JP2003218317A - Semiconductor power conversion device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To inexpensively provide a small and highly reliable semiconductor power conversion device having an excellent function. <P>SOLUTION: In the semiconductor power conversion device, two control substrates 109a and 109b are electrically connected by a flexible substrate 110, they are bent and are separately arranged above and below in a case 106. The two control substrates 109a and 109b are made into one continuous substrate at first. They are electrically connected by the flexible substrate 110. Necessary electronic parts are mounted on it and they are divided and separated. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a modularized semiconductor device, and more particularly to a modularized semiconductor power conversion device including a control circuit.

[0002]

2. Description of the Related Art In recent years, a semiconductor power conversion device having a built-in semiconductor switching element and modularized has been known as a MOSFE device.
With the high performance of power switching elements such as T and the conversion of home appliances into inverters, their applications have expanded, and there has been a strong demand for higher functionality and lower prices.

In particular, the demand for higher functionality is increasing year by year. Recently, not only power semiconductor modules having semiconductor switching elements built therein but also IPMs (Inteligent Power Modules) having control circuits built therein, and further, Higher functionality is being advanced to semiconductor power converters with built-in control circuits including even microcomputers.

In the case of such a semiconductor power conversion device, the power circuit portion on which the semiconductor switching element is mounted is usually made of a metal having a high thermal conductivity in consideration of a large amount of heat generated by the semiconductor switching element. In general, it is composed of a metal substrate, an insulating substrate made of a material having high thermal conductivity and high electric insulation, and an electric circuit having a pattern formed thereon.

At this time, ceramics, which are expensive but have high thermal conductivity, are mainly used as an insulating substrate in products having a large amount of heat and medium to large capacity. An inexpensive synthetic resin having a low conductivity is used.

On the other hand, since the control circuit portion mainly handles a small amount of electric power required for signal processing and control processing, a multi-layer glass epoxy substrate which is inexpensive and can easily form multi-layer wiring is often used.

With the recent demand for higher functionality, the area of the control circuit portion is relatively larger than that of the power circuit portion. On the other hand, the size of such control circuit portion is large. Contrary to the increase in the area, it is required to reduce the installation area, which is remarkable especially in applications where the installation place is limited such as for vehicles.

Here, FIG. 8 shows an example of a semiconductor power conversion device module according to the prior art. As shown in the figure, this module is a heat sink 101 which is a metal substrate.
Is made of aluminum, and the resin insulation layer 10
2 is used, and an electric circuit 103 having a conductor pattern is provided on the surface of the resin insulating layer 102.
03, a power semiconductor element 104 such as a MOSFET is solder-bonded to the upper surface of the power supply terminal 03 via a heat diffusion plate 802.

In the case of the module of FIG. 8, an electric circuit 1
The thickness of the conductor pattern forming No. 03 is generally about 70 μm, and the thickness of the heat diffusion plate 802 is about 0.7 to 1.2 mm in consideration of the effect of heat diffusion. Then, the electric circuit 103 is connected by the thin metal wire 105,
Further, wiring is provided to the external connection terminal 108 provided on the case 106.

On the other hand, the control board 109 is formed of a glass epoxy multi-layer board.
The control boards 109a and 109b are divided into two pieces for the purpose of reducing the installation area. First, the control board 109b is bonded to the resin insulating layer 102 with an adhesive or the like, and is connected to the semiconductor switching element 104. The control board 109b is electrically joined by the thin metal wires 105.

The control board 109b has an internal terminal 8 embedded and locked in the case 106 by insert formation.
One end of 01 is soldered and the other end is soldered to the control board 109a, so that they are electrically connected to each other.

Then, a soft gel resin 107 is injected into the case 106 to fill it, and then the case 106 is sealed with a relatively hard resin 112. Here, the reason why the soft resin is used is to avoid applying a stress due to thermal deformation of the case 106 to the semiconductor switching element, the thin metal wire, the terminal, etc. inside, and the soft resin 107 is mainly used. Silicon gel is used for.

On the other hand, an epoxy resin is mainly used as the hard resin 112. Here, the hard resin is used because the lid of the case 106 is formed by the hard resin 112. .

Here, the reason why the heat diffusion plate 802 is provided below the semiconductor switching element 104 is because the heat generation of the semiconductor switching element 104 is taken into consideration. Is difficult to hold, which may cause malfunction.

Further, the reason why the control board is divided into two is to reduce the installation area of the module. This is because the power circuit portion is improved as the function of the semiconductor power converter becomes higher as described above. This is because it corresponds to the fact that the area of the control circuit portion is relatively large.

Here, these control boards 109a, 10
9b is mainly composed of a multilayer glass epoxy substrate,
At this time, in order to reduce the installation area, a part or the whole of the control board where the amount of heat generated by the mounted components is small should be installed spatially apart from the heat dissipation plate 101, which is a metal board, on the power circuit part. Is effective.

Further, since electronic components such as microcomputers generally have a lower allowable operating temperature than semiconductor switching elements, spatially separating from a metal substrate whose temperature rises due to heat generation of the semiconductor switching elements causes malfunction. Prevent
It is convenient for high reliability.

Here, in the module shown in FIG.
As described above, since the internal terminal 801 is mechanically locked to the case 106, stress is generated in the soldered portion of the control board due to vibration during use, thermal deformation of the resin due to thermal agitation, etc. Need to be relaxed.

Therefore, in this module, the internal terminal 8
Bend (bending part) is formed on 01, and by this, upper and lower 2
No stress is applied to the connecting portions (soldering portions) of the control boards 109a and 109b divided into the sheets.

If the stress applied to the soldering portion is large, the soldering portion may be broken, which may cause malfunction of the switching element, which may cause damage in some cases.
When the control board is connected by the internal terminals, the bend is essential in order to reduce the distortion generated in the soldered portion.

By the way, according to the internal terminal provided with the bend as described above, a high reliability of the solder joint can be obtained, but on the other hand, there is a problem that the complicated bend shape is required and the cost becomes high.

Further, even with such an internal terminal, the control board can be automatically assembled using a robot, but
Since the number of times of solder joining is increased as compared with the case where the control board is not divided into two, that is, the case where the installation area is large, there is a problem in that it is also expensive in this respect.

Therefore, a module is known in which a connector is used to connect two control boards, and an example thereof will be described with reference to FIG. 9. In this module, first, one control board 109a is provided with an electronic signal. The male connector 901a is soldered at the same time as the parts, and the other control board 1
Similarly, the female connector 901 and the electronic component
Solder b.

Then, the control board 109b is connected to the resin insulation layer 1
02, the male connector 901a is replaced by the female connector 90.
By positioning the control board 109a on the control board 109b so that the control boards 109a and 1b are inserted, electrical connection between the control boards can be obtained.

In the case of the module shown in FIG. 9, the connector can be soldered at the same time as the electronic parts, so that the number of times of soldering can be reduced and the cost can be reduced as compared with the module of FIG. In some cases, the function of relieving the stress of the solder portion caused by vibration or thermal deformation cannot be obtained, so there is a problem in terms of reliability.

[0026]

As described above, the semiconductor power converter comprises a power circuit section including a semiconductor switching element that generates a large amount of heat and a control circuit section that handles small signals such as control and signals. Therefore, in order to reduce the installation area, in the above conventional technique, a part or all of the control circuit unit is spatially separated from the metal substrate.

Therefore, in order to obtain a highly reliable semiconductor power conversion device, it is necessary to dissipate heat from the semiconductor switching element and to relieve stress in the soldering portion. It is necessary to simplify, especially reduce the number of times of soldering as much as possible, but these requirements cannot be satisfied by the above-mentioned conventional technology.

By the way, as a method of reducing the installation area of the module and relaxing the stress of the soldered portion of the control board, there is a method of using a flat cable as shown in FIG. 10, which requires particularly reliability. As shown in the figure, the control board 109a, 109b made of a double-sided multi-layer glass epoxy board is connected by a flat cable 1002.

Here, in FIG. 10, the control board 109 is used.
The electronic components mounted on a and 109b are omitted, and the whole is fixedly housed in a metal case 1001 and modularized, and an electrical connection with the outside is made of a glass epoxy substrate. The connector 111 soldered on the control board 109b is used, and the connecting portion of this is connected to the case 10
It is realized by putting it out of 01.

Flat cable 1 used here
002 can be flexibly bent and has a lower rigidity than the internal terminal 801 of FIG. 8, so that the distortion generated in the soldering portion is small even when the vehicle is vibrated, the engine is rotated, or the like. Further, the flat cable can be assembled at low cost because it can be mounted in the through hole portion of the glass epoxy substrate similarly to the dip component.

However, in this case, since the glass epoxy substrate has a large thermal resistance and cannot radiate the heat generated by the semiconductor switching element, the power circuit section cannot be provided. Further, at this time, as shown by the symbol P, the terminal portion of the flat cable 1002 protrudes from the through hole of the glass epoxy substrate to the opposite side, so that it cannot be installed in close contact with the case 1001.

Therefore, in the module of FIG. 10,
Since the glass epoxy substrate is thermally isolated from the case, it is not suitable for mounting a semiconductor switching element that requires heat dissipation, and cannot be applied to a semiconductor power conversion device.

Next, as a method of similarly reducing the installation area of the module and relieving the stress of the soldered portion of the control board, there is a method of using a flexible board as shown in FIG. Here, this flexible substrate has been conventionally used for consumer use such as a video camera or a thin display, especially when miniaturization is required.

FIG. 11 shows a thin display 110.
1 and the circuit board 109 made of glass epoxy are electrically connected to each other by soldering via the flexible board 110. In this case, electronic parts such as the driver IC 1102 are provided on the flexible board 110. It is supposed to be installed.

Since the flexible substrate 110 can be easily deformed in shape, it has a high degree of freedom in mounting and, like a flat cable, has low rigidity.
There is little concern about disconnection due to thermal shock. Moreover, unlike flat cables, parts can be mounted, which is effective for downsizing.

However, since the flexible substrate has a high thermal resistance like the glass epoxy substrate, it cannot be provided with a power circuit including a semiconductor switching element and cannot be applied to a semiconductor power converter.

An object of the present invention is to provide a highly functional semiconductor power conversion device which is small in size and has high reliability at low cost.

[0038]

The above object is to provide a semiconductor power converter having a plurality of control boards in the same case as a semiconductor switching element, wherein the plurality of control boards are initially one board. The flexible substrate is electrically connected between the respective portions to be the control boards, and the portions to be the control boards are separated.

At this time, at least a part of the flexible substrate has a longitudinal elastic modulus of 1 × 10 ^{−3 to 2} MPa.
The above-mentioned object can be achieved even if it is sealed with resin and at least one of the plurality of control boards is spatially separated from the heat dissipation plate of the semiconductor switching element.

Here, the plurality of control boards are
The substrate may be separated by breaking it, and at this time, the one substrate is provided with a groove, a slit, or a plurality of small holes arranged in a stitch shape at a predetermined position. May be given by dividing by any of these parts.

[0041]

BEST MODE FOR CARRYING OUT THE INVENTION A semiconductor power converter according to the present invention will be described in detail below with reference to the embodiments shown in the drawings. However, the present invention is not limited to the embodiments described below.

First, FIG. 1 shows an embodiment in which the present invention is applied to a semiconductor power conversion device including a three-phase power circuit composed of MOSFETs and a control circuit for controlling the power circuit. 1A shows a planar structure, and then FIG. 1B shows a cross section taken along the line AA of FIG. 1A.

In these figures, the semiconductor power converter according to this embodiment is provided with one surface of the heat sink 101 (see FIG. 1).
A resin insulating layer 102 serving as an insulating substrate is provided on the upper surface in (b), an electric circuit 103 is bonded onto the resin insulating layer 102, and the semiconductor switching element 104 is bonded to the predetermined position by soldering. There is.

Here, also in this embodiment, the control board is divided into the upper control board 109a and the lower control board 109b, both of which are made of a multi-layer glass epoxy board, respectively, such as a microcomputer. Electronic components and signal terminals are mounted (in the figure, electronic components mounted on the control board 109 are omitted).

Next, the semiconductor switching element 104, the electric circuit 103, the control board 109b, and the external connection terminal 1 insert-formed in the case 106 made of synthetic resin.
The thin wires 105 are connected to each other through the wire 08. The thin wires 105 are made of aluminum alloy wire having a diameter of about 300 to 500 μmφ.

In this embodiment, particularly, FIG. 1 (b)
As is clearly shown in FIG. 1, the case 106 is bonded to the heat sink 101 via the resin insulating layer 102.
One control board 109b is mechanically joined to the heat dissipation plate 101 via a resin insulating layer 102, and the other control board 109a is mounted on a pedestal 114 formed on some inner wall surfaces of the case 106. And the sealing resin 112 described later.
Fixed by.

Here, reference numeral 110 is a flexible substrate, and this flexible substrate 110 has a structure in which a circuit pattern made of a metal thin plate (foil) is sandwiched between thin plates of synthetic resin such as polyimide, and a plurality of metal thin plates and resin are used. A multilayer circuit board can also be formed by stacking thin plates.

A circuit pattern for connection is formed on the flexible substrate 110. One end of this circuit pattern is soldered to the circuit pattern of the control board 109b, and the other end is soldered to the control board. By soldering to the circuit pattern 109a, two control boards 10
The necessary electrical connection is provided between 9a and 109b.

At this time, a connector 111 for connecting to the outside is mounted on the control board 109a, and the connector 111 and the flexible board 1 are mounted on the control board 109a.
By mounting 09 separately on one surface and the other surface, electrical connection to both the outside and the power circuit unit can be obtained.

Here, these control boards 109a, 109
The details of b will be described with reference to FIG. 2. First, as shown in the drawing, these control boards 109a and 109b are initially formed as left and right parts in the figure of one board 109, and their boundary lines are formed. A groove G having a V-shaped cross section is formed on both sides.

Next, in this state, the flexible boards 109 are arranged on the respective control boards 109 so as to straddle the groove G.
After soldering to the portions to be a and 109b, one control board, for example, the portion to be the control board 109b is bent as shown by an arrow X in the figure, and the other control board 10 is formed in the groove G.
A force is applied from the portion to be 9a until it is broken, and the respective portions are separated while being connected by the flexible substrate 109.

Then, in this way, the flexible substrate 1
Control board 109 separated into two in a state of being connected by 10
After joining the joint surface J of the control substrate 109b, which is one of a and 109b, onto the resin insulating layer 102 with an adhesive or the like,
The other control board 109a is bent at the portion of the flexible board 110 so as to overlap with the control board 109b, and as shown in FIG. 1, it is turned upside down on the power circuit section and placed on the pedestal 114 and fixed. It is.

Here, in this way, the control board 109a
As described above, the gel-like resin 107 is injected into the case 106 to mount the semiconductor switching element 104, the thin metal wire 105, and the electric circuit 103 before the case is mounted on the case 106.

As already described in the means for solving the problem, in order to relieve the stress applied to the solder portion due to vibration and thermal deformation in the flexible board having low rigidity, the rigidity of the flexible board is low. It is necessary to use a resin so that the flexible substrate can move freely to some extent.

Therefore, at the time of sealing and use, the thin metal wire 1 is used.
05, the semiconductor switching element 104, and the flexible substrate 110 are prevented from being adversely affected, a relatively soft material such as silicon gel is used for the sealing resin 107. , Generally in the range of 1 × 10 ^{−3 to 2} Mpa.

However, as long as the elastic modulus is small and the thermal stress buffering effect of the flexible substrate 110 can be sufficiently exerted, the sealing resin 107 is not limited to silicon gel.

Then, the sealing resin 112 is further applied to the case 10.
The semiconductor power conversion device is completed by filling the inside of the substrate 6 and hardening it, fixing the control substrate 109a, and forming the lid of the case 106. Here, the points that are not particularly described are the same as in the case of the conventional technique described in FIG.

Next, details of the embodiment of FIG. 1 will be described more specifically. First, the heat sink 101 is made of aluminum or an aluminum alloy when weight reduction is important. This is because the heat sink has a relatively large volume in the semiconductor power conversion device, and thus aluminum, which has a smaller specific gravity and is lighter than copper, is suitable. However, when considering heat dissipation,
Copper or a copper alloy having a higher thermal conductivity is used.

The heat dissipation plate 101 is designed to sufficiently reduce the thermal resistance due to the spread of heat inside.
It is desirable that the thickness is 1.5 to 5.0 mm. Further, the heat dissipation plate 101 is provided with a mounting hole 113, which allows the semiconductor power conversion device to be easily mounted on a cooling fin (not shown).

Next, the resin insulation layer 102 is required to have low thermal resistance and high insulation, and therefore, an epoxy resin in which a filler is dispersed is used. Here, as the filler, one made of an inorganic compound having a high thermal conductivity such as silicon oxide or aluminum oxide is used.

At this time, as the content of the filler increases,
Although the thermal resistance of the resin insulating layer 102 can be reduced, the amount of filler that can be dispersed in the epoxy resin is limited, so the filler content is usually in the range of 75 to 95%. In this case, the thermal conductivity of the resin insulating layer 102 is in the range of 2-5 W / mK.

On the other hand, another effective method for reducing the thermal resistance of the resin insulating layer 102 is to make it thinner.
However, if the resin insulation layer 102 is made thin, the dielectric strength is reduced accordingly, and pinholes and the like are likely to occur in the resin insulation layer 102, which may lead to a decrease in reliability. There is a limit to the lower limit of the thickness of the resin insulating layer 102, and depending on the required withstand voltage, it is 50
The lower limit is about 250 μm.

Further, since the resin insulating layer 102 insulates the electric circuit 103 from the heat sink 101, a creeping distance corresponding to the withstand voltage is required around the electric circuit 103, which is sufficient. It is necessary to make up for the lack by covering the surface of the heat sink 101 with an insulating layer. Therefore, in this embodiment, the resin insulating layer 102 is provided on the entire surface including the surface of the heat sink 101 on the electric circuit 103 side.

Next, the electric circuit 103 is made of copper or aluminum and adhered to the surface of the resin insulating layer 102. Here, the plate thickness is considerably thicker than 0.7 mm. Then, as a result of making the electric circuit 103 thick (0.7 mm or more) in this way, in this embodiment, sufficient heat spread can be obtained here. Therefore, as in the prior art described in FIG. , Heat diffusion plate 8
It is not necessary to provide 02.

Here, this electric circuit 103 is shown in FIG.
, It has a planar shape corresponding to the circuit pattern, but at this time, since copper or aluminum is excellent in workability, even if the plate thickness is 0.7 mm or more, press working Can be produced in any shape according to the circuit pattern.

By the way, in the above embodiment, for example, nickel, silver, platinum, tin,
At least one metal having good solder wettability selected from the group consisting of antimony, lead, copper, zinc, palladium, and gold, or nickel, silver, platinum, tin, antimony, lead, copper,
At least 2 selected from the group of zinc, palladium and gold
You may make it coat | cover the alloy with good solder wettability containing a kind of metal.

Next, a MOSFET is used as the semiconductor switching element 104, and this is soldered to the electric circuit 103. The MOSFET is connected to the electric circuit 103 and the control board 109b by a thin metal wire 105.

Next, the flexible substrate 110 is soldered and electrically connected to the control substrates 109a and 109b in advance. Then, these control boards 109a, 10
9b is made of glass epoxy.

Flexible substrate 1 used here
Since 10 has a lower rigidity than the internal terminal 801 used in the conventional technique of FIG. 8, for example, the stress caused by thermal deformation can be sufficiently relaxed, and the strain generated in the soldered portion can be reduced.

At this time, the flexible board does not have the function of supporting the control board like the internal terminals in the prior art, but in this embodiment, the control board 109a is positioned and supported by the pedestal 114 on the case 106, and Since it is fixed to the case 106 by the sealing resin 112 above, there is no risk of the control board moving due to vibration, and distortion occurring in the soldered portion can be reliably suppressed.

On the other hand, the flexible substrate is, for example, as shown in FIG.
The rigidity is higher than that of the conventional flat cable shown in FIG. Therefore, unlike the flat cable, automatic surface mounting is possible, and as described later, an inexpensive manufacturing process with a small number of soldering times can be applied.

Next, the control board 109a is a multi-layer board made of glass epoxy, on which electronic parts are mounted on both sides.
On the other hand, electronic parts are mounted only on the upper surface of the control board 109b. As described above, in FIG. 1, electronic components on the control board 109 are omitted.

Therefore, in this embodiment, the control board 10
When electronic components are mounted on 9a and 109b, the flexible board 110 is also mounted at the same time as described with reference to FIG. 2 so that the internal terminals 801 and the control boards 109a and 109b required in the conventional technique shown in FIG. The soldering process can be eliminated.

Further, in this embodiment, as shown in FIG. 1B, since most of the flexible substrate 110 is sealed with soft silicon gel, the case 106 is soldered due to thermal deformation or vibration. The stress applied to the joint can be relaxed.

This semiconductor power conversion device is completed by sealing with a sealing resin 112. This sealing resin 112 forms the lid of the case 106, and the control board 109a is placed on the case 106. A relatively hard epoxy resin is used because it has a fixing function.

According to this embodiment, since the flexible board is used for the electrical connection between the two control boards, the installation area is small and the reliability is high in spite of having a high function. It is possible to provide the semiconductor power conversion device at low cost.

Further, according to this embodiment, a plurality of control boards, that is, two control boards are initially made as one control board, and after the necessary electronic components are mounted, they are required to be divided. It is possible to obtain as many as two control boards.

Therefore, according to this embodiment, initially 1
Since it can be formed as a single board, it can be easily surface-mounted by automatic mounting in spite of having a plurality of control boards, and as a result, it can be manufactured by an inexpensive process with a small number of times of soldering. be able to.

Next, another embodiment of the present invention will be described. First, a semiconductor power conversion device according to a second embodiment of the present invention will be described with reference to FIG.

First, as shown in FIG. 3A, the control board 1
09 is prepared. Here, this control board 109 is shown in FIG.
As described above, the electronic component forming the control circuit is mounted on the control substrate 109 made of the multi-layer glass epoxy substrate. At this time, the flexible substrate 110 and the connector 11 are mounted on different surfaces. .

The glass epoxy board is provided with a V-shaped groove G, which is a notch in advance, and can be easily divided into a plurality of parts only by bending after mounting electronic parts. Therefore, the flexible board 110 and the connector are individually provided. After soldering the electronic component including 111, the number of times of soldering can be reduced by breaking the solder.

Therefore, the control substrate 109a and the control substrate 109b have a constraint that they must first have the same thickness and the same number of conductive layers. The melting points of the solders used for mounting the electronic components must be the same on the lower side of 109a (FIG. 3) and the upper side of the control board 109b, and the compositions must be the same. Furthermore, on the side on which the connector 111 is mounted, However, there is a constraint that electronic parts should not be mounted on the portion beyond the groove G, that is, on the joint surface J.

On the other hand, the flexible substrate 110 needs to have a certain degree of rigidity in order to enable automatic surface mounting, that is, to prevent deformation during vacuum suction by an automatic mounter. As a result, there are restrictions on the length.

Here, the length of the flexible board 110 means the distance from one connection point to the other connection point, and if this length becomes long, there is also a portion where the flexible boards come into contact with each control board surface. Because it will be long
This is because the area where electronic parts cannot be mounted becomes large.

Therefore, in the assembling process of this embodiment, since a flexible substrate having a smaller size and a smaller degree of bending flexibility is used as compared with the flat cable, it is necessary to devise a method for minimizing the bending of the flexible substrate. is there.

First, since the length of the flexible substrate 110 is limited, the control substrate 109a and the control substrate 109 are
The length of b is limited to about 60 to 90%.
That is, if an attempt is made to increase the distance between the control boards, as shown in FIG. 1B, when the flexible board 110 is bent, the surfaces of the control boards 109a and 109b come into contact with each other so that electronic components cannot be mounted. The storage area becomes large. Therefore, the length of the flexible substrate 110,
As a result, it is important to reduce the distance between the control board 109a and the control board 109b.

On the other hand, as a result of the rigidity and size restrictions of the flexible substrate 110, there is no fear that the flexible substrate 110 will be twisted even after the semiconductor power conversion device is completed. That is, each control board 10
After being split into 9a and the control board 109b, there is no fear that the split ends will be out of parallel.

Returning to FIG. 3, in this embodiment, a part of the outer wall of the case 106 is formed separately so that the bending of the flexible substrate 110 can be reduced.

When the control board 109 as shown in FIG. 3 (a) is prepared, then as shown in FIG. 3 (b), the control board 109 is formed with a case 106 in which external terminals 108 are insert-formed. At the same time, it is adhered to the heat dissipation plate 101. Here, it is a matter of course that the semiconductor switching element 104 and the like are soldered to the heat dissipation plate 101.

Then, at this time, the joint surface J of the portion of the control substrate 109 to be the control substrate 109b is not covered with the resin insulating layer 10.
The case 106 at this time is partially formed as a wall portion 302 (described later) and is open, as described above.
As shown in (c), the control board 109 can be adhered as it is without being broken, and the external terminal 108 formed by the fine metal wire 105 can be used.
The semiconductor switching element 104, the electric circuit 103, and the control board 109b can be connected to each other.

Thereafter, the control board 109 is divided by the groove G, and the flexible board 110 is folded back as shown by an arrow X,
The control board 109a is placed on the pedestal 114. And
After the wall portion 302 is adhered, it is sealed with a silicon gel resin 107, and further sealed with a sealing resin 112, and the control board 1
09a is fixed to the case 106, and the case 10
If the lid 6 is covered, the semiconductor power conversion device shown in FIG. 3D can be obtained.

Next, FIG. 4 shows another embodiment of the present invention in which the division of the case 106 is changed. In this case,
First, as shown in FIG. 4A, the control board 109 is divided by the groove G so as to be divided into a control board 109a and a control board 109b, and then only one control board 109b is placed on the bottom step portion 106b of the case 106. To bond.

In this embodiment, the height of the wall of the case 106 on the side of the control board is sufficient to fill the silicon gel resin 107, and at this time, the step portion 106b corresponds to the length of the flexible board 110. The possible height, that is, the height between the control boards 109a and 109b is a predetermined value.

Next, as shown in FIG. 4B, the thin metal wire 1
Connect with 05. Then, after this, as shown in FIG. 4C, the flexible substrate 110 is folded back, and the control substrate 109a is positioned on the pedestal 114 of the case 106 by engaging with the pawl 115.

Then, as shown in FIG. 4 (d),
After sealing with the silicon gel resin 107, the lid 401 including a part of the wall on the control substrate side is bonded from above the case 106 to complete the semiconductor power conversion device. At this time, although not clearly shown, the lid 401 has a hole through which the connector 111 is inserted. In the above embodiment, the control board is divided into two. The present invention may be carried out by dividing it into a larger number, and FIG. 5 shows an embodiment of the present invention when the control board is divided into three.

In this embodiment, first, as shown in FIG. 5 (a), two grooves for bending and separating, that is, a groove G and a groove Ga, are provided on the control board 109, whereby the control board 109a and the control board 109 are provided. The control board 109c and that portion are further formed between the portions to be 109b. At this time, unlike the embodiment described with reference to FIGS. 1 to 4, as illustrated, the flexible substrate 110 and the connector 111 are connected.
Are mounted on the same surface of the control board 109.

Further, in this embodiment, as shown in FIG. 5B, a case 106c having only a bottom portion and no side wall portion is used. First, the control board 109 is controlled on a part of the case 106c. The portion to be the substrate 109b is adhered, and then the metal thin wire 105 is connected.

Next, as also shown in FIG. 5 (b), the control board 109 is divided at two places, the groove G and the groove Ga, and the portion to be the control board 109c is overlapped on the control board 109b. After being bent at the portion of (4), the control board 109a is folded back to the opposite side so that the control board 109a overlaps the control board 109c at the groove Ga.

At this time, although not shown, the case 106
A pedestal and a boss pin are set up in the peripheral portion of c, and correspondingly, a hole is formed at a predetermined position on the control board 109, and by combining these, a control as shown in FIG. The positioning of the substrates 109a, 109b, 109c is obtained.

Then, after that, cover the lid 401a,
After the case 106c is adhered, the silicon gel resin 107 is internally injected from the silicon gel injection hole 502 provided in the lid 401a, the resin is cured to a predetermined viscosity, and the cap 501 is adhered to complete the semiconductor power conversion device.

According to the embodiment of FIG. 5, in a module having a small installation area, the area of the control board can be further ensured, and it is possible to sufficiently cope with the further high function of the semiconductor power conversion device. it can.

Next, FIG. 6 is also an embodiment of the present invention, which corresponds to the modification of FIG. 5, and as shown in FIG. 6 (a), a portion serving as a control board 109a on which the connector 111 is mounted is It is arranged in the center and control boards 109b and 109c are provided on both sides thereof. At this time, regarding the connector 111,
The flexible substrate 111 mounted on the surface opposite to the flexible substrates 110 and 110a and connecting the control substrate 109a and the control substrate 109b has a certain length.
It is a long one.

As a result, all the control boards have the same bending direction, and the control boards 109a and 109b are arranged as if the control board 109c is wrapped between them. The rest is the same as that of the embodiment of FIG. 5, and a high-performance semiconductor power conversion device can be obtained with a small installation area.

Next, the division of the control board, which is one of the features of the present invention, will be described in more detail. First, in each of the above embodiments, the board is divided by the groove G and divided into a plurality of boards. ing.

Therefore, in this case, as shown in FIG. 7A, a groove G (Ga) having a V-shaped cross section is formed on both sides of the glass epoxy substrate 109, and the groove G (Ga) is bent to show the structure. , For example, two control boards 109
Therefore, in this case, after cutting, in this case, a corner drop as a trace of the V groove remains at the end of each substrate.

Next, FIG. 7 (b) shows a plurality of slits (rectangular holes) S with the connecting portion R left instead of the V-grooves to form a dividing line, and FIG. 7 (c) shows a plurality of slits. The small holes H of are formed in a seam shape,
This is used as a dividing line, and in any case, when the board is bent, stress is concentrated on the remaining connecting portion, and at this portion, the stress is broken, so that it is separated into two sheets as shown in the figure. Will be. Then, after breaking, the protrusion P remains as a trace.

Therefore, in any case, traces remain after the chips are divided, but since the area is small when viewed from the entire control board, there is no particular problem. Further, no component can be mounted on the cut portion, but this is also a small area when viewed from the entire control board, so this is not a particular problem.

According to the above embodiment, although it has a plurality of control boards, it can be initially formed as one board, and as a result, it can be easily surface-mounted by automatic mounting. It can be easily manufactured by an inexpensive process with a small number of times of soldering.

According to the above embodiments, the module installation area is suppressed to be small, the reliability of the connection portion is not deteriorated due to the stress, and the heat dissipation of the semiconductor switching element is sufficient, although the function is high. Thus, it is possible to provide a semiconductor power conversion device that can sufficiently reduce the cost by reducing the assembly process.

[0110]

According to the present invention, since a plurality of control boards are obtained from one board, there is no fear that the reliability of the connection portion will be deteriorated by stress, and the heat dissipation of the semiconductor switching element will be sufficient. While having high reliability by being obtained,
A sufficient cost reduction can be obtained by reducing the assembly process, and as a result, a highly functional semiconductor power conversion device having a small installation area can be easily provided at low cost.

[Brief description of drawings]

FIG. 1 is a plan view and a cross-sectional view showing a first embodiment of a semiconductor power conversion device according to the present invention.

FIG. 2 is an explanatory diagram showing an example of a control board according to the embodiment of the present invention.

FIG. 3 is a cross-sectional view showing an assembly process of the first embodiment of the semiconductor power conversion device according to the present invention.

FIG. 4 is a cross-sectional view showing an assembly process of the second embodiment of the semiconductor power conversion device according to the present invention.

FIG. 5 is a cross-sectional view showing the assembly process of the third embodiment of the semiconductor power conversion device according to the present invention.

FIG. 6 is a cross-sectional view showing the assembly process of the fourth embodiment of the semiconductor power conversion device according to the present invention.

FIG. 7 is an explanatory diagram of a method of dividing the control board according to the embodiment of the present invention.

FIG. 8 is a cross-sectional view for explaining a first example of the related art.

FIG. 9 is a sectional view for explaining a second example of the conventional technique.

FIG. 10 is a sectional view for explaining a third example of the conventional technique.

FIG. 11 is a sectional view of a reservoir for explaining a fourth example of the related art.

[Explanation of symbols]

101 heat sink 102 resin insulation layer 103 electric circuit 104 Semiconductor switching element (MOSFET) 105 thin metal wire 106 cases 107 Resin for sealing 108 External connection terminal 109 control board 110 Flexible substrate 111 Connector for external connection 112 Resin for sealing 113 mounting holes 114 pedestal 115th 301 Notch 302 wall 401 lid 501 cap 502 Gel injection hole 801 internal terminal 802 Heat diffusion plate 901 connector 1001 metal case 1002 flat cable 1101 thin display 1102 Driver IC

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Akira Bando             3-1-1 Sachimachi, Hitachi City, Ibaraki Prefecture Stock Association             Hitachi, Ltd. Inverter Promotion Division F term (reference) 5H740 BA12 MM08 PP01 PP02 PP03                       PP06

Claims (4)

[Claims] 1. A semiconductor power conversion device comprising a plurality of control boards in the same case as a semiconductor switching element, wherein the plurality of control boards are initially made from one board, and each control board is provided. A semiconductor power conversion device, characterized in that, after electrical connection is made by a flexible substrate between the portions to be the control board, the portions to be the control board are separated. 2. The invention according to claim 1, wherein at least a part of the flexible substrate is sealed with a resin having a longitudinal elastic coefficient of 1 × 10 ^{−3 to 2} MPa, and the flexible substrate is one of the plurality of control substrates. At least one sheet is spatially separated from the heat radiating plate of the semiconductor switching element. 3. The semiconductor power conversion device according to claim 1, wherein the plurality of control boards are separated by breaking the one board. 4. The invention according to claim 3, wherein the one substrate is provided with a groove, a slit, or a plurality of small holes arranged in a stitch shape at a predetermined position, and the separation includes A semiconductor power conversion device characterized by being given by being divided by any part.