WO2012098585A1

WO2012098585A1 - Three-phase inverter for driving variable-speed electric machine

Info

Publication number: WO2012098585A1
Application number: PCT/JP2011/000305
Authority: WO
Inventors: Shouichi Tanaka
Original assignee: Three Eye Co., Ltd.
Priority date: 2011-01-21
Filing date: 2011-01-21
Publication date: 2012-07-26

Abstract

A first three-phase inverter is connected to one ends of three phase windings of a motor. A second three-phase inverter is connected to the other ends of the three phase windings. The two inverters have a full bridge mode and a star connection mode. The full bridge mode is operated in a high speed range, and the star connection mode is operated in a low speed range. The first inverter has lower resistance than the second inverter. Preferably, the second inverter is not PWM-switched in both of the star connection mode and the full bridge mode. A three-phase inverter is driven with a one-pulse-method instead of a PWM-method in a low speed area, if a required motor torque is large.

Description

THREE-PHASE INVERTER FOR DRIVING VARIABLE-SPEED ELECTRIC MACHINE

Cross-Reference to Related Application

This application claims benefit, for example under 35 U.S.C.119, from Japan Patent Application No.2010-40324 filed on Feb/25/2010, which has the title of SIX-LEG THREE-PHASE INVERTER APPARATUS DRIVING VARIABLE-SPEED MOTOR, of which the entire content is incorporated herein reference. Moreover, the contents of the above application should be used to consider this application.

Background of the Invention

1. Field of the Invention
The present invention relates to a three-phase inverter for driving a three-phase electric machine of variable-speed type, in particular to a three-phase inverter capable of increasing a low-speed torque of a variable-speed motor.

2. Description of the Related Art
A stator winding of a three-phase electric machine has an induced voltage, which is approximately proportional to the machine speed. In many variable-speed applications, the three-phase machine must have a large torque in a low speed range and an adequate torque in a high speed range. A low-speed-torque can be increased by means of increasing a turn number of a stator winding. However, an induced voltage being a generation voltage of the stator winding is increased by means of increasing a turn number of the stator winding. It is well-known that a stator winding with a large turn number can not produce a motor torque in a high speed range.

PCT/EP2003/007698 discloses the dual-three-phase inverter, the DTP inverter. In the DTP inverter shown in Figure 1, three phase windings 1000 are driven by a first three-phase inverter 1001 and a second three-phase inverter 1002. Each of phase windings U, V and W are driven by each of three single-phase full-bridge inverters, the three H-bridges. A high potential DC link line 1003 is connected to a high potential DC terminal 1004 of the second inverter 1002 via a mode-changing switch 1005. A low potential DC link line 1006 is connected to a low potential DC terminal 1007 of the second inverter 1002 directly. The above dual-three-phase inverter has two operation modes.

According to the first mode, the mode-changing switch 1005 and three lower switches of inverter 1002 are turned-off. The three-phase winding 1000 is star-connected, because turned-on three upper switches of the second inverter 1002 constitute a neutral point of the star-connection. According to the second mode, the mode-changing switch 1005 is turned-on. The three-phase winding 1000 is driven by three H-bridges independently.

However, the above DTP inverter needs thirteen pairs of a power transistor and a free-wheeling diode. PCT/EP2003/007698 requires that each pair is equal to each other. Accordingly, the DTP inverter of PCT/EP2003/007698 has a high cost, which is equal to 233% of the conventional three-phase inverter. Moreover, the mode-changing-switch 1005 connecting the high potential DC terminal 1002 and the high potential DC link line 1003 requires a high gate voltage. Electrical insulation of a gate controller applying the high gate voltage needs a high cost.

Japan unexamined patent publication No. 2010-081786 applied by the inventor discloses a similar DTP inverter for driving a three-phase motor, too.

PCT/EP2003/007698 Japan unexamined patent publication No. 2010-081786

One object of the invention is to provide a three-phase inverter capable of increasing a low-speed torque of a variable-speed electric machine. Another object of the invention is to provide an economical three-phase inverter of increasing a low-speed torque of a variable-speed electric machine. Another object of the invention is to increase the low-speed-torque of a three-phase motor without a high expense. Another object of the invention is to provide a six-leg three-phase inverter capable of driving a variable-speed electric machine in a high speed zone.

According to a first aspect of the invention, a six-leg three-phase inverter has a first inverter (4A), a second inverter (4B) and a mode-changing switch (7). The six-leg inverter has a star connection mode and a full bridge mode. The second inverter (4B) constitutes a neutral point of a star-connected three phase windings (10U, 10V and 10W) in the star connection mode, when the mode-changing switch (7) is turned off. The inverters (4A and 4B) constitute three single-phase full bridges in the full bridge mode, when the mode-changing switch (7) is turned on. The star connection mode is operated in higher torque area (304) than an operating area (305) of the full bridge mode, when a rotor speed is lower than a predetermined value. Accordingly, the torque in the low speed area is increased without increasing a weight of the machine.

According to a preferred embodiment, the second inverter (4B) has larger switching loss than the first inverter (4A) in same condition. Moreover, the first inverter (4A) is PWM-switched in the star connection mode. Accordingly, the switching loss of the second inverter is reduced, even though the second inverter (4B) has larger switching loss. Because the second inverter 4(B) forms a neutral point of the star-connected three-phase winding. As the result, a production cost of the second inverter (4B) is decreased. Furthermore, it is prevented to increase a power loss of the second inverter even though the second inverter is PWM-switched in the full bridge mode. Because a stator current of the full bridge mode operated in the high speed area (305) is smaller.

According to a preferred embodiment, the first inverter (4A) is PWM-switched in both of the star connection mode and the full bridge mode, and the second inverter (4B) is not PWM-switched in both of the star connection mode and the full bridge mode. Accordingly, the power loss of the second inverter (4B) is reduced largely, even though the resistance of the second inverter is increased.

According to another preferred embodiment, the second inverter (4B) is switched at each timing when a new phase voltage of three phase voltages (Vu, Vv and Vw) becomes larger than another two phase voltages. Accordingly, the full bridge mode is operated without PWM-switching of the second inverter.

According to a preferred embodiment, the star connection mode is operated in a higher torque area (304), and the full bridge mode is operated in a lower torque area (304 and 305). As the result, the PWM-switching loss of the second inverter (4B) is reduced, even though the second inverter (4B) has the large switching loss, because the second inverter (4B) being the neutral point is not PWM-switched both of lower-speed areas (303 and 304).

According to a preferred embodiment, switches (41, 42, 51, 52, 61 and 62) of the second inverter (4B) have larger switching loss and smaller constant conductive loss than switches (11, 12, 21, 22, 31 and 32) of the first inverter (4A) in same condition. As the result, the power loss of the second inverter (4B) is reduced. For example, each switch has an IGBT. The switching loss and the constant conductive loss of the IGBT can be controlled by means of controlling a life time of remaining careers. The switching loss of the IGBT is decreased by means of decreasing the life time, and the constant conductive loss of the IGBT is decreased by means of increasing the life time.

According to another preferred embodiment, each switch (11, 12, 21, 22, 31 and 32) of the first inverter (4A) consists of more semiconductor chips connected in parallel to each other than each switch (41, 42, 51, 52, 61 and 62) of the second inverter (4B). As the result, all of switch of both inverters can be made with same chips.

According to another preferred embodiment, the mode-changing switch (7) connects the negative DC terminal (103) of the second inverter (4B) to a negative terminal (105) of the DC power supply (8). Accordingly, a turned-on resistance of the mode-changing switch (7) is decreased, even though a gate voltage of the mode-changing switch (7) is not increased.

According to another preferred embodiment, the mode-changing switch (7) is switched, after three lower switches (42, 52 and 62) of the second inverter (4B) are turned off. Accordingly, the power loss of the mode-changing switch (7) is decreased. Particularly, a relay as the mode-changing switch (7) prevents the sparking.

According to another preferred embodiment, the inverters (4A and 4B) are connected to the DC power supply (8A) via a boost DC/DC converter 8(B). The boost DC/DC converter 8(B) and the first inverter (4A) are operated with a single-phase-switching-mode in the star connection mode. Accordingly, a power loss of the first inverter is reduced largely in the star connection mode.

According to another preferred embodiment, the inverters (4A and 4B) are operated with a two-phase-switching-mode in the full bridge mode. Accordingly, a power loss of the first inverter is reduced largely in the full bridge mode.

According to another preferred embodiment, both of the inverters (4A and 4B) are operated with a one-pulse-method in a low speed area, when a required motor is increased. Accordingly, a large motor torque is produced in the low speed area. The power loss of the inverter is reduced, too. The one-pulse-method is often called the rectangular-pulse method.

According to another preferred embodiment, a duty ratio of the inverter (4A) switched with the PWM-method is increased, just after when the star connection mode is started instead of the full bridge mode. The duty ratio of the inverter (4A) switched with the PWM-method is decreased, just after when the full bridge mode is started instead of the star connection mode. As the result, the torque difference before and after the mode-changing is reduced.

According to a second aspect of the invention, a three-phase inverter (4) for driving a variable-speed electric machine is operated with a PWM-method, if a required motor torque is smaller than a predetermined value. The inverter (4) is operated with a one-pulse-method in a low speed area (308), if the required motor torque is larger than the predetermined value. Accordingly, the large motor torque is produced in the low speed area. The power loss of the inverter is reduced, too.

Figure 1 is a block circuit diagram showing a dual-three-phase inverter of the prior arts. Figure 2 is a block circuit diagram showing a dual-three-phase inverter of the first embodiment of the invention. Figure 3 is a schematic diagram showing four stages of the dual three-phase inverter shown in Figure 2. Figure 4 is a schematic diagram showing the other two stages of the dual three-phase inverter shown in Figure 2. Figure 5 is a timing chart showing sinusoidal curves of a three-phase voltage. Figure 6 is a diagram showing a current and a resistance of a transistor, when the transistor is turned on and turned off. Figure 7 is an arranged circuit diagram showing one arranged DTP inverter. Figure 8 is a diagram showing a torque-speed characteristic of a three-phase motor. Figure 9 is another arranged circuit diagram showing another arranged DTP inverter of driving the three-phase generator. Figure 10 is a circuit diagram showing an actually-designed DTP inverter. Figure 11 is a circuit diagram for showing a switching timing of the mode-changing switch. Figure 12 is another circuit diagram for showing a switching timing of the mode-changing switch. Figure 13 is a circuit diagram showing a dual-three-phase inverter of the second embodiment of the invention. Figure 14 is a schematic diagram showing six stages of the dual three-phase inverter shown in Figure 13, when the star connection mode with the single-phase-switching-mode is employed. Figure 15 is a table diagram showing six stages A-F. Figure 16 is a timing chart showing sinusoidal curves of a three-phase voltage. Figure 17 is a timing chart showing the biggest inter-phase voltage Vx applied to the three-phase inverter. Figure 18 is a schematic diagram showing four stages of the dual three-phase inverter shown in Figure 13. Figure 19 is a schematic diagram showing the other two stages of the dual three-phase inverter shown in Figure 13. Figure 20 is a diagram showing a torque-speed characteristic of the second embodiment. Figure 21 is a circuit diagram showing a conventional three-phase motor-driving apparatus of the second embodiment. Figure 22 is a timing chart showing a 120 degrees method of the three-phase inverter of the second embodiment. Figure 23 is a timing chart showing waveforms of a U-phase one-pulse voltage and the three-phase voltage.

Proffered Embodiments of the Invention

(The first embodiment)
The first embodiment is explained referring to Figure 2. Figure 2 is a block circuit diagram showing a dual-three-phase inverter, the DTP inverter, which is equal to a six-leg three-phase inverter. The DTP inverter applies a three-phase voltage to a three-phase winding wound on the EV traction motor. The DTP inverter has a first three-phase inverter 4A, a second three-phase inverter 4B and a mode-changing switch 7. The DTP inverter consists of six legs 1-6 constituted by a half-bridges each. A dc current power supply 8 supplies a direct current power to the

inverters

4A and 4B controlled by a controller (not shown). The EV traction motor consists of a three-phase synchronous motor or a three-phase asynchronous motor. Inverter 4A consists of the three legs 1-3. Inverter 4B consists of the three legs 4-6.

U-phase leg 1 consists of an upper switch 11 and a lower switch 12 connected in series to each other. V-phase leg 2 consists of an upper switch 21 and a lower switch 22 connected in series to each other. W-phase leg 3 consists of an upper switch 31 and a lower switch 32 connected in series to each other. U- phase leg 4 consists of an upper switch 41 and a lower switch 42 connected in series to each other. V- phase leg 5 consists of an upper switch 51 and a lower switch 52 connected in series to each other. W-phase leg 6 consists of an upper switch 61 and a lower switch 62 connected in series to each other.

Each switch of six legs 1-6 consists of a pair of a power transistor and a free-wheeling diode, which are connected in parallel to each other. A positive terminal 100 of inverter 4A and a positive terminal 102 of inverter 4B are connected to a positive terminal 104 of the DC power supply 8. A negative terminal 101 of inverter 4A is connected to a negative terminal 105 of DC power supply 8. A negative terminal 103 of inverter 4B is connected to the negative terminal 105 via the mode-changing switch 7 consisting of a pair of an IGBT and a free-wheeling diode D. A collector electrode of the IGBT is connected to the negative terminal 103. An emitter electrode of the IGBT is connected to negative terminal 105.

The stator winding of the EV traction motor consists of a U-phase winding 10U, a V-phase winding 10V and a W-phase winding 10W. U-phase winding 10U is connected to both output terminals of

U-phase legs

1 and 4. V-phase winding 10V is connected to both output terminals of V-

phase legs

2 and 5. W-phase winding 10W is connected to both output terminals of W-

phase legs

3 and 6. A U-phase full-bridge consisting of

U-phase legs

1 and 4 applies a U-phase voltage to U-phase winding 10U. A V-phase full-bridge consisting of V-

phase legs

2 and 5 applies a V-phase voltage to V-phase winding 10V. A W-phase full-bridge consisting of W-

phase legs

3 and 6 applies a W-phase voltage to W-phase winding 10W. The controller has a plurality of operation modes for driving

inverters

4A and 4B. One operation mode is called the star connection mode, and another operation mode is called the full bridge mode.

The star connection mode is explained referring to Figure 2. According to the star connection mode, DC power supply 8 is separated from inverter 4B by means of turning-off the mode-changing switch 7. Moreover, three

upper switches

41, 51 and 61 of inverter 4B are turned off. Three

lower switches

42, 52 and 62 of inverter 4B are turned on. Accordingly, a neutral point of the star-connected three-

phase windings

10U, 10V and 10W consists of turned-on

lower transistors

42, 52, 62 of inverter 4B. The first inverter 4A is switched with a well-known PWM method in order to supply a three-phase alternative current to three-

phase windings

10U, 10V and 10W.

The full bridge mode is explained referring to Figures 3-5. According to the full bridge mode, DC power supply 8 applies the DC voltage to inverter 4B by means of turning-on the mode-changing switch 7. The

inverters

4A and 4B are switched to apply the three-phase voltage to three-

phase windings

10U, 10V and 10W. The three full-bridges apply a three-phase voltage Vu, Vv ad Vw with predetermined wave pattern, preferably three-phase sinusoidal wave pattern. In the full bridge mode,

lower switches

42, 52, 62 and 7 can have a low turned-on-resistance, because an emitter electrode or a source electrode of each lower transistor of

lower switches

42, 52, 62 and 7 is connected to the low potential terminal of DC power supply 8.

Figure 3 shows four stages TA-TD of

inverters

4A and 4B. Figure 4 shows two stages TE-TF of

inverters

4A and 4B. The stages TA-TF are shown in Figure 5 showing waveforms of three phase voltages Vu, Vv and Vw applied to three

phase windings

10U, 10V and 10W respectively. It is well-known that each angle difference among three

phase voltages

10U, 10V and 10W is an electric angle of 120 electrical degrees each.

The stage TA is a period from 60 electrical degrees to 120 electrical degrees. The stage TB is a period from 120 electrical degrees to 180 electrical degrees. The stage TC is a period from 180 electrical degrees to 240 electrical degrees. The stage TD is a period from 240 electrical degrees to 300 electrical degrees. The stage TE is a period from 300 electrical degrees to 0 electrical degrees. The stage TF is a period from 0 electrical degrees to 60 electrical degrees.

In the stage TA from 60 electrical degrees to 120 electrical degrees, an absolute amplitude of U-phase voltage Vu becomes the biggest. The

switches

42, 51 and 61 of inverter 4B are always turned on. The

switches

41, 52 and 62 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

In the stage TB from 120 electrical degrees to 180 electrical degrees, absolute amplitude of V-phase voltage Vv becomes the biggest. The

switches

42, 51 and 62 of inverter 4B are always turned on. The

switches

41, 52 and 61 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

In the stage TC from 180 electrical degrees to 240 electrical degrees, an absolute amplitude of W-phase voltage Vw becomes the biggest. The

switches

41, 51 and 62 of inverter 4B are always turned on. The

switches

42, 52 and 61 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

In the stage TD from 240 electrical degrees to 300 electrical degrees, an absolute amplitude of U-phase voltage Vu becomes the biggest. The

switches

41, 52 and 62 of inverter 4B are always turned on. The

switches

42, 51 and 61 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

In the stage TE from 300 electrical degrees to 360 electrical degrees, an absolute amplitude of V-phase voltage Vv becomes the biggest. The

switches

41, 52 and 61 of inverter 4B are always turned on. The

switches

42, 51 and 62 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

In the stage TF from 360 electrical degrees to 60 electrical degrees, an absolute amplitude of W-phase voltage Vw becomes the biggest. The

switches

42, 52 and 61 of inverter 4B are always turned on. The

switches

41, 51 and 62 of inverter 4B are always turned off. The first inverter 4A is PWM-switched in order to produce the three-phase voltage shown in Figure 5.

Consequently, all of six switches of the second inverter 4B are not PWM-switched in both of the star connection mode and the full bridge mode. In the other words, the legs 4-6 become the fixed legs. Three legs 1-3 of inverter 4A become the switched legs, because three legs 1-3 are PWM-switched in both of the star connection mode and the full bridge mode.

A power loss of the inverter 4B is reduced largely in comparison with the first inverter 4A. Because almost power loss of a transistor is a switching loss produced in a turning-on period Tr and a turning-off period Td as shown in Figure 6. Figure 6 is a schematic diagram showing configurations of a current and a resistance of a transistor, for example the IGBT. Almost power loss is generated in the turning-on period Tr and the turning-off transient period Td, while the gate voltage is changing. As the result, inverter 4B without PWM-switching has a small power loss in comparison with inverter 4A in both of the star connection mode and the full bridge mode.

(One arrangement)
Figure 7 is a circuit diagram showing one arranged DTP inverter. A relay 7A is employed instead of the IGBT 7 shown in Figure 2. The relay 7A can stop a bi-directional current. However, the relay 7A has a spark problem. In order to avoid the spark of relay 7A, relay 7A are turned on and turned off in a predetermined period just after three

lower switches

42, 52 and 62 of the second inverter 4B have been turned off. Accordingly, the spark of the relay 7A is prevented.

Upper switches

41, 51 and 61 of the second inverter 4B are turned off, if relay 7A can be connected to the high potential DC link bus line.

Differences between the full bridge mode and the star connection mode are explained. Three

phase windings

10U, 10V and 10W have predetermined impedance each. The full-bridge mode applies a voltage value of DC power supply 8 to phase

windings

10U, 10V and 10W each. The star-connection mode applies the voltage value of DC power supply 8 to series-connected two of three

phase windings

10U, 10V and 10W.

A mode-changing method is explained referring to Figure 8. Figure 8 is a torque-speed diagram showing a torque-speed characteristic of the EV traction motor. The line 301 shows a torque-speed characteristic of the star connection mode. The line 302 shows a torque-speed characteristic of the full bridge mode. Accordingly, three operation areas 303-305 are formed by the lines 301-302. The operation area 303 is a low-speed-low-torque area, in which the star connection mode and the full bridge mode can be operated. The operation area 304 is a low-speed-high-torque area, in which the star connection mode can be operated. The area 305 is a high-speed-low-torque area, in which the full bridge mode can be operated. Either one of the star connection mode and the full bridge mode is selected in accordance with a predetermined parameter, for example a total efficiency. Preferably, the full bridge mode is employed in the operation area 303. Consequently, the EV traction motor can be operated in wider torque-speed range. A compact motor can be employed, because the low-speed torque is improved.

Furthermore, the mode-changing method of this embodiment has

other operation areas

308 and 309. The operation area 308 has more torque than the operation area 304 with the star connection mode. The operation area 309 has more speed than the operation area 305 with the full bridge mode. In the

operation areas

308 and 309, the first inverter 4A is driven with one-pulse-method instead of the PWM method. The one-pulse-method means that each switch of the first inverter 4A is switched in turn in accordance with a rotation angle of a rotor. In the other words, both

inverters

4A and 4B are switched in turn in accordance with a rotation angle of a rotor. It is known that the one-pulse-method includes the current-supplying method of the 120 electrical degrees type, the current-supplying method of the 180 degrees type and the current-supplying method of the 150 degrees type. The current-supplying method of the 120 degrees type is illustrated in Figure 22. It is well-known that the one-pulse-method produces larger torque than the PWM-method under the equal voltage of the DC supply.

But, the one-pulse-method produces larger vibration and a larger power loss, too. However, the very-high-torque area 308 and the very-high-speed area 309 are not operated frequently. As the result, it is acceptable to increasing of the vibration, the acoustic noise and the power loss. Consequently, the compact motor can have further wider torque-speed area. Moreover, Figure 8 shows a continues-rating torque-speed characteristic. In a short period, the torque-speed characteristic shown in Figure 7 can be enlarged.

(Another arrangement)
Figure 9 is a circuit diagram showing another arranged DTP inverter of a three-phase alternative-current generator. A three-phase full-bridge diode rectifier 4A is employed instead of the first three-phase inverter 4A shown in Figure 2. The DTP inverter shown in Figure 9 has the star connection mode and the full bridge mode in the electric power generation.

Figure 10 is a circuit diagram showing an actually-designed DTP inverter shown in Figure 2. The DTP inverter shown in Figure 10 drives a three-phase EV traction motor, of which a rating power is more than 60 kW. Each of six

switches

11, 12, 21, 22, 31 and 32 of the first inverter 4A consists of two chip pairs 400 each. Each chip pair 400 consists of a transistor chip and a free-wheeling diode chip connected in parallel to each other. Two chip pairs 400 of each switch are connected in parallel to each other. Each of six

switches

41, 42, 51, 52, 61 and 62 of the second inverter 4B consists of one chip pair 400, which consists of one transistor chip and one free-wheeling diode chip connected in parallel to each other. Each transistor chip is same to each other. Each free-wheeling diode chip is same to each other. Accordingly, a turned-on-resistance of each switch of inverter 4B has a resistance value of two times in comparison with each switch of inverter 4A. However, a cost of the inverter 4B is reduced largely.

A power loss of the DTP inverter is not increased largely in both of the star connection mode and the full bridge mode, even though the resistance of the second inverter is increased. Because all of six

switches

41, 42, 51, 52, 61 and 62 of the second inverter 4B are not PWM-switched in both of the star connection mode and the full bridge mode. The PWM-switching less inverter 4B has less power loss than the PWM-switched inverter 4A.

Furthermore, the motor current flowing through the inverter 4B is small in the full bridge mode, because the full bridge mode is operated in the low-torque-high-speed area 305 as shown in Figure 8. The inverter loss is largely depending on the current. Accordingly, the power loss of the inverter 4B is reduced, even though the inverter 4B is PWM-switched in the full bridge mode. Furthermore, six switches of the first inverter 4A are switched with the PWM-method in both of the star-connection mode and the full bridge mode. However, all switches of the first inverter 4A have a low value of the turned-on resistance, because the switches of inverter 4A can have a low on-resistance.

Preferably, each transistor chip consists of an IGBT, the insulated gate bipolar transistor. Each IGBT of the second inverter 4B has larger switching loss and smaller constant conductive loss than Each IGBT of the first inverter 4A in same condition. As the result, the power loss of the second inverter 4B is reduced. The IGBT of the second inverter 4B has longer life time of remaining careers than the IGBT of the first inverter 4A. An IGBT with longer life time of the remaining careers can have larger switching loss and smaller constant conductive loss than an IGBT with shorter life time of the remaining careers.

Preferable switching methods of the mode-changing switch are explained referring to Figures 11-12. Figure 11 is a circuit diagram showing the DTP inverter having a low-side mode-changing switch 7. Figure 12 is a circuit diagram showing the DTP inverter having a high-side mode-changing switch 7H. In Figure 11,

lower switches

42, 52 and 62 are turned off, just before the low-side mode-changing switch 7 is turned off. In Figure 12,

upper switches

41, 51 and 61 are turned off, just before the high-side mode-changing switch 7H is turned off. It is preferable to control to generate an equal torque before and after when the mode is changed.

It is important to decrease a torque difference between two torques before and after the mode-changing. Mechanical shock and acoustic noise are generated by the mode-changing, when the torque difference is large. A duty ratio of the inverter 4A switched with the PWM-method is increased, just after when the star connection mode is started instead of the full bridge mode. The duty ratio of the inverter 4A switched with the PWM-method is decreased, just after when the full bridge mode is started instead of the star connection mode. As the result, the torque difference before and after the mode-changing is reduced.

(The second embodiment)
The second embodiment is explained referring to Figure 13. Figure 13 is a circuit diagram showing the DTP inverter for driving the three-phase EV traction motor. The DC power supply 8 has a bi-directional boost/down DC/DC converter 8B boosting a battery voltage Vb of the battery 8A. The DTP inverter having the first inverter 4A, the second inverter 4B and the mode-changing switch 7 is essentially same as the DTP inverter shown in Figure 2. The positive terminal 102 of inverter 4B is connected to an upper switch 8E of converter 8B. Converter 8B consists of a reactor 8C, the upper switch 8E and a lower switch 8F.

A contact point connecting the switches 8E and 8F is connected to the positive terminal 104 of battery 8A through the reactor 8C. The lower terminal 103 of inverter 4B is connected to a negative terminal 105 of battery 8A via the mode-changing switch 7. A smoothing capacitor 50 connects the positive terminal 104 of the battery 8A and the positive terminal 102 of

inverters

4A and 4B.

Operation of the bi-directional boost/down DC/DC converter 8B is well known. The lower switch 8F and a freewheeling diode D2 constitute the boost converter. The upper switch 8E and a freewheeling diode D1 constitute the down converter. In a motor mode, converter 8B applies a boost voltage to

inverters

4A and 4B. In a generator mode, a step-down voltage of converter 8B charges battery 8A. The full-bridge mode is operated, when the switch 7 is turned on. The star connection mode is operated, when the switch 7 is turned off. It is same as the first embodiment explained above.

The star connection mode of the second embodiment is explained referring to Figures 14-17. In this star connection mode, converter 8B outputs a boost voltage having a three-phase-full-wave-rectified-waveform. The mode-changing switch 7 and

upper switches

41, 51 and 61 are turned off.

Lower switches

42, 52 and 62 are turned on. Furthermore, an upper switch of one of three legs 1-3 of inverter 4A is turned on. A lower switch of another one of three legs 1-3 is turned on. The other one of three legs 1-3 is switched with the PWM method. This switching method is called the single-phase switching method, the SPSM.

In the single-phase switching method of the star-connection mode, inverter 4A and converter 8B applies a three-phase voltage to three

phase windings

10U, 10V and 10W. The DC/DC converter 8B changes a battery voltage Vb to a periodically-changed voltage Vx. Converter 8B applies the voltage Vx to three-phase inverter 4A. The voltage Vx has the three-phase-full-wave-rectified-waveform.

Figure 14 is a schematic diagram showing six stages A-F shown in Figures 15 and 16. The stage A is a period from 30 electrical degrees to 90 electrical degrees. The stage B is a period from 90 electrical degrees to 150 electrical degrees. The stage C is a period from electrical 150 degrees to 210 electrical degrees. The stage D is a period from 210 electrical degrees to 270 electrical degrees. The stage E is a period from 270 electrical degrees to 330 electrical degrees. The stage F is a period from 330 electrical degrees to 30 electrical degrees. In each of six stages A-F, two of three legs 1-3 keep a constant state, and one of three legs is PWM-switched for forming a smaller inter-phase voltage Vy with the sinusoidal waveform.

The switched leg can be switched with a well-known PWM method or the allowable-band-switching method. In Figure 14, V-phase leg 2 is the switched leg in the stages A and D. W-phase leg 3 is the switched leg in the stages B and E. U-phase leg 1 is the switched leg in the stages C and F. U-phase leg 1 outputs U-phase voltage Vu. V-phase leg 2 outputs V-phase voltage Vv. W-phase leg 3 outputs W-phase voltage Vw. A voltage between two phase voltages selected among three phase voltages Vu, Vv and Vw is called the inter-phase voltage. The inter-phase voltage with the largest amplitude is called as the biggest inter-phase voltage Vx.

In the stage A, upper switch 11 of U-phase leg 1 and lower switch 32 of W-phase leg 3 are turned-on. The biggest inter-phase voltage Vx is applied to U-phase winding 10U and W-phase winding 10W. In the other words, U-phase leg 1 and W-phase leg 3 are the fixed legs in the stage A. V-phase leg 2 is the switched leg switched with the PWM-method. Switching patterns of each of legs 1-3 are shown in Figure 15. Stages A-F are decided in accordance with a rotor angle detected by the rotor angle sensor. Figure 15 shows six states of

switches

11, 12, 21, 22, 31 and 32 of inverter 4A in the stages A-F.

The gate voltage UU is applied to upper switch 11. The gate voltage UL is applied to lower switch 12. The gate voltage VU is applied to upper switch 21. Gate voltage VL is applied to lower switch 22. Gate voltage WU is applied to upper switch 31. The gate voltage WL is applied to lower switch 32. One of two switches of the switched leg has a duty ratio changing from 0% to 100% in the period of 60 electrical degrees excessively. The other one of two switches of the switched leg has the duty ratio changing from 100% to 0% in the period of the above 60 electrical degrees excessively.

The SPSM operation of DC/DC converter 8B is further explained referring to Figures 16 and 17. Figure 16 shows a three-phase sinusoidal wave voltage applied to three

phase windings

10U, 10V and 10W. Converter 8B outputs the biggest inter-phase voltage Vx to the inverter 4A. The controller calculates a duty ratio Dx of converter 8B in turn in order to output the biggest inter-phase voltage Vx with the sinusoidal waveform in accordance with the detected rotor angle and an instruction value of a motor torque. The controller calculates the PWM duty ratio Dy of the switched leg of inverter 4A in order to output the smaller inter-phase voltage Vy with the sinusoidal waveform in accordance with the detected rotor angle and the instruction value of the motor torque.

The biggest inter-phase voltage Vx is changed in each electrical angle of 60 electrical degrees as shown in Figure 16 in turn. In the stage A from 30 electrical degrees to 90 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vu-Vw. In the stage B from 90 electrical degrees to 150 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vu-Vv. In the stage C from 150 electrical degrees to 210 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vw-Vv. In the stage D from 210 electrical degrees to 270 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vw-Vu. In the stage E from 270 electrical degrees to 330 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vv-Vu. In the stage F from 330 electrical degrees to 30 electrical degrees, the biggest inter-phase voltage Vx is the inter-phase voltage Vv-Vw.

The biggest inter-phase voltage Vx has a waveform shown in Figure 17. The waveform of the biggest inter-phase voltage Vx is equal to the three-phase full-wave-rectified-waveform. When the biggest value of one phase-voltage is 1, the value of the biggest inter-phase voltage Vx is 1.5-1.73.

The biggest inter-phase voltage Vx has a part of sinusoidal waveform in each stage A-F. As the result, only one leg is PWM-switched for outputting the three-phase sin waveforms shown in Figure 16. A value of the smaller inter-phase voltage Vy alternatively changes from 0% to 100% and from 100% to 0% of a value of the biggest inter-phase voltage Vx.

The full bridge mode is explained referring to Figures 18-19. In this full bridge mode, DC/DC converter 8B outputs a boost voltage Vx' having a three-phase full-wave-rectified-waveform. The mode-changing switch 7 is turned on. As shown in Figure 18 and 19, one of three legs 1-3 of inverter 4A is the fixed leg, and the other two of three legs 1-3 of inverter 4A is the switched legs switched with the PWM method. Three legs 4-6 of inverter 4B are the fixed legs having the same operation as the inverter 4B shown in Figures 3-4, which shows the full bridge mode of the first embodiment. As the result, the power loss of the DTP inverter is reduced, because only two legs of the six legs of the DTP inverter are PWM-switched. This switching method is called the two-phase switching method, the TPSM.

Figures 18-19 are schematic diagrams showing six stages TA-TF shown in Figure 5. The stage TA is the period from 60 electrical degrees to 120 electrical degrees. The stage TB is the period from 120 electrical degrees to 180 electrical degrees. The stage TC is the period from 180 electrical degrees to 240 electrical degrees. The stage TD is the period from 240 electrical degrees to 300 electrical degrees. The stage TE is the period from 300 electrical degrees to 360 electrical degrees. The stage TF is the period from 360 electrical degrees to 60 electrical degrees.

In the two-phase switching method, the TPSM, of the full bridge mode, inverters 4A-4B apply a three-phase voltage to three

phase windings

10U, 10V and 10W. The DC/DC converter 8B changes a battery voltage Vb to a periodically-changed voltage Vx'. Converter 8B applies the voltage Vx' to three-

phase inverters

4A and 4B. Converter 8B periodically changes the amplitude of the voltage Vx'. The amplitude of voltage Vx' having the three-phase-full-wave-rectified-waveform becomes 0.865 at 0 electrical degrees, at 60 electrical degrees, at 120 electrical degrees, at electrical 180 degrees, at 240 electrical degrees and at 300 electrical degrees, if the amplitude of voltage Vx' becomes 1.0 at 30 electrical degrees, at 90 electrical degrees, at 150 degrees, at 210 electrical degrees, at 270 electrical degrees and at 330 electrical degrees.

The TPSM operation of

inverters

4A and 4B is explained referring to Figures 18-19. Figure 18 shows four switching stages TA-TD of

inverters

4A and 4B. Figure 19 shows two switching stages TE-TF of

inverters

4A and 4B. An electric angle of 360 electrical degrees is divided to the six stages TA-TF.

Six stages TA-TF shown in Figures 18-19 are essentially same as six stages TA-TF shown in Figures 3-4. In the stage TA, upper switch 11 is turned on, and lower switch 12 is turned off. Accordingly, the boosted voltage Vx' applied by converter 8B is applied to U-phase winding 10U directly. The boosted voltage Vx' has same waveform as the U-phase voltage Vu as shown in Figure 5. Another switches 21, 22, 31 and 32 of

legs

2 and 3 are PWM-switched in order to produce V-phase voltage Vv and W-phase voltage Vw shown in Figure 5 from boosted voltage Vx'.

In the stage TB, upper switch 21 is turned off, and lower switch 12 is turned on. Accordingly, boosted voltage Vx' is applied to V-phase winding 10V directly. Boosted voltage Vx' has same waveform as V-phase voltage Vv as shown in Figure 5.

Legs

1 and 3 are PWM-switched in order to produce U-phase voltage Vu and W-phase voltage Vw shown in Figure 5 from boosted voltage Vx'.

In the stage TC, upper switch 31 is turned on, and lower switch 32 is turned off. Accordingly, boosted voltage Vx' is applied to W-phase winding 10W directly. Boosted voltage Vx' has same waveform as W-phase voltage Vw as shown in Figure 5.

Legs

1 and 2 are PWM-switched in order to produce U-phase voltage Vu and V-phase voltage Vv shown in Figure 5 from boosted voltage Vx'.

In the stage TD, upper switch 11 is turned off, and lower switch 12 is turned on. Accordingly, boosted voltage Vx' is applied to U-phase winding 10U directly. Boosted voltage Vx' has same waveform as U-phase voltage Vu as shown in Figure 5.

Legs

2 and 3 are PWM-switched in order to produce V-phase voltage Vv and W-phase voltage Vw shown in Figure 5 from the boosted voltage Vx'.

In the stage TE, upper switch 21 is turned on, and lower switch 22 is turned off. Accordingly, boosted voltage Vx' is applied to V-phase winding 10V directly. Boosted voltage Vx' has same waveform as V-phase voltage Vv as shown in Figure 5.

Legs

In the stage TF, upper switch 31 is turned off, and lower switch 32 is turned on. Accordingly, boosted voltage Vx' is applied to W-phase winding 10W directly. Boosted voltage Vx' has same waveform as W-phase voltage Vw as shown in Figure 5.

Legs

1 and 2 are PWM-switched in order to produce U-phase voltage Vu and V-phase voltage Vv shown in Figure 5 from boosted voltage Vx'. In each of stages TA-TF, only two legs of inverter 4A are PWM-switched. The power loss of inverter 4A is reduced. Moreover, the power loss of inverter 4B is reduced, because inverter 4B is not PWM-switched in the full bridge mode.

(The third embodiment)
The third embodiment is explained referring to Figures 20-23. Figure 21 is a circuit diagram showing a conventional motor-driving apparatus for driving a three-phase EV traction motor. The boosted voltage Vx of the DC power supply 8 is applied to a conventional three-phase inverter 4. The inverter 4 applies three phase voltages Vu, Vv and Vw to three

phase windings

10U, 10V and 10W of the traction motor.

Figure 8 is a torque-speed diagram showing a torque-speed characteristic of the EV traction motor shown in Figure 21. A line 310 shows a torque-speed characteristic of the conventional traction motor. An operation area 303 surrounding by the line 310 is a conventional operation area of a conventional traction motor.

In this embodiment, the inverter 4 has

other operation areas

308 and 309. The operation area 308 has more torque than the conventional operation area 304. The operation area 309 has more speed than the conventional operation area 303. In the conventional operation area, the inverter 4 is driven with a well-known PWM method. In the

operation areas

308 and 309, the inverter 4 is driven with one-pulse-method instead of the PWM method explained above.

For example, inverter 4 is driven with the current-supplying method of the 120 electrical degrees type shown in Figure 22. Figure 23 shows a U-phase voltage Vu' of the one-pulse-method and a three-phase voltage Vu, Vv and Vw having sinusoidal waveforms. It is well-known that the one-pulse-method produces larger torque than the PWM-method under an equal voltage of the DC supply. But, the one-pulse-method produces larger vibration and a larger power loss, too. However, the high-torque area 308 and the high-speed area 309 are not operated frequently. As the result, it is acceptable to increasing of the vibration, the acoustic noise and the power loss. Consequently, the compact motor can have further wider torque-speed area.

Other arrangements are explained hereinafter. It is preferable that the resistance of the second inverter (4B) is 180%-510% of the resistance of the first inverter (4B). Accordingly, the production cost of the second inverter is reduced without increasing of the power loss of the second inverter. It is preferable that the first inverter consists of more semiconductor chips connected in parallel to each other than the second inverter. As the result, all of switch of both inverters can be made with same chips.

It is preferable that the full bridge mode is operated in the low speed range for a predetermined short period, when bigger motor torque is required. Accordingly, the motor torque is boosted in a short period. It is preferable that the first inverter is PWM-switched, and the second inverter (4B) is not PWM-switched. Accordingly, the power loss of the second inverter is reduced largely, even though the resistance of the second inverter is increased.

It is preferable that three legs (4-6) of the second inverter (4B) is switched in turn at each timing, when a new phase voltage of three phase voltages (Vu, Vv and Vw) becomes larger than another two phase voltages. Accordingly, the full bridge mode is operated without PWM-switching of the second inverter. It is preferable that the mode-changing switch (7) is switched, after three lower switches (42, 52 and 62) of the second inverter (4B) are turned off. Accordingly, the power loss of the mode-changing switch (7) is decreased. Particularly, a relay as the mode-changing switch (7) prevents the sparking.

Claims

A six-leg three-phase inverter for driving a variable-speed electric machine:
comprising a first inverter (4A), a second inverter (4B) and a mode-changing switch (7);
the first inverter (4A) having three phase legs (1-3) connects one ends of three phase windings (10U, 10V and 10W) of the variable-speed electric machine to a DC power supply (8);
the second inverter (4B) having three phase legs (4-6) connects the other ends of three phase windings (10U, 10V and 10W) to the DC power supply (8) via the mode-changing switch (7);
wherein the six-leg three-phase inverter has a star connection mode and a full bridge mode;
the second inverter (4B) constitutes a neutral point of a star-connected three phase windings (10U, 10V and 10W) in the star connection mode, when the mode-changing switch (7) is turned off;
the inverters (4A and 4B) constitutes three single-phase full bridges in the full bridge mode, when the mode-changing switch (7) is turned on; and
the star connection mode is operated in higher torque area (304) than an operating area (305) of the full bridge mode.
The six-leg three-phase inverter according to claim 2, wherein the first inverter (4A) is PWM-switched in both of the star connection mode and the full bridge mode, and the second inverter (4B) is not PWM-switched in both of the star connection mode and the full bridge mode.
The six-leg three-phase inverter according to claim 2, wherein the second inverter (4B) is switched at each timing when a new phase voltage of three phase voltages (Vu, Vv and Vw) becomes larger than another two phase voltages.
The six-leg three-phase inverter according to claim 1, wherein the star connection mode is operated in a higher torque area (304); and
the full bridge mode is operated in a lower torque area (304 and 305).
The six-leg three-phase inverter according to claim 1, wherein switches (41, 42, 51, 52, 61 and 62) of the second inverter (4B) has larger switching loss and smaller constant conductive loss than switches (11, 12, 21, 22, 31 and 32) of the first inverter (4A) in same condition.
The six-leg three-phase inverter according to claim 1, wherein each switch (11, 12, 21, 22, 31 and 32) of the first inverter (4A) consists of more semiconductor chips connected in parallel to each other than each switch (41, 42, 51, 52, 61 and 62) of the second inverter (4B).
The six-leg three-phase inverter according to claim 1, wherein the mode-changing switch (7) connects the negative DC terminal (103) of the second inverter (4B) to a negative terminal (105) of the DC power supply (8).
The six-leg three-phase inverter according to claim 7, wherein the mode-changing switch (7) is switched, after three lower switches (42, 52 and 62) of the second inverter (4B) are turned off.
The six-leg three-phase inverter according to claim 1, wherein the inverters (4A and 4B) are connected to the DC power supply (8A) via a boost DC/DC converter 8(B);
the boost DC/DC converter 8(B) and the first inverter (4A) are operated with a single-phase-switching-mode in the star connection mode;
the boost DC/DC converter 8(B) applies a biggest inter-phase voltage (Vx) to the first inverter (4A) in the single-phase-switching-mode;
one leg of the first inverter (4A) is a switched leg;
other two legs of the first inventor (4A) are fixed legs;
the switched leg is switched with a PWM method;
the fixed leg is not switched with a PWM method; and
one switched leg and two fixed legs of the first inverter (4A) are changed in turn in the single-phase-switching-mode.
The six-leg three-phase inverter according to claim 10, wherein the inverters (4A and 4B) are operated with a two-phase-switching-mode in the full bridge mode;
the boost DC/DC converter applies a biggest inter-phase voltage (Vx') to the inverters (4A and 4B) in the two-phase-switching-mode;
the first inverter (4A) has one switched leg and two fixed legs, which are changed in turn; and
the second inverter (4B) has three fixed legs;
The six-leg three-phase inverter according to claim 1, wherein both of the inverters (4A and 4B) are operated with a one-pulse-method instead of the PWM-switching method, when a required torque is larger than a torque produced by the PWM-method in the series connection mode.
The six-leg three-phase inverter according to claim 1, wherein a duty ratio of the inverter (4A) switched with the PWM-method is increased, just after when the star connection mode is started instead of the full bridge mode; and
the duty ratio of the inverter (4A) switched with the PWM-method is decreased, just after when the full bridge mode is started instead of the star connection mode.