Method for acquiring pick-up sequence of semiconductor assembly device

Abstract

The invention relates to a method for determining a pick-up sequence of a semiconductor mounting device, which is designed to remove a semiconductor from a semiconductor reservoir according to the pick-up sequence and to mount the semiconductor on a substrate, the method comprises the following steps: providing a value (2) of at least one power parameter for each semiconductor from the semiconductor reservoir; and ascertaining a pick-up sequence on the basis of the provided value of the at least one power parameter such that the substrate (3) can be optimally assembled with respect to the at least one power parameter on the basis of the pick-up sequence.

Description

Technical Field

[0001] The invention relates to a method for determining a pick-up sequence of a semiconductor assembly device and in particular to a method for determining a pick-up sequence of a semiconductor assembly device in such a way that the assembled electronic circuits are optimized with respect to at least one performance parameter based on the pick-up sequence. Background Art

[0002] In power electronics applications, the individual semiconductor chips required are often combined into so-called power modules, that is, the required semiconductor chips are mounted on a carrier such as a printed circuit board or a substrate, such as a ceramic, and are connected to one another via printed conductors. The assembly of the printed circuit board with the semiconductor chips or semiconductors is usually carried out by a so-called semiconductor assembly device.

[0003] Such semiconductor assembly devices, which are also often referred to as pick-and-place devices, are used to remove semiconductors one after another from a processed and pre-measured wafer consisting of a plurality of semiconductors and to place them in an ordered sequence on a carrier, for example a ceramic substrate carrier of a power module, a reel, a tape and reel package, a circuit board or the like, wherein the individual semiconductors mounted on the substrate are usually subsequently connected to one another via printed conductors.

[0004] In order to enable an optimal and high-performance operation of the power module or of the semiconductors thus mounted on a substrate and connected in parallel, it should be ensured that the semiconductors mounted successively on the substrate of the power module are close to one another in terms of their dimensions.

[0005] Document EP 1 480 507 B1 discloses a system in which, when mounted on a substrate, a semiconductor is provided on a wafer stage, where it is received, transported and placed on a substrate by a semiconductor assembly device, which is placed on a support surface of the substrate stage. The wafer stage is oriented at a predetermined angle to the support surface of the substrate stage, wherein the wafer stage is partially located below the substrate stage. The semiconductor assembly device also includes a slide with a pivot arm, which carries a welding head, wherein the pivot arm can be pivoted back and forth between two predetermined rotational positions, wherein in a first rotational position, the longitudinal axis of the pivot arm encloses a predetermined angle with a normal to the support surface of the substrate stage, and wherein in a second rotational position, the longitudinal axis of the pivot arm extends perpendicularly to the support surface of the substrate stage. Summary of the invention

[0006] The object of the present invention is therefore to provide an optimized picking sequence or order for a semiconductor assembly device, based on which semiconductors are removed from a semiconductor storage device and mounted on a substrate or carrier by the semiconductor assembly device, so as to enable optimal and high-performance operation of a power module having semiconductors thus mounted on a substrate and connected in parallel.

[0007] This object is achieved by a method for acquiring a pick-up sequence of a semiconductor assembly device according to the features of claim 1 .

[0008] Furthermore, the object is also achieved by a controller for acquiring a picking sequence of a semiconductor assembly device according to the features of claim 7 .

[0009] According to one embodiment of the present invention, the object is achieved by a method for obtaining a picking order of a semiconductor assembly device, which is configured to take out semiconductors from a semiconductor storage device according to a picking order and mount the semiconductors on a substrate, wherein a value of at least one power parameter is provided for each semiconductor from the semiconductor storage device, and the picking order is obtained based on the value of the at least one power parameter provided so that the substrate can be optimally assembled in terms of at least one power parameter based on the picking order.

[0010] In this context, power parameters refer to parameters that describe the specifications of a semiconductor.

[0011] Furthermore, substrate in this case refers in particular to a measured and processed wafer.

[0012] Being able to optimally assemble the substrate with respect to at least one power parameter further means that the substrate is assembled based on the corresponding values ​​of at least one power parameter of the individual semiconductors from the semiconductor reservoir so that the module finally or ultimately assembled by the semiconductor assembly device has the highest possible performance with respect to the desired application.

[0013] Based on the corresponding value of at least one power parameter of each semiconductor from the semiconductor storage, the substrate is assembled in such a way that the module finally or ultimately assembled by the semiconductor assembly device has the maximum possible performance with respect to the desired application, which has the advantage that costs can be saved. For example, fewer failures will occur when the assembled substrate is subsequently used. In addition, the number of semiconductors required for a specific application or desired power module can be reduced by selecting semiconductors in an optimized manner, which in turn produces a smaller power module. In addition, the sorting of the consumption of semiconductors in production can also be avoided. Therefore, an optimized picking sequence of the semiconductor assembly device is generally provided, which can achieve optimal and high-performance operation of the power module or the semiconductors mounted on the substrate and connected in parallel in this way.

[0014] In this case, the method furthermore comprises, for each semiconductor from the semiconductor store, respectively measuring or detecting a corresponding value of at least one power parameter.

[0015] Measuring the individual values ​​here means in particular that they are detected based on corresponding sensors. A sensor, also called a detector or (measurement) detector, is a technical component that can qualitatively or quantitatively detect specific physical or chemical properties and / or material properties of its surroundings as a measured variable.

[0016] Thus, circumstances outside the actual data processing system on which the picking sequence is acquired can be taken into account and incorporated into the method. Although additional effort may be incurred here due to the corresponding sorting or labeling of the individual semiconductors, this additional effort is compensated many times over by the optimized picking sequence.

[0017] In one embodiment, the step of obtaining the picking sequence based on the provided value of at least one power parameter comprises: obtaining the picking sequence based on the provided value of at least one power parameter and a potential moving time of the semiconductor assembly device.

[0018] The movement time of the semiconductor assembly device refers to the time required to bring the semiconductor assembly device or a corresponding arm of the semiconductor assembly device into the position required for picking up or receiving the corresponding or desired semiconductor.

[0019] This allows the finished assembled substrate or the finished power module to be optimized with regard to its performance and costs, while at the same time also minimizing the movement time required for assembling the substrate by the semiconductor assembly device and thus also the time required overall for assembling the corresponding substrate.

[0020] The method for obtaining a picking sequence is designed in such a way that fluctuations of at least one power parameter on the substrate are minimized in the following way: for example, an average value of the power parameter of different batches is calculated and the distribution of the components is set in such a way that at the end of the wafer picking process, the chips are positioned as close to the average value of the batch as possible.

[0021] A linear or sinusoidal distribution of the chips in the reel is obtained using values ​​close to the average value of at least one power parameter at the beginning and at the end of a plurality of batches.

[0022] Furthermore, the step of obtaining the picking order based on the provided value of at least one power parameter can also include: obtaining the picking order based on a machine learning algorithm.

[0023] Machine learning algorithms are based on the use of statistical methods to train a data processing device so that it can perform a specific task without having to be explicitly programmed for this purpose. The goal of machine learning is to create an algorithm that can learn and make predictions from data.

[0024] Machine learning algorithms have the advantage that they can make relatively precise predictions with relatively few resources and relatively little computing effort, which proves to be particularly advantageous when more than one performance parameter is to be taken into account or when the picking sequence is to be optimized based on more than one performance parameter.

[0025] The at least one power parameter can also be a threshold voltage.

[0026] The threshold voltage, which is also called flux voltage, cut-off voltage, forward voltage or collector saturation voltage, indicates the voltage that must be applied to the barrier of a rectifier or diode in order for the current to be significantly greater than the cut-off current, or the voltage above which a semiconductor diode conducts in the penetration direction.

[0027] Based on the threshold voltage, it is therefore possible to infer whether semiconductors arranged one after the other on the substrate are adapted to one another.

[0028] In this case, the power parameter is a threshold voltage, a connection resistance or a breakdown voltage, but this is only one possible embodiment. The at least one power parameter can then also be, for example, a breakdown voltage or a combination of a plurality of power parameters.

[0029] Another embodiment of the present invention also describes a method for assembling a substrate using semiconductors using a semiconductor assembly device, wherein the semiconductor assembly device is configured to take out semiconductors from a semiconductor storage device in a picking order and assemble the substrate using the taken out semiconductors, wherein the method comprises: obtaining a picking order using the method for obtaining a picking order of a semiconductor assembly device described above; and assembling the substrate using the semiconductor assembly device based on the obtained picking order.

[0030] Therefore, a method for assembling a substrate with semiconductors by means of a semiconductor assembly device is described, which is based on an optimized picking sequence and a picking sequence that can in particular achieve optimal and high-performance operation of a power module or a semiconductor mounted on a substrate and connected in parallel. The substrate is assembled based on the corresponding value of at least one power parameter of each semiconductor from a semiconductor storage so that the substrate finally or finally assembled by the semiconductor assembly device has the maximum possible performance in terms of the desired application, which has the advantage that costs can be saved. For example, fewer faults will occur when the assembled substrate is subsequently used. In addition, by selecting semiconductors in an optimized manner accordingly, the number of semiconductors required for a specific application or a desired power module can be reduced, which in turn produces a smaller power module. In addition, the classification of the consumption of semiconductors in production can also be avoided.

[0031] Another embodiment of the present invention also describes a controller for acquiring a picking order of a semiconductor assembly device, which is configured to take out semiconductors from a semiconductor storage device in a picking order and mount the semiconductors on a substrate, wherein the controller has a providing unit and an acquiring unit, wherein the providing unit is configured to provide a value of at least one power parameter for each semiconductor from the semiconductor storage device, and the acquiring unit is configured to acquire the picking order based on the value of the at least one power parameter provided, so that the substrate can be optimally assembled in terms of at least one power parameter based on the picking order.

[0032] A controller for obtaining an optimized pick-up sequence and a pick-up sequence that enables in particular an optimal and high-performance operation of a power module or a semiconductor mounted on a substrate and connected in parallel is thus described. The substrate is assembled based on the value of at least one power parameter for each semiconductor from a semiconductor storage device so that the substrate finally or ultimately assembled by the semiconductor assembly device has the highest possible performance with respect to the desired application, which has the advantage that costs can be saved. For example, fewer faults will occur when the assembled substrate is subsequently used. In addition, the number of semiconductors required for a specific application or a desired power module can be reduced by a correspondingly optimized selection of semiconductors, which in turn results in a smaller power module. In addition, the wasteful sorting of semiconductors in production can also be avoided.

[0033] In this case, the controller can also have a detection unit which is designed to measure or detect the value of at least one power parameter for each semiconductor from the semiconductor storage. Thus, situations outside the actual data processing device on which the picking sequence is acquired can be taken into account and incorporated into the method. Although additional costs may be incurred here as a result of the sorting or labeling of the individual semiconductors, these additional costs are compensated many times over by the optimized picking sequence.

[0034] In one embodiment, the acquisition unit is further configured to acquire the picking sequence based on the provided value of at least one power parameter and the potential or corresponding movement time of the semiconductor assembly device. This allows the final assembled substrate or the final power module to be optimized in terms of its performance and cost, while also minimizing the movement time of the semiconductor assembly device required for assembling the substrate and thus also minimizing the time required for assembling the corresponding substrate overall.

[0035] Furthermore, the detection unit can also be designed to detect the picking sequence based on a machine learning algorithm. A machine learning algorithm has the advantage that a relatively precise prediction can be made with relatively few resources and relatively little computational effort, which proves to be particularly advantageous when more than one power parameter is to be taken into account or when the picking sequence is to be optimized based on more than one power parameter.

[0036] The at least one power parameter can also be a threshold voltage. Based on the threshold voltage, it can be concluded whether the semiconductors arranged one after the other on the substrate are adapted to one another.

[0037] Here, the fact that the power parameter is a threshold voltage is also only one possible embodiment. Thus, at least one power parameter is, for example, also a breakdown voltage or a combination of a plurality of power parameters.

[0038] In addition, another embodiment of the present invention also describes a semiconductor assembly device, which is configured to take out semiconductors from a semiconductor storage device in a picking order and mount the semiconductors on a substrate, wherein the semiconductor assembly device has a receiving unit, which is configured to receive a picking order obtained by the controller described above for obtaining the picking order of the semiconductor assembly device, and wherein the semiconductor assembly device is configured to assemble the substrate based on the received picking order.

[0039] Therefore, a semiconductor assembly device for assembling substrates with semiconductors based on an optimized picking sequence and a picking sequence that can especially achieve optimal and high-performance operation of power modules or semiconductors installed on substrates and connected in parallel is described. The substrate is assembled based on the corresponding value of at least one power parameter of each semiconductor from a semiconductor storage so that the substrate finally or finally assembled by the semiconductor assembly device has the maximum possible performance with respect to the desired application, which has the advantage that costs can be saved. For example, fewer faults will occur when the assembled substrate is used later. In addition, by selecting semiconductors in an optimized manner, the number of semiconductors required for a specific application or a desired power module can be reduced, which in turn produces a smaller power module. In addition, the classification of the consumption of semiconductors in production can also be avoided.

[0040] In summary, it can be stated that the present invention provides a method for determining a pick-up sequence of a semiconductor assembly device in such a way that a substrate or power module assembled on the basis of the pick-up sequence is optimized with respect to at least one performance parameter.

[0041] The described embodiments and developments can be combined with one another as desired.

[0042] Further possible embodiments, improvements and implementations of the present invention also include combinations of features of the present invention described previously or subsequently with respect to the exemplary embodiments that are not explicitly mentioned. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Any accompanying drawings should be intended to help further understand the embodiments of the present invention. These drawings illustrate the embodiments and are used to explain the principles and schemes of the present invention in conjunction with the description.

[0044] Further embodiments and many of the advantages mentioned ensue with reference to the drawings.The elements shown in the drawings are not necessarily shown true to scale relative to one another.

[0045] in:

[0046] Figure 1A flow chart showing a method for assembling a substrate using a semiconductor by a semiconductor assembly apparatus according to an embodiment of the present invention;

[0047] Figure 2 A schematic block diagram of a controller for obtaining a picking sequence of a semiconductor assembly device according to an embodiment of the present invention is shown.

[0048] In the figures of the drawings, identical reference signs denote identical or functionally identical elements, components or parts, unless otherwise indicated. DETAILED DESCRIPTION

[0049] Figure 1 A flow chart of a method 1 for mounting a substrate with semiconductors by a semiconductor mounting apparatus according to an embodiment of the present invention is shown.

[0050] Semiconductor assembly devices, which are also often referred to as pick-and-place devices, are used to remove semiconductors one after another from a store of semiconductors, i.e. a plurality of semiconductors, or from a large number of semiconductors of the same type, which are usually situated next to one another on a carrier, of measured and processed wafers, and to mount them one after another on a substrate or printed circuit board, wherein the individual semiconductors mounted on the substrate are usually subsequently connected in parallel and connected to one another via printed conductors.

[0051] In order to enable optimal and high-performance operation of the power modules or semiconductors thus mounted on a substrate and connected in parallel, it should be ensured that the semiconductors mounted one after the other on the substrate of the power module and the individual modules are close to one another in terms of their size or power.

[0052] In addition to the required minimum distance and the corresponding structural size, the required or maximum number of semiconductors to be placed on such a substrate for the application is also highly dependent on the maximum power expected or the maximum required power.

[0053] In the case of modules with asymmetry, aging of components, degradation of the effective power or performance and even destruction of the entire module may also occur.

[0054] In order to be able to achieve an optimized performance, it should therefore be ensured that the semiconductors mounted one after the other on a substrate or the individual substrates do not differ significantly from one another with respect to at least one performance parameter.

[0055] Figure 1A method is shown here, in which in step 2, the value of at least one power parameter is provided for each semiconductor from a semiconductor storage device, and in a subsequent step 3, a picking order is obtained based on the provided value of the at least one power parameter in such a way that the substrate can be optimally assembled with respect to at least one power parameter based on the obtained picking order.

[0056] Based on the corresponding value of at least one power parameter of each semiconductor from the semiconductor storage, the substrate is assembled in such a way that the substrate finally or ultimately assembled by the semiconductor assembly device has the maximum possible performance in terms of the desired application, which has the advantage that costs can be saved. For example, fewer failures will occur when the assembled substrate is used later. In addition, by selecting semiconductors in an optimized manner accordingly, the number of semiconductors required for a specific application or a desired power module can be reduced, which in turn produces a smaller power module. In addition, the classification of semiconductor consumption in production can also be avoided. In addition, the classification of semiconductor consumption in production can also be avoided. Therefore, an optimized picking sequence for a semiconductor assembly device is generally provided, which can achieve optimal and high-performance operation of power modules or semiconductors mounted on a substrate and connected in parallel in this way.

[0057] Based on the pick-up sequence determined in this way, it can be ensured that there are only slight, in particular negligible differences between two semiconductors mounted in close succession or between two assembled substrates with respect to at least one power parameter. Based on the determined pick-up sequence, identical power modules or power modules with slight, tolerable differences can also be subsequently constructed.

[0058] In this case, the pick-up sequence can begin and / or end with an increasing or decreasing value of at least one power parameter. In particular, the pick-up sequence can also be configured such that a specific desired shape, such as a parabolic, sinusoidal or sawtooth-shaped course, is produced with respect to the series of values ​​for at least one power parameter of semiconductors installed one after the other.

[0059] Furthermore, the acquired picking sequence can then be made available, in particular, in the form of a data file, for example a text file.

[0060] like Figure 1 As further shown, the method 1 also has a step 4 of detecting the value of at least one power parameter for each semiconductor from the semiconductor storage. In particular, the semiconductors can be divided into classes or semiconductors with the same value of at least one power parameter can be identified. The detection of the individual values ​​can be achieved, for example, by wafer measurement, KGD (Known Good Die) measurement or similar methods.

[0061] Furthermore, further optimization can be achieved by preselecting corresponding or initially available batches and wafers.

[0062] according to Figure 1 In an embodiment, step 3 of acquiring a picking order based on the provided value of at least one power parameter further comprises: acquiring a picking order based on the provided value of at least one power parameter and a potential moving time of the semiconductor assembly device.

[0063] Here, for example, the respective movement times or corresponding movement paths of the semiconductor assembly device, in particular the corresponding parameters of the various possible movement paths, can be considered based on the tolerance level of the value of at least one power parameter. In particular, the larger the tolerance range, the more semiconductors are usually available.

[0064] In addition, according to Figure 1 The step 3 of obtaining the picking sequence based on the provided value of at least one power parameter of the embodiment further comprises: obtaining the picking sequence based on a machine learning algorithm. In this case, for example, the machine learning algorithm, such as a neural network, can be trained on data about the picking sequence known from other batches, and on the corresponding values ​​of the value of at least one power parameter, and on the properties of the corresponding last assembled or finally obtained substrate, and in particular on the correlation between these values.

[0065] Furthermore, at least one power parameter is a threshold voltage.

[0066] The threshold voltage describes, for example, the switching threshold of a metal oxide semiconductor field effect transistor, wherein time differences caused by different switching thresholds of semiconductors mounted one after the other are undesirable since they can lead to different loads on the individual semiconductors, in particular when the corresponding module is switched on or off.

[0067] Figure 2 A schematic block diagram of a controller 10 for acquiring a picking sequence of a semiconductor assembly device according to an embodiment of the present invention is shown.

[0068] like Figure 2 As shown, the controller 10 has a providing unit 11 and an acquisition unit 12, wherein the providing unit is configured to provide a value of at least one power parameter for each semiconductor from the semiconductor storage, and the acquisition unit is configured to acquire a picking sequence based on the value of the at least one power parameter provided, so that the substrate can be optimally assembled in terms of at least one power parameter based on the picking sequence.

[0069] The providing unit can be, for example, a receiver which is designed to receive corresponding values ​​from a detection unit or one or more sensors. The acquisition unit can also be implemented, for example, based on a code which is stored in a memory and can be executed by a processor.

[0070] according to Figure 2 In an embodiment of the present invention, the control unit 10 further comprises a detection unit 13 which is designed to detect the value of at least one power parameter for each semiconductor from the semiconductor reservoir.

[0071] The detection unit can be realized here, for example, on the basis of corresponding sensors.

[0072] Furthermore, the detection unit 12 shown is designed to detect a pick-up sequence based on the provided value of the at least one power parameter and a potential or corresponding movement time of the semiconductor assembly device.

[0073] Furthermore, the depicted acquisition unit 12 is designed to acquire a picking sequence based on a machine learning algorithm.

[0074] according to Figure 2 In an implementation manner, at least one power parameter is a threshold voltage.

Claims

1. A method for obtaining a pick-up sequence of a semiconductor assembly device, the semiconductor assembly device being configured to take out a semiconductor from a semiconductor storage device according to the pick-up sequence and to mount the semiconductor on a substrate, in, The method comprises the following steps: - providing a value (2) of at least one power parameter for each semiconductor from the semiconductor storage; as well as - the picking sequence is determined based on the provided value of at least one power parameter in such a way that the substrate (3) can be optimally mounted based on the picking sequence with respect to the at least one power parameter.

2. The method according to claim 1, in, The method further comprises the following steps: - a respective value (4) of the at least one power parameter is detected for each semiconductor from the semiconductor store.

3. The method according to claim 1 or 2, in, The step (3) of acquiring the picking sequence based on the provided value of at least one power parameter comprises: acquiring the picking sequence based on the provided value of at least one power parameter and a potential moving time of the semiconductor assembly device.

4. The method according to any one of claims 1 to 3, in, The step (3) of obtaining the picking order based on the provided value of at least one power parameter comprises: obtaining the picking order based on a machine learning algorithm.

5. The method according to any one of claims 1 to 4, in, The at least one power parameter is a threshold voltage, an access resistance or a breakdown voltage.

6. A method for mounting a substrate using a semiconductor by a semiconductor mounting device, the semiconductor mounting device being configured to take out a semiconductor from a semiconductor storage in accordance with the pick-up sequence and to mount the semiconductor on a substrate, in, The method (1) comprises the following steps: - obtaining the picking sequence by the method for obtaining the picking sequence of a semiconductor assembly device according to any one of claims 1 to 6; and - Mounting the substrate (5) by the semiconductor mounting device based on the acquired picking sequence.

7. A controller for obtaining a pick-up sequence of a semiconductor assembly device, the semiconductor assembly device being configured to take out a semiconductor from a semiconductor storage according to the pick-up sequence and to mount the semiconductor on a substrate, in, The controller (10) has a providing unit (11) and an acquiring unit (12), wherein the providing unit is configured to provide at least one power parameter value for each semiconductor from the semiconductor storage, and the acquiring unit is configured to acquire the picking sequence based on the provided at least one power parameter value so that the substrate can be optimally assembled in terms of at least one power parameter based on the picking sequence.

8. The controller according to claim 7, in, The control unit also has a detection unit (13) which is designed to detect a respective value of at least one power parameter for each semiconductor from the semiconductor store.

9. The controller according to claim 7 or 8, in, The acquisition unit (12) is configured to acquire the picking sequence based on the provided value of at least one power parameter and a potential movement time of the semiconductor assembly device.

10. A controller according to any one of claims 7 to 9, in, The acquisition unit (12) is configured to acquire the picking sequence based on a machine learning algorithm.

11. A controller according to any one of claims 7 to 10, in, The at least one power parameter is a threshold voltage.

12. A semiconductor assembly device configured to take out a semiconductor from a semiconductor storage in a picking order and mount the semiconductor on a substrate, in, The semiconductor assembly device has a receiving unit for receiving a picking order acquired by a controller for acquiring a picking order of the semiconductor assembly device according to any one of claims 7 to 11, and wherein the semiconductor assembly device is configured to assemble substrates based on the received picking order.

Images