The present application claims priority from Japanese application JP2005-283260 filed on Sep. 29, 2005, the content of which is hereby incorporated by reference into this application.
1. Field of the Invention
The present invention relates to the lead connection with the substrate and the semiconductor element mounted thereon in semiconductor modules such as power modules mounted on electric apparatuses for automobile.
2. Description of the Related Art
Lately electronic devices have been introduced for controlling various electric apparatuses in automobiles and other vehicles. An example of electric apparatus incorporating an electronic device is an electric power steering unit shown in
The power module is constituted as a so-called semiconductor module with a power element such as IGBT (Insulated Gate Bipolar Transistor) suitable for controlling electric apparatuses driven by a relatively large current such as an electric power steering device mounted thereon. This type of power module is also called “an in-vehicle module” as it is mounted on vehicles.
The control unit of the above-mentioned electric power steering unit and other power modules is required to discharge efficiently the heat generated thereby from the power module to control large current with a power element (semiconductor element) mounted thereon. In order to improve the dissipation of heat from the power module, semiconductor element represented by MOS (Metal-Oxide Semiconductor) are connected by die bonding with the power module in the form of bare chips (not sealed by resin and the like) on the substrate,
As the form of electrical connection with semiconductor element mounted on the substrate of power modules, the Al wire bonding connection method used in other semiconductor modules has been used in the past.
As the bonding wire used for connecting chips by A1 wire bonding shown in
In order to realize such a lead connection of such semiconductor modules, the following technology is now under consideration. For example, Patent Document 1 teaches the method of connecting in layers a lead, semiconductor element, and a heat sink on the substrate by using a plurality of solders with different melting points, and that of connecting them in one stroke by using a Sn solder paste (containing mainly tin). This method simplifies the assembly process that will be necessarily complex by the wire bonding connection method. And Patent Document 2 examines how to improve the reliability of soldering by using a heat sink, an insulating substrate, silicon chips and a material with a linear expansion coefficient difference of 10×10−6 or less. And Non-Patent Document 1 examines how to reduce thermal stress applied by the lead to the chips and to prevent the breakdown of chip electrodes by making the ends of the lead comb-teeth shaped.
[Patent Document 2]
[Non-patent Document 1]
The patent documents each referred above or later are listed as follows.
For realizing electrical connections with this MOS chip 3, in order to prevent any drop in the electric characteristics of the whole module, the wiring must be connected with all the source pads (source electrodes) 8 of the MOS chip 3 formed in a split shape on the front surface. According to the A1 wire bonding method, the connection of A1 wire 4 with each of these source pads 8 leads to a connection that prevents any drop in electric characteristics. For realizing a lead connection, the lead 6 must have a width sufficient to reach all the source pads 8 and must be soldered.
For supplying solder between the chip 3 and the lead 6, the method of inserting a soldering sheet between them is suitable in view of its applicability. At this time, due to the variation of position of mounting the soldering sheet and the lead on the front surface of the MOS chip (a plurality of source pads 8), and the displacement of the soldering sheets and lead from the source pads (source electrodes) 8 resulting from the flux accompanying the reflow of the soldering sheets, a part of the source pad 8 (for example, its extremity) may not be wetted with the solder even after the solder has melted. In other words, in the front surface of the MOS chip 3, the wetting condition with the solder of the source pad 8 becomes uneven. The solder deforms in such a way that its surface may be smaller due to its surface tension.
As described above, when the solder wetting coverage in the source pads 8 and the amount of solder wetting the source pad 8 are unevenly distributed on the front surface of the MOS chip 3, the lead 6 due to be connected with the source pad 8 by the solder moves on the front surface of the MOS chip in response to the solder wetting coverage and the distribution of the solder. As a result, the position of the lead 6 in relation to the source pad 8 deviates from a position suitable for electrical connection between the lead 6 and the MOS chip 3.
An example of displacement actually occurred of the lead 6 in relation to the pad 8 due to a variation in the wetting condition of solders (appearance of a non-wetting region) on the front surface of the pad (electrode) 8 connected with the lead 6 is shown in
Such a problem perceived by the inventors of the present invention has not been examined in the technical field related with the mounting of semiconductor modules. Patent Documents 1 and 2 and Non-patent Document 1 respectively propose a connecting method by using a lead for the purpose of simplifying the respective process, a lead material and a form of lead having a high reliability of connection. However, references to these documents are not sufficient to find a technology capable of solving the unevenness of wetting condition of solder on the front surface of pads (electrodes) and the displacement of the lead vis-a-vis the pad discovered by the inventors.
In view of the problem described above, the basic structure of the lead that characterizes semiconductor modules conceived by the inventors is shown in
In the main surface of the substrate 1 extending on the x-y plane, the electrodes 8 of the semiconductor element 3 to which a part of the lead 6 (for example an end of the lead 6) is connected are juxtaposed and extend along the x-y plane. And the other electrodes 1d to which the other parts of the lead 6 (parts other than the part mentioned above, for example, the other end of the lead 6) are connected extend along the x-y plane. The lead 6 is described as a plate or a film formed by a electrically conductive material having a main surface that extends along the x-y plane with one part mentioned above connected with the electrode 8 of the semiconductor element 3 and the other part mentioned above connected with the other electrodes id formed in the main surface of the substrate 1, and a thickness in the z axis direction (direction that crosses with the main surface). One side of the main surface of the lead 6 having such a form (in
As shown in
The width a of the lead 6 needs not meet the above-mentioned relationship (a≧b) with regards to the above-mentioned width b, and it is enough that the relationship is fulfilled at least in a part of the Region A. In other words, a part of the electrodes 8 joined with the lead 6 (Region A) may not be covered by the lead 6 and be exposed. When the main surface of the Region A of the lead 6 is extended from one end nearest to its Region B to the other end farthest from its Region B, the width a of the lead 6 is defined as the dimension of main surface of the Region A along the direction crossing the extending direction (here the X axis direction), and its value needs only be larger than or equal to the width b of the electrodes 8 mentioned above at least on one side from either one end side or the other end side of the main surface of the Region A. And when at least one of a plurality of conductive layers constituting an electrode 8 arranged in one direction (X axis direction) extends in a direction (Y axis direction) crossing with the juxtaposition direction (X axis direction), the Region A of the lead 6 is preferably formed so as to fulfill the relationship (a≧b) mentioned above at least on one side of both ends along the extending direction. When the Region A of the lead 6 is formed so as to fulfill the relationship (a≧b) mentioned above on both ends, a part of the electrode 8 mentioned above may be exposed, and for example a U-shaped notch may be formed between the both ends. The width a of the Region A of the lead 6 will be narrower than the width b mentioned above by the part where the notch has been formed.
When an electrode 8 is constituted by a plurality of conductive layers juxtaposed in one direction (X axis direction), the width b of the electrode 8 is defined along the juxtaposition direction of a plurality of conductive layers (the above-mentioned X axis direction). The width b of the electrode is defined as the total sum of, for example, the width of each of a plurality of conductive layers along the juxtaposition direction and the gaps separating a plurality of pairs of neighboring conductive layers along the juxtaposition direction. And the width b of the electrode is also defined as the distance between a pair of sides located on both ends defined along the juxtaposition side of the plurality of conductive layers. The “side” in each of a pair of conductive layers forming both ends of the electrode 8 mentioned above is defined as a side that does not face the other of the plurality of conductive layers and that extends in the direction crossing the juxtaposition direction mentioned above. When one end and the other end of the main surface of a semiconductor element (on which the conductive layer is formed) separated in the juxtaposition direction of a plurality of conductive layers are taken as the reference, the width b of the electrode is defined as the distance separating in the juxtaposition direction one side located the on one end side of the conductive layer nearest to the one end of the main surface from the other side located on the other end side of the other conductive layer nearest to the other end of the main surface.
On the other hand, when the lead B above is formed to extend from one end of the Region A side to the other end of the Region B side above along the extending direction, at least a notch extending to the main surface from the other end of the lead along the extending direction may be formed in the Region B of the lead to split the lead into a plurality of parts. The plurality of parts formed in the Region B of the lead are lined up in the direction crossing their extending direction in the main surface of the lead. Corresponding to the Region B of the lead thus split up, the electrode 1d to be connected therewith may be split up, and each of them may be soldered to one of a plurality of parts constituting the Region B of the above-mentioned lead corresponding thereto. At this time, the electrode 1d may be split up into a n number of conductive films corresponding to a n number of parts (n is an integer of three or more) formed by splitting the Region B of the lead.
One of the representative structures of semiconductor modules according to the present invention includes a substrate (1a) having the main surface wherein a wiring pattern (1c) including a first conductive film (1d) and a second conductive film (1d) is formed, a semiconductor element (3) electrically connected with the first conductive film (1d) and mounted on the main surface of the substrate (1a), and a lead (6) electrically connecting the second conductive film (1d) and the semiconductor element 3), and is characterized in that the lead (6) is disposed with its main surface facing the main surface of the substrate (6), the main surface of the lead (6) includes a first region (Region A) facing the main surface of the opposite side of “the main surface of the substrate (1a)” and a second region (Region B) facing the second conductive film of the semiconductor element (3), an electrode (8) is formed on the main surface of the semiconductor element (3) to be connected with solder (2) with the first region (Region A) of the main surface of the lead (6), the first region (Region A) of the main surface of the lead (6) extends in a first direction (Y axis direction) from its one end near the second region (Region B) to its other end far from the second region (Region B), and the first width (a) of at least a part of the first region (Region A) of the main surface of the lead (6) along the second direction (X axis direction) crossing the first direction is larger than or equal to the second width (b) of the electrode (8) soldered to the first region (Region A) along the second direction (X axis direction).
A typical semiconductor modules according to the present invention are described as follows.
According to the structure of the semiconductor module according to the present invention described above, even if the wetting condition of the solder 2 is uneven in the electrode 8 formed on the main surface of the semiconductor element 3 mentioned above, as soon as the solder 2 begins reflowing, the solder 2 that had been localized in a part of the electrodes 8 spreads to other parts of the electrodes 8 in “a defective wetting” condition through the main surface of the Region A of the lead 6. As a result, the difference in wetting expansion of solder 2 in the contact surface between the electrodes 8 and the lead 6 is sufficiently reduced before the force applied by the reflowed solder 2 to the lead 6 grows larger. This is considered due to the phenomenon of an expansion of the width of at least a part of the lead 6 to more than or equal to the width of the electrode 8 accelerates the wetting expansion of the solder 2 in the main surface (Region A) of the lead 6 prior to the reflow of the solder 2 between the electrodes 8 and the lead 6. Therefore, it is not exaggerating to say that the wetting of the solder 2 begins spreading evenly throughout the electrodes 8 at the time when the reflow of the solder 2 turns full-fledged between the electrode 8 and the lead 6. By causing a genuine reflow of wet solder 2 in a part of the electrodes 8, it is possible to limit the phenomenon of the lead 6 moving with the other parts of the electrodes 8 with insufficient wetting of the solder 2 serving as a fulcrum, and the displacement of the lead 6 in relation to the electrodes 8 (semiconductor element 3) resulting from the phenomenon can be dissolved. And as the ratio of the region contributing to the electrical connection with the lead 6 in the main surface of the electrodes 8 facing the lead 6 rises drastically, the connection resistance between the lead 6 and the electrodes 8 can be contained. Therefore, even if the current controlled by the semiconductor element 3 increases, any excessive rise in temperature of the semiconductor module can be contained.
The effect of the semiconductor module of the present invention described above is made more apparent as the Region B of the main surface of the lead 6 is split into a plurality of parts. In particular, when the main surface of the lead 6 extends from its Region A to its Region B, the reflow of the solder 2 between the Region B of the lead 6 and the electrodes 1d on the main surface of the substrate is thought to aid the “displacement” of the Region B of the lead 6 in relation to the electrodes 8. The split up of the Region B of the lead 6 into a plurality of parts enables to restrain the rotation of the lead 6 in the main surface of the substrate 1 due to the reflow of the solder 2 between the Region B and the electrodes 1d on the main surface of the substrate. This effect is enhanced by splitting up the electrode 1d on the main surface of the substrate in accordance with the shape of the Region B of the lead 6. For example, when each of the n number (n is an integer of three or more) of parts constituting the comb-teeth shaped Region B of the lead 6 is joined with one corresponding to each part of the n number of conductive films constituting the electrode 1d on the main surface of the substrate, not only the “displacement” of the Region A of the lead 6 vis-a-vis the electrodes 8 but also the “displacement” of the Region B of the lead 6 vis-a-vis the electrodes 1d can be dissolved.
In any of the semiconductor modules according to the present invention, it is possible to obtain power modules having a lead connection part free from displacement of semiconductor element vis-a-vis electrodes and suitable in particular for controlling large current thanks to the characteristic structure of the lead 6 and the characteristic structure of the electrode 1d joined to the Region B of the lead.
The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings in which:
We will describe below the specific structures characterizing the semiconductor modules according to the present invention with reference to the related drawings for the following embodiment. In the semiconductor module described for this embodiment, we will refer from time to time
In this embodiment, based on the following examination related to five types of lead 6 produced on the trial basis in order to prevent respectively uneven distribution of solder 2 wetting in the electrodes 8 (hereinafter referred to as “source pad 8”) formed on the main surface of the semiconductor element 3 and joined with the lead 6 by the solder 2 and the displacement of the lead 6 in relation with the source pad 8, we will describe the specific structure of the semiconductor module according to the present invention.
Each of five types of lead 6 produced on the trial basis is common in terms of the shape of the main surface as shown in
Regarding the five types of lead 6 belonging to the “lead 1” group used in this embodiment, in order to study the impact of the width of the lead 6 (the width a of the Region A referred to with reference to
In this embodiment, we produced the semiconductor module mentioned above by the procedure shown in
We chose the size of the soldering sheet 2b corresponding to the area of the whole source pads (including the gaps among the source pads 8) so that it may cover the whole area of the plurality of source pads 8 split formed on the upper surface of the TEG chip 3. And to study the impact of variations in the position of supplying the soldering sheet 2b on the source pads 8, we shifted the supply position of the soldering sheet by 0 mm (no displacement), 0.2 mm and 0.5 mm in relation to the source pads 8.
After applying flux 10 on the soldering sheet 2b, we arranged the sample of lead 6 described above on the substrate (metal base substrate) 1 by adjusting the position of its Region A on the soldering sheet 2b with flux 10 applied thereon and the position of its Region B on the soldering paste 2a applied on the other one of the wiring 1c (the second electrode) by putting them together respectively. We placed the lead 6 on the substrate 1 by shifting the position of its supply to the source pad 8 respectively by 0 mm (no displacement), 0.2 mm and 0.5 mm in order to study the impact of the displacement on the source pads 8 of its Region A.
We prepared five samples of the semiconductor module produced on the trial basis for this embodiment for each group of different positions of soldering sheet 20b in relation to the source pad 8 mentioned above and different position of lead to each of the above-mentioned five sample types of lead 6. After the lead 6 was placed on its substrate 1, the samples of these semiconductor modules were placed in a reflow furnace, and the joining between the chip 3 and the first electrode 1c by the reflow of the soldering paste 2a above, the joining between the lead 6 and the second electrode 1c and the joining between the chip 3 and the lead 6 by the reflow of the soldering sheet 2b above were completed. The reflow conditions in the reflow furnace included a maximum temperature of 245° C., and we continued its heating with a melting point of the solder of 217° C. or more for 3 minutes and 20 seconds.
In view of the above, when the lead 6 has a width a (=W) larger than or equal to the width b of the whole of the electrode 8 joined thereto by the solder 2b in its Region A, it has become clear that the problem of uneven solder wetting in the electrodes 8 of the semiconductor element 3 mounted on the semiconductor module is solved by using the lead 6 in the semiconductor module, and that the outbreak of any displacement of the lead 6 vis-a-vis the electrodes 8 can be contained.
On the other hand, when the Region B of a lead 6 extending in the Y axis direction from one end of the Region A side to the other end of the Region B side is separated by a “notch” extending in the Y axis direction from the other end, and is split into two parts juxtaposed in the X axis direction, the effect described above was observed more clearly. And the effect described above could also be clearly observed by, in addition to the Region B thus formed of the lead 6, splitting the second electrode 1d of the substrate 1 joined with the Region B into two conductive films juxtaposed in the X axis direction in the main surface of the substrate 1. Moreover, it was confirmed that the effect of the lead 6 characterized by the shape of the Region A above can be enhanced by splitting the Region B of the lead 6 into n number (n is an integer of three or more) of parts and forming a so-called comb-teeth shaped part and by soldering each of the “parts” with each one corresponding to the “parts” of a n number of the conductive films juxtaposed in the X axis direction on the substrate 1 and constituting the second electrodes 1d.
Therefore, a semiconductor module (power module) suitable for controlling a large current can be obtained by using the lead 6 described in this embodiment.
[Embodiment 2]
In this embodiment, in the place of a lead of the “lead I” group produced on the trial basis for the embodiment 1, we produced samples of lead 6 having a main surface (x-y plane) of different shapes from the same classifying them according to their shape into the groups “lead II” (see
The lead 6 examined in the embodiment 1 fulfills the required relationship that the width a of Region A of its main surface extending in the X axis direction is larger than or equal to the width b (=4.9 mm) along the X axis direction of the electrodes (source pads) 8 of the semiconductor element 3 to be soldered to the Region A in the “whole area” extending from “the nearest end to the Region B” of the Region A to “the other end farther from the Region B.” However, in the samples # (1) and (2) of the “lead II” and the samples # (2) and (3) of the “lead III” examined in this embodiment, in respective Region A, the above-mentioned relationship (a≧b) is not partially fulfilled. In other words, when the Region A of these samples is divided into one end side whose length in the Y axis direction from its Region B side is L1, the central part whose length in the Y axis direction from its Region B side is L2, and the other end side whose length in the Y axis direction from its Region B side is L1, the width a (=W−W3) of the central part of the sample # (1) of the “lead II” is 4.5 mm, the width a of the central part of the sample # (2) of the “lead II” and the width a (=W−2×W3) of the central part of the sample # (2) of the “lead III” are 4.7 mm, and the width a of the central part of the sample # (3) of the “lead III” is 4.4 mm, and they are all smaller than the width b of the electrodes 8 of the semiconductor element 3. The characteristic common to these leads 6 (samples) is that the width (hereinafter referred to as “the first width”) a along the X axis direction in the Region A of the main surface in “at least one part” of one end side near the Region B (one end side part) and the other end side far from the Region B (the other end side part) is larger than or equal to the width (hereinafter referred to as “the second width”) b along the X axis direction of the electrodes 8 of the semiconductor element 3 to be soldered to the Region A, and the first width a in the part other than “at least one part” of the Region A of the main surface of the lead is smaller than the second width b above. Incidentally, the sample # (1) of the “lead III” is equal to the sample # (1) of the “lead I” in terms of each of thickness (dimension in the t and x axis direction), the total length (L) along the Y axis direction, the length of the Region A (2×L1+L2), the length and the split structure of the Region B.
Regarding the samples # (1)-(3) of the “lead II”, the samples # (2) and (3) of the “lead III” and the sample # (1) of the “lead IV” examined in this embodiment, as the comparison of the central part, the one end part and the other end part in respective Region A shows clearly, the first width a along the X axis direction of the Region A of the main surface of the lead is larger than or equal to the second width b along the X axis direction of the electrodes 8 of the semiconductor element 3 to be soldered to the Region A, respectively in one end side and the other end side, and at least a notch extending in the X axis direction is formed in the other part sandwiched between one end side and the other end side of the Region A of the main surface of the lead (the central part mentioned above) In the semiconductor module with such leads 6 (samples), as shown in
It is recommended to create the “notch” in the main surface of the lead (Region A) at a position where the control electrode (gate pad) 7 of the semiconductor element 3 formed in the main surface faces the notch through a gap in the X direction within the main surface (x-y plane) of the semiconductor element 3 where the electrodes (source pads) 8 are formed. For example, when the wiring (conductive film, not shown in any figure) other than the first electrode 1c (the first conductive film) 1d formed in the main surface of the substrate 1 and the second electrode 1c (the second conductive film) 1d and the control electrode 7 of the semiconductor element 3 are electrically connected by wire bonding, the notch formed in the Region A of the main surface of the lead facilitate the positioning of the wire bonder in the main surface of the control electrode 7 and reduces the risk of short circuits between the bonding wire (not shown) and the lead 6.
Like the embodiment 1, as samples of semiconductor module examined in this embodiment, we prepared three types with different position of supplying the soldering sheet 20b to the source pads 8 for each seven types of lead 6 used for the same, and we prepared three types with different arrangement of the Region A of the lead 6 for the source pads 8 for each condition of three types with different positioning between the soldering sheet 20b and the source pads 8. We prepared five samples for each of 63 types of semiconductor modules thus classified, and counted the number of defective samples that occurred for each type and summarized the results in a table shown in
From the facts described above, it has become clear that uneven solder wetting in the electrodes 8 of the semiconductor element 3 mounted on a semiconductor module will be dissolved even if the first width a in the Region A of the lead 6 is neither larger than nor equal to the second width b of the electrodes 8 connected therewith by the solder 2b in the whole Region A as long as the required relationship (a≧b) is fulfilled at least in a part of the Region A. And it has become clear that the occurrence of displacement of the lead 6 vis-a-vis the electrodes 8 of the semiconductor element 3 will be contained in the semiconductor module shown in this embodiment also.
And the above-mentioned effect was more clearly recognized when, in any main surface extending in the Y axis direction from one end on the Region A side to the other end on the Region B side of the three types of lead 6 examined in this embodiment, the Region B was separated by a “notch” extending from the other end in the Y axis direction and was split into two parts juxtaposed in the X axis direction in the main surface of the substrate 1. And the above-mentioned effect was clearly recognized when the second electrode 1d of the substrate 1 to be connected with the Region B was split into two conductive films juxtaposed in the X axis direction corresponding to the Region B of the lead 6 thus formed. And it was also confirmed that the effect of the lead 6 characterized by the shape of the Region A will be enhanced by splitting the Region B of the lead 6 in a n number (n is an integer of three or more) of parts and soldering each of the “parts” with each one corresponding to the “parts” of a n number of conductive films juxtaposed in the X axis direction in the substrate 1 and constituting the second electrode 1d.
Therefore, various types of lead 6 described in this embodiment may be used to obtain semiconductor modules (power modules) suitable for controlling large current.
In the power module mounted on in-vehicle electric apparatuses, the lead connection for mounting the plate-shaped or film-shaped lead as mentioned above on the substrate by facing the main surface of the lead against the main surface of the substrate tends to increase in the future. Therefore, the problem of uneven solder wetting condition on the front surface of the electrodes (conductive films) soldered on the main surface of the lead to which the inventors paid attention and that of displacement of lead vis-a-vis the electrodes are likely to occur frequently in the future. The present invention provides a suitable and effective lead structure for solving these problems.
While we have shown and described several embodiments in accordance with the present invention, it is understood that the same is not limited thereto but is susceptible of numerous changes and modifications as known to those skilled in the art, and we therefore do not wish to be limited to the details shown and described herein but intend to cover all such changes and modifications as are encompassed by the scope of the appended claims.
Number | Date | Country | Kind |
---|---|---|---|
2005-283260 | Sep 2005 | JP | national |