Power module

1. Field of the Invention

The present invention relates to a power module comprising a plurality of switching elements connected in series, and in particular to a power module capable of reducing the radiation noise.

2. Background Art

In a power module, such as an inverter, a switching element, such as an IGBT (Insulated Gate Bipolar Transistor), a MOSFET (Metal Oxide Semiconductor Field Effect Transistor), and a diode, is mounted. The size of the switching element is increased or decreased corresponding to the required capacity. When the capacity is large, the element size is also increased. However, if the element size is extremely large, the yield is lowered, and the manufacturing becomes difficult. Therefore, there is a certain limitation in the largest element size. For this reason, if the capacity is insufficient with just one element, a plurality of elements are connected in series (for example, refer to Japanese Patent Application Laid-Open No. 08-195471).

When the parasitic inductances of a plurality of wirings each connecting a plurality of switching elements to the terminals, respectively differ, high-frequency oscillation is formed between the plurality of elements. When the difference in the performances of a plurality of elements is large, the switching performances of a plurality of elements are different. In this case also, high-frequency oscillation occurs between a plurality of elements. There is a problem wherein the radiation noise level at the oscillation frequency is elevated by the high-frequency oscillation, and various EMI (electromagnetic interference) regulations cannot be satisfied.

In view of the above-described problems, an object of the present invention is to provide a power module capable of reducing the radiation noise.

According to the present invention, a power module comprises: first and second terminals; first and second switching elements having a first electrode and a second electrode which is connected to the second terminal; first and second wirings respectively connecting the first electrodes of the first and second switching elements to the first terminal; and a third wiring directly connecting the first electrode of the first switching element to the first electrode of the second switching element, wherein parasitic inductances of the first and second wiring are different or switching characteristics of the first and second switching elements are different.

The present invention makes it possible to reduce the radiation noise.

Other and further objects, features and advantages of the invention will appear more fully from the following description.

FIG. 1 is a circuit diagram showing a three-phase inverter.

FIG. 2 is a top view showing a power module according to the first embodiment.

FIG. 3 is a sectional view along the line A-A′ on FIG. 2 .

FIG. 4 is a top view showing a power module according to the first comparative example.

FIG. 5 is a sectional view taken along the line A-A′ in FIG. 4 .

FIG. 6 is an equivalent circuit of a power module according to the first comparative example.

FIG. 7 is the result of simulation for the first comparative example.

FIG. 8 is an equivalent circuit of the power module according to the first embodiment.

FIG. 9 is the result of simulation for the first embodiment.

FIG. 10 is a top view showing a power module according to the second embodiment.

FIG. 11 is a sectional view taken along line A-A′ in FIG. 10 .

FIG. 12 is a top view showing a power module according to the second comparative example.

FIG. 13 is a sectional view taken along the line A-A′ in FIG. 12 .

FIG. 14 is the result of simulation for the second comparative example.

FIG. 15 is the result of simulation for the second comparative example.

FIG. 16 is a top view showing a power module according to the third embodiment.

FIG. 17 is a sectional view taken along with the line A-A′ in FIG. 16 .

FIG. 18 is a top view showing a power module according to the fourth embodiment.

FIG. 19 is a sectional view taken along with the line A-A′ in FIG. 18 .

FIG. 20 is a top view showing a power module according to the fifth embodiment.

FIG. 21 is a sectional view taken along with the line A-A′ in FIG. 20 .

A power module according to the embodiments of the present invention will be described with reference to the drawings. The same components will be denoted by the same symbols, and the repeated description thereof may be omitted.

First Embodiment

FIG. 1 is a circuit diagram showing a three-phase inverter. In a U-phase arm, a transistor Q 1 and diodes D 1 to D 4 are connected in parallel between the P-terminal and the U-terminal; and a transistor Q 2 and diodes D 5 to D 8 are connected in parallel between the U terminal and the N-terminal. In a V-phase arm, a transistor Q 3 and diodes D 9 to D 12 are connected in parallel between the P-terminal and the V-terminal; and a transistor Q 4 and diodes D 13 to D 16 are connected in parallel between the V terminal and the N-terminal. In a W-phase arm, a transistor Q 5 and diodes D 17 to D 20 are connected in parallel between the P-terminal and the W-terminal; and a transistor Q 6 and diodes D 21 to D 24 are connected in parallel between the W terminal and the N-terminal. An inductance load (L load) is connected between the P-terminal and the U-terminal. Although the transistors Q 1 to Q 6 are IGBTs, these can be MOSFETs or the like.

FIG. 2 is a top view showing a power module according to the first embodiment, and FIG. 3 is a sectional view along the line A-A′ on FIG. 2 . The power module corresponds to the U-phase arm of the three-phase inverter shown in FIG. 1 .

A conductive board 10 is formed between the P-terminal and the U-terminal; and a conductive board 12 is formed between the N-terminal and the U-terminal. A transistor Q 1 and diodes D 1 to D 4 are mounted on the conductive board 10 . A transistor Q 2 and diodes D 5 to D 8 are mounted on the conductive board 12 .

The cathode 14 of the diodes D 1 to D 4 and the collector electrode 16 of the transistor Q 1 are connected to the conductive board 10 ; and the cathode 14 of the diodes D 5 to D 8 and the collector electrode 16 of the transistor Q 2 are connected to the conductive board 12 .

Al wires L 1 to L 4 connect the anodes 18 of the diodes D 1 to D 4 to the U-terminal, respectively. Al wires L 5 to L 8 connect the anodes 18 of the diodes D 5 to D 8 to the N-terminal, respectively.

Al wires L 9 to L 12 connect the anodes 18 of the diodes D 1 to D 4 to the emitter electrodes 20 of the transistor Q 1 , respectively. Al wires L 13 to L 16 connect the anodes 18 of the diodes D 5 to D 8 to the emitter electrodes 20 of the transistor Q 2 , respectively. Al wires L 17 to L 20 connect the conductive board 10 to the P-terminal; and A 1 wires L 21 to L 24 connect the conductive board 12 to the U-terminal. In the drawing, although each of Al wires L 1 to L 16 connected to the diode is shown by one line, since the current that can flow in a wire is limited, a plurality of wires are practically used corresponding to the capacity.

The diodes D 1 to D 4 are placed so that the size of the conductive board 10 is reduced. Therefore, the length of Al wires L 1 and L 3 differs from the length of Al wires L 2 and L 4 , and the parasitic inductances of them are also different.

In the present embodiment, Al wires L 25 to L 28 are added to directly connect the anodes 18 of the diodes D 1 , D 3 , D 5 , and D 7 to the anodes 18 of the diodes D 2 , D 4 , D 6 , and D 8 . Since the Al wires L 25 to L 28 connect between the diodes in a short distance, the parasitic inductance thereof is lower than the parasitic inductance of the Al wires L 1 to L 8 .

Since no load current flows in Al wires L 25 to L 28 , the use of a plurality of wires is not required. Therefore, the resistance value of the Al wires L 25 to L 28 is higher than the resistance value of the Al wires L 1 to L 8 . In addition, since the frequency of high-frequency oscillation elevates, the resistance component of Al wires L 25 to L 28 is also elevated by the skin effect.

Next, the effect of the first embodiment will be described in comparison with the first comparative example. FIG. 4 is a top view showing a power module according to the first comparative example. FIG. 5 is a sectional view taken along the line A-A′ in FIG. 4 . No Al wires L 25 to L 28 are present in the first comparative example.

FIG. 6 is an equivalent circuit of a power module according to the first comparative example; and FIG. 7 is the result of simulation for the first comparative example. The simulation was carried out at the time of turning ON of the transistor Q 2 in the circuit of FIG. 6 . Since the parasitic inductance of Al wires L 1 and L 3 is different from the parasitic inductance of Al wires L 2 and L 4 , the trailing speeds of free-wheel currents flowing in the diodes D 1 and D 3 , and D 2 and D 4 are different; and timing when the recovery current flows is also different. As a result, high-frequency oscillation occurs between diodes D 1 and D 4 connected in parallel, and current flows. The current cannot be observed by observing load current or the collector current of the transistor Q 2 . In addition, high-frequency oscillation is not attenuated in the first comparative example immediately.

FIG. 8 is an equivalent circuit of the power module according to the first embodiment; and FIG. 9 is the result of simulation for the first embodiment. The simulation was carried out at the time of turning ON of the transistor Q 2 in the circuit of FIG. 8 . In comparison with the first comparative example, the frequency of high-frequency oscillation is high, and the high-frequency signals are attenuated immediately.

In the first comparative example, the current generated by high-frequency oscillation flows through L 1 to L 4 ; however, in the first embodiment, the current generated by high-frequency oscillation flows through Al wires L 25 and L 26 . Thereby, in the first embodiment, the frequency of the high-frequency oscillation is elevated. Therefore, the frequency band of the high-frequency oscillation can be transited to a band having practically no problems (out of the band regulated by various EMI standards), the radiation noise can be reduced. The resonance frequency is determined by the parasitic inductance of wiring and the inter-terminal capacity of the element.

Furthermore, since the resistance values of Al wires L 25 and L 26 are large, the energy of the high-frequency oscillation is inverted to a Joule heat in the Al wires L 25 and L 26 . Therefore, since high-frequency oscillation is attenuated in a short time, the radiation noise can be reduced.

Second Embodiment

FIG. 10 is a top view showing a power module according to the second embodiment; and FIG. 11 is a sectional view taken along line A-A′ in FIG. 10 . Unlike the first embodiment, since diodes D 1 to D 4 are arrayed in line so that the lengths of Al wires L 1 to L 4 become the same, the parasitic inductances of the Al wires L 1 to L 4 are the same. However, the switching characteristics (recovery characteristics) of the diodes D 1 and D 3 are different from the switching characteristics of the diodes D 2 and D 4 .

Next, the effect of the second embodiment will be described in comparison to the second comparative example. FIG. 12 is a top view showing a power module according to the second comparative example, and FIG. 13 is a sectional view taken along the line A-A′ in FIG. 12 . No Al wires L 25 to L 28 are present in the second comparative example.

FIG. 14 is the result of simulation for the second comparative example. In the circuit shown in FIG. 6 , diodes D 1 to D 4 have the same switching characteristics, and the simulation was carried out making the parasitic inductance of each of the wirings L 1 to L 4 be 4 nH. Since the parasitic inductance of each of the wirings L 1 to L 4 connecting the diodes D 1 to D 4 with the U-terminal is the same, and in addition, since the switching characteristics are the same the wave shapes of the current flowing through diodes D 1 to D 4 match each other.

FIG. 15 is the result of simulation for the second comparative example. Unlike with FIG. 14 , the simulation was carried out for the case wherein the switching characteristics of the diodes D 1 and D 3 are slower than the switching characteristics of the diodes D 2 and D 4 . The free-wheel current of the transistor Q 2 at the time of turning ON substantially agrees; however, the timing when the recovery current reaches the peak becomes different. Therefore, after the recovery current has flowed, high-frequency oscillation occurs between the diodes D 1 and D 4 connected in parallel.

Whereas in the second embodiment, since the frequency band of the high-frequency oscillation can be transferred to the band with practically no problem by adding Al wires L 25 to L 28 , as in the first embodiment, the radiation noise can be reduced.

Third Embodiment

FIG. 16 is a top view showing a power module according to the third embodiment, and FIG. 17 is a sectional view taken along with the line A-A′ in FIG. 16 . Although the arrangement of the diodes D 1 to D 4 is the same as in the first embodiment, one metallic board 22 is used in place of Al wires L 1 to L 4 , L 9 to L 12 , L 25 , and L 26 . The metallic board 22 is connected to the diodes D 1 to D 4 with a solder 24 or the like.

By the use of the metallic board 22 , since the parasitic inductance of wires between adjacent diodes is reduced more than the wire connection in the first embodiment, and the high-frequency oscillation frequency among the diodes D 1 to D 4 transits to a higher frequency, the radiation noise can be further reduced.

Fourth Embodiment

FIG. 18 is a top view showing a power module according to the fourth embodiment, and FIG. 19 is a sectional view taken along with the line A-A′ in FIG. 18 . Although the arrangement of the diodes D 1 to D 4 is the same as in the second embodiment, one metallic board 22 is used in place of Al wires L 1 to L 4 , L 9 to L 12 , L 25 , and L 26 . In this case also, the radiation noise can be reduced in the same manner as in the third embodiment.

Fifth Embodiment

FIG. 20 is a top view showing a power module according to the fifth embodiment, and FIG. 21 is a sectional view taken along with the line A-A′ in FIG. 20 . The Al wire L 1 connects the anode 18 of the diode D 1 to the U-terminal. The Al wire L 25 connects the anode 18 of the diode D 1 to the anode 18 of the diode D 2 . The Al wires L 1 and L 25 connect the anode 18 of the diode D 2 to the U-terminal. The Al wire L 3 connects the anode 18 of the diode D 3 to the U-terminal. The Al wire L 26 connects the anode 18 of the diode D 1 to the anode 18 of the diode D 4 . The Al wires L 3 and L 26 connect the anode 18 of the diode D 4 to the U-terminal. Thereby, since the inductances between adjacent diodes (between D 1 and D 2 , between D 3 and D 4 ) are reduced, and the high-frequency oscillation frequency among the diodes D 1 to D 4 transits to a higher frequency, the radiation noise can be further reduced.

In the first to fifth embodiments described above, although diodes are connected in parallel, the present invention is also effective when transistors, such as IGBTs and MOSFETs, are connected in parallel.

The switching element is not limited to those formed by silicon, but may be those formed by a wide-band-gap semiconductor having a larger band gap than silicon. The wide-band-gap semiconductors include, for example, silicon carbide, gallium nitride-based materials, or diamond.

Since high-frequency oscillation occurs more easily in the transistor formed by a wide-band-gap semiconductor than in the transistor formed by silicon, the high-frequency oscillation between elements connected in parallel at the time of parallel connection frequently causes problems. Therefore, the present invention is particularly effective in the case wherein the switching element is formed of a wide band-gap semiconductor.

In addition, since an element formed of a wide band-gap semiconductor has a high breakdown voltage and allowable current density, the size of the element can be reduced. By using a down-sized element, the size of the power module equipped with such elements can also be reduced. Since the heat resistance of the element is high, the size of the heat dissipation fin can be reduced, and since the water-cooled parts can be cooled by the air, the power module can be further down-sized. Furthermore, since the power loss of the element is low and the element is highly efficient, the efficiency of the power module can be elevated. Although it is desirable that both the transistor and the diode be formed of wide-band-gap semiconductor, the above-described effect can be obtained if either one of elements is formed of a wide-band-gap semiconductor.

Obviously many modifications and variations of the present invention are possible in the light of the above teachings. It is therefore to be understood that within the scope of the appended claims the invention may be practiced otherwise than as specifically described.

The entire disclosure of a Japanese Patent Application No. 2010-175372, filed on Aug. 4, 2010 including specification, claims, drawings and summary, on which the Convention priority of the present application is based, are incorporated herein by reference in its entirety.