JP5541393B2 - Method for manufacturing power semiconductor device - Google Patents

Description

  The present invention relates to a method for manufacturing a power semiconductor device including a power module and a cooler.

  The power semiconductor device includes a power module and a cooler attached to the power module. The power module is obtained by sealing a power element such as an IGBT by a transfer molding method or the like. The power module includes a heat dissipation surface exposed from the mold resin for heat dissipation. The heat radiating surface is provided in order to release heat of the power element to the outside of the power module, and is made of a metal having good heat conduction. Such a heat radiating surface and a cooler are fixed by soldering. Attaching a cooler to the heat radiating surface is effective in improving the thermal resistance of the power module.

  Here, in some cases, heat conduction grease is used for fixing the heat radiation surface of the power module and the cooler. In order to reduce the thermal resistance of the power module, the thermal conductive grease is preferably formed thin. However, the heat radiation surface of the power module and the plane of the cooler have certain irregularities, and the minimum film thickness of the grease is limited by the degree of the irregularities. Therefore, if the thermal resistance is to be reduced when the thermal conductive grease is used, it is necessary to reduce the unevenness of the heat dissipation surface and the cooler plane. In addition, since the power element of the power module is silicon, ceramic, copper, or aluminum is used for the substrate and wiring, so thermal expansion is not uniform and a certain warp occurs due to temperature change. For this reason, it has been difficult to reduce the aforementioned unevenness. Further, since the thermal conductivity of the thermal conductive grease is significantly lower than that of Cu or Si, it is difficult to reduce the thermal resistance when the thermal conductive grease is used for fixing the heat radiation surface and the cooler.

  Therefore, as described above, when the heat radiation surface of the power module and the cooler are fixed by soldering, the thermal resistance can be reduced as compared with the case where the thermal conductive grease is used. Patent Documents 1-4 describe a technique related to heat dissipation of a power module.

JP 2004-022973 A JP 2007-096252 A JP 2007-043041 A JP 2001-308266 A

  In order to solder the heat radiating surface of the power module and the cooler, since the heat capacity of the cooler is high, an apparatus for heating the solder material to a temperature higher than the melting point becomes large. As a result, there is a problem that the manufacturing cost increases. In particular, when the power module has a plurality of soldering surfaces, there is a problem that it is difficult to ensure an appropriate reflow profile for each soldering surface.

  The power semiconductor device described above performs X-ray inspection in order to ensure that there is no defect such as a void at the joint between the heat radiation surface of the power module and the soldering surface of the cooler during the manufacturing process. However, since the cooler usually has a gap inside, there is a problem that it is difficult to find voids by X-ray inspection.

  A plurality of power modules may be fixed to both the front and back surfaces of the cooler so as to sandwich the cooler. In that case, first, the power module is fixed to the surface of the cooler and necessary wire bonding is performed, and then the power semiconductor device is inverted and the same processing is performed on the back surface. FIG. 12 shows an example of such a power semiconductor device. In FIG. 12, a silicon nitride plate 504 is fixed to the surface of a cooler 502 having a coolant flow path 500. Further, a metal circuit board 506 is disposed on the silicon nitride plate 504, and the semiconductor element 530 is fixed on the metal circuit board 506. Wire bonding is performed at predetermined positions of the semiconductor element 530 and the metal circuit board 506. When these series of steps are completed, the power semiconductor device is inverted, and the same processing is performed with the back surface of the cooler facing upward. In order to invert the power semiconductor device described above, a special mechanism is required. Therefore, there has been a problem that an apparatus used for manufacturing the power semiconductor device is enlarged. In addition, if this reversal is performed manually, there is a problem that the work becomes complicated.

  In many cases, the power module enhances insulation by sealing the vicinity of the power element disposed in the package with silicone gel. However, when the power modules are arranged on both sides of the cooler as described above, the power module on the cooler surface side is filled with silicone gel and cured, and then the power semiconductor device is turned over to reverse the power on the back side of the cooler. The same processing is performed for the module. Since filling and curing of the silicone gel use gravity, there is a problem that the processing of the power module on the front surface and the back surface of the cooler cannot be performed at the same time, which hinders productivity.

  For example, a power semiconductor device may be attached to a moving body such as a hybrid car, an electric car, or a train. At this time, the silicone gel of the power module located below the cooler causes a phase lag with respect to vibration, and moves differently from the casing (package) of the power module and damages the wires built in the power module. was there. As a result, there is a problem that the period during which the reliability of the power semiconductor device can be guaranteed is shortened.

  In order to avoid this problem, it is conceivable not to use a silicone gel. In this case, there is a problem that the apparatus becomes large in order to ensure the creeping insulation distance.

  The present invention has been made to solve the above-described problems, and provides a method for manufacturing a power semiconductor device that can facilitate manufacture, reduce the size of the device, and improve reliability. Objective.

  In the method for manufacturing a power semiconductor device according to the present invention, the heat dissipation surface of the first power module having the heat dissipation surface that is sealed with the mold resin and exposed from the mold resin, and the surface of the first cooling part are first. Solder the first step of soldering with solder, the heat dissipation surface of the second power module having the same configuration as the first power module, and the surface of the second cooling unit having the same configuration as the first cooling unit with the second solder And after the first step and the second step, the first cooling portion and the second cooling portion are formed such that the first cooling portion and the second cooling portion form a gap serving as a refrigerant passage. And a third step of fixing the cooling unit.

  According to the present invention, a small and highly reliable power semiconductor device can be easily manufactured.

1 is a diagram illustrating an appearance of a power semiconductor device according to a first embodiment. It is a figure explaining the flowchart which manufactures the semiconductor device for electric power. It is sectional drawing explaining the structure of a 1st cooling part. It is sectional drawing explaining the structure of a 1st power module. It is sectional drawing explaining the state which soldered the 1st cooling part and the 1st power module. It is sectional drawing explaining the state which soldered the 2nd cooling part and the 2nd power module. It is sectional drawing explaining the adhering of a 1st cooling part and a 2nd cooling part. It is a figure explaining the semiconductor device with which the several power module was fixed to the surface and back surface of a cooler. It is a figure explaining the semiconductor device for electric power by which the power module of the surface of a cooler is shifted by half pitch with respect to the power module of the back surface of a cooler. It is a figure explaining the external appearance of the semiconductor device for electric power of Embodiment 2. FIG. It is a figure explaining the external appearance of the semiconductor device for electric power of Embodiment 3. FIG. It is a figure explaining the external appearance of a known semiconductor device for electric power. It is a figure explaining the semiconductor device for electric power of Embodiment 4. FIG. It is a figure explaining a straight fin. It is sectional drawing explaining the adhering of a 1st cooling part and a 2nd cooling part. It is a figure explaining a seal part.

Embodiment 1
This embodiment will be described with reference to FIGS. In addition, even if it is a different figure number, the same code | symbol may be attached | subjected to the same or corresponding component, and description of multiple times may be abbreviate | omitted. The same applies to other embodiments.

  FIG. 1 is a diagram for explaining the external appearance of the power semiconductor device 10 of this embodiment. The power semiconductor device 10 includes a cooler 30 having a first cooling unit 18 and a second cooling unit 28. The first power module 12 is soldered to the first cooling unit 18. The second power module 22 is soldered to the second cooling unit 28. Both the first power module 12 and the second power module 22 include power elements inside. The first power module 12 includes a main terminal 14 and a control terminal 16, and the second power module 22 includes a main terminal 24 and a control terminal 26. Hereinafter, a method for manufacturing the power semiconductor device 10 of the present embodiment will be described.

  FIG. 2 is a flowchart illustrating a method for manufacturing the power semiconductor device 10. At the start of this flowchart, the cooler is divided into a first cooling unit 18 and a second cooling unit 28. First, in step 40, the first cooling unit 18 and the first power module 12 are soldered. The configuration of the first cooling unit 18 will be described with reference to FIG. FIG. 3 is a cross-sectional view of the first cooling unit 18. The first cooling unit 18 is in contact with the power module on the front surface, and the refrigerant is scheduled to pass through on the back surface. A plurality of pin fins 50 are arranged on the back surface of the first cooling unit 18. The pin fin 50 increases the heat radiation area from the power module to the refrigerant. It also has a refrigerant stirring action. Further, a seal portion 52 is disposed on the back surface of the first cooling portion 18. The seal portion 52 is a portion that is fixed to a seal portion of the second cooling portion 28 described later. The seal portion 52 is disposed so as to surround the pin fin 50. The seal portion 52 has a tapered cross section having an angle of about 30 ° and a width of about 2 mm on the pin fin 50 side, for example. A concave portion is provided on the outer peripheral side opposite to the pin fin 50. The first cooling unit 18 has a thickness of about 7 mm, for example, and is made of Al. In the portion where the pin fins 50 are formed, the plate-like portion of the first cooling unit 18 out of the thickness of 7 mm is about 3 mm, and the portion of the pin fins 50 is formed with a thickness of less than 4 mm. In addition, a portion thicker than the seal portion 52 is formed outside the seal portion 52.

  FIG. 4 is a cross-sectional view illustrating the configuration of the first power module 12. In the first power module 12 of the present embodiment, a power element 66 is fixed to the surface of a metal plate 64 made of Cu having a thickness of approximately 1 mm to 3 mm via a solder 68. The power element 66 is wired between the power elements, the main terminal 14, and the control terminal 16 by wires 70. The power element 66 is mainly made of silicon, but may be a compound semiconductor such as SiC or GaN. Further, a metal layer 62 is formed on the back surface of the metal plate 64 with an insulating layer 60 interposed. The metal layer 62 is, for example, a Cu foil having a thickness of about 100 μm to 200 μm.

  The above-described configuration is covered with the mold resin 72 so that the main terminal 14, the control terminal 16, and the metal layer 62 are exposed to the outside. The material and content of the filler of the mold resin 72 are adjusted so that the linear expansion coefficient is close to Cu. For example, a material such as Al may be used as the metal layer 62. In that case, the linear expansion coefficient of the mold resin 72 is set to a value close to that of Al or the like.

  Thus, in the first power module 12 of the present embodiment, the metal layer 62 is fixed to the insulating layer 60. Therefore, the insulation inside the mold resin 72 is maintained and the metal layer 62 can be soldered.

  FIG. 5 is a cross-sectional view illustrating a state where the first cooling unit 18 and the first power module 12 are soldered in Step 40. In step 40, the surface of the first cooling unit 18 and the metal layer 62 of the first power module 12 are soldered. Solder 80 is used for this soldering.

  When step 40 is completed, the process proceeds to step 42. In step 42, the second cooling unit 28 and the second power module 22 are soldered. The configuration of the second cooling unit 28 is the same as that of the first cooling unit 18. The second power module 22 has the same configuration as the first power module 12.

  FIG. 6 is a cross-sectional view for explaining the step 42, and the metal layer 62 of the second power module 22 is fixed to the second cooling part 28 by the solder 90 as shown in the figure. Since the fixing of the second power module 22 and the second cooling unit 28 performed in step 42 is the same as the fixing of the first power module 12 and the first cooling unit 18 in step 40, detailed description thereof is omitted.

  When step 42 is completed, the process proceeds to step 44. In step 44, the first cooling unit 18 and the second cooling unit 28 are fixed so as to form a gap serving as a refrigerant passage. FIG. 7 is a cross-sectional view illustrating step 44. As shown in FIG. 7, the first cooling part 18 and the second cooling part 28 are integrated by a liquid packing 100 applied to the seal part 52 of both. When the first cooling unit 18 and the second cooling unit 28 are fixed by the liquid packing 100, a gap 102 that is a refrigerant passage is formed. Further, for example, a part of the first cooling unit 18 may be provided with a port for separately fixing the inlet and outlet of the refrigerant, and the pipe joint may be fixed by the liquid packing 100 at the same time.

  The supply amount of the liquid packing 100 described above is smaller than the volume of the gap between the seal part 52 of the first cooling part 18 and the seal part 52 of the second cooling part 28. More details are as follows. First, the liquid packing 100 is disposed at a substantially central portion of the seal portion 52. When the first cooling unit 18 and the second cooling unit 28 are bonded and integrated, the liquid packing 100 is pushed out to the concave portion and the tapered portion of the seal portion 52. However, since the amount of the liquid packing 100 supplied is smaller than the volume of the gap between the seal portion 52 of the first cooling portion 18 and the seal portion 52 of the second cooling portion 28, the liquid packing 100 protrudes outside the seal portion 52. Can be suppressed. Therefore, the shape of the coolant channel can be stabilized even after the first cooling unit 18 and the second cooling unit 28 are bonded together. Thus, the 1st cooling part 18 and the 2nd cooling part 28 are integrated, the cooler 30 is formed, and the process of step 44 is finished. The power semiconductor device is completed through the above steps.

  According to the configuration of the power semiconductor device of the present embodiment, the first cooling unit 18 and the first power module 12 are soldered before the cooler 30 is configured. Moreover, the 2nd cooling part 28 and the 2nd power module 22 are soldered. Therefore, the problem that the cooler having a high heat capacity becomes an obstacle when the solder material is heated to the melting point or more can be solved.

  Moreover, when soldering the power module to the front and back surfaces of the cooler, it is difficult to perform soldering while holding the power modules on both sides. In addition, it is difficult to maintain an appropriate reflow profile for each soldering surface. For this reason, there is a problem that the equipment for manufacturing the power semiconductor device is enlarged, but the configuration of the present embodiment can solve this problem. That is, in the present embodiment, the power module is soldered in a state where the cooler is divided, so that the front and back surfaces of the cooler are not soldered simultaneously. Therefore, since the power semiconductor device of this embodiment can be manufactured by a normal assembly device, it is excellent in terms of cost and ease of manufacture.

  The power semiconductor device of this embodiment can determine the presence or absence of a void or the like at the joint by performing an X-ray inspection at the stage where the metal layer 62 that is the heat dissipation surface of the power module and the first cooling unit 18 are soldered. That is, since the X-ray inspection can be performed on the first cooling unit 18 or the second cooling unit 28 that does not have a gap, it is possible to improve the determination accuracy of the presence or absence of defects such as voids. In addition, it is preferable to use a jig (not shown) having a thickness that makes the total X-ray attenuation with the first cooling unit the same as the X-rays passed during the X-ray inspection. Such a jig is simple in that the shape is transferred with the same material as that of the first cooling part and the thickness becomes constant when combined.

  The power semiconductor device of this embodiment does not require a mechanism for inverting the cooler. That is, since the cooler 30 is divided into the first cooling part 18 and the second cooling part 28 in the manufacturing process of the power semiconductor device, both of them are the same with the surface to be soldered to the power module facing upward. Can be assembled in the process. Therefore, since a mechanism for reversing the apparatus is not required, an increase in the size of the manufacturing apparatus can be avoided, and simple assembly using gravity can be performed.

  Since the power semiconductor device of this embodiment does not use silicone gel, it is possible to avoid a decrease in reliability of the power semiconductor device due to vibration of the silicone gel. Moreover, the factor which impedes productivity which waits for completion | finish of the process of the surface of a cooler, and processes the back surface can also be eliminated.

The power semiconductor device of this embodiment is excellent from the viewpoint of downsizing the device. That is, disposing the power modules on the front and back surfaces of the cooler 30 contributes to downsizing of the power semiconductor device. Further, it is possible to avoid the increase in the size of the device in order to secure the creeping insulation distance by protecting the power element and the like with the mold resin, which contributes to the reduction in size. In the present embodiment, the heat transfer coefficient of about 20000 to 30000 [W / m ² · K] can be realized by the pin fin 50 while realizing miniaturization as described above.
In addition, when the pin fin 50 is used, the rigidity of the cooler can be reduced as compared with the case where a straight fin described later is used. Therefore, the long-term reliability of the connection by the solder 80 and the solder 90 can be improved.

  Further, the material and content of the filler are adjusted so that the mold resin 72 of the first power module 12 and the second power module 22 has a linear expansion coefficient close to that of the metal layer 62. Therefore, it is possible to manufacture a power semiconductor device with high reliability by suppressing warpage of the power module due to heat.

  As described above, the power semiconductor device according to the present embodiment solves the problem when the power module is attached to the front and back surfaces of the cooler in order to reduce the size of the power semiconductor device and facilitates its manufacture. Specifically, it is possible to obtain benefits such as simplification of the heating mechanism by soldering the power module in a state where the cooler is divided.

  The configuration of the present embodiment can be variously modified without departing from the scope of the present invention. For example, the first cooling unit 18 and the second cooling unit 28 are sealed with the liquid packing 100, but a gasket may be used instead of the liquid packing 100. In addition, other methods may be used as long as the first cooling unit 18 and the second cooling unit 28 can be fixed so as to form a gap serving as a refrigerant passage.

  For example, as shown in FIG. 8, a plurality of power modules (112, 114, 116) are soldered to the first cooling section 18 with solder 110, and a plurality of power modules (118, 120, 122) may be soldered.

  For example, as shown in FIG. 9, the power modules may be shifted by half a pitch on the front and back sides of the cooler 30. With such an arrangement, thermal interference between the front and back surfaces of the cooler 30 can be suppressed. That is, the place where the temperature is highest in the cooler 30 is a joint (fixed) portion with the power module. The temperature distribution in the cooler 30 is a concentric semicircular distribution centering on the joint between the cooler 30 and the power module. Therefore, if the power module is fixed to the back side of the cooler immediately below the power module fixed to the cooler surface, the temperature rise of the refrigerant in the region sandwiched between the power modules becomes remarkable, and the power module is not sufficiently cooled. Can be considered. And since the cooling capacity of the refrigerant | coolant of the area | region which is not pinched | interposed into a power module will not be utilized, the cooling capacity of the cooler 30 cannot be utilized to the maximum, and it is unpreferable. However, as shown in FIG. 9, if the power modules are shifted by half a pitch on the front and back of the cooler 30, there is no region where the temperature rise of the refrigerant is significant, and the temperature difference between the refrigerant and the power module is large. Since the state can be maintained, the cooling effect of the power module can be enhanced. Therefore, by disposing the power modules above and below the cooler as shown in FIG. 9, it is possible to maintain a large temperature difference between the refrigerant and the power module and to efficiently cool the power module. Therefore, it is possible to avoid an increase in size and cost of the power semiconductor device due to an increase in the size of the cooler and enhancement of the cooling function. Moreover, the flow path of a refrigerant | coolant can be simplified and the weight of the pump which circulates a refrigerant | coolant can be reduced.

Embodiment 2
This embodiment will be described with reference to FIG. FIG. 10 is a diagram illustrating an appearance of a power semiconductor device in which a main wiring block or the like is added to the power semiconductor device described in the first embodiment.

  The main terminal 14 of the first power module 12 and the main terminal 24 of the second power module 22 extend in the same direction. That is, the main terminals of the power module are concentrated on one side of the cooler 30. The main wiring block 200 is fixed to the “one side surface of the cooler 30”. The main wiring block 200 has predetermined wiring inside. Two electrodes extend from the main wiring block 200 and are joined to the main terminal 14 and the main terminal 24 at the joint portions 202 and 212, respectively. This joining is performed by a technique such as TIG welding from the distal direction of the main terminals 14 and 24, for example.

  The control terminal 16 of the first power module 12 is fixed to the control board 206, and necessary electrical connection is performed. Similarly, the control terminal 26 of the second power module 22 is fixed to the control board 216 and necessary electrical connection is performed. Further, a shield plate 204 is fixed to the surface opposite to the surface where the first power module 12 is fixed to the cooler 30. Similarly, a shield plate 214 is fixed to the surface opposite to the surface where the second power module 22 is fixed to the cooler 30. The shield plates 204 and 214 are disposed between the circuit loop of the main wiring and the control boards 206 and 216 so as to protrude in the extending direction of the main terminals 14 and 24 so as to block both. Since the shield plates 204 and 214 are arranged in this way, it is impossible to reach the control board linearly from the main circuit.

  According to the configuration of the present embodiment, the main terminal 14 and the main terminal 24 extend in the same direction. Therefore, welding with the electrodes of the main wiring block 200 can be performed from the same direction for the plurality of main terminals. Therefore, productivity can be improved.

  According to the configuration of the present embodiment, the power semiconductor device can be reduced in size. In general, it is necessary to secure an insulation distance for an electrode exposed portion such as a main terminal in order to ensure insulation from the surroundings. In this embodiment, the main wiring block 200 and the main terminals 14 and 24 are close to each other, and necessary wiring is accommodated in the main wiring block 200, so that the electrode exposed portion itself is extremely small. Therefore, it is not necessary to increase the size of the power semiconductor device in order to secure the insulation distance, and the power semiconductor device can be reduced in size and weight.

  According to the configuration of the present embodiment, it is possible to reduce the influence of noise received by the control board. Generally, in a power semiconductor device, a high voltage is turned on and off in a short period of several μs, so that high dV / dt and high dI / dt occur, and peripheral circuits are affected by noise. For example, when a voltage change occurs in a closed circuit inside the main wiring unit or the power module, a magnetic field is generated in a direction perpendicular to the closed circuit, and a voltage crossing occurs in the circuit crossing the magnetic field, which may cause adverse effects such as malfunction. Therefore, in this embodiment, the shield plates 204 and 214 are arranged so that the magnetic field cannot reach the control boards 206 and 216 linearly from the main circuit. Therefore, since the control boards 206 and 216 are not easily influenced by noise due to the electromagnetic shielding effect of the shield plates 204 and 214, adverse effects such as malfunction can be avoided.

  The configuration of the present embodiment can be variously modified without departing from the scope of the present invention. For example, the shield plate may be disposed away from the power module by any holding means. In this case, since the creeping distance between the main terminal of the power module and the shield plate can be increased, the power semiconductor device can be used with a high operating voltage without increasing the size of the power semiconductor device. In addition, at least a modification corresponding to the first embodiment can be made.

Embodiment 3
This embodiment will be described with reference to FIG. FIG. 11 is a diagram illustrating the appearance of a power semiconductor device in which a smoothing capacitor or the like is added to the power semiconductor device described in the second embodiment.

  An electrode extending from the main wiring block 200 (referred to as an electrode 306 of the main wiring block) is connected to the large current wiring portion 302. The large current wiring portion 302 connects a power semiconductor device and a smoothing capacitor 300 described later. Smoothing capacitor 300 is connected to high-current wiring portion 302 by an electrode extending from smoothing capacitor 300 (referred to as smoothing capacitor electrode 304). Therefore, the electrode 306 of the main wiring block is connected to the electrode 304 of the smoothing capacitor via the large current wiring portion 302.

  As is clear from FIG. 11, the smoothing capacitor electrode 304, the main terminals 14 and 24, and the main wiring block electrode 306 are drawn out on the same plane. That is, they are wired with parallel plates.

  Generally, the power module has a P terminal, an N terminal, and an output terminal (U terminal, V terminal, W terminal), and the P terminal and the N terminal are connected to a smoothing capacitor. The surge during the off operation of the power element is absorbed by the smoothing capacitor. Here, since the loss of the power semiconductor device increases when the surge voltage is large, the surge voltage is preferably reduced. In order to reduce the surge voltage, it is effective to reduce the closed loop inductance of the P terminal N terminal and the smoothing capacitor. In order to reduce the inductance, the smoothing capacitor 300 is made as close as possible to the PN terminal, and the wiring is insulated while the space gap is reduced and the surface area is increased. The wiring having such a configuration is called a parallel plate. Here, even if the parallel plate forms a curved surface, an inductance reduction effect can be obtained by maintaining a predetermined insulation distance, but the inductance increases when the curved surface on the inside and outside of the curved surface are different.

  According to the configuration of the present embodiment, since the parallel plates are arranged by laminating planes, it is possible to reduce inductance and suppress surges while avoiding the harmful effects of configuring curved surfaces. And since a surge can be suppressed, an energy loss can be reduced and a loss can be suppressed. Therefore, since the heat radiation to the cooler 30 can be suppressed, the cooler 30 can be downsized, and the power semiconductor device can be downsized.

  The configuration of the present embodiment can be variously modified without departing from the scope of the present invention. For example, in a hybrid vehicle or an electric vehicle, the amount of battery mounted may be limited in terms of weight and cost, and a driving voltage higher than the battery voltage may be used in combination with a booster circuit. In such a case, the above-described effects can be obtained by arranging the terminals of the reactor and the boosting capacitor like the electrode 304 of the smoothing capacitor in this embodiment. As described above, the wiring paths through which a large current passes are concentrated on the same plane, and the parallel plate is utilized to the maximum, whereby the power semiconductor device can be reduced in size and weight. In addition, at least a modification corresponding to the first embodiment can be made.

Embodiment 4
This embodiment will be described with reference to FIGS. 13-16. FIG. 13 is a view for explaining the appearance of the power semiconductor device of this embodiment. However, the inside of the cooler 600 is described for convenience of explanation. The 1st cooling part 602 is fixed with the 2nd cooling part 604 so that the space | gap used as a refrigerant path may be formed. Then, instead of the pin fins of the first embodiment, straight fins 606 are arranged in the first cooling unit 602. Similarly, straight fins 608 are disposed in the second cooling unit 604.

  FIG. 14 is a perspective view of the straight fin 608. Here, the double-headed arrow in FIG. 14 is the flow direction of the refrigerant. As can be seen from FIG. 14, the straight fins 608 are formed to extend in parallel with the flow direction of the refrigerant. Further, the height of the straight fin 608 is determined so as not to contact the straight fin 606 when the first cooling portion 602 and the second cooling portion 604 are fixed with a gasket or packing. Accordingly, the sum of the heights of the straight fin 608 and the straight fin 606 is smaller than the height of the gap through which the refrigerant passes. In the present embodiment, the height of the gap through which the refrigerant passes is 3 mm, the width of the straight fin 608 is 1.5 mm, and the distance between the straight fin and the adjacent straight fin is about 2 mm. Such straight fins 608 are formed, for example, by die casting. The straight fin 606 is formed in the same manner.

15 and 16 are diagrams for explaining the connection between the first cooling unit 602 and the second cooling unit 604. FIG. 15 illustrates a state in which the first cooling unit 602 and the second cooling unit 604 are connected by the seal unit 52. Further, FIG. 16 shows a state in which the liquid packing 100 is applied to the seal portion 52. Since these have been described in the previous embodiment, a detailed description thereof will be omitted. The power semiconductor device of this embodiment has the above-described configuration. In this embodiment, the heat transfer coefficient of about 10000 [W / m ² · K] can be realized by using the straight fins 606 and 608.

  According to the configuration of the present embodiment, the cooling function of the cooler 600 can be enhanced by the straight fins 606 and the straight fins 608. In addition, the straight fin 606 and the straight fin 608 contribute to increasing the rigidity of the cooler 600. Therefore, the rigidity of the cooler 600 can be secured even if the members constituting the cooler 600 are thinned. Thereby, the member of the side wall of the cooler 600 can be made thin, for example. Therefore, the cooler 600 can be reduced in size and weight.

  The configuration of the present embodiment can be variously modified without departing from the scope of the present invention. For example, intermittent straight fins may be used instead of the straight fins without any gaps as in this embodiment. With such a configuration, the rigidity of the cooler can be increased and the heat transfer rate can be increased. In addition, at least a modification corresponding to the first embodiment can be made.

  DESCRIPTION OF SYMBOLS 10 Power semiconductor device, 12 1st power module, 18 1st cooling part, 22 2nd power module, 28 2nd cooling part, 30 cooler, 50 pin fin, 52 seal part, 100 liquid packing, 200 main wiring block, 204 Shield Plate, 214 Shield Plate, 206 Control Board, 216 Control Board, 300 Smoothing Capacitor, 302 Large Current Wiring Section, 606 Straight Fin, 608 Straight Fin

Claims (2)

A first step of soldering the heat-dissipating surface of the first power module having a heat-dissipating surface that is sealed with a mold resin and exposed from the mold resin;
A second step of soldering the heat radiation surface of the second power module having the same configuration as the first power module and the surface of the second cooling unit having the same configuration as the first cooling unit with a second solder;
After the first step and the second step, the first cooling unit and the second cooling unit are fixed so that the first cooling unit and the second cooling unit form a gap that serves as a refrigerant passage. A method for manufacturing a power semiconductor device comprising three steps. Before the third step, performing X-ray inspection to determine the presence or absence of voids in the first solder;
2. The method for manufacturing a power semiconductor device according to claim 1, further comprising a step of performing an X-ray inspection to determine the presence or absence of voids in the second solder before the third step.

Images