DC-DC converter with snubber circuit

FIG. 2 is a circuit configuration diagram showing a DC-DC converter with a snubber circuit of Example 1. The DC-DC converter with the snubber circuit shown in FIG. 2 is formed of a multiphase, transformer-linked, step-up chopper circuit.

The DC-DC converter with the snubber circuit includes a DC power supply Vi, a transformer T 1 (first transformer), a transformer T 2 (second transformer), a reactor Lr 1 (first reactor), a reactor Lr 2 (second reactor), a reactor L 1 (third reactor), a switch Tr 1 (first switch), a switch Tr 2 (second switch), diodes D 1 , D 2 , D 3 and D 4 , a snubber diode Ds 1 (first snubber diode), a snubber diode Ds 2 (second snubber diode), a snubber resistor Rs, a snubber capacitor Cs, a smoothing capacitor Co and a control circuit 10 .

The transformer T 1 has a primary winding 1 a (n 1 turns), an additional winding 1 b (n 3 turns), and a secondary winding 1 c (n 2 turns) electromagnetically coupled to the primary winding 1 a . The transformer T 2 is configured in the same way as the transformer T 1 , hence having a primary winding 2 a (n 4 turns), an additional winding 2 b (n 6 turns), and a secondary winding 2 c (n 5 turns) electromagnetically coupled to the primary winding 2 a.

The primary winding 1 a of the transformer T 1 and the collector and emitter of the switch Tr 1 formed of an IGBT (insulated gate bipolar transistor) are connected to both ends of the DC power supply Vi. The primary winding 2 a of the transformer T 2 and the collector and emitter of the switch Tr 2 formed of an IGBT are connected to both ends of the DC power supply Vi.

A first series circuit, which includes the reactor Lr 1 , the additional winding 1 b of the transformer T 1 , the diode D 1 and the smoothing capacitor Co, is connected to both ends of the switch Tr 1 . A second series circuit, which includes the reactor Lr 2 , the additional winding 2 b of the transformer T 2 , the diode D 3 and the smoothing capacitor Co, is connected to both ends of the switch Tr 2 .

The diode D 2 is connected between one end (collector) of the switch Tr 1 and one end (positive terminal) of the smoothing capacitor Co. The diode D 4 is connected between one end (collector) of the switch Tr 2 and the one end (positive terminal) of the smoothing capacitor Co. The reactor L 1 is connected to both ends of a series circuit in which the secondary winding 1 c of the transformer T 1 and the secondary winding 2 c of the transformer T 2 are connected in series.

A third series circuit, which includes the snubber capacitor Cs and the snubber resistor Rs, is connected to both ends of the smoothing capacitor Co. The snubber diode Ds 1 is connected to a node at which the snubber capacitor Cs and the snubber resistor Rs are connected, and to a node at which the reactor Lr 1 and the additional winding 1 b of the transformer T 1 are connected. The snubber diode Ds 2 is connected to the node at which the snubber capacitor Cs and the snubber resistor Rs are connected, and to a node at which the reactor Lr 2 and the additional winding 2 b of the transformer T 2 are connected.

The snubber capacitor Cs, the snubber resistor Rs and the snubber diodes Ds 1 and Ds 2 constitute a snubber circuit.

Based on an output voltage Vo across the smoothing capacitor Co, the control circuit 10 performs control in such a way that it turns on the switch Tr 2 after the switch Tr 1 is turned on but before the switch Tr 1 is turned off, and that it turns on the switch Tr 1 before the switch Tr 2 is turned off. That is, in every ½ cycle, there is an overlapping period in which the switch Tr 1 and the switch Tr 2 are both on.

The transformer T 1 , the reactor Lr 1 , the diode D 1 , the diode D 2 and the switch Tr 1 constitute a first converter. The transformer T 2 , the reactor Lr 2 , the diode D 3 , the diode D 4 and the switch Tr 2 constitute a second converter.

According to the DC-DC converter with the snubber circuit configured as described above, the reactor Lr 1 is connected to the primary winding 1 a and the additional winding 1 b of the transformer T 1 , and the reactor Lr 2 is connected to the primary winding 2 a and the additional winding 2 a of the transformer T 2 . Accordingly, recovery losses of the diodes D 1 to D 4 as well as switching losses at the time of turning on the switches Tr 1 and Tr 2 can be suppressed.

In the following, described is the operation of the snubber circuit, which is a feature of the DC-DC converter with the snubber circuit of Example 1. The snubber circuit operates to demonstrate its surge-voltage suppressing effect mainly when the diode D 1 is turned off (same applies to the diode D 3 ), and also when the switch Tr 1 is turned off (same applied to the switch Tr 2 ). Thus, by referring to the drawings, description will be given of the operation at the time of turning off the diode D 1 and the operation at the time of turning off the switch Tr 1 .

First, the operation at the time of turning off the diode D 1 will be described using FIGS. 3 and 5 . Note that the operation at the time of turning off the diode D 3 is the same as the operation at the time of turning off the diode D 1 .

In FIG. 3 , the diode D 1 having been turned off to be in a reverse-blocking state is illustrated equivalently as a depletion-layer capacitor Cd 1 . In a mode M 1 shown in FIG. 3( a ), when the switch Tr 1 is turned on, an electric current D 1 i in the diode D 1 and an electric current Lr 1 i in the reactor Lr 1 decrease. In a mode M 2 shown in FIG. 3( b ), the diode D 1 is turned off, and a voltage D 1 v across the diode D 1 rises.

Next, in a mode M 3 shown in FIG. 3( c ), the snubber capacitor Cs having the snubber resistor Rs connected thereto in series is connected to an output voltage Vo in parallel. The snubber capacitor Cs is charged up to the output voltage Vo and is clamped at that voltage. Thus, no further discharge is performed. Accordingly, there is no power loss, which would otherwise occur along with unnecessary charge and discharge of the snubber capacitor Cs. The power loss occurring in this clamp-type snubber circuit is not associated with the snubber capacitor Cs. Moreover, there is no time required for charge and discharge. Hence, high-speed operation is possible.

Then, when the voltage Vd 1 across the capacitor Cd 1 with the diode D 1 turned off reaches or exceeds a voltage obtained by adding the output voltage Vo and the voltage across the additional winding 1 b (n 3 ) because of the resonance between the reactor Lr 1 and the capacitor Cd 1 , the potential of the cathode of the snubber diode Ds 1 becomes lower than that of the ground. This turns on the snubber diode Ds 1 and causes an electric current Ds 1 i to flow through the snubber diode Ds 1 . A resonant electric current flows and is charged into the snubber capacitor Cs, and a surge voltage is absorbed by the snubber circuit. The energy of the surge voltage absorbed by the snubber circuit is released to the ground side through the snubber resistor Rs. For this reason, resonance with a parasitic inductance of the wire at the time of the discharge of the snubber capacitor Cs is attenuated by the snubber resistor Rs, so that no noise is generated. Then, as the electric current Ds 1 i in the snubber diode Ds 1 becomes zero, the mode changes to a mode M 4 shown in FIG. 3( d ) whereby only the switch Tr 1 is on.

FIG. 4 shows a timing chart at the time of turning off the diode D 1 , in a case without the snubber circuit. As compared to the case without the snubber circuit in FIG. 4 , a case with the snubber circuit in FIG. 5 shows significant suppression of the surge voltage generated across the diode D 1 .

Next, by using FIGS. 6 and 8 , description will be given of an effect on the surge voltage which is generated across the switch Tr 1 due to the resonance between a parasitic inductance Lp of the wire generated when the switch Tr 1 is turned off, and the parasitic capacitance (unillustrated) of the switch Tr 1 .

Note that the operation at the time of turning off the switch Tr 2 is the same as the operation at the time of turning off the switch Tr 1 .

First, in a mode M 1 shown in FIG. 6( a ), the switch Tr 1 is turned on, and an electric current Tr 1 i flows through the switch Tr 1 . Next, in a mode M 2 shown in FIG. 6( b ), the switch Tr 1 is turned off, and a voltage Tr 1 v across the switch Tr 1 rises. Then, electric currents D 2 i and Csi flow through a path from Vi to Vi through 1 a (n 1 ), D 2 , Cs, and Rs. That is, the electric currents D 2 i and Csi flow into the snubber circuit, and a surge voltage is thus absorbed by the snubber circuit.

Next, in a mode M 3 shown in FIG. 6( c ), electric currents flow through a path from Vi to Vi through 1 a (n 1 ), D 2 , Cs, and Rs, a path from Vi to Vi through 1 a (n 1 ), D 2 , Lp, and Vo, and a path from Vi to Vi through 1 a (n 1 ), Lr 1 , 1 b (n 3 ), D 1 , and Vo.

In this case too, the energy of the surge voltage absorbed by the snubber circuit is released to the ground side through the snubber resistor Rs. For this reason, resonance with the parasitic inductance Lp of the wire at the time of discharge of the snubber capacitor Cs is attenuated by the snubber resistor Rs, so that no noise is generated.

FIG. 7 shows a timing chart at the time of turning off the switch Tr 1 in a case without the snubber circuit. As compared to the case without the snubber circuit in FIG. 7 , a case with the snubber circuit in FIG. 8 shows significant suppression of the surge voltage generated when the switch Tr 1 is turned off.

FIG. 9 shows the waveforms of the voltage across and the electric current in the diode D 1 in the case without the snubber circuit. FIG. 10 shows the waveforms of the voltage across and the electric current in the diode D 1 in the case with the snubber circuit. FIG. 11 shows the waveforms of the voltage across and the electric current in the switch Tr 1 in the case without the snubber circuit. FIG. 12 shows the waveforms of the voltage across and the electric current in the switch Tr 1 in the case with the snubber circuit. FIGS. 9 to 12 show the results obtained by operating the respective circuits in experiments.

A comparison between FIGS. 9 and 10 shows that the peak voltage, which is 464 V in the case without the snubber circuit, is suppressed down to 276 V by providing the snubber circuit, meaning that the peak voltage across the diode D 1 i is reduced by 88 V. A comparison between FIGS. 11 and 12 shows that the peak voltage, which is 320 V in the case without the snubber circuit, is suppressed down to 308 V by providing the snubber circuit.

FIG. 13 is a circuit configuration diagram showing a DC-DC converter with the snubber circuit of Example 2. In Example 1, the reactor Lr 1 is connected to the additional winding 1 b in series, and the reactor Lr 2 is connected to the additional winding 2 b in series. In contrast, Example 2 is characterized in that the reactor Lr 1 is connected to the switch Tr 1 in series, and the reactor Lr 2 is connected to the switch Tr 2 in series.

The configuration of the snubber circuit constituted of the snubber capacitor Cs, the snubber resistor Rs and the snubber diodes Ds 1 and Ds 2 is the same as the configuration in Example 1 shown in FIG. 2 , and thus description thereof will be omitted.

In the following, described is the operation and effect of the DC-DC converter with the snubber circuit having the reactors Lr 1 and Lr 2 connected to the respective switches Tr 1 and Tr 2 in series.

First, the switch Tr 1 is turned on by a gate signal of the switch Tr 1 from the control circuit 10 . Then, an electric current flows through a path from the positive side of Vi to the negative side of Vi through 1 a , Lr 1 , and Tr 1 . Accordingly, the electric current flowing through the primary winding 1 a of the transformer T 1 increases. At the same time, a voltage is generated also across the secondary winding 1 c of the transformer T 1 , and an electric current flows through the reactor L 1 by flowing through a path from 1 c to 1 c through 2 c and L 1 .

This electric current flows in accordance with the law of equal ampere-turns of transformers, causing energy to be stored in the reactor L 1 . At the same time, the same electric current flows through the secondary winding 2 c of the transformer T 2 . Accordingly, across the primary winding 2 a of the transformer T 2 and across the additional winding 2 b thereof, voltages corresponding to the respective numbers of turns are induced.

When an additional winding ratio of the transformer T 2 is A=(n 4 +n 6 )/n 4 , an electric current, which is 1/A of the electric current in the switch Tr 1 , flows through the diode D 3 by flowing through a path from the positive side of Vi to the negative side of Vi through 2 a , 2 b , D 3 , and Co. The electric current D 3 i in the diode D 3 flows until a time at which the switch Tr 2 is turned on. The output voltage Vo across the smoothing capacitor Co is the sum of a voltage across the DC power supply Vi (an input voltage), a voltage generated across the primary winding 2 a of the transformer T 2 , and a voltage generated across the additional winding 2 b of the transformer T 2 .

When a duty factor of the switch Tr 1 is D (D=Ton/T), the voltage generated across the transformer T 2 is A·Vi·D, where Ton is a period of time during which the switch Tr 1 is on, and T is a cycle in which the switch Tr 1 is switched. The output voltage Vo across the smoothing capacitor Co is Vo=Vi(1+A·D). Accordingly, the output voltage Vo can be controlled by changing the duty factor D.

Next, the switch Tr 2 is turned off by the gate signal from the control circuit 10 . The voltage across the collector and the emitter of the switch Tr 2 rises. Then, first, an electric current flows through a path from the positive side of Vi to the negative side of Vi through 2 a , Lr 2 , D 4 , and Co. Accordingly, an electric current flows through the diode D 4 .

However, due to the voltage applied across the additional winding 2 b of the transformer T 2 , the electric current in the reactor Lr 2 is commutated to the diode D 3 . Accordingly, the electric current flowing through the diode D 3 increases. Along with this, the electric current in the diode D 4 decreases gently. Once the electric currents in the primary winding 2 a and the additional winding 2 b of the transformer T 2 finish being commutated to the diode D 3 , the diode D 4 is turned off. Since the diode D 4 is turned off along with the gentle decrease of the electric current therein, the occurrence of a recovery loss at the diode D 4 is suppressed. Then, after the transformer's electric current is completely commutated to the diode D 3 , the electric current is in a state where it is passing only through the diode D 3 and outputted.

Next, the switch Tr 2 is turned on by the gate signal of the switch Tr 2 from the control circuit 10 . Then, the electric currents in the primary winding 2 a and the additional winding 2 b of the transformer T 2 start being commutated from the diode D 3 to the switch Tr 2 .

At this time, due to the reactor Lr 2 , the electric current in the switch Tr 2 increases gently. Thereby, a zero-current turn on operation can be achieved. Along with this, the electric current in the diode D 3 decreases gently as well, hence suppressing the occurrence of a recovery loss at the time of turn off.

The electric current flows through a path from the positive side of Vi to the negative side of Vi through 2 a , Lr 2 , and Tr 2 . Accordingly, the electric current flowing through the primary winding 2 a of the transformer T 2 increases. At the same time, a voltage is also generated across the secondary winding 2 c of the transformer T 2 , and an electric current flows through the reactor L 1 by flowing through a path from 2 c to 2 c through L 1 and 1 c.

This electric current flows in accordance with the law of equal ampere-turns of transformers, causing energy to be stored in the reactor L 1 . At the same time, the same electric current flows through the secondary winding 1 c of the transformer T 1 . Accordingly, across the primary winding 1 a of the transformer T 1 and across the additional winding 1 b thereof, voltages corresponding to the respective numbers of turns are induced.

When an additional winding ratio of the transformer T 1 is A=(n 1 +n 3 )/n 1 , an electric current, which is 1/A of the electric current in the switch Tr 2 , flows through the diode D 1 by flowing through a path from the positive side of Vi to the negative side of Vi through 1 a , 1 b , D 1 , and Co. The electric current in the diode D 1 flows until a time at which the switch Tr 1 is turned on. The output voltage Vo across the smoothing capacitor Co is the sum of a voltage across the DC power supply Vi (an input voltage), a voltage generated across the primary winding 1 a of the transformer T 1 , and a voltage generated across the additional winding 1 b of the transformer T 1 .

When a duty factor of the switch Tr 2 is D (D=Ton/T), the voltage generated across the transformer T 1 is A·Vi·D, where Ton is a period of time during which the switch Tr 2 is on, and T is a cycle in which the switch Tr 2 is switched. The output voltage Vo across the smoothing capacitor Co is Vo=Vi(1+A·D). Accordingly, the output voltage Vo can be controlled by changing the duty factor D.

Next, the switch Tr 1 is turned off by the gate signal from the control circuit 10 . The voltage across the collector and the emitter of the switch Tr 1 rises. Then, first, an electric current flows through a path from the positive side of Vi to the negative side of Vi, 1 a , Lr 1 , D 2 , and Co. Accordingly, an electric current flows through the diode D 2 .

However, due to the voltage applied across the additional winding 1 b of the transformer T 1 , the electric current in the reactor Lr 1 is commutated to the diode D 1 . Accordingly, the electric current flowing through the diode D 1 increases. Along with this, the electric current in the diode D 2 decreases gently. Once the electric currents in the primary winding 1 a and the additional winding 1 b of the transformer T 1 finish being commutated to the diode D 1 , the diode D 2 is turned off. Since the diode D 2 is turned off along with the gentle decrease of the electric current therein, the occurrence of a recovery loss at the diode D 2 is suppressed. Then, after the transformer's electric current is completely commutated to the diode D 1 , the electric current is in a state where it is passing only through the diode D 1 and outputted.

Next, when the switch Tr 1 is turned on, the electric currents in the primary winding 1 a and the additional winding 1 b of the transformer T 1 start being commutated from the diode D 1 to the switch Tr 1 .

At this time, due to the reactor Lr 1 , the electric current in the switch Tr 1 increases gently. Thereby, a zero-current turn on operation can be achieved. Along with this, the electric current in the diode D 1 decreases gently as well, hence suppressing the occurrence of a recovery loss at the time of turn off.

Meanwhile, like Example 1, the snubber circuit constituted of the snubber capacitor Cs, the snubber resistor Rs, the snubber diode Ds 1 and the snubber diode Ds 2 is provided in Example 2. Thus, the operation in Example 2 is performed substantially similarly to the operation shown in FIGS. 6( a ) to 6 ( d ) in Example 1. As a result, Example 2 can achieve advantageous effects similar to the advantageous effects of Example 1.

That is, it is possible to reduce a surge voltage and suppress noise generation to thereby prevent elements such as the switches from being broken. Moreover, since the voltage across the snubber capacitor Cs is clamped at the output voltage, no further discharge is performed. Accordingly, there is no power loss, which would otherwise occur along with unnecessary charge and discharge of the snubber capacitor Cs. The power loss occurring in this clamp-type snubber circuit is not associated with the snubber capacitor Cs. Moreover, there is no time required for charge and discharge. Hence, high-speed operation is possible.