips (9), sampling of the chuck temperature may be performed by applying a constant sampling period continuously.

[0111] During inspection of the semiconductor chip (91), the chuck temperature may exceed the allowable temperature range, so adjustment of the chuck temperature control parameter is performed. The chuck temperature control parameter setting unit (104) shown in FIG. 5 derives a chuck temperature control parameter corresponding to the change in chuck temperature shown as graph G10 in FIG. 7, and sets the derived chuck temperature control parameter to the chuck temperature adjustment unit (20a).

[0112] In the inspection of the semiconductor chip (92) shown as the second contact, a new chuck temperature control parameter is set, and temperature control of the wafer chuck (20) is performed. Graph G12 shows the change in chuck temperature obtained during the inspection of the semiconductor chip (92).

[0113] The chuck temperature during inspection of the semiconductor chip (92) is improved compared to the chuck temperature during inspection of the semiconductor chip (91), but there are cases where it exceeds the allowable temperature range. The chuck temperature control parameter setting unit (104) derives a chuck temperature control parameter corresponding to the change in chuck temperature shown as graph G12 of FIG. 7, and sets the acquired chuck temperature control parameter to the chuck temperature adjustment unit (20a).

[0114] In the inspection of the semiconductor chip (93) shown as the third contact, a new chuck temperature control parameter is set, and temperature control of the wafer chuck (20) is performed. Graph G14 shows the change in chuck temperature obtained during the inspection of the semiconductor chip (93).

[0115] The chuck temperature during inspection of the semiconductor chip (93) is improved compared to the chuck temperature during inspection of the semiconductor chip (92) and falls within the allowable temperature range. In this way, the chuck temperature control parameter is updated each time contact is made, and the chuck temperature control parameter is optimized. As a result, the chuck temperature control parameter that is optimal for the wafer to be inspected is selected, and the temperature control of the wafer chuck (20) is performed.

[0116] FIG. 8 is a schematic diagram of the adjustment of chuck temperature control parameters according to the prior art. FIG. 8 illustrates the change in chuck temperature by applying a graph format. In addition, the horizontal axis of graphs G1, G2, and G3 shown in FIG. 8 represents time, and the vertical axis of graph G1, etc. represents the chuck temperature.

[0117] In the temperature control of a wafer chuck (20) according to the prior art, an operator identifies a change in the chuck temperature and, based on the operator's empirical perspective, applies a PID control parameter (K) as a chuck temperature control parameter. p , K i , K d Adjustments of ) were being carried out.

[0118] For example, as shown in graph G1, when chuck temperatures (Tc11, Tc12, and Tc13) are acquired at sampling timings t11, t12, and t13, respectively, a PID control parameter (K) already set as a chuck temperature control parameter p , K i , K d p-parameter K of ) p , i parameter K i and d parameter K d For each of, an adjustment coefficient C based on the operator's empirical perspective p , adjustment coefficient C i and adjustment coefficient C d is derived, and the adjustment coefficient (C p , C i, C d A new PID control parameter (C) that is multiplied by ) p ×K p , C i ×K i , C d ×K d ) is set.

[0119] In addition, as shown in graph G2, when the chuck temperatures (Tc21, Tc22, and Tc23) are acquired at sampling timings t21, t22, and t23, respectively, the d parameter K d Regarding this, the adjustment coefficient C based on the operator's empirical perspective d ...is derived. In addition, as shown in graph G3, when chuck temperatures Tc31, Tc32, and Tc33 are acquired at sampling timings t31, t32, and t33, respectively, the i parameter K i Regarding this, the adjustment coefficient C based on the operator's empirical perspective i ...is derived. That is, in chuck temperature control according to the prior art, one or more chuck temperatures obtained during inspection of the wafer (W) were used, and chuck temperature control parameters were set based on the operator's empirical perspective.

[0120] [Explanation of the learning model applied to the classification section]

[0121] The chuck temperature control unit (21) according to the present embodiment performs automatic adjustment of chuck temperature control parameters by applying machine learning. Below, the machine learning applied to the chuck temperature control unit (21) is described in detail.

[0122] Although three types of chuck temperature changes are exemplified in FIG. 8, there are infinitely many types of chuck temperature changes. It is difficult to perform machine learning that targets all of the infinitely many chuck temperature changes. Therefore, a learned model is created to classify chuck temperature changes into multiple patterns, and the learned model is applied to the classification unit (102) shown in FIG. 5.

[0123] FIG. 9 is an explanatory diagram of the first pattern of change in chuck temperature. FIG. 9 illustrates the change in chuck temperature by applying a graph format in which the horizontal axis represents time and the vertical axis represents chuck temperature. Each of the graphs G20, G22, G24, and G26 shown in FIG. 9 represents the first pattern of chuck temperature hunting.

[0124] FIG. 10 is an explanatory diagram of a second pattern of chuck temperature change. FIG. 10 illustrates the change in chuck temperature by applying a graph format similar to that of FIG. 9. Each of graphs G30, G32, and G34 shown in FIG. 10 represents a second pattern of chuck temperature change with a long progression time.

[0125] FIG. 11 is an explanatory diagram of the third pattern of chuck temperature change. FIG. 11 illustrates the change in chuck temperature by applying a graph format similar to FIG. 9 and FIG. 10. Each of graphs G40, G42, G44, and G46 shown in FIG. 11 represents the third pattern in which the change in chuck temperature is large.

[0126] FIG. 12 is an explanatory diagram of the fourth pattern of change in chuck temperature. FIG. 12 illustrates the change in chuck temperature by applying a graph format similar to FIG. 9 to FIG. 11. Graph G50 shown in FIG. 12 represents the fourth pattern in which the chuck temperature is stabilized.

[0127] FIG. 13 is a schematic diagram of the creation of a completed model applied to the classification of patterns of chuck temperature change. First, a plurality of temperature change samples representing changes in chuck temperature are represented using a determined number of feature quantities, and the patterns of temperature change of chuck temperature are labeled. FIG. 13 illustrates, as a plurality of temperature change samples, temperature change sample SP1, temperature change sample SP2, temperature change sample SP3, temperature change sample SP4, temperature change sample SP5, and temperature change sample SP6.

[0128] As perspectives for characterizing the change in temperature, the direction of the onset of the temperature change, the time until the temperature change converges, the area of ​​the temperature outside the allowable temperature range, the number of inflection points of the temperature change, the frequency of the temperature change, and the number of times the temperature has deviated from the allowable temperature range can be cited.

[0129] The direction of the onset of the temperature change is derived from the temperature measurement at the first sampling and the temperature measurement at the second sampling. The direction of the onset of the temperature change can be defined as the direction in which the temperature rises as the positive direction and the direction in which the temperature falls as the negative direction.

[0130] The area of ​​temperature outside the allowable temperature range is the area of ​​the portion of the curve representing temperature change that deviates from the allowable temperature range PR. The number of inflection points of temperature change is the number of inflection points on the curve representing temperature change. The frequency of temperature change is the frequency on the curve representing temperature change, and is derived by performing a Fourier transform on the curve representing temperature change.

[0131] As shown in FIG. 13, temperature change samples SP1 and SP2 are labeled with a first pattern of hunting for changes in chuck temperature. Likewise, temperature change samples SP3 and SP4 are labeled with a second pattern of long changes in chuck temperature, and temperature change samples SP5 and SP6 are labeled with a third pattern of large changes in chuck temperature.

[0132] The learning completion model (200) is generated by performing learning on a pair of temperature change patterns classified as a change in chuck temperature and a feature quantity representing the change in temperature as learning data. When a feature quantity representing the change in chuck temperature is input to the learning completion model (200), a temperature change pattern corresponding to the feature quantity representing the change in chuck temperature is output.

[0133] Examples of supervised machine learning algorithms applied to the trained model (200) include k-neighborhood methods, decision trees such as classification trees, random forests, non-linear SVMs, and neural networks. Also, SVM is an abbreviation for Support-Vector Machine.

[0134] A learning completed model (200) is generated by performing learning for each type of wafer (W). For a plurality of learning completed models (200) corresponding to each type of a plurality of wafers (W), the type of wafer (W) is assigned as an index.

[0135] The chuck temperature control unit (21) shown in FIG. 5 may be equipped with a feature derivation unit that derives a feature quantity corresponding to a change in chuck temperature from a change in chuck temperature. That is, the classification unit (102) may be equipped with a feature derivation unit and, when a change in chuck temperature is input, may output a temperature change pattern corresponding to a change in chuck temperature. The feature derivation unit may be a component of the classification unit (102).

[0136] Additionally, the chuck temperature control unit (21) may be provided with a chuck temperature change derivation unit that derives a change in chuck temperature from a plurality of chuck temperatures acquired in a time series order. That is, the classification unit (102) may be provided with a chuck temperature change derivation unit, and when a plurality of chuck temperatures are input in a time series order, it may output a temperature change pattern corresponding to a change in chuck temperature.

[0137] That is, the classification unit (102) may be provided with a chuck temperature change derivation unit that derives a chuck temperature change corresponding to the format of input data representing the input chuck temperature. The learning completed model (200) can be identified as a learning completed model that has learned the relationship between the characteristics of the chuck temperature change and the pattern of the chuck temperature change. Furthermore, the learning completed model (200) described in the embodiment is an example of a learning completed model that has learned the correspondence between the characteristics of the chuck temperature change and the temperature change pattern classified by the chuck temperature change.

[0138] For each of the multiple patterns of chuck temperature change, an adjustment coefficient applied to the chuck temperature control parameter is associated. In the example shown in FIG. 13, for the first pattern of hunting of the chuck temperature change, the chuck temperature control parameter (K p , K i , K d1 Adjustment coefficient (C) applied to ) p1 , C i1 , C d1 ) is corresponded. Likewise, for the second pattern with a long processing time for the change in chuck temperature, the adjustment coefficient (C p2 , C i2 , C d2 ) is corresponded, and in the third pattern where the change in chuck temperature is large, the adjustment coefficient (C p3 , C i3 , C d3 ) is corresponded.

[0139] The relationship between the pattern of change in chuck temperature and the adjustment coefficient is, for each pattern of change in chuck temperature, the chuck temperature control parameter (K p , Ki , K d By repeatedly performing adjustments, the final chuck temperature is derived from the relationship between the chuck temperature control parameter, which has stabilized the chuck temperature, and the initial value of the chuck temperature control parameter. In addition, the procedure for generating the learning completed model (200) described in the embodiment is an example of a method for generating a learning model.

[0140] [Operation of the trained model]

[0141] FIG. 14 is a schematic diagram of the classification of patterns of change in chuck temperature to which the learning completed model shown in FIG. 13 is applied. FIG. 14 shows the steps of a chuck temperature control method that classifies the chuck temperature into one of a plurality of predetermined temperature change patterns and derives an adjustment coefficient corresponding to the temperature change pattern.

[0142] In the chuck temperature acquisition process S10, the chuck temperature is acquired using the chuck temperature acquisition unit (100) shown in FIG. 5. In the classification process S12, the classification unit (102) shown in FIG. 5 outputs a temperature change pattern corresponding to the change in chuck temperature. FIG. 14 shows an example in which a score indicating certainty is output for each temperature change pattern as the output (210) of the temperature change pattern.

[0143] In the chuck temperature control parameter acquisition process S14, the chuck temperature control parameter setting unit (104) outputs a chuck temperature control parameter corresponding to the temperature change pattern output from the classification unit (102) in the classification process S12. In FIG. 14, as a temperature change pattern, a first pattern is adopted in which a change in chuck temperature having the highest value of the score is hunted, and an adjustment coefficient (C) corresponding to the first pattern is p , C i , C d Shows an example where ) is output.

[0144] The chuck temperature control parameter setting unit (104) is an adjustment coefficient (C p , C i , Cd Heating control parameters and cooling control parameters to which ) are applied are set for the chuck temperature control unit. The process of setting heating control parameters and cooling control parameters for the chuck temperature control unit may be included in the chuck temperature control method as a chuck temperature control parameter setting process.

[0145] After the chuck temperature acquisition process S10 shown in FIG. 14, a process for deriving a change in chuck temperature from a plurality of chuck temperatures acquired in a time series order may be performed. The process for deriving a change in chuck temperature may be performed integrally with the chuck temperature acquisition process S10.

[0146] After the process of deriving the change in chuck temperature, a process of deriving a feature quantity corresponding to the change in chuck temperature from the change in chuck temperature may be executed. The process of deriving the feature quantity may be executed integrally with the process of deriving the change in chuck temperature.

[0147] Before the classification process S12, a learning model selection process may be executed to determine the type of wafer (W) to be measured and to select a learning completed model (200) according to the type of wafer (W). The learning completed model selection process may be executed integrally with the classification process S12.

[0148] [Specific example of chuck temperature control]

[0149] FIG. 15 is a schematic diagram showing a specific example of chuck temperature control. FIG. 15 illustrates a case in which multiple semiconductor chips (9) included in one wafer (W) are inspected in sequence, similar to FIG. 7.

[0150] For the inspection of the first semiconductor chip (9) shown as the first contact, the initial value (K) of the chuck temperature control parameter p0 , K i0 , K d0) is set. In the first contact, for example, if the change in chuck temperature is classified into a first hunting pattern, an adjustment coefficient (C) corresponding to the first pattern is set. p1 , C i1 , C d1 ) is acquired.

[0151] In the second contact, the chuck temperature control parameter (C p1 ×K p0 , C i1 ×K i0 , C d1 ×K d0 ) is set. In the second contact, for example, if the change in chuck temperature is classified into a second pattern with a long processing time, an adjustment coefficient (C) corresponding to the second pattern is set. p2 , C i2 , C d2 ) is acquired.

[0152] In the third contact, the chuck temperature control parameter (C p1 ×C p2 ×K p0 , C i1 ×C i2 ×K i0 , C d1 ×C d2 ×K d0 ) is applied. FIG. 15 illustrates a case where, for example, in the third contact, the chuck temperature is stabilized and classified into a fourth pattern.

[0153] In the contact from the 4th time onwards, the chuck temperature control parameter (C p1 ×C p2 ×K p0 , C i1 ×C i2 ×K i0 , C d1 ×C d2 ×K d0 For ), an adjustment coefficient (1, 1, 1) corresponding to the fourth pattern in which the chuck temperature is stable is applied. The adjustment coefficient (1, 1, 1) is applied when the chuck temperature control parameter of the previous contact is not changed.

[0154] That is, the chuck temperature control parameter (K) corresponding to the change in chuck temperature acquired during the first contact. p0 , K i0 , K d0 Regarding ), finally, the adjustment coefficient (C p1 ×C p2 , C i1 ×C i2 , C d1 ×C d2 When ) is applied, chuck temperature control is performed to realize a stable chuck temperature.

[0155] Therefore, if the learning completed model (200) is retrained using the change in chuck temperature acquired during the first contact and the adjustment coefficient of the chuck temperature control parameter that realizes a stable chuck temperature, high-precision chuck temperature control can also be realized. Automatic adjustment of the chuck temperature control parameter to which machine learning is applied can be handled by increasing the explanatory variable even when the combination of PID control loops changes or the number of PID control loops increases.

[0156] The first contact shown in FIG. 15 is identified as the nth contact when n is an integer. The second contact shown in FIG. 15 is identified as the jth contact when j is an integer exceeding n. The third contact shown in FIG. 15 is identified as the kth contact when k is an integer exceeding j. The chuck temperature control parameter for the nth contact is the initial value (K p0 , K i0 , K d0 For ), the adjustment coefficient (C p , C i , C d It is acceptable if ) is applied.

[0157] [Variations of Chuck Temperature Control]

[0158] In the inspection of the wafer (W) shown in FIG. 7, the inspection of the semiconductor chip (94) is performed after the inspection of the semiconductor chip (93), and when a change in the chuck temperature shown as graph G14 is obtained during the inspection of the semiconductor chip (94), the chuck temperature control unit (21) shown in FIG. 5 may not perform the update of the chuck temperature control parameter.

[0159] That is, the chuck temperature control unit (21) determines whether the chuck temperature obtained during inspection of the wafer (W) is outside the specified allowable temperature range, and if the chuck temperature is outside the specified allowable temperature range, it may update the chuck temperature control parameter. In addition, if the chuck temperature is within the specified allowable temperature range, it may not update the chuck temperature control parameter.

[0160] [Example of the structure of a program executed by a computer]

[0161] Various functions of the chuck temperature control unit (21) shown in FIG. 5 are realized by a computer executing a program. Examples of various functions realized by the computer include a function to acquire the chuck temperature, a function to classify changes in the chuck temperature into a defined pattern, and a function to acquire chuck temperature control parameters based on the pattern of changes in the chuck temperature. The program is non-temporary and is stored in a computer-readable memory medium.

[0162] [Effect of the embodiment]

[0163] The chuck temperature control applied to the probe according to the embodiment makes it possible to obtain the following effects.

[0164] [1]

[0165] Using a learned model (200) that has learned the characteristics of the temperature change of the wafer chuck (20) and the temperature change pattern of the wafer chuck (20), the characteristics of the temperature change of the wafer chuck (20) are classified into one of a plurality of predefined temperature change patterns. A temperature control parameter corresponding to the classified temperature change pattern is derived. The derived temperature control parameter is applied to the chuck temperature adjustment unit (20a) that adjusts the chuck temperature. By doing so, automatic temperature adjustment of the wafer chuck (20) according to the temperature change of the wafer chuck (20) is realized.

[0166] [2]

[0167] As a characteristic of the temperature change of the wafer chuck (20), a feature quantity representing the characteristic of the temperature change of the wafer chuck (20) is applied. The learning model learns the correspondence between the feature quantity representing the characteristic of the temperature change of the chuck temperature and a defined temperature change pattern. When the learning completed model receives the feature quantity representing the characteristic of the temperature change of the chuck temperature, a temperature change pattern corresponding to the feature quantity is output. By doing so, the countless temperature changes of the chuck temperature are classified into a pre-defined temperature change pattern.

[0168] [3]

[0169] Among the completed learning models corresponding to each of multiple wafers of different types, the completed learning model corresponding to the wafer under inspection is selected. Accordingly, the chuck temperature change according to the type of wafer under inspection is classified.

[0170] [4]

[0171] PID control parameters are applied as chuck temperature control parameters. The chuck temperature control parameters are the already set PID control parameters (K p , K i , K d Adjustment coefficient (C) for ) p , C i , C d The PID control parameter (C) multiplied by ) p×K p , C i ×K i , C d ×K d ) is applied. By doing so, chuck temperature control is realized in which new PID control parameters based on already set PID control parameters are applied.

[0172] [5]

[0173] As a chuck temperature control parameter, the PID control parameter (K p , K i , K d The adjustment coefficient (C) multiplied for ) p , C i , C d ...is derived. Accordingly, chuck temperature control parameters suitable for PID control are derived.

[0174] The embodiments of the present invention described above may be modified, added, or deleted as appropriate without departing from the spirit of the present invention. The present invention is not limited to the embodiments described above, and many modifications are possible by those skilled in the art within the technical scope of the present invention. Furthermore, embodiments, modifications, and applications may be implemented in appropriate combinations. Explanation of the symbols

[0175] 9 semiconductor chip 9a electrode pad 10 Prober 12 Base 13 Y Stage 14 Y Moving Section 15 X Stage 16 X Moving Part 17 Zθ Stage 18 Zθ Moving Part 20 wafer chuck 20a chuck temperature control unit 20b Chuck temperature sensor 21 Chuck temperature control unit 21a Chuck temperature control unit 23 Support 24 Head Stage 25 Card Holder 25a retention hole 26 probe card 29 Wafer positioning camera 30 Up and down stage 31 Needle position alignment camera 32 Cleaning plate 34 Temperature sensor 35 Probe needle 40 Control unit 42 Card temperature acquisition unit 44 Needle position acquisition unit 46 Wafer information acquisition unit 52 Movement control unit 91 Semiconductor chip 92 Semiconductor Chip 93 Semiconductor Chip 94 semiconductor chip 100 ship temperature acquisition unit 102 Classification Section 102a Classification Section 103 Learning Completed Model Memory 104 Chuck temperature control parameter setting section 104a Chuck temperature control parameter setting unit 110 Heating control parameter derivation unit 110a Heating control parameter derivation unit 112 Cooling control parameter derivation unit 112a Cooling Control Parameter Derivation Unit 200 Training Completed Model 210 Output G1 Graph G2 graph G3 graph G10 graph G12 graph G14 graph G20 graph G22 graph G24 graph G26 graph G30 graph G32 graph G34 graph G40 graph G42 graph G44 graph G46 graph G50 Graph SP1 Temperature Change Sample SP2 Temperature Change Sample SP3 Temperature Change Sample SP4 Temperature Change Sample SP5 Temperature Change Sample SP6 Temperature Change Sample W Wafer

Claims

Claim 1 A temperature control device comprising: a chuck temperature acquisition unit for acquiring a chuck temperature representing the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon; a learning completion model applied to the characteristics of the temporal change of the chuck temperature and the corresponding relationship of a specified number of temperature change patterns classified according to the characteristics of the temporal change of the chuck temperature as a temperature change pattern representing temperature behavior based on the temporal change of the chuck temperature, wherein when the chuck temperature is input, the classification unit outputs the temperature change pattern according to the temporal change of the input chuck temperature; and a temperature control parameter setting unit for deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter, and controlling the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter set using the temperature control parameter setting unit. Claim 2 In claim 1, the temperature control device comprises a classification unit having a chuck temperature change derivation unit that derives a change in the chuck temperature from a plurality of chuck temperatures acquired at different timings. Claim 3 A temperature control device according to claim 1, wherein the learning completion model learns a correspondence relationship between a feature quantity representing a characteristic of the temporal change of the chuck temperature and a temperature change pattern classified according to the characteristic of the temporal change of the chuck temperature, and the classification unit acquires the feature quantity as a characteristic of the temporal change of the chuck temperature and outputs the temperature change pattern corresponding to the feature quantity. Claim 4 In any one of claims 1 to 3, the temperature control parameter setting unit is a temperature control device that derives the temperature control parameter using an adjustment coefficient corresponding to the temperature change pattern. Claim 5 A temperature control device according to any one of claims 1 to 3, comprising a wafer information acquisition unit for acquiring wafer information including the type of wafer to be measured, and a plurality of the learning completed models generated by the learning for each type of wafer, wherein the classification unit selects one learning completed model from the plurality of the learning completed models according to the type of wafer included in the wafer information acquired using the wafer information acquisition unit. Claim 6 A temperature control method comprising: a process of acquiring a chuck temperature representing the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon; a process of using a classification unit to which a learned model is applied, wherein the classification unit is used to learn the characteristics of the temporal change of the chuck temperature and the corresponding relationship of a specified number of temperature change patterns classified according to the characteristics of the temporal change of the chuck temperature as a temperature change pattern representing temperature behavior based on the temporal change of the chuck temperature, and outputting the temperature change pattern according to the temporal change of the chuck temperature when the chuck temperature is input; a process of deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter; and a process of controlling the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter. Claim 7 A non-transient and computer-readable recording medium that records a program for realizing a function of acquiring a chuck temperature representing the temperature of a wafer chuck holding a wafer having a plurality of semiconductor chips formed thereon, a function of outputting the temperature change pattern according to the time change of the chuck temperature when the chuck temperature is input, using a classification unit to which a learned completed model is applied as a temperature change pattern representing temperature behavior based on the time change of the chuck temperature and a temperature change pattern representing temperature behavior based on the time change of the chuck temperature, a function of deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit and setting the temperature control parameter when the chuck temperature is input to the classification unit, and a function of controlling the operation of a chuck temperature adjustment unit that adjusts the chuck temperature by applying the temperature control parameter. Claim 8 A prober comprising a wafer chuck for holding a wafer having a plurality of semiconductor chips formed thereon, a probe card having a probe needle, a relative movement unit for moving the wafer chuck relative to the probe needle, a chuck temperature adjustment device for adjusting the temperature of the wafer chuck, and a temperature control device for controlling the operation of the chuck temperature adjustment device by applying a temperature control parameter, wherein the temperature control device comprises a chuck temperature acquisition unit for acquiring a chuck temperature representing the temperature of the wafer chuck, a learning completed model applied to the characteristics of the temporal change of the chuck temperature and the corresponding relationship of a specified number of temperature change patterns classified according to the characteristics of the temporal change of the chuck temperature as a temperature change pattern representing temperature behavior based on the temporal change of the chuck temperature, a classification unit for outputting the temperature change pattern according to the temporal change of the input chuck temperature when the chuck temperature is input, and a temperature control parameter setting unit for deriving a temperature control parameter corresponding to the temperature change pattern output from the classification unit when the chuck temperature is input to the classification unit and setting the temperature control parameter, and the chuck that adjusts the chuck temperature by applying the temperature control parameter set using the temperature control parameter setting unit. A probe that controls the operation of a temperature control device. Claim 9 A method for generating a learning model that generates a learning completed model by learning the characteristics of the temporal change of a chuck temperature, which is the temperature of a wafer chuck holding a wafer having multiple semiconductor chips formed thereon, and the corresponding relationship of a specified number of temperature change patterns classified according to the characteristics of the temporal change of the chuck temperature, as a temperature change pattern representing temperature behavior based on the temporal change of the chuck temperature.

Images