JPWO2020017163A1 - Switching power supply - Google Patents

Abstract

In the LLC type switching power supply, a feedback value indicating a load condition is supplied and has a control unit that forms a drive signal for the switching element. The control unit has a switching frequency according to the feedback value in a first region where the load is heavy. In the second region where the load is lighter than that in the first region, burst control is performed in which the switching frequency is fixed and a switching ON section and a switching OFF section are provided. This is a switching power supply in which the ON time ratio is continuously changed according to the load condition by controlling both the number of switching ON times and the OFF time during the burst cycle. FIG. 15

Description

The present technology relates to an LLC type switching power supply.

Conventionally, an LLC type switching power supply (DC-DC converter) that uses two Ls and one C is known. Such a switching power supply is a soft switching system, has high efficiency and low noise, and is widely used. On the other hand, since it has a characteristic that the regulation range is narrower than that of other methods, it is not suitable for applications where the output voltage variable range is wide or where the input voltage fluctuation is large.

In the case of a charger for charging a secondary battery, since the output voltage fluctuation range is wide, a PWM control type switching power supply has been conventionally used. Recently, there have been many demands for larger capacities (several hundred watts or more), and using a conventional PWM control type switching power supply is inefficient and disadvantageous in terms of size and cost. Therefore, if a charger can be realized by an LLC switching power supply, a low-cost and highly efficient charger can be realized. However, as described above, due to the characteristic that the regulation range of the LLC type switching power supply is narrow, there is a problem in the behavior of the charger in the light load (low voltage and small current) region.

In the LLC type switching power supply, when the load becomes small, the ratio of the exciting current to the current output on the secondary side becomes large, and the efficiency decreases. For this reason, there arises a problem that the power consumption of the electronic device during standby becomes large. That is, in addition to the current that becomes the energy to be transmitted to the secondary side, the exciting current that flows only on the primary side due to resonance flows. The exciting current due to this resonance continues to flow regardless of the current consumed by the load. Therefore, when the load is light, the decrease in efficiency due to the exciting current due to resonance becomes relatively large.

In Patent Document 1, in order to solve such a problem, it is possible to set a normal mode in which the oscillator is continuously operated to control the power supply and a burst mode in which the oscillator is intermittently operated to control the power supply. Are listed. When the burst mode is set, the power supply control is intermittently stopped by detecting the output voltage on the secondary side, so that the power consumption during standby can be reduced.

Further, Patent Document 2 describes that the power consumption of the switching power supply is further reduced. In Patent Document 2, paying attention to the fact that the switching control of the switch element is unnecessary during the switching stop period, the supply of the control power to the control means is stopped to reduce the power consumption.

Japanese Unexamined Patent Publication No. 2009-189108 Japanese Unexamined Patent Publication No. 2013-038857

As described in Patent Document 1 or Patent Document 2, it is possible to widen the regulation range of the light load region of the LLC type switching power supply by using the burst mode (intermittent oscillation mode) at the time of light load. The one described in Patent Document 1 or Patent Document 2 is a burst control technique for suppressing power consumption during standby, and a burst frequency of about several tens of Hz to several hundreds of Hz is usually used.

However, in the case of a low burst frequency of several tens of Hz to several hundreds of Hz, the ripple current or ripple voltage of the output becomes large because the switching is stopped for a long time. In the case of a charger, there is a demand to suppress the ripple current for charging even when the load is light, and at a low burst frequency, the ripple current is large, so that the required specifications of the battery may not be satisfied. The ripple current can be reduced by increasing the burst frequency to several tens of kHz, but simply increasing the burst frequency makes it difficult to fine-tune the ON time ratio during the burst period, making it difficult to operate stably. There is a problem of becoming.

Therefore, an object of the present technology is to provide a switching power supply capable of reducing the ripple current (or ripple voltage) during burst operation.

This technology is an LLC switching power supply.
It has a control unit that is supplied with a feedback value indicating the load condition and forms a drive signal for the switching element.
The control unit
In the first region where the load is heavy, frequency control is performed to change the switching frequency according to the feedback value.
In the second region where the load is lighter than that of the first region, burst control is performed by fixing the switching frequency and providing a switching ON section and a switching OFF section.
In burst control, it is a switching power supply in which the ON time ratio is continuously changed according to the load condition by controlling both the switching ON frequency and the OFF time.

According to at least one embodiment, the ON time ratio of the burst period can be continuously changed, and the OFF time can always be controlled to the minimum while ensuring the required minimum OFF time at the time of burst. Therefore, the ripple current (or ripple voltage) during burst operation can be minimized. The effects described here are not necessarily limited, and may be any of the effects described in the present technology or an effect different from them.

FIG. 1 is a connection diagram of an LLC type switching power supply. FIG. 2 is a block diagram showing a configuration of constant voltage control. FIG. 3 is a block diagram showing a configuration of constant current control. FIG. 4 is a block diagram showing a configuration in which both constant voltage control and constant current control are performed. FIG. 5 is a connection diagram showing a configuration of constant voltage control. FIG. 6 is a connection diagram showing a configuration of constant current control. FIG. 7 is a waveform diagram showing a waveform of a drive signal of an LLC type switching power supply. FIG. 8 is a waveform diagram showing a waveform of a drive signal in a burst mode of an LLC type switching power supply. FIG. 9 is a waveform diagram showing the relationship between the length of the OFF section and the magnitude of the ripple. FIG. 10 is a waveform diagram showing the length of the OFF section. FIG. 11 is a waveform diagram showing the relationship between the switching frequency and the magnitude of the ripple. FIG. 12 is a timing chart showing the relationship between the load and the ON time ratio. FIG. 13 is a timing chart showing the relationship between the load and the ON time ratio. FIG. 14 is a timing chart for explaining the change in the ON time ratio when changing from the non-burst mode to the burst mode. FIG. 15 is a timing chart used for explaining the burst mode according to the present technology. FIG. 16 is a timing chart for explaining control when the ON time ratio is high. FIG. 17 is a timing chart for explaining control when the ON time ratio is low. FIG. 18 is a timing chart for explaining control when the ON time ratio is high. FIG. 19 is a timing chart for explaining control when the ON time ratio is low. FIG. 20 is a flowchart for explaining the control of the normal mode. FIG. 21 is a flowchart for explaining control in the case of the burst mode (switching number n ≧ 2). FIG. 22 is a flowchart for explaining control in the case of the burst mode (switching number n = 1). FIG. 23 is a flowchart for explaining the control of the burst mode when the table is used. FIG. 24 is a diagram showing an example of a table for controlling the burst mode. FIG. 25 is a diagram showing an example of a table for controlling the burst mode. FIG. 26 is a graph showing the relationship between the ON time ratio and the burst frequency. FIG. 27 is a graph showing the relationship between the ON time ratio and the OFF time. FIG. 28 is a graph showing the relationship between the ON time ratio and the number of ON times. FIG. 29 is a timing chart for explaining the soft start and the soft end. FIG. 30 is a timing chart for explaining soft start and soft end. FIG. 31 is a timing chart used for explaining control when the soft start and the soft end are not used. FIG. 32 is a timing chart for explaining a modification of the burst mode according to the present technology.

Hereinafter, embodiments and the like of the present technology will be described with reference to the drawings.
The embodiments described below are suitable specific examples of the present technology, and the contents of the present technology are not limited to these embodiments.

FIG. 1 shows an example configuration of an LLC type switching power supply to which the present invention is applied. The configuration of FIG. 1 specifies a parasitic element. Vin is an input power source, Q1 is a high-side MOSFET, and Q2 is a low-side MOSFET. A diode D1 and a capacitance C1 exist in parallel as parasitic elements between the drain and the source of the MOSFET Q1. A diode D2 and a capacitance C2 exist in parallel as parasitic elements between the drain and the source of the MOSFET Q2. A drive signal is supplied from the control unit to each gate of the MOSFET Q1 and the MOSFET Q2, and the MOSFET Q1 and the MOSFET Q2 perform a switching operation.

The inductance L0, the primary coil L1 of the transformer TR, and the capacitance C3 are connected in series between the connection point of the source of the MOSFET Q1 and the drain of the MOSFET Q2 and the source of the MOSFET Q2. The secondary coil of the transformer TR is divided into two inductances L2a and L2b, one end of the secondary coil is connected to the output terminal t1 via the diode D3a, and the other end of the secondary coil is connected to the output terminal t1 via the diode D3b. Is connected with. The connection midpoint of the secondary coil is taken out as the output terminal t2, and the capacitance C4 is connected between the output terminals t1 and t2. The output voltage Vout is taken out from the output terminals t1 and t2.

In the above-mentioned LLC switching power supply, drive signals having opposite phases are supplied to the gates of MOSFET Q1 and MOSFET Q2, and these MOSFETs Q1 and Q2 perform a differential switching operation.

The LLC switching power supply usually outputs a constant voltage, and the output voltage is controlled to a constant value by feedback control. This is generally called constant voltage control or CV control. The structure of the feedback is as shown in FIG. The output voltage (or its partial pressure value) and the reference voltage are input to the error amplifier 11, and a feedback (denoted as FB in the figure) signal is formed at the output of the error amplifier 11. This feedback signal is supplied to the control unit 12. When the output and the control unit 12 are insulated, the feedback signal is supplied to the control unit 12 through an insulating element 13 such as a photocoupler. The drive signal output to the switching elements (MOSFETs Q1 and Q2) is obtained from the control unit 12. The output voltage is controlled by the drive signal output. In an error amplifier, negative feedback is applied so that the two inputs (output voltage or its divided voltage value and reference voltage) are equal, and as a result, the output voltage can be controlled to be constant. it can. Control methods include burst control, frequency control, dead time control, and the like.

When an LLC switching power supply is used for a charger, as shown in FIG. 3, the output current is usually converted into a voltage value by the IV conversion amplifier 24 and input to the error amplifier 21. In the error amplifier 21, it is compared with the reference voltage corresponding to the control current value. The output of the error amplifier is supplied to the control unit 12. Since the error amplifier is subjected to negative feedback so that the two inputs are equal, the output current is controlled to a constant value by the feedback control. This is generally called constant current control or CC control. The operation itself of the control unit 12 is the same as that of the control unit 12 in the constant voltage (CV) control.

For example, in the case of a lithium ion secondary battery, constant current constant voltage charging is performed, so that constant current (CC) control and constant voltage (CV) control are used in combination. As shown in FIG. 4, the output of the error amplifier 11 and the output of the error amplifier 21 are added via the diodes 22 and 23 to form a feedback signal. The control unit 12 is controlled by this feedback signal.

FIG. 5 shows a more specific configuration of an LLC type switching power supply that performs constant voltage control. The configuration of FIG. 5 specifies a parasitic element. Vin is an input power source, Q1 is a high-side MOSFET, and Q2 is a low-side MOSFET. A diode D1 and a capacitance C1 exist in parallel as parasitic elements between the drain and the source of the MOSFET Q1. A diode D2 and a capacitance C2 exist in parallel as parasitic elements between the drain and the source of the MOSFET Q2. Drive signals H-DRV and L-DRV are supplied to the respective gates of MOSFET Q1 and MOSFET Q2, and MOSFET Q1 and MOSFET Q2 perform a switching operation.

The inductance L0, the primary coil L1 of the transformer TR, and the capacitance C3 are connected in series between the connection point of the source of the MOSFET Q1 and the drain of the MOSFET Q2 and the source of the MOSFET Q2. The secondary coil of the transformer TR is divided into two inductances L2a and L2b, one end of the secondary coil is connected to the output terminal t1 via the diode D3a, and the other end of the secondary coil is connected to the output terminal t1 via the diode D3b. Is connected with. The connection midpoint of the secondary coil is taken out as the output terminal t2, and the capacitance C4 is connected between the output terminals t1 and t2. The output voltage for the load 10 (for example, a lithium ion secondary battery) is taken out from the output terminals t1 and t2. In the above-mentioned LLC type switching power supply, drive signals H-DRV and L-DRV having opposite phases are supplied to the gates of MOSFET Q1 and MOSFET Q2, and these MOSFETs Q1 and Q2 perform a differential switching operation.

The output voltage is divided by the resistors R1 and R2, the divided voltage is input to the error amplifier 11, compared with the reference voltage, and negative feedback (negative feedback) is applied so that these are equal values. The feedback signal from the error amplifier 11 is supplied to the control unit 12 through the photocoupler 13. The output unit 15 is connected to the control unit 12, and the drive signals H-DRV and L-DRV for the MOSFETs Q1 and Q2 are output from the output unit 15.

FIG. 6 shows a specific configuration of an LLC switching power supply that performs constant current control. The configuration of the switching power supply is the same as in the case of constant voltage control. The output current is detected by the detection resistor R0 and is supplied to the error amplifier 21 via the current amplifier 16. The error amplifier 21 compares the control current values, and negative feedback is applied so that these values are equal. The feedback signal from the error amplifier 21 is supplied to the control unit 12 through the photocoupler 13. The output unit 15 is connected to the control unit 12, and the drive signals H-DRV and L-DRV for the MOSFETs Q1 and Q2 are output from the output unit 15.

"Drive signal in LLC switching power supply"
In the LLC type switching power supply, the drive signals H-DRV and L-DRV are shown in FIG. These drive signals are anti-phase pulses. The MOSFET turns on during the high level period of the drive signal. One cycle of the drive signals H-DRV and L-DRV is referred to as a switching cycle. Further, one cycle of the switching cycle is defined as one switching frequency. The reciprocal of the switching cycle is the switching frequency, and in the case of the LLC method, constant voltage control or constant current control is possible by changing the switching frequency by feedback control.

As shown in FIG. 8, the burst mode (intermittent oscillation mode) is a mode in which there is a switching ON section and a switching OFF section. The switching ON section and the switching OFF section are collectively called the burst period, and the reciprocal of the burst frequency is the burst frequency. In the burst mode, constant current control or constant voltage control is performed according to the time ratio between the switching ON section and the switching OFF section.

The ripple current (ripple voltage) can be reduced as the switching OFF section is shorter. Comparing the waveform of the upper drive signal and the waveform of the lower drive signal in FIG. 9, the lower waveform has a shorter OFF section. As a result, the shorter the OFF section of the output ripple current (or ripple voltage), the smaller the value.

From the viewpoint of minimizing the ripple current (or ripple voltage), the shorter the switching OFF section is, the better. However, in the LLC method, the switching OFF section has a required minimum OFF time. This is because there is a period during which current flows through the body diodes (diodes D1 and D2) of the MOSFET even when switching is turned off. If switching is started during this period, half-bridge will occur during the reverse recovery period of the body diode. A through current will flow through the diode, which is not preferable. The minimum required OFF time is slightly different depending on the exciting current of the circuit and the load condition, but as shown in FIG. 10, it is about one switching cycle from the result of the experiment.

Further, in the case of the same number of switching times, the higher the switching frequency, the shorter the switching OFF time can be. The switching frequency of the lower waveform is higher than that of the upper waveform of FIG. Since the switching OFF time can be shortened as the switching frequency is higher, the output ripple current (output ripple voltage) can be reduced.

"Existing burst control method"
An existing method of burst control in such an LLC type switching power supply will be described with reference to FIG. The basic control method of burst is to adjust and control the switching ON time ratio at the time of burst. When the load becomes lighter, the ratio of switching ON time is lowered. In the case of an LLC switching power supply, one switching results in "ON of high-side MOSFET Q1 and OFF of low-side MOSFET Q2" → "OFF of high-side MOSFET Q1 and ON of low-side MOSFET Q2". Since the ON time is adjusted by the number of switchings, the ON time becomes "switching cycle × number of ONs" depending on the number of switchings, and takes discrete values (see FIG. 7).

On the other hand, the OFF time has a minimum required OFF time to be secured, which is approximately one switching cycle (see FIG. 10). As explained with reference to FIG. 9, when the OFF time is lengthened, the ripple current increases. Therefore, in order to suppress the ripple current, it is better to shorten the OFF time as much as possible while ensuring the necessary OFF period. ..

Further, the ON time ratio at the time of burst must be changed as continuously as possible according to the load, or stable regulation characteristics cannot be obtained. Since the ON time at the time of burst becomes a discrete value according to the number of switchings, if the burst frequency is fixed in a high frequency burst, the ON time ratio fluctuates greatly with a one-step change in the number of ONs. , It is not possible to finely control the ON time ratio due to load fluctuations, and stable operation cannot be obtained.

An example of a simple high-frequency burst (fixed burst frequency) (see FIG. 13)
For example, switching count 3 times, OFF = switching 1 cycle (ON time ratio in this case is 0.75) → (1 step down when the load becomes lighter) → switching count 2 times, OFF = switching 2 cycles (in this case) ON time ratio is 0.5)
If the burst frequency is fixed in a high-frequency burst, the ON time will change significantly with one-step adjustment, and there will be a problem that stable operation will not be performed and ripple will increase.

In addition, if the burst frequency is fixed in a high-frequency burst, the ON time ratio fluctuates greatly at the boundary from the non-burst operation to the burst operation, and stable operation is performed under load conditions corresponding to this boundary. You will not be able to.

An example of a simple high-frequency burst (fixed burst frequency) (see FIG. 14)
For example, when burst cycle = 4 times the switching cycle (ON time ratio in non-burst operation is 1.0) → (when the load becomes lighter) → switching count 3 times, OFF = switching 1 cycle (ON in this case) Time ratio is 0.75)
In the case of a high-frequency burst with a fixed burst frequency, there is a large jump in the ON time ratio between the state in which the burst is entered in the minimum OFF time and the non-burst operation, and stable operation is not performed under load conditions corresponding to this boundary.

"Burst control method using this technology"
In the present technology, in the burst control of the LLC method, the number of switching ON times and the OFF time of the burst period are controlled so that the ON time ratio of the burst period can be continuously changed according to the load condition. By continuously controlling the OFF time (steps finer than one switching cycle, stepless, etc.) by this control method, the ON time ratio during burst operation can be continuously changed, and the burst can be changed. Optimal control of the OFF time becomes possible. Hereinafter, the burst mode by the present technology will be described. In addition, "continuous" is expressed as continuous including variable in relatively small steps that do not make a big leap or variable in steps.

The LLC method is a frequency control method. The lighter the load, the higher the switching frequency. An upper limit set value (fmax1) is set for this switching frequency, and for a light load region higher than that, the upper limit switching frequency (fmax1) is fixed, the burst control is entered, and the control is performed according to the ratio of the switching ON time. ..

FIG. 15 shows the burst control of the present technology, and shows the control from a heavy load to an ultra-light load. In the present technology, the upper limit value fmax1 of the switching frequency for frequency control is set. Up to the upper limit value fmax1, the switching frequency fsw is controlled according to the feedback value (denoted as FB value) indicating the weight of the load. That is, when the load becomes lighter, the switching frequency fsw is increased. This control is in the range of non-burst control. As an example, fmax1 is set to less than 150 kHz. This frequency is below the regulation band of the noise terminal voltage, and the cost of the AC filter can be reduced.

When the switching frequency fsw reaches the upper limit set value fmax1, the process shifts to burst control. The switching frequency fsw is fixed to the upper limit set value fmax1, and the number of switching times and the OFF time are controlled by the FB value. The lighter the load, the less the number of ONs. Further, the burst control by the present technology is divided into a case of (switching number ≥ 2) and a case of (switching number = 1) during the burst period.

In the case of (switching number ≥ 2) (ON time ratio ≥ 0.5), as shown in FIGS. 16 and 17, the ON time ratio is adjusted by controlling the switching number and the OFF time by feedback. To do. The lighter the load, the less the number of ONs. It is expressed as ON time = switching cycle x number of switchings. The OFF time in FIGS. 16 and 17 shows an optimum value of 1 switching cycle or more and less than 2 switching cycles.

In the case of (switching number = 1), the OFF time is controlled as shown in FIGS. 18 and 19. The lighter the load, the longer the OFF time. As shown in FIG. 18, since the ON time ratio is close to 0.5, the number of switchings is reduced. The OFF time in FIG. 18 shows an optimum value of 1 switching cycle or more and less than 2 switching cycles.

In the burst control in the present technology described above, the points when the number of switching times ≥ 2 are as follows.
The higher the ON time ratio (closer to 1), the larger the number of switchings during the burst period is controlled. The OFF time is controlled to an optimum value of, for example, the required minimum OFF time (= about 1 switching cycle) or more and less than 2 switching cycles. As a result, the burst frequency is low.

The lower the ON time ratio (closer to 0.5), the less the number of switchings during the burst period is controlled. The OFF time is controlled to an optimum value of, for example, the required minimum OFF time (= about 1 switching cycle) or more and less than 2 switching cycles. As a result, the burst frequency is high. As the load becomes lighter, the number of switchings during the burst period is reduced, and when the ON time ratio is 0.5, the number of switchings = 1 and the OFF time = 1 switching cycle (= minimum OFF time).

As described above, ON time = switching cycle x number of switchings, and takes discrete values, but by controlling the OFF time continuously (steps or steps that are somewhat finer than one switching cycle), the ON time The ratio can be continuously fine-tuned.

In the burst control in the present technology described above, when the number of switching times is 1, the number of times of switching is fixed at 1, and the OFF time is controlled at the required minimum OFF time (= about 1 cycle of switching) or more to set the ON time ratio. adjust.

When the number of switching times = 1 (ON time ratio <0.5) and the ON time ratio is high (when the ON time ratio is close to 0.5), the OFF time becomes the minimum OFF time as shown in FIG. It will be close.

When the number of switching times = 1 and the ON time ratio is low (when the ON time ratio is close to 0), the OFF time becomes an optimum value of one switching cycle or more as shown in FIG.

In the burst control in the present technology described above, the points when the number of switching times = 1 are as follows.
The number of switchings is fixed to one, and the OFF time is controlled to an optimum value equal to or larger than the required minimum OFF time (= about 1 switching cycle).

When the ON time ratio is high (close to 0.5), the OFF time is close to the required minimum OFF time (= about 1 switching cycle), and as a result, the burst frequency is high.

When the ON time ratio is low (close to 0), the OFF time becomes long, and as a result, the burst frequency becomes low.

By continuously controlling the OFF time (steps finer than one switching cycle, stepless, etc.), the ON time ratio can be continuously finely adjusted.

From the viewpoint of minimizing the ripple current, it is preferable that the OFF time is short even when the number of switching times is 1. The method of shortening the OFF time when the number of switching times = 1 will be described later.

"Explanation of feedback control in normal mode (frequency control)"
The control operation of the control unit 12 will be described. An example of the feedback control in the normal mode will be described with reference to FIG. In this example, it is assumed that the upper limit value and the lower limit value are set for the switching frequency fsw.

Step S1: It is determined whether or not the value of the feedback signal (FB value) is high. Here, a high FB value means that the output is insufficient.
Step S2: When it is determined that the FB value is high, it is determined whether or not the switching frequency is higher than the lower limit value.
Step S3: If it is determined in step S2 that the switching frequency is higher than the lower limit value, the switching frequency is lowered. Then, the process returns to the determination process of step S1.
Step S4: If it is determined in step S2 that the switching frequency is equal to or lower than the lower limit value, the switching frequency is operated at the lower limit value. Then, the process returns to the determination process of step S1.

Step S5: In step S1, if the FB value is not high, that is, if it is determined that the output is excessive, it is determined whether the switching frequency is less than the upper limit value.
Step S6: If it is determined in step S5 that the switching frequency is not less than the upper limit value, the burst mode is set.
Step S7: If it is determined in step S5 that the switching frequency is less than the upper limit value, the switching frequency is increased and the process returns to the determination process in step S1.

"Explanation of burst mode (switching count n ≧ 2)"
Next, an example of feedback control in the burst mode (switching number n ≧ 2) will be described with reference to FIG. In this example, the upper limit is set for the number of switchings in one burst cycle.
Step S11: Determine whether the FB value is high. Here, a high FB value means that the output is insufficient.
Step S12: When it is determined that the FB value is high, it is determined whether or not the OFF time is the lower limit value.
Step S13: When the OFF time is determined to be the lower limit value, it is determined whether the number of switchings is the upper limit value.
Step S14: If it is determined in step S13 that the number of switchings is not the upper limit value, the number of switchings is increased. Then, the process returns to the determination process of step S1.
Step S15: When the number of switchings is determined to be the upper limit value in step S13, the frequency control (continuous mode) is started.

Step S16: In step S12, if it is determined that the OFF time is not the lower limit value, the OFF time is reduced by the lower limit value or more. Then, the control returns to the FB value determination (step S11).
Step S17: In the FB value determination in step S11, if it is determined that the FB value is not high (excessive output), it is determined whether or not the OFF time is the upper limit value. That is, it is determined whether or not the OFF time is (<(T · n) / (n-1)). Here, T indicates a switching cycle, and n indicates the number of switchings in one burst cycle.
Step S18: In step S17, if it is determined that the OFF time is not the upper limit value, the OFF time is increased below the upper limit. Then, the control returns to the FB value determination (step S11).

Step S19: In step S17, when the OFF time is determined to be the upper limit value, it is determined whether or not the number of switchings is more than 2.
Step S20: In step S19, if it is determined that the number of switchings is more than 2, the number of switchings is reduced. Then, the control returns to the FB value determination (step S11).
Step S21: In step S19, when it is determined that the number of switchings is 2 or less, the control shifts to the mode in which the number of switchings = 1.

"Explanation of burst mode (switching number n = 1)"
Next, the feedback control of the burst mode (switching number n = 1) will be described with reference to FIG.
Step S31: Determine whether the FB value is high. Here, a high FB value means that the output is insufficient.
Step S32: When it is determined that the FB value is high, it is determined whether or not the OFF time is the lower limit value.
Step S33: If it is determined that the OFF time is not the lower limit value, the OFF time is reduced below the lower limit value. Then, the process returns to the determination process of the FB value in step S31.
Step S34: In step S32, when it is determined that the OFF time is the lower limit value, the number of switchings is set to 2 and the control proceeds to (n ≧ 2).
Step S35: If it is determined in step S31 that the FB value is not high (that is, the output is excessive), the OFF time is increased and the process returns to the FB value determination process.

"Burst mode using table"
The switching frequency is fixed at the upper limit, and the OFF time is T or more. However, T is a switching cycle.
As shown in the flowchart of FIG. 23, when a table is used, as will be described later, a table of the number of switching times and the OFF time according to the ON time ratio is prepared, and the ON time ratio is set according to the FB value. Change.

Step S41: Determine whether the FB value is high. Here, a high FB value means that the output is insufficient.
Step S42: When it is determined that the FB value is high, it is determined whether or not the ON time ratio of the table is less than the upper limit.
Step S43: If it is determined that the ON time ratio is less than the upper limit, the ON time ratio is increased. Then, the process returns to the determination process in step S41.
Step S44: If it is determined in step S42 that the ON time ratio is not less than the upper limit, the frequency control mode (continuous operation) is entered.
Step S45: In step S41, if it is determined that the FB value is not high, the ON time ratio is lowered.

"Example of table"
An example of changing the ON time ratio, the number of ON times, and the OFF time of the burst is shown in FIGS. 24 and 25 in the form of a table. These two tables are a series of tables that follow from the table of FIG. 24 to the table of FIG. 25, and the load is assumed to be lighter from the top to the bottom of the table. That is, the top row in FIG. 24 is the value when the load is the heaviest, and the bottom row in FIG. 25 is the value when the load is the lightest. As can be seen from the examples of FIGS. 24 and 25, the number of ONs is reduced as the load becomes lighter. By adjusting the OFF time in steps finer than one switching cycle or in a stepless manner, the leap in the ON time ratio can be eliminated and stable regulation characteristics can be realized. The OFF time is expressed in 0.1 increments for convenience, but in reality, it may not be in 0.1 increments or may be stepless.

Further, the points of the tables shown in FIGS. 24 and 25 will be described.
This table controls the number of switching times and the OFF time in one burst cycle, and by controlling the OFF time in steps smaller than one switching cycle to some extent, the ON time ratio during the burst period can be changed without jumping. It shows that you can do it.

When adjusting the ON time ratio, by controlling the number of switching times and the OFF time in combination, the control is shown in which the number of switching times is reduced rather than simply increasing the OFF time. By optimally controlling the OFF time, the ripple current (or ripple voltage) can be minimized. This is an example of optimally controlling the OFF time for the purpose of minimizing the ripple current (or ripple voltage).

The relationship between the number of switching times n shown in this table and the maximum OFF time in the number of switching times is expressed by an expression as follows. When the number of switching times n ≧ 2, if the maximum value of the OFF time is × and the switching period is T, the following equation holds from the relationship of the ON time ratio (where n is an integer of 2 or more).

{T · (n-1)} / {T · (n-1) + T} <T · n / {T · n + ×}

When this equation is solved for × (maximum value of OFF time), the following equation is obtained.

X <T ・ n / (n-1)

Therefore, if the number of switching times n and the switching cycle T are determined, the maximum value of the OFF time is determined. If the OFF time is not extended beyond this maximum value and the output becomes excessive, the number of switchings may be reduced by one. By controlling the number of switchings and the OFF time as in this equation, the OFF time can be optimally controlled, and in the case of this control equation, the ripple current (or ripple voltage) can be minimized. The following is a summary including the case where the number of switching times = 1.

When the number of switching times n ≧ 2, the OFF time X is feedback-controlled so that the number of switching times n and the OFF time × are as follows.

T <X <T · n / (n-1) (T: switching cycle, n: number of stitches)

If the output is insufficient, n → (n + 1)
If the output is too high, n → (n-1)
When n = 1, shift to control of n = 1.

When the number of switching times n = 1, the OFF time T is feedback-controlled so that the OFF time X becomes as follows.

T <X

If the output is too high, lengthen x. If the output is insufficient, shorten x in the range of T <. If X = T and the output is insufficient, set n = 2 and the number of switching times n ≧ 2. Move to control

When implementing such control in actual hardware, a logic circuit may be constructed and hardware may be constructed from the above relational expression, or a table as shown above may be created. It may be controlled based on the table.

FIG. 26 shows the relationship between the ON time ratio and the burst frequency when the above-mentioned control is performed, FIG. 27 shows the relationship between the ON time ratio and the OFF time, and FIG. 28 shows the relationship between the ON time ratio and the number of ON times.

In the case of burst control, sound squeal may actually occur, and the burst frequency is often selected to be 20 kHz or more, which is equal to or higher than the audible band, or a low frequency that is hard to hear. However, in this control, since the burst frequency fluctuates depending on the load condition, there may be a case where the burst frequency falls into the audible band of 20 kHz or less. In that case, as shown in FIG. 29, at the start of the burst, the switching frequency starts to oscillate from a high place and the frequency is gradually lowered, so-called soft start, and at the end of the burst, the switching frequency is gradually raised and then turned off. It is effective to use so-called soft OFF (or soft end).

In addition, using soft start also has a secondary benefit. As the load becomes lighter, the number of ON times of switching at the time of burst decreases, so in the end, only the soft start part remains, and finally, when the number of switching times = 1 time, the action of soft start Then, the switching frequency is automatically increased (see FIG. 30).

When the same number of switching times is once, the gain decreases when the switching frequency is increased, and the OFF time becomes shorter under the same load conditions. Therefore, as a side effect of soft start, the ripple current can be reduced. It will be possible.

Further, from this idea, when soft start or soft OFF is not used, the following control is also effective from the viewpoint of minimizing the ripple current. That is, when the number of switching times = 1 and the soft start is not used, the following method is effective as a device for reducing the ripple current.

The control in the burst operation shown in FIG. 15 described above is divided into two modes, a case where the number of switching times ≥ 2 and a case where the number of times switching = 1, but the case where the number of times switching = 1 is further divided into two modes. (See FIGS. 31 and 32).

1. 1. When the number of switchings becomes 1, the number of switchings is fixed at 1 and the OFF time is fixed at the minimum OFF time.
2. The upper limit switching frequency fmax2 in this mode is set higher than fmax1, and is controlled between fmax1 and fmax2 by feedback.
3. 3. When fmax2 is reached and unregulated (that is, excessive output) occurs at a lighter load, fmax2 is fixed and the control shifts to OFF time control.

<4. Modification example>
Although one embodiment of the present technology has been specifically described above, the present technology is not limited to the above-mentioned one embodiment, and various modifications based on the technical idea of the present technology are possible. .. Further, the configurations, methods, processes, shapes, materials, numerical values, etc. mentioned in the above-described embodiments are merely examples, and different configurations, methods, processes, shapes, materials, numerical values, etc. may be used as necessary. May be good.

The present technology can also have the following configurations.
(1)
LLC switching power supply
It has a control unit that is supplied with a feedback value indicating the load condition and forms a drive signal for the switching element.
The control unit
In the first region where the load is heavy, frequency control is performed to change the switching frequency according to the feedback value.
In the second region where the load is lighter than that of the first region, burst control is performed by fixing the switching frequency and providing a switching ON section and a switching OFF section.
In the burst control, a switching power supply in which the ON time ratio is continuously changed according to a load condition by controlling both the number of switching ON times and the OFF time.
(2)
The switching power supply according to (1), wherein when the ON time ratio is lowered, the OFF time is controlled to an optimum value, and the number of switchings in one burst cycle is reduced as the load becomes lighter.
(3)
The switching power supply according to (2), wherein when the number of switching times becomes one, the OFF time is controlled to be extended as the load becomes lighter.
(4)
The switching power supply according to any one of (1) to (3), wherein a soft start and a soft end are combined in the burst control.
(5)
The switching power supply according to any one of (1) to (3), wherein when the number of times of switching becomes one in the burst control, the number of times of switching and the OFF time are fixed and frequency control is performed again.
(6)
The switching power supply according to (5), wherein the frequency is fixed and the OFF time is controlled when the frequency control cannot be stabilized due to excessive output.
(7)
The switching power supply according to (1), wherein the load is a secondary battery.

Q1, Q2 ... MOSFET, TR ... transformer, t1, t2 ... output terminal,
11,21 ... Error amplifier, 12 ... Control unit

Claims (7)

LLC switching power supply
It has a control unit that is supplied with a feedback value indicating the load condition and forms a drive signal for the switching element.
The control unit
In the first region where the load is heavy, frequency control is performed to change the switching frequency according to the feedback value.
In the second region where the load is lighter than that of the first region, burst control is performed by fixing the switching frequency and providing a switching ON section and a switching OFF section.
In the burst control, a switching power supply in which the ON time ratio is continuously changed according to a load condition by controlling both the number of switching ON times and the OFF time. The switching power supply according to claim 1, wherein when the ON time ratio is lowered, the OFF time is controlled to an optimum value, and the number of switchings in one burst cycle is reduced as the load becomes lighter. The switching power supply according to claim 2, wherein when the number of switching times becomes one, the OFF time is controlled to be extended as the load becomes lighter. The switching power supply according to claim 1, wherein a soft start and a soft end are combined in the burst control. The switching power supply according to claim 1, wherein in the burst control, when the number of times of switching becomes one, the number of times of switching and the OFF time are fixed, and frequency control is performed again. The switching power supply according to claim 5, wherein in the frequency control, when the output becomes excessive and stabilization becomes impossible, the frequency is fixed and the OFF time is controlled. The switching power supply according to claim 1, wherein the load is a secondary battery.

Images