Power conversion apparatus

FIG. 2A is an outline perspective view of the charging-discharging capacitor 5 according to Embodiment 1 of the present invention, and FIG. 2B is a bottom view of the charging-discharging capacitor 5 . As illustrated in FIG. 2A and FIG. 2B , the charging-discharging capacitor 5 includes a first capacitor element 501 a , a second capacitor element 501 b , a first wiring 502 a , a second wiring 502 b , a first outflow-inflow portion 503 a , a second outflow-inflow portion 503 b , a first electrode portion 504 a , a first electrode portion 504 b , a second electrode portion 505 a , and a second electrode portion 505 b.

Further, a center point of the first capacitor element 501 a and a center point of the second capacitor element 501 b are arranged on a straight line 21 at a distance D 1 therebetween, and are connected in parallel between the terminals A and B via the first wiring 502 a and the second wiring 502 b . Specifically, the first outflow-inflow portion 503 a is formed on the first wiring 502 a and connected to the junction point A, and the second outflow-inflow portion 503 b is formed on the second wiring 502 b and connected to the junction point B. Further, the first electrode portion 504 a and the first electrode portion 504 b are formed in the first wiring 502 a , and the second electrode portion 505 a and the second electrode portion 505 b are formed in the second wiring 502 b.

Further, one end of the first capacitor element 501 a is connected to the first wiring 502 a via the first electrode portion 504 a , and another end thereof is connected to the second wiring 502 b via the second electrode portion 505 a . One end of the second capacitor element 501 b is connected to the first wiring 502 a via the first electrode portion 504 b , and another end thereof is connected to the second wiring 502 b via the second electrode portion 505 b.

Here, the first capacitor element 501 a and the second capacitor element 501 b are assumed to have the same film material permittivity, film thickness (distance between internal electrodes), film width (distance between external electrodes), and film wound length, and to have the same capacitance, equivalent series resistance (ESR), and equivalent series inductance (ESL).

Further, the first wiring 502 a and the second wiring 502 b are assumed to have the same length and cross section, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are assumed to have the same length and cross section, and the first electrode portion 504 a , the first electrode portion 504 b , the second electrode portion 505 a , and the second electrode portion 505 b are assumed to have the same length and cross section.

The first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed at locations on a straight line (dotted line) 25 at a distance D 2 from a straight line (dot-and-dash line) 22 indicating a center of the charging-discharging capacitor 5 . Specifically, a loop area of a current path 11 from the junction points A and B corresponding to the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b , respectively, to the first capacitor element 501 a and a loop area of a current path 12 from the junction points A and B to the second capacitor element 501 b are different from each other, the latter loop area being larger than the former.

Therefore, an inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and an inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter inductance component being larger than the former. The first capacitor element 501 a , the second capacitor element 501 b , the first wiring 502 a , and the second wiring 502 b are housed in a case (not shown) formed of a polyphenylene sulfide (PPS) resin or the like. The case is filled with an epoxy resin or the like.

FIG. 3 is a graph for showing frequency characteristics of an absolute value |Z| of an impedance Z of the charging-discharging capacitor 5 according to Embodiment 1 of the present invention. In this case, there are shown an impedance 31 a on the side of the first capacitor element 501 a , an impedance 31 b on the side of the second capacitor element 501 b , and an impedance 31 of the charging-discharging capacitor 5 that is a parallel circuit of the first capacitor element 501 a and the second capacitor element 501 b . As shown in FIG. 3 , the impedance 31 a has a series resonance point Y 1 at a series resonance frequency fs 1 , the impedance 31 b has a series resonance point Y 2 at a series resonance frequency fs 2 , and the impedance 31 has a parallel resonance point X 1 at a parallel resonance frequency fp 1 .

The series resonance frequency fs 1 is expressed by Expression (1) below using an inductance component L 1 and a capacitance component C 1 of the current path 11 to the first capacitor element 501 a.

fs 1=1/{2π( L 1· C 1) 1/2 }  Expression (1)

Further, the series resonance frequency fs 2 is expressed by Expression (2) below using an inductance component L 2 and a capacitance component C 2 of the current path 12 to the second capacitor element 501 b . It is to be noted that characteristics of the first capacitor element 501 a and of the second capacitor element 501 b are the same, so that C 1 =C 2 is found.

fs 2=1/{2π( L 2 ·C 2) 1/2 }  Expression (2)

As described above, the relationship of L 1 <L 2 holds, and thus, fs 1 >fs 2 . In a frequency band between the series resonance frequencies fs 2 and fs 1 , the first capacitor element 501 a acts as a capacitor and the second capacitor element 501 b acts as an inductor. Thus, a parallel resonance of the first capacitor element 501 a and the second capacitor element 501 b occurs. As a result, the impedance 31 of the charging-discharging capacitor 5 is increased. Specifically, as shown in FIG. 3 , the impedance 31 of the charging-discharging capacitor 5 takes a maximum value at the parallel resonance point X 1 at the time when the frequency is the parallel resonance frequency fp 1 (fs 2 <fp 1 <fs 1 ).

FIG. 4 is a graph for showing a relationship between the absolute value of the impedance 31 of the charging-discharging capacitor 5 according to Example 1 and the respective frequency bands, where assuming that N is a positive number and the driving frequency band is fsw to fsw×N, the first capacitor element 501 a is designed so as to satisfy fs 1 >fsw×N, and the second capacitor element 501 b is designed so as to satisfy fs 2 >fsw×N.

Assuming that a noise frequency band required to be attenuated is fn 1 to fn 2 (fn 1 <fn 2 ), the first capacitor element 501 a is designed so as to satisfy fs 1 >fn 2 , and the second capacitor element 501 b is designed so as to satisfy fs 2 <fn 1 .

Therefore, the charging-discharging capacitor 5 does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band. It is to be noted that dimensions of the wirings, the outflow-inflow portions, and the electrode portions are appropriately adjusted depending on the driving frequency band and the noise frequency band. With N having a small value, a relatively low frequency band in the vicinity of the driving frequency can be covered as a noise frequency band.

FIG. 5 is a graph for showing a difference in frequency characteristics of the absolute values of the impedances in the presence of a parallel resonance point of the charging-discharging capacitor according to Embodiment 1 of the present invention and in the absence of the parallel resonance point of the charging-discharging capacitor. Here, there are shown the impedance 31 (solid line) of the charging-discharging capacitor 5 in the presence of the parallel resonance point due to displaced locations of the outflow-inflow portions (the present invention) and an impedance 32 (dotted line) of the charging-discharging capacitor 5 in the absence of the parallel resonance point due to non-displaced locations of the outflow-inflow portions (distance D 2 is zero).

As shown in FIG. 5 , the impedance 32 has a series resonance point Y 3 at a series resonance frequency fs 3 . Compared with the impedance 32 , the impedance 31 is larger in the vicinity of the parallel resonance point X 1 . Therefore, compared with the impedance 32 , the impedance 31 can attenuate more noise in the noise frequency band in the vicinity of the parallel resonance point.

While as an example, both the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are arranged so as to be displaced from the straight line 22 by the distance D 2 , only any one of the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b may be arranged so as to be displaced from the straight line 22 by the distance D 2 . Specifically, insofar as the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b may be arranged arbitrarily within the manufactural restraints of the capacitor.

Thus, in the power conversion apparatus including the plurality of semiconductor switching elements, the reactor, the charging-discharging capacitor, and the control unit configured to insert the charging-discharging capacitor in series between the input and the output thereof through driving control of the plurality of semiconductor switching elements at a predetermined driving frequency, the charging-discharging capacitor is formed by connecting the first capacitor element in parallel with the second capacitor element. Further, the charging-discharging capacitor includes the first wiring, the second wiring, the first outflow-inflow portion, the second outflow-inflow portion, the plurality of first electrode portions, and the plurality of second electrode portions. The first outflow-inflow portion is formed in the first wiring and is connected to the output side. The second outflow-inflow portion is formed in the second wiring and is connected to the input side. The one end of the first capacitor element and the one end of the second capacitor element are connected to the first wiring via the first electrode portions, respectively, and the another end of the first capacitor element and the another end of the second capacitor element are connected to the second wiring via the second electrode portions, respectively. The outflow-inflow portions of the charging-discharging capacitor are arranged so as to be displaced from the center of the capacitor elements.

This makes the loop areas of the current paths to the respective capacitor elements are different from each other, and makes the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions are different from each other. Therefore, the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band, and the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 6A is a perspective view of Example 2 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 6B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 6A and FIG. 6B is different from the charging-discharging capacitor 5 illustrated in FIG. 2A and FIG. 2B in that the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , and in that the first electrode portion 504 b and the second electrode portion 505 b are longer than the first electrode portion 504 a and the second electrode portion 505 a , respectively, by a distance L. In other words, those electrode portions have mutually different lengths.

Therefore, the loop area of the current path 11 from the junction points A and B to the first capacitor element 501 a and the loop area of the current path 12 from the junction points A and B to the second capacitor element 501 b illustrated in FIG. 6A are different from each other, the latter loop area being larger than the former. In other words, the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter inductance component being larger than the former. Except for the locations of the outflow-inflow portions and the lengths of the electrode portions of the charging-discharging capacitor 5 , this Example 2 is similar to Example 1 illustrated in FIG. 2A and FIG. 2B , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

As described above, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , and the first electrode portion 504 b and the second electrode portion 505 b are longer than the first electrode portion 504 a and the second electrode portion 505 a , respectively, by the distance L. Therefore, the inductance component L 2 of the current path 12 to the second capacitor element 501 b is larger than the inductance component L 1 of the current path 11 to the first capacitor element 501 a.

Therefore, similarly to Example 1, the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the first capacitor element 501 a , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the second capacitor element 501 b , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 .

Meanwhile, the impedance in the case where all the electrode portions have the same length can be thought to have characteristics similar to those of the impedance 32 shown in FIG. 5 . Therefore, it can be found that at the time when the electrode portions have different lengths, more noise in the noise frequency band can be attenuated in the vicinity of the parallel resonance point.

While in this case, as an example, the first electrode portion 504 b and the second electrode portion 505 b are longer than the first electrode portion 504 a and the second electrode portion 505 a , respectively, by the distance L, only the first electrode portion 504 b may have a larger length. Specifically, insofar as the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the first electrode portion 504 a , the first electrode portion 504 b , the second electrode portion 505 a , and the second electrode portion 505 b may have any length within the manufactural restraints of the capacitor.

In this way, by adjusting at least one of the plurality of first electrode portions and the plurality of second electrode portions to have a different length, the loop areas of the current paths to the respective capacitor elements are different from each other, and the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions are different from each other. Therefore, the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band, and the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 7A is a perspective view of Example 3 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 7B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 7A and FIG. 7B is different from the charging-discharging capacitor 5 illustrated in FIG. 2A and FIG. 2B in that the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , and in that the first electrode portion 504 a and the second electrode portion 505 a are formed on a straight line 23 indicating a center of the first capacitor element 501 a , but the first electrode portion 504 b and the second electrode portion 505 b are formed at locations that are displaced from a straight line 24 indicating a center of the second capacitor element 501 b by a distance D 3 .

Therefore, the loop area of the current path 11 from the junction points A and B to the first capacitor element 501 a and the loop area of the current path 12 from the junction points A and B to the second capacitor element 501 b illustrated in FIG. 7A are different from each other, the latter loop area being larger than the former. Therefore, the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter inductance component being larger than the former.

Except for the locations of the outflow-inflow portions and the electrode portions of the charging-discharging capacitor 5 , this Example 3 is similar to Example 1 of the charging-discharging capacitor 5 illustrated in FIG. 1 , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

As described above, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , the first electrode portion 504 a and the second electrode portion 505 a are formed on the straight line 23 indicating the center of the first capacitor element 501 a , and the first electrode portion 504 b and the second electrode portion 505 b are formed at the locations that are displaced from the straight line 24 indicating the center of the second capacitor element 501 b by the distance D 3 . Thus, the inductance component L 2 of the current path 12 to the second capacitor element 501 b is larger than the inductance component L 1 of the current path 11 to the first capacitor element 501 a.

Therefore, similarly to Example 1 of the charging-discharging capacitor 5 , the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the first capacitor element 501 a , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the second capacitor element 501 b , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 .

Meanwhile, the impedance at the time when all the electrode portions are formed at the respective centers of the capacitor elements can be thought to have characteristics similar to those of the impedance 32 shown in FIG. 5 . Therefore, it can be found that when the electrode portions are arranged so as to be displaced from the center of the capacitor element, more noise in the noise frequency band can be attenuated in the vicinity of the parallel resonance point.

In this case, as an example, the first electrode portion 504 a and the second electrode portion 505 a are formed on the straight line 23 indicating the center of the first capacitor element 501 a , and the first electrode portion 504 b and the second electrode portion 505 b are formed at the locations that are displaced from the straight line 24 indicating the center of the second capacitor element 501 b by the distance D 3 . However, the first electrode portion 504 a and the second electrode portion 505 a may be formed on the straight line 23 , the first electrode portion 504 b may be formed on the straight line 24 , and the second electrode portion 505 b may be formed at the location that is displaced from the straight line 24 by the distance D 3 .

Specifically, insofar as the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the distances from the straight line 23 to the first electrode portion 504 a and the second electrode portion 505 a , respectively, and the distances from the straight line 24 to the first electrode portion 504 b and the second electrode portion 505 b , respectively, may be any lengths within the manufactural restraints of the capacitor.

In this way, by arranging at least one of the plurality of first electrode portions and the plurality of second electrode portions so as to be displaced from the center of the capacitor element, the loop areas of the current paths to the respective capacitor elements are made different from each other. Then, the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions are different from each other. As a result, the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band, and the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 8A is a perspective view of Example 4 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 8B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 8A and FIG. 8B is different from the charging-discharging capacitor 5 of Example 1 illustrated in FIG. 2A and FIG. 2B in that the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , and in that a capacitance of the first capacitor element 501 a and a capacitance of the second capacitor element 501 b are different from each other. In this case, by adjusting a wound length of a film material of the latter to be larger than that of the former, the capacitances of the capacitor elements are made different from each other.

Therefore, the capacitance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the capacitance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter capacitance component being larger than the former. Except for the locations of the outflow-inflow portions and the capacitances of the capacitor elements of the charging-discharging capacitor 5 , this Example 4 is similar to Example 1 illustrated in FIG. 2A and FIG. 2B , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

As described above, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 , and the capacitance of the second capacitor element 501 b is larger than the capacitance of the first capacitor element 501 a . Therefore, similarly to Example 1, the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the first capacitor element 501 a , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the second capacitor element 501 b , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 .

Meanwhile, the impedance in the case where the capacitances of the capacitor elements are the same can be thought to have characteristics similar to those of the impedance 32 shown in FIG. 5 . Therefore, when the capacitances of the capacitor elements are different from each other, more noise in the noise frequency band can be attenuated in the vicinity of the parallel resonance point.

In this case, as an example, the capacitances of the capacitor elements are made different from each other by adjusting the wound lengths of the films to differ from each other. However, the film widths (distances between external electrodes), the film material permittivities, the film thicknesses (distances between internal electrodes), or the like may differ from each other.

In this way, by adjusting the wound lengths of the two capacitor elements to differ from each other, the capacitance components at the time when the capacitor elements are viewed from the outflow-inflow portions are different from each other. Thus, the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band. Therefore, the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 9A is a perspective view of Example 5 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 9B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 9A and FIG. 9B is different from the charging-discharging capacitor 5 of Example 1 illustrated in FIG. 2A and FIG. 2B in the following: A third capacitor element 501 c having characteristics similar to those of other capacitor elements is added outside the first capacitor element 501 a , the third capacitor element 501 c and the other capacitor elements 501 a and 501 b are arranged on the straight line 21 with the same distance D 1 therebetween, a first electrode portion 504 c and a second electrode portion 505 c are added, and the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on the straight line 22 indicating the center of the charging-discharging capacitor 5 .

In this case, the straight line 23 indicates the center of the first capacitor element 501 a , the straight line 24 indicates the center of the second capacitor element 501 b , and a straight line 25 indicates a center of the third capacitor element 501 c . The first electrode portion 504 a and the second electrode portion 505 a are arranged on the straight line 23 , the first electrode portion 504 b and the second electrode portion 505 b are arranged on the straight line 24 , and the first electrode portion 504 c and the second electrode portion 505 c are arranged on the straight line 25 . Further, all of the electrode portions have the same length.

Therefore, the loop area of the current path 12 from the junction points A and B to the second capacitor element 501 b and the loop area of a current path 13 from the junction points A and B to the third capacitor element 501 c illustrated in FIG. 9A are the same. Meanwhile, the loop area of the current path 11 from the junction points A and B to the first capacitor element 501 a is smaller than and differs from the loop areas in the cases of the other two capacitor elements 501 b and 501 c.

In other words, the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B and the inductance component at the time when the third capacitor element 501 c is viewed from the junction points A and B are the same. Meanwhile, the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B is different from (is smaller than) the inductance components in the cases of the other two capacitor elements 501 b and 501 c.

Except for the locations of the outflow-inflow portions and the numbers of the capacitor elements and the electrode portions of the charging-discharging capacitor 5 , this Example 5 is similar to Example 1 illustrated in FIG. 2A and FIG. 2B , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

Similarly to Example 1, the series resonance frequency fs 1 of the current path 11 to the first capacitor element 501 a is expressed by Expression (1) above, and the series resonance frequency fs 2 of the current path 12 to the second capacitor element 501 b is expressed by Expression (2) above. Further, the inductance component of the current path 12 and the inductance component of the current path 13 are the same, so that it can be found that a series resonance frequency of the current path 13 to the third capacitor element 501 c is expressed by Expression (2) above.

Therefore, similarly to the case of Example 1, the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the first capacitor element 501 a , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the second capacitor element 501 b as well as on the side of the third capacitor element 501 c , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 . Specifically, more noise in the noise frequency band can be attenuated in the vicinity of the parallel resonance point.

By connecting the three capacitor elements in parallel and arranging the outflow-inflow portions at the center of the charging-discharging capacitor, the loop area of the current path to the first capacitor element is different from the loop areas of the current paths to the other two capacitor elements. Thus, the inductance component at the time when the first capacitor element is viewed from the outflow-inflow portions is different from the inductance components at the time when the other two capacitor elements are viewed from the outflow-inflow portions.

Therefore, the charging-discharging capacitor 5 does not have a parallel resonance point in the driving frequency band and has a parallel resonance point in the noise frequency band, and the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 10A is a perspective view of Example 6 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 10B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 10A and FIG. 10B is different from the charging-discharging capacitor 5 of Example 5 illustrated in FIG. 9A and FIG. 9B in that the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed on a straight line 26 that is displaced, by the distance D 2 , from the straight line 22 indicating the center of the charging-discharging capacitor 5 .

Therefore, the loop area of the current path 11 from the junction points A and B to the first capacitor element 501 a , the loop area of the current path 12 from the junction points A and B to the second capacitor element 501 b , and the loop area of the current path 13 from the junction points A and B to the third capacitor element 501 c illustrated in FIG. 10A are different from one another. Further, the loop areas become larger in the order of 11→13→12.

Therefore, the inductance components become larger in the order of the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B, the inductance component at the time when the third capacitor element 501 c is viewed from the junction points A and B, and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B. Except for the structure of the charging-discharging capacitor 5 , this Example 6 is similar to Example 1 illustrated in FIG. 2A and FIG. 2B , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

FIG. 11 is a graph for showing frequency characteristics of the absolute value |Z| of the impedance Z of the charging-discharging capacitor 5 according to this Example 6. In this case, there are shown the impedance 31 a on the side of the first capacitor element 501 a , the impedance 31 b on the side of the second capacitor element 501 b , an impedance 31 c on the side of the third capacitor element 501 c , and the impedance 31 of the charging-discharging capacitor 5 . As shown in FIG. 11 , the impedance 31 a has the series resonance point Y 1 at the series resonance frequency fs 1 , the impedance 31 b has the series resonance point Y 2 at the series resonance frequency fs 2 , and the impedance 31 c has the series resonance point Y 3 at the series resonance frequency fs 3 . Further, the impedance 31 has the three series resonance points described above, the parallel resonance point X 1 at the parallel resonance frequency fp 1 , and a parallel resonance point X 2 at a parallel resonance frequency fp 2 , respectively.

Similarly to Example 1, the series resonance frequency fs 1 of the current path 11 to the first capacitor element 501 a is expressed by Expression (1) above, and the series resonance frequency fs 2 of the current path 12 to the second capacitor element 501 b is expressed by Expression (2) above. Further, the series resonance frequency fs 3 is expressed by Expression (3) below using an inductance component L 3 and a capacitance component C 3 of the current path 13 to the third capacitor element 501 c , where the characteristics of the capacitor elements are the same, so that C 1 =C 2 =C 3 is found.

fs 3=1/{2π( L 3 C 3) 1/2 }  Expression (3)

As described above, a relationship of L 1 <L 3 <L 2 holds for the inductance component, and thus, fs 2 <fs 3 <fs 1 is found. In a frequency band between fs 2 and fs 3 , the first capacitor element 501 a and the third capacitor element 501 c act as a capacitor and the second capacitor element 501 b acts as an inductor, and thus, parallel resonance occurs between the capacitor elements. As a result, the impedance 31 of the charging-discharging capacitor 5 is increased. As shown in FIG. 11 , the impedance 31 of the charging-discharging capacitor 5 takes a maximum value at the parallel resonance point X 2 when the frequency is the parallel resonance frequency fp 2 (fs 2 <fp 2 <fs 3 ).

Further, in a frequency band between fs 3 and fs 1 , the second capacitor element 501 b and the third capacitor element 501 c act as an inductor and the first capacitor element 501 a acts as a capacitor, and thus, parallel resonance occurs between the capacitor elements. As a result, the impedance of the charging-discharging capacitor 5 is increased. As shown in FIG. 11 , the impedance 31 of the charging-discharging capacitor 5 takes a maximum value at the parallel resonance point X 1 when the frequency is the parallel resonance frequency fp 1 (fs 3 <fp 1 <fs 1 ).

FIG. 12 is a graph for showing a relationship between the absolute value 121 of the impedance 2 of the charging-discharging capacitor according to Example 6 and the respective frequency bands. Assuming that N is a positive number and the driving frequency band is fsw to fsw×N, the first capacitor element 501 a is designed so as to satisfy fs 1 >fsw×N, the second capacitor element 501 b is designed so as to satisfy fs 2 >fsw×N, and the third capacitor element 501 c is designed so as to satisfy fs 3 >fsw×N.

Assuming that the noise frequency band required to be attenuated is fn 1 to fn 2 (fn 1 <fn 2 ), the first capacitor element 501 a is designed so as to satisfy fs 1 >fn 2 , and the second capacitor element 501 b is designed so as to satisfy fs 2 <fn 1 .

Therefore, the charging-discharging capacitor 5 does not have a parallel resonance point in the driving frequency band and has a parallel resonance point in the noise frequency band. Dimensions of the wirings, the outflow-inflow portions, and the electrode portions are appropriately adjusted depending on the driving frequency band and the noise frequency band. Further, with N having a small value, a relatively low frequency band in the vicinity of the driving frequency can be covered as a noise frequency band.

FIG. 13 is a graph for comparing frequency characteristics of the absolute values of the impedances in the presence of a parallel resonance point of the charging-discharging capacitor according to Example 6 and in the absence of the parallel resonance point of the charging-discharging capacitor. Here, there are shown the impedance 31 of the charging-discharging capacitor 5 in the presence of the parallel resonance point and the impedance 32 of the charging-discharging capacitor 5 in the absence of the parallel resonance point. As shown in FIG. 13 , compared with the impedance 32 , the impedance 31 of Example 6 is large in the vicinity of the parallel resonance points X 1 and X 2 . Therefore, compared with the impedance 32 , the impedance 31 according to Example 6 can attenuate more noise in the noise frequency band in the vicinity of the parallel resonance points.

In this way, by connecting the three capacitor elements in parallel and arranging the outflow-inflow portions so as to be displaced from the center of the charging-discharging capacitor, the loop areas of the current paths to the respective capacitor elements are different from one another. Therefore, the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions are different from one another, and the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has parallel resonance points in the noise frequency band. Therefore, the impedance becomes larger in the vicinity of the parallel resonance point. Therefore, noise in a desired frequency band including a relatively low frequency band in the vicinity of the driving frequency of the semiconductor switching elements can be reduced without the addition of a noise filter outside the apparatus.

FIG. 14A is a perspective view of Example 7 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 14B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 14A and FIG. 14B is different from the charging-discharging capacitor 5 of Example 1 illustrated in FIG. 2A and FIG. 2B in that the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are formed at the center of the charging-discharging capacitor 5 , and in that the third capacitor element 501 c having a small capacitance is arranged in a gap between the first capacitor element 501 a and the second capacitor element 501 b.

Further, the third capacitor element 501 c is arranged at the point of intersection of the straight line 25 that is displaced, by the distance L, from the straight line 21 indicating the center of the first capacitor element 501 a and the second capacitor element 501 b as well as the straight line 22 indicating the center of the charging-discharging capacitor 5 . In this case, by adjusting a wound length of a film of the third capacitor element 501 c to be smaller than those of the other two capacitor elements, the capacitance of the third capacitor element 501 c is reduced such that the capacitances are different from one another.

Therefore, the capacitance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the capacitance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are the same, and the capacitance component at the time when the third capacitor element 501 c is viewed from the junction points A and B becomes smaller than the capacitance components in the cases of the other two capacitor elements 501 a and 501 b . The inductance components at the time when the capacitor elements are viewed from the junction points A and B are assumed to be the same. Except for the structure of the charging-discharging capacitor 5 , this Example 7 is similar to Example 1 illustrated in FIG. 2A and FIG. 2B , and thus, the description of the structure and the operation of the entire apparatus will be hereafter omitted.

When the series resonance frequency of the current path 11 to the first capacitor element 501 a is represented by fs 2 , fs 2 is expressed by Expression (2) above. The inductance components of the current paths 11 and 12 are the same and the capacitance components of the current paths 11 and 12 are the same, and thus, the series resonance frequency of the current path 12 to the second capacitor element 501 b is also expressed by Expression (2) above. Further, at the time when the series resonance frequency of the current path 13 to the third capacitor element 501 c is represented by fs 1 , fs 1 is expressed by Expression (1) above.

Therefore, similarly to Example 1, the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the third capacitor element 501 c , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the first capacitor element 501 a and on the side of the second capacitor element 501 b , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 .

Specifically, more noise in the noise frequency band can be attenuated in the vicinity of the parallel resonance point. The third capacitor element 501 c is assumed to be arranged in an empty space, and thus, the capacitance thereof is extremely small compared with those of the other two capacitor elements 501 a and 501 b . Therefore, compared with the charging-discharging capacitors 5 of Examples 1 to 6 described above, the series resonance frequency fs 1 becomes higher. Thus, the charging-discharging capacitor 5 of Example 7 can operate as a wider band noise filter.

In this case, the loop areas of the current paths 11 to 13 are the same, but the loop areas may be different from one another by adjusting the dimensions of the wirings, the outflow-inflow portions, and the electrode portions.

In this way, by arranging the third capacitor element 501 c having the smaller capacitance component among the three capacitor elements in the gap between the first capacitor element 501 a and the second capacitor element 501 b , in addition to the effect of Examples 1 to 6, it is advantageous that the charging-discharging capacitor can function as a wider band noise filter without increasing the size.

FIG. 15A is a perspective view of Example 8 of the charging-discharging capacitor used in the power conversion apparatus of the present invention and FIG. 15B is a bottom view of the charging-discharging capacitor.

The charging-discharging capacitor 5 illustrated in FIG. 15A and FIG. 15B is different from the charging-discharging capacitor 5 of Example 1 illustrated in FIG. 2A and FIG. 2B in that the capacitance of the first capacitor element 501 a and the capacitance of the second capacitor element 501 b are different from each other. In this case, by adjusting the wound length of the film material of the latter to be larger than that of the former, the capacitances of the capacitor elements are made different from each other. Specifically, the capacitance of the second capacitor element 501 b is larger to some extent than that of the first capacitor element 501 a . Therefore, the capacitance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the capacitance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter capacitance component being larger than the former. In this case, as an example, the capacitances of the capacitor elements are made different from each other by making the wound lengths of the films different from each other. However, the film widths (distances between external electrodes), the film material permittivities, the film thicknesses (distances between internal electrodes), or the like may be different from each other.

Further, the loop area of the current path 11 from the junction points A and B corresponding to the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b , respectively, to the first capacitor element 501 a and the loop area of the current path 12 from the junction points A and B to the second capacitor element 501 b are different from each other, the latter loop area being larger than the former. Therefore, the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the latter inductance component being larger than the former. While as an example, both the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b are arranged so as to be displaced from the straight line 22 by the distance D 2 , only any one of the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b may be arranged so as to be displaced from the straight line 22 by the distance D 2 . Specifically, insofar as the inductance component at the time when the first capacitor element 501 a is viewed from the junction points A and B and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are different from each other, the first outflow-inflow portion 503 a and the second outflow-inflow portion 503 b may be arbitrarily arranged within the manufactural restraints of the capacitor.

Specifically, the capacitance component and the inductance component at the time when the second capacitor element 501 b is viewed from the junction points A and B are larger than those at the time when the first capacitor element 501 a is viewed from the junction points A and B. Therefore, similarly to Example 1, the absolute value of the impedance of the charging-discharging capacitor 5 has characteristics similar to those of the impedance 31 shown in FIG. 5 , i.e., has the series resonance point Y 1 at the series resonance frequency fs 1 on the side of the first capacitor element 501 a , has the series resonance point Y 2 at the series resonance frequency fs 2 on the side of the second capacitor element 501 b , and has the parallel resonance point X 1 at the parallel resonance frequency fp 1 between fs 2 and fs 1 . However, compared with the cases of the charging-discharging capacitors 5 of Examples 1 to 4, the difference between fs 1 and fs 2 becomes larger.

In this way, by making the capacitance components at the time when the capacitor elements are viewed from the outflow-inflow portions different from each other and making the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions different from each other, the difference in frequency between the series resonance points is increased. Thus, in addition to the effect of Examples 1 to 4, the charging-discharging capacitor can operate as a wider band noise filter.

Other Modified Examples of Charging-Discharging Capacitor

Other than the cases of Examples 1 to 8 described above, for example, by adjusting at least one of cross sections of the plurality of first electrode portions and cross sections of the plurality of second electrode portions different from each other, the loop areas of the current paths to the respective capacitor elements can be made different from each other. Therefore, the inductance components at the time when the capacitor elements are viewed from the outflow-inflow portions differ from each other, and the charging-discharging capacitor does not have a parallel resonance point in the driving frequency band but has a parallel resonance point in the noise frequency band. Thus, the impedance can be large in the vicinity of the parallel resonance point. Therefore, an effect similar to those described above can be obtained.

While in the cases of Examples 1 to 8 described above, two or three capacitor elements are used, even when four or more capacitor elements are used, a similar effect can be obtained for reasons similar to those described above.

Further, while in the cases of Examples 1 to 8 described above, a film capacitor is applied to the charging-discharging capacitor, other ordinary capacitors such as an electrolytic capacitor and a ceramic capacitor may also be used. Further, capacitors of a plurality of kinds such as a combination of a film capacitor and a ceramic capacitor may be connected in parallel. Further, it goes without saying that when a stacked capacitor is used, the capacitances may be made different from each other by causing the numbers of stacked layers to differ from each other.

Further, while in the cases of Examples 1 to 8 described above, the capacitor elements are arranged between the first wiring and the second wiring, the first wiring and the second wiring may be stacked via an insulator such as insulating paper therebetween and the capacitors may be arranged on or under the insulator.

Further, while in the cases of Examples 1 to 8 described above, the plurality of first electrode portions are formed on the first wiring and the plurality of second electrode portions are formed on the second wiring, the electrode portions may be formed of a conductive material different from that of the wirings and one ends thereof may be connected to the wirings and another ends thereof may be connected to the capacitor elements, respectively.

Still further, while in the cases of Examples 1 to 8, the capacitor elements are arranged so that electrode surfaces thereof are arranged in a vertical direction, the electrode surfaces may be arranged in a horizontal direction.

Further, while in the cases of Examples 1 to 8, each of the semiconductor switching elements 2 a to 2 d includes an IGBT and a diode, a MOSFET or a JFET may be substituted for the IGBT. When a MOSFET is used, a MOSFET body diode may be substituted for the diode. Further, each of the semiconductor switching elements 2 a to 2 d may be formed using a wide band gap semiconductor having a larger band gap than silicon, for example, silicon carbide (SiC), a gallium nitride-based material, or diamond.

It goes without saying that the present invention is not limited to Examples 1 to 8 described above and encompasses all possible combinations thereof.