A continuous mode bridgeless power factor correction circuit and a control method thereof

Abstract

The application discloses a continuous mode bridgeless power factor correction circuit and a control method thereof, which comprises an alternating voltage source AC, a diode D1, a diode D2, a main switch Q1, a main switch Q2, an output capacitor C1, a positive half cycle control circuit and a negative half cycle control circuit. In the application, the freewheeling diode is reversely biased before the main switch is turned on, and the current is transferred to the control circuit auxiliary super-fast diode branch. The body diode of the freewheeling diode has no reverse current when the main switch is turned on, and can work at high frequency. Through proper dead zone setting, the freewheeling diode can realize zero voltage switching. The control circuit only works for a short time, the control method is simple, the average power is small, and the cost is low.

Description

Technical Field

[0001] This invention relates to the field of power supply technology, and more specifically, to a totem-pole bridgeless power factor correction circuit and its control method using a continuous current mode employing a common silicon MOSFET. Background Technology

[0002] The rapid development of high-frequency power electronics technology has led to a significant increase in the number of power supply products connected to the AC power grid. In order to limit and reduce the harmonic pollution of the power grid caused by these power supply products, the harmonic standard IEC61000-3-2 has imposed strict limits on the harmonics generated by power supply products connected to the power grid. For power supply products with an input power greater than 75W connected to the power grid, power factor correction (PFC) must be added to meet the harmonic requirements.

[0003] There are two types of power factor correction schemes: passive power factor correction and active power factor correction. Active power factor correction schemes have been widely used due to their small size and high efficiency, and related technologies have also developed and improved rapidly. In order to achieve a high cost performance, power factor correction of low-power power supply products connected to the grid generally adopts the critical mode, while power supply products connected to the grid generally adopt the continuous mode.

[0004] Active power factor correction technology has undergone rapid development from bridged to bridgeless. Bridgeless PFC technology has also been subdivided into various forms. Among them, the continuous mode totem pole bridgeless PFC topology has been widely used in high-power power supply products because it has fewer components, higher efficiency, and less interference, and can overcome some of the shortcomings of other bridgeless PFC topologies.

[0005] Because the reverse recovery time of the body diode of ordinary high-voltage silicon MOSFETs is relatively long, existing continuous-mode bridgeless PFC converters generally use IGBTs, high-speed silicon MOSFETs, silicon carbide MOSFETs, and gallium nitride MOSFETs with short reverse recovery times due to internal anti-parallel diodes. However, the operating frequency of conventional IGBTs is generally limited to below 40kHz, resulting in large PFC inductors and high losses. Compared with silicon MOSFETs, IGBTs have higher on-resistance and diode conduction in reverse, resulting in high conduction losses. Silicon carbide MOSFETs have extremely short reverse recovery times and low on-resistance, making them suitable for high-power, high-frequency applications, but their high price makes them unacceptable for ordinary applications. Gallium nitride MOSFETs also have extremely short reverse recovery times and low on-resistance, making them suitable for high-frequency applications, but they are mainly used in low-power power supplies. To make them into commonly used high-power PFCs, multi-phase interleaving is required, which also increases the cost.

[0006] Taking IGBT as an example, here's a brief introduction to the principle of continuous mode totem-pole bridgeless PFC. The circuit is as follows... Figure 1As shown in Figure 2, for ease of explanation, the internal diodes of the IGBT are drawn externally and connected in anti-parallel with the drain and source of the IGBT. During the positive half-cycle of AC, Q2 is the main diode and Q1 is the freewheeling diode. When Q2 is turned on, the AC power supply charges L1 through the Q2-D2 branch, increasing the current in L1 and storing energy. When Q2 is turned off, the inductor L1 charges C1 through the D2-AC power supply-L1-DQ1 branch, releasing its own energy. The Q2 drive and inductor L1 current waveforms are shown in Figure 2. Because it is a continuous current mode, the conduction current of the Q1 anti-parallel diode is not zero. When Q2 is turned on, a reverse bias voltage is applied to DQ1. Due to the reverse recovery, the current suddenly reverses phase. The high voltage on the capacitor is applied to the drain and source of Q2 through the low-resistance DQ1. Because there is a reverse recovery current in this branch, there is a large transient power between the drain and source of Q2 at the moment of turn-on. The Q1 anti-parallel diode is completely blocked only after the reverse recovery is completed. In the negative half-cycle of AC, the working mechanism is similar to that of the positive half-cycle. Q1 is the main diode and Q2 is the freewheeling diode. When Q1 is turned on, the AC power supply charges L1 through D1 and the Q1 branch, increasing the current of L1 and storing energy. When Q1 is turned off, the inductor L1 charges C1 through the DQ2-L1-AC power supply-D1 branch, releasing its own energy. The Q1 drive and inductor L1 current waveforms are shown in Figure 3. When Q1 is turned on, because it is in continuous current mode, the conduction current of DQ2 is not zero. After the reverse bias voltage is applied, due to the reverse recovery, the current suddenly reverses. The high voltage on the capacitor is applied to the drain and source of Q1 through the low-resistance DQ2. Because there is a reverse recovery current in this branch, there is a large transient power between the collector and emitter of Q1 at the moment of turn-on. After the reverse recovery is completed, the anti-parallel diode Q2 is completely blocked.

[0007] Clearly, only a sufficiently short inversion recovery time allows the circuit to function properly. If the inversion recovery time is too long, the main switching transistor will be unable to withstand power consumption exceeding its own limit and will burn out. This is also the main reason why ordinary silicon MOSFETs with long inversion recovery times cannot be directly used in the high-frequency branch of the totem-pole bridgeless PFC in continuous mode. Summary of the Invention

[0008] To address the drawbacks of conventional IGBT devices in continuous-mode totem-pole bridgeless PFC topologies, such as low operating frequency, large inductor size, and high losses, as well as the high cost of silicon MOSFETs with internally parallel ultrafast recovery diodes and gallium nitride MOSFETs, this invention proposes a continuous-mode totem-pole bridgeless PFC circuit using ordinary silicon MOSFETs. This circuit avoids the long reverse recovery time of silicon MOSFETs while offering advantages such as high-frequency operation, low cost, and simple control.

[0009] On the other hand, this invention also proposes a control method for a totem-pole bridgeless PFC using a continuous mode ordinary silicon MOS transistor. The logic and timing are simple and clear, require fewer components, have low cost, and are easy to implement.

[0010] Based on the above points, the present invention proposes the following technical solution:

[0011] Firstly, if a totem-pole bridgeless PFC converter in continuous mode were to use ordinary silicon MOSFETs, the forward current of the freewheeling diode's body diode must be diverted to another branch before the main conductor turns on. This requires the body diode to be reverse-biased, a crucial but obvious point. The difficulty lies in how to achieve this action in a cost-effective and simplified manner. Adding more components would result in higher costs and implementation difficulties, leading to higher overall cost and lower reliability. Therefore, it might be more efficient to use silicon carbide or gallium nitride MOSFETs directly.

[0012] Based on the above, the present invention provides a continuous mode totem pole bridgeless PFC converter, as shown in Figure 4, characterized by including an AC voltage source AC, diode D1, diode D2, main switch Q1, main switch Q2, boost inductor L1, output capacitor C1, positive half-cycle control module, and negative half-cycle control module.

[0013] The gate of the main switch Q1 is connected to the drive 1 circuit. The positive half-cycle control module consists of a forward auxiliary transformer T1, an auxiliary ultrafast diode D3, an auxiliary MOSFET Q3, a drive 3 circuit, a magnetic reset circuit D4, R2, and C2. The secondary winding of auxiliary transformer T1 is connected to the anode of auxiliary ultrafast diode D3. The secondary winding of auxiliary transformer T1 is connected to the source of main switch Q1. The cathode of auxiliary ultrafast diode D3 is connected to the drain of main switch Q1. The primary winding of auxiliary transformer T1 is connected to the anode of capacitor C1. The primary winding of auxiliary transformer T1 is connected to the drain of auxiliary MOSFET Q3. The source of auxiliary MOSFET Q3 is connected to the cathode of capacitor C1. The gate of auxiliary MOSFET Q3 is connected to driver circuit 3. R2 and C2 are connected in parallel, one end connected to the primary winding of auxiliary transformer T1, and the other end connected to the cathode of ultrafast diode D4. The anode of D4 is connected to the primary winding of auxiliary transformer T1, forming a magnetic reset circuit. For ease of explanation, the internal diode of Q1 is drawn externally, i.e., DQ1.

[0014] The gate of the main switch Q2 is connected to the drive 2 circuit. The negative half-cycle control module consists of a forward auxiliary transformer T2, an auxiliary ultrafast diode D5, an auxiliary MOSFET Q4, a drive 4 circuit, and a magnetic reset circuit D6, C3, and R3. The secondary winding of auxiliary transformer T2 (non-same-name terminal) is connected to the negative terminal of auxiliary ultrafast diode D5. The secondary winding of auxiliary transformer T2 (same-name terminal) is connected to the drain of main switch Q2. The positive terminal of auxiliary ultrafast diode D5 is connected to the source of main switch Q2. The primary winding of auxiliary transformer T2 (same-name terminal) is connected to the positive terminal of capacitor C1. The primary winding of auxiliary transformer T1 (non-same-name terminal) is connected to the drain of auxiliary MOSFET Q4. The source of auxiliary MOSFET Q4 is connected to the negative terminal of capacitor C1. The gate of auxiliary MOSFET Q4 is connected to driver circuit 4. R3 and C3 are connected in parallel, one end connected to the primary winding of auxiliary transformer T2 (same-name terminal), and the other end connected to the negative terminal of ultrafast diode D6. The positive terminal of D6 is connected to the primary winding of auxiliary transformer T2 (non-same-name terminal), forming a magnetic reset circuit. For ease of explanation, the internal diode of Q2 is drawn externally, i.e., DQ2.

[0015] The working principle of this invention is as follows: During the positive half-cycle of the AC power supply, before the main switch Q2 is turned on, the inductor discharges in three stages. In the first stage before these three stages, to avoid shoot-through, there must be a dead zone between the upper and lower switches. Q2 is turned off for 100 nanoseconds, and then Q1 is turned on. Because the body diode DQ1 of Q1 first provides freewheeling current to the inductor L1 and conducts, Q1 is turned on with zero voltage. The inductor L1 charges C1 through D2-AC-L1-Q1 source-Q1 drain, releasing its own energy. This stage occupies most of the freewheeling phase of the inductor. When this stage ends, Q1 is turned off, and the second stage begins. The inductor current continues to charge C1 through the body diode DQ1 of Q1. Because DQ1 is conducting, Q2 is turned off with zero voltage. After 60 nanoseconds (adjusted according to actual conditions to ensure that Q1 is completely turned off), the drive circuit 3 turns on the auxiliary MOSFET Q3. Through voltage conversion, the secondary side of T1 receives 8V. The voltage (the secondary voltage can be set according to the characteristics of the MOS transistor; the body diode should be turned off quickly, but the secondary voltage should also be as low as possible to reduce losses) is applied through diode D3, forcibly applying a reverse bias voltage to the body diode of Q1. Although the body diode of Q1 has a reverse recovery current, the low reverse bias voltage results in low losses, and the freewheeling current flowing through the body diode quickly drops to zero, ending the second stage and starting the third stage. The freewheeling current is transferred to the secondary side of T1 and the branch of D3 to charge C1. Q3 conducts for 400 nanoseconds before turning on the main MOS transistor Q2 (depending on the characteristics of the MOS transistor used, this setting time should be as short as possible while ensuring that the body diode of Q1 is completely turned off to improve efficiency), ending the third stage. Because D3 is an ultrafast diode, the losses caused by reverse recovery are very small. The auxiliary MOS transistor Q3 is turned off 100 nanoseconds after the main MOS transistor Q2 is turned on. At this time, due to the high voltage reverse bias, D3 is already turned off. Because the magnetic reset circuit of T1 is working, T1's magnetic reset is ensured.

[0016] During the positive half-cycle of the AC power supply, the relevant switching transistor drive and inductor current waveforms are shown in Figure 5. i_L1 is the inductor current, Vg_Q2 is the drive waveform of the main transistor Q2, Vg_Q1 is the drive waveform of the freewheeling transistor Q1, i_DQ1 is the body diode current waveform of the freewheeling transistor, and i_SD1 is the current flowing through SD in the freewheeling transistor. In Figure 6, Vg_Q3 is the drive waveform of the auxiliary transistor, and i_D3 is the current waveform of the auxiliary ultrafast recovery diode D3. According to the time, i_DQ1 is already zero before the main transistor Q2 is turned on, thus avoiding the reverse recovery of DQ1.

[0017] During the negative half-cycle of the AC power supply, Q1 is the main switching transistor, Q2 is the freewheeling transistor, and the relevant control circuit of T2 is working. Its working mechanism is similar to that of the positive half-cycle. Before the main switch Q1 is turned on, the inductor discharges in three stages. In the first stage, to avoid shoot-through, there must be a dead time between the upper and lower switches. Q2 turns on 100 nanoseconds after Q1 is turned off. Because the body diode DQ2 of Q2 first provides freewheeling current to the inductor L1 and conducts, Q2 is turned on at zero voltage. The inductor L1 charges C1 through the source-drain-L1-AC-D1 branch of Q2. This stage occupies most of the time of the inductor freewheeling stage. When this stage ends, Q2 turns off and the second stage begins. The inductor L1 charges C1 through the DQ2-L1-AC-D1 branch. Because DQ2 is turned on, Q2 is turned off at zero voltage. After 60 nanoseconds (adjust according to the actual situation to ensure that Q2 is completely turned off), the drive circuit 4 turns on the auxiliary MOSFET Q4. Through voltage conversion, the secondary side of T2 receives 8V (the actual voltage can be adjusted according to the MOSFET). The secondary voltage is set to ensure the body diode is quickly cut off, but the secondary voltage should also be as low as possible to reduce losses. Through diode D5, a reverse bias voltage is forcibly applied to the body diode of Q2. Although the body diode of Q2 has a reverse recovery current, the low reverse bias voltage results in minimal losses, and the freewheeling current flowing through the body diode quickly drops to zero, ending the second stage and starting the third stage. The inductor freewheeling current transfers to the secondary side of T2 and the D5 branch to maintain charging of C1. Q4 conducts for 400 nanoseconds before turning on the main MOSFET Q1 (depending on the characteristics of the MOSFET used, this setting time should be as short as possible while ensuring the body diode of Q1 is completely cut off to improve efficiency), ending the third stage. Because D5 is an ultrafast diode, the losses due to reverse recovery are minimal. The auxiliary MOSFET Q4 turns off 100 nanoseconds after the main MOSFET Q1 is turned on. At this point, due to the high reverse bias, D5 is already cut off. Because the magnetic reset circuit of T2 is working, T2's magnetic reset is ensured.

[0018] During the negative half-cycle of the AC power supply, the relevant switching transistor drive and inductor current waveforms are shown in Figure 7. i_L1 is the inductor current, Vg_Q1 is the drive waveform of the main transistor Q1, Vg_Q2 is the drive waveform of the freewheeling transistor Q2, i_DQ2 is the body diode current waveform of the freewheeling transistor, and i_SD2 is the current flowing through SD in the freewheeling transistor. In Figure 8, Vg_Q4 is the drive waveform of the auxiliary transistor, and i_D5 is the current waveform of the auxiliary ultrafast recovery diode D5. According to the time, i_DQ2 is already zero before the main transistor Q1 is turned on, thus avoiding the reverse recovery of DQ2.

[0019] The main switch should be driven to turn off in as short a time as possible. Since the MOS has an output capacitor, L1 charges it, and the voltage rises in time. This allows for low-voltage turn-off.

[0020] In this embodiment, Driver 1, Driver 2, Driver 3, and Driver 4 are driving circuits for each MOS transistor. Since major brand manufacturers have mature and reliable silicon MOS transistor driving devices, this embodiment uses existing mature driving devices. Therefore, the driving circuit structure will not be described in detail in this invention.

[0021] The circuit operates within the range of 60-150kHz, making it easy to handle conducted radiation. Existing magnetic core materials, capacitors, power transistors, etc., can operate with high reliability and low cost, achieving a high cost-performance ratio.

[0022] The beneficial effects of this invention are as follows.

[0023] 1. When the main switch is turned on, the freewheeling current has been transferred from the body diode of the freewheeling diode to the auxiliary ultrafast diode and the secondary branch of the auxiliary transformer. This effectively avoids the problem of long reverse recovery time of the body diode of the freewheeling diode and the problem of large capacitor voltage being directly applied to the main switch, causing large losses and damage to the main switch. It can operate at high frequency in continuous mode.

[0024] 2. The body diode is already conducting when the freewheeling diode is turned on and off, thus achieving zero-voltage switching of the freewheeling diode.

[0025] 3. During the freewheeling period of inductor L1, the freewheeling diode is reverse-conducting for most of the time, with low on-state resistance, resulting in low voltage drop and low conduction loss.

[0026] 4. The control circuit has low cost and simple control scheme. Compared with traditional IGBT or silicon carbide MOSFET control, only two sets of auxiliary circuits with very small average power are added. It can be implemented by a general microcontroller. The original loop control algorithm remains basically unchanged. Only a freewheeling diode reverse bias timing control is added, which is easy to understand and implement.

[0027] 5. The auxiliary transformer is a high-low voltage converter. The secondary transformer has a large current and low voltage, while the primary side has a high voltage and small current. This allows the auxiliary transistor to be a low-power MOSFET, which reduces the cost.

[0028] 6. Because the auxiliary transformer, auxiliary switching transistor, and auxiliary ultrafast diode have very short conduction times throughout the entire high-frequency operating cycle, these devices have very low average power, low cost, and are readily available.

[0029] 7. Compared with IGBTs, silicon carbide MOSFETs, gallium nitride MOSFETs, and silicon MOSFETs have simple, mature driving circuits and low cost.

[0030] A continuous-mode totem-pole bridgeless PFC control method, employing the bridgeless PFC converter, includes control methods for the main switching transistor and the freewheeling transistor.

[0031] In continuous mode, the upper and lower switches of the totem-pole bridgeless PFC bridge arm operate in complementary conduction modes to ensure no shoot-through. Furthermore, appropriate dead-time settings enable zero-voltage switching of the freewheeling diode.

[0032] During the positive half-cycle of the AC power supply, the main switch is Q2, the freewheeling diode is Q1, and the auxiliary diode is Q3. The control method for each switch is as follows: When the driver 2 circuit sends a low level, 100 nanoseconds after the main switch Q2 is turned off (this delay should be determined based on the characteristics of the driver and the main switch, and should be as short as possible while ensuring that Q1 turns on at zero voltage), the driver 1 circuit sends a high level to turn on the freewheeling diode Q1. Since the freewheeling diode DQ1 conducts first, Q1 turns on at zero voltage. The turn-on time of Q1 can be set according to the turn-on time of Q2 calculated by the loop and the set dead time. After this time, the driver 1 circuit sends a low level to turn off the freewheeling diode Q1. Since the freewheeling diode DQ1 is still conducting, Q1 turns off at zero voltage. After Q1 goes low for 60 nanoseconds, the driver circuit 3 sends a high level, Q3 turns on, DQ1 is forcibly reverse-biased and cut off, and the freewheeling current is transferred to the D3 branch. After 400 nanoseconds, the driver circuit 2 sends a high level, turning on the main switch Q2. After another 100 nanoseconds, the driver circuit 3 sends a low level, turning off the auxiliary switch Q3. When the MCU calculates the turn-on time, the driver circuit 2 sends a low level, turning off Q2.

[0033] Optionally, the current of the freewheeling diode DQ1 can be detected. When Q3 is turned on and DQ1 is forcibly reverse-biased, if the current of DQ1 is zero, the main switch Q2 can be turned on with an appropriate delay. See Figure 10.

[0034] During the negative half-cycle of the AC power supply, the main switch is Q1, the freewheeling diode is Q2, and the auxiliary diode is Q4. The control method for each switch is as follows: When the driver circuit 1 sends a low level, 100 nanoseconds after the main switch Q1 turns off (this delay should be determined based on the characteristics of the driver and the main switch, and should be as short as possible while ensuring that Q2 turns on at zero voltage), the driver circuit 2 sends a high level to turn on the freewheeling diode Q2. Since the freewheeling diode DQ2 conducts first, Q2 turns on at zero voltage. The turn-on time of Q2 can be set according to the turn-on time of Q1 calculated by the loop and the set dead time. After this time, the driver circuit 2 sends a low level to turn off the freewheeling diode Q2. Since the freewheeling diode DQ2 is still conducting, Q2 turns off at zero voltage. After Q2 goes low for 60 nanoseconds, the driver circuit 4 outputs a high level, Q4 turns on, DQ2 is forcibly reverse-biased and cut off, and the freewheeling current is transferred to the D5 branch. After 400 nanoseconds, the driver circuit 1 outputs a high level, turning on the main switch Q1. After another 100 nanoseconds, the driver circuit 4 outputs a low level, turning off the auxiliary switch Q4. When the turn-on time of Q1 calculated by the MCU expires, the driver circuit 1 outputs a low level, turning off Q1.

[0035] Optionally, the current of the freewheeling diode DQ2 can be detected. When Q4 is turned on and DQ2 is forcibly reverse-biased, if the current of DQ2 is zero, the main switch Q1 can be turned on with an appropriate delay, as shown in Figure 10.

[0036] The above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit the application of this technical solution to totem-pole bridgeless PFC topology. Those skilled in the art should understand that modifications or equivalent substitutions to the technical solution of the present invention without departing from the core of the technical solution should be covered within the scope of the claims of the present invention. For example, in a single-phase bidirectional grid-connected inverter, as shown in Figure 11, D1 and D2 are replaced with ordinary silicon MOSFETs Q5 and Q6, referred to as low-frequency arms. During AC-DC conversion, it is a totem-pole bridgeless PFC circuit. Q5 and Q6 have the same function as D1 and D2, providing a low-frequency path, but the use of MOSFETs reduces conduction losses. During DC-AC conversion, in the positive half-cycle of AC, Q6 is on and Q5 is off. During each high-frequency operating cycle, before the freewheeling phase of L1, Q2 is on, discharging the AC voltage source. This phase accounts for the majority of the freewheeling phase. After the freewheeling phase of L1, Q2 turns off at zero voltage. After a 60-nanosecond delay, Q4 turns on, forcibly reverse-biasing DQ2, transferring the freewheeling current to the D5 branch. 400 nanoseconds after Q4 turns on, Q1 turns on, ending the freewheeling and beginning to charge L1 and the AC voltage source branch. This also avoids the problem of long inversion recovery time associated with ordinary silicon MOSFETs. The situation is similar in the negative half-cycle of AC and will not be elaborated further.

[0037] Optionally, D3, Q1, T1 and D5, Q2, T2 can be connected as shown in Figure 9 to achieve the function of the present invention. However, since the secondary winding of the auxiliary transformer is connected in series with the drain of the main switching transistors Q1 and Q2, the losses of the auxiliary transformer, especially the copper losses, are greatly increased. The power capacity of the auxiliary transformer must be increased. This method is not the optimal solution, but since its core is also to use the transformer to step down the voltage and force the reverse bias of the body diode to transfer the freewheeling current, it should also be within the scope of the claims of the present invention. Attached Figure Description

[0038] Figure 1 shows the basic circuit diagram of a typical totem-pole bridgeless PFC converter using IGBTs in the existing technical solution.

[0039] Figure 2 shows the inductor current waveform and the drive waveform of the main switch Q2 in the positive half-cycle of the existing technical solution.

[0040] Figure 3 shows the inductor current waveform and the drive waveform of the main switch Q1 during the negative half-cycle in the existing technical solution.

[0041] Figure 4 is a basic circuit diagram of a totem-pole bridgeless PFC converter using silicon MOS transistors in a specific embodiment of the present invention.

[0042] Figure 5 is a comparison of the following waveforms in a specific embodiment of the present invention: AC positive half-cycle inductor current waveform IL1, main switch Q2 driving waveform Vg_q2, freewheeling diode Q1 driving waveform Vg_q1, freewheeling diode Q1 body diode current i_DQ1 waveform, and freewheeling diode Q1 source-drain current i_sd1 waveform.

[0043] Figure 6 is a comparison of the AC positive half-cycle inductor current waveform IL1, the auxiliary switch Q3 driving waveform Vg_q3, the freewheeling diode Q1 driving waveform Vg_q1, the freewheeling diode Q1 body diode current i_DQ1 waveform, and the auxiliary ultrafast recovery diode D3 current i_D3 waveform in a specific embodiment of the present invention.

[0044] Figure 7 is a comparison of the following waveforms in a specific embodiment of the present invention: AC negative half-cycle inductor current waveform IL1, main switch Q1 driving waveform Vg_q1, freewheeling diode Q2 driving waveform Vg_q2, freewheeling diode Q2 body diode current i_DQ2 waveform, and freewheeling diode Q2 source-drain current i_sd2 waveform.

[0045] Figure 8 is a comparison of the AC negative half-cycle inductor current waveform IL1, the auxiliary switch Q4 driving waveform Vg_q4, the freewheeling diode Q2 driving waveform Vg_q2, the freewheeling diode Q2 body diode current i_DQ2 waveform, and the auxiliary ultrafast recovery diode D5 current i_D5 waveform in a specific embodiment of the present invention.

[0046] Figure 9 shows another optional modified circuit of the present invention that falls within the scope of this patent.

[0047] Figure 10 shows the totem pole bridgeless PFC circuit diagram with current detection added according to the present invention.

[0048] Figure 11 is a circuit diagram of the single-phase bidirectional inverter of the present invention.

[0049] The dead zone in the figure has been magnified for easier observation; the actual dead zone time is very short. Detailed Implementation

[0050] To address the drawbacks of conventional IGBT devices in continuous-mode totem-pole bridgeless PFC topologies, such as low operating frequency, large inductor size, and high losses, as well as the high cost of silicon MOS transistors with internally parallel ultrafast recovery diodes and gallium nitride MOS transistors, this invention proposes a continuous-mode totem-pole bridgeless PFC circuit using ordinary silicon MOS transistors. This circuit avoids the long reverse recovery time of silicon MOS transistors while offering advantages such as high-frequency operation, low cost, and simple control.

[0051] On the other hand, this invention also proposes a control method for a totem-pole bridgeless PFC using a continuous mode ordinary silicon MOS transistor. The logic and timing are simple and clear, require fewer components, have low cost, and are easy to implement.

[0052] Based on the above points, the present invention proposes the following technical solution:

[0053] Firstly, if a totem-pole bridgeless PFC converter in continuous mode were to use ordinary silicon MOSFETs, the forward current of the freewheeling diode's body diode must be diverted to another branch before the main conductor turns on. This requires the body diode to be reverse-biased, a crucial but obvious point. The difficulty lies in how to achieve this action in a cost-effective and simplified manner. Adding more components would result in higher costs and implementation difficulties, leading to higher overall cost and lower reliability. Therefore, it might be more efficient to use silicon carbide or gallium nitride MOSFETs directly.

[0054] Based on the above, the present invention provides a continuous mode totem pole bridgeless PFC converter, as shown in Figure 4, characterized by including an AC voltage source AC, diode D1, diode D2, main switch Q1, main switch Q2, boost inductor L1, output capacitor C1, positive half-cycle control module, and negative half-cycle control module.

[0055] The gate of the main switch Q1 is connected to the drive 1 circuit. The positive half-cycle control module consists of a forward auxiliary transformer T1, an auxiliary ultrafast diode D3, an auxiliary MOSFET Q3, a drive 3 circuit, a magnetic reset circuit D4, R2, and C2. The secondary winding of auxiliary transformer T1 is connected to the anode of auxiliary ultrafast diode D3. The secondary winding of auxiliary transformer T1 is connected to the source of main switch Q1. The cathode of auxiliary ultrafast diode D3 is connected to the drain of main switch Q1. The primary winding of auxiliary transformer T1 is connected to the anode of capacitor C1. The primary winding of auxiliary transformer T1 is connected to the drain of auxiliary MOSFET Q3. The source of auxiliary MOSFET Q3 is connected to the cathode of capacitor C1. The gate of auxiliary MOSFET Q3 is connected to driver circuit 3. R2 and C2 are connected in parallel, one end connected to the primary winding of auxiliary transformer T1, and the other end connected to the cathode of ultrafast diode D4. The anode of D4 is connected to the primary winding of auxiliary transformer T1, forming a magnetic reset circuit. For ease of explanation, the internal diode of Q1 is drawn externally, i.e., DQ1.

[0056] The gate of the main switch Q2 is connected to the drive 2 circuit. The negative half-cycle control module consists of a forward auxiliary transformer T2, an auxiliary ultrafast diode D5, an auxiliary MOSFET Q4, a drive 4 circuit, and a magnetic reset circuit D6, C3, and R3. The secondary winding of auxiliary transformer T2 (non-same-name terminal) is connected to the negative terminal of auxiliary ultrafast diode D5. The secondary winding of auxiliary transformer T2 (same-name terminal) is connected to the drain of main switch Q2. The positive terminal of auxiliary ultrafast diode D5 is connected to the source of main switch Q2. The primary winding of auxiliary transformer T2 (same-name terminal) is connected to the positive terminal of capacitor C1. The primary winding of auxiliary transformer T1 (non-same-name terminal) is connected to the drain of auxiliary MOSFET Q4. The source of auxiliary MOSFET Q4 is connected to the negative terminal of capacitor C1. The gate of auxiliary MOSFET Q4 is connected to driver circuit 4. R3 and C3 are connected in parallel, one end connected to the primary winding of auxiliary transformer T2 (same-name terminal), and the other end connected to the negative terminal of ultrafast diode D6. The positive terminal of D6 is connected to the primary winding of auxiliary transformer T2 (non-same-name terminal), forming a magnetic reset circuit. For ease of explanation, the internal diode of Q2 is drawn externally, i.e., DQ2.

[0057] The working principle of this invention is as follows: During the positive half-cycle of the AC power supply, before the main switch Q2 is turned on, the inductor discharges in three stages. In the first stage before these three stages, to avoid shoot-through, there must be a dead zone between the upper and lower switches. Q2 is turned off for 100 nanoseconds, and then Q1 is turned on. Because the body diode DQ1 of Q1 first provides freewheeling current to the inductor L1 and conducts, Q1 is turned on with zero voltage. The inductor L1 charges C1 through D2-AC-L1-Q1 source-Q1 drain, releasing its own energy. This stage occupies most of the freewheeling phase of the inductor. When this stage ends, Q1 is turned off, and the second stage begins. The inductor current continues to charge C1 through the body diode DQ1 of Q1. Because DQ1 is conducting, Q2 is turned off with zero voltage. After 60 nanoseconds (adjusted according to actual conditions to ensure that Q1 is completely turned off), the drive circuit 3 turns on the auxiliary MOSFET Q3. Through voltage conversion, the secondary side of T1 receives 8V. The voltage (the secondary voltage can be set according to the characteristics of the MOS transistor; the body diode should be turned off quickly, but the secondary voltage should also be as low as possible to reduce losses) is applied through diode D3, forcibly applying a reverse bias voltage to the body diode of Q1. Although the body diode of Q1 has a reverse recovery current, the low reverse bias voltage results in low losses, and the freewheeling current flowing through the body diode quickly drops to zero, ending the second stage and starting the third stage. The freewheeling current is transferred to the secondary side of T1 and the branch of D3 to charge C1. Q3 conducts for 400 nanoseconds before turning on the main MOS transistor Q2 (depending on the characteristics of the MOS transistor used, this setting time should be as short as possible while ensuring that the body diode of Q1 is completely turned off to improve efficiency), ending the third stage. Because D3 is an ultrafast diode, the losses caused by reverse recovery are very small. The auxiliary MOS transistor Q3 is turned off 100 nanoseconds after the main MOS transistor Q2 is turned on. At this time, due to the high voltage reverse bias, D3 is already turned off. Because the magnetic reset circuit of T1 is working, T1's magnetic reset is ensured.

[0058] During the positive half-cycle of the AC power supply, the relevant switching transistor drive and inductor current waveforms are shown in Figure 5. i_L1 is the inductor current, Vg_Q2 is the drive waveform of the main transistor Q2, Vg_Q1 is the drive waveform of the freewheeling transistor Q1, i_DQ1 is the body diode current waveform of the freewheeling transistor, and i_SD1 is the current flowing through SD in the freewheeling transistor. In Figure 6, Vg_Q3 is the drive waveform of the auxiliary transistor, and i_D3 is the current waveform of the auxiliary ultrafast recovery diode D3. According to the time, i_DQ1 is already zero before the main transistor Q2 is turned on, thus avoiding the reverse recovery of DQ1.

[0059] During the negative half-cycle of the AC power supply, Q1 is the main switching transistor, Q2 is the freewheeling transistor, and the relevant control circuit of T2 is working. Its working mechanism is similar to that of the positive half-cycle. Before the main switch Q1 is turned on, the inductor discharges in three stages. In the first stage, to avoid shoot-through, there must be a dead time between the upper and lower switches. Q2 turns on 100 nanoseconds after Q1 is turned off. Because the body diode DQ2 of Q2 first provides freewheeling current to the inductor L1 and conducts, Q2 is turned on at zero voltage. The inductor L1 charges C1 through the source-drain-L1-AC-D1 branch of Q2. This stage occupies most of the time of the inductor freewheeling stage. When this stage ends, Q2 turns off and the second stage begins. The inductor L1 charges C1 through the DQ2-L1-AC-D1 branch. Because DQ2 is turned on, Q2 is turned off at zero voltage. After 60 nanoseconds (adjust according to the actual situation to ensure that Q2 is completely turned off), the drive circuit 4 turns on the auxiliary MOSFET Q4. Through voltage conversion, the secondary side of T2 receives 8V (the actual voltage can be adjusted according to the MOSFET). The secondary voltage is set to ensure the body diode is quickly cut off, but the secondary voltage should also be as low as possible to reduce losses. Through diode D5, a reverse bias voltage is forcibly applied to the body diode of Q2. Although the body diode of Q2 has a reverse recovery current, the low reverse bias voltage results in minimal losses, and the freewheeling current flowing through the body diode quickly drops to zero, ending the second stage and starting the third stage. The inductor freewheeling current transfers to the secondary side of T2 and the D5 branch to maintain charging of C1. Q4 conducts for 400 nanoseconds before turning on the main MOSFET Q1 (depending on the characteristics of the MOSFET used, this setting time should be as short as possible while ensuring the body diode of Q1 is completely cut off to improve efficiency), ending the third stage. Because D5 is an ultrafast diode, the losses due to reverse recovery are minimal. The auxiliary MOSFET Q4 turns off 100 nanoseconds after the main MOSFET Q1 is turned on. At this point, due to the high reverse bias, D5 is already cut off. Because the magnetic reset circuit of T2 is working, T2's magnetic reset is ensured.

[0060] During the negative half-cycle of the AC power supply, the relevant switching transistor drive and inductor current waveforms are shown in Figure 7. i_L1 is the inductor current, Vg_Q1 is the drive waveform of the main transistor Q1, Vg_Q2 is the drive waveform of the freewheeling transistor Q2, i_DQ2 is the body diode current waveform of the freewheeling transistor, and i_SD2 is the current flowing through SD in the freewheeling transistor. In Figure 8, Vg_Q4 is the drive waveform of the auxiliary transistor, and i_D5 is the current waveform of the auxiliary ultrafast recovery diode D5. According to the time, i_DQ2 is already zero before the main transistor Q1 is turned on, thus avoiding the reverse recovery of DQ2.

[0061] The main switch should be driven to turn off in as short a time as possible. Since the MOS has an output capacitor, L1 charges it, and the voltage rises in time. This allows for low-voltage turn-off.

[0062] In this embodiment, Driver 1, Driver 2, Driver 3, and Driver 4 are driving circuits for each MOS transistor. Since major brand manufacturers have mature and reliable silicon MOS transistor driving devices, this embodiment uses existing mature driving devices. Therefore, the driving circuit structure will not be described in detail in this invention.

[0063] The circuit operates within the range of 60-150kHz, making it easy to handle conducted radiation. Existing magnetic core materials, capacitors, power transistors, etc., can operate with high reliability and low cost, achieving a high cost-performance ratio.

[0064] The beneficial effects of this invention are as follows.

[0065] 1. When the main switch is turned on, the freewheeling current has been transferred from the body diode of the freewheeling diode to the auxiliary ultrafast diode and the secondary branch of the auxiliary transformer. This effectively avoids the problem of long reverse recovery time of the body diode of the freewheeling diode and the problem of large capacitor voltage being directly applied to the main switch, causing large losses and damage to the main switch. It can operate at high frequency in continuous mode.

[0066] 2. The body diode is already conducting when the freewheeling diode is turned on and off, thus achieving zero-voltage switching of the freewheeling diode.

[0067] 3. During the freewheeling period of inductor L1, the freewheeling diode is reverse-conducting for most of the time, with low on-state resistance, resulting in low voltage drop and low conduction loss.

[0068] 4. The control circuit has low cost and simple control scheme. Compared with traditional IGBT or silicon carbide MOSFET control, only two sets of auxiliary circuits with very small average power are added. It can be implemented by a general microcontroller. The original loop control algorithm remains basically unchanged. Only a freewheeling diode reverse bias timing control is added, which is easy to understand and implement.

[0069] 5. The auxiliary transformer is a high-low voltage converter. The secondary transformer has a large current and low voltage, while the primary side has a high voltage and small current. This allows the auxiliary transistor to be a low-power MOSFET, which reduces the cost.

[0070] 6. Because the auxiliary transformer, auxiliary switching transistor, and auxiliary ultrafast diode have very short conduction times throughout the entire high-frequency operating cycle, these devices have very low average power, low cost, and are readily available.

[0071] 7. Compared with IGBTs, silicon carbide MOSFETs, gallium nitride MOSFETs, and silicon MOSFETs have simple, mature driving circuits and low cost.

[0072] A continuous-mode totem-pole bridgeless PFC control method, employing the bridgeless PFC converter, includes control methods for the main switching transistor and the freewheeling transistor.

[0073] In continuous mode, the upper and lower switches of the totem-pole bridgeless PFC bridge arm operate in complementary conduction mode to ensure no shoot-through. Furthermore, appropriate dead-time settings enable zero-voltage switching of the freewheeling diode.

[0074] During the positive half-cycle of the AC power supply, the main switch is Q2, the freewheeling diode is Q1, and the auxiliary diode is Q3. The control method for each switch is as follows: When the driver 2 circuit sends a low level, 100 nanoseconds after the main switch Q2 is turned off (this delay should be determined based on the characteristics of the driver and the main switch, and should be as short as possible while ensuring that Q1 turns on at zero voltage), the driver 1 circuit sends a high level to turn on the freewheeling diode Q1. Since the freewheeling diode DQ1 conducts first, Q1 turns on at zero voltage. The turn-on time of Q1 can be set according to the turn-on time of Q2 calculated by the loop and the set dead time. After this time, the driver 1 circuit sends a low level to turn off the freewheeling diode Q1. Since the freewheeling diode DQ1 is still conducting, Q1 turns off at zero voltage. After Q1 goes low for 60 nanoseconds, the driver circuit 3 sends a high level, Q3 turns on, DQ1 is forcibly reverse-biased and cut off, and the freewheeling current is transferred to the D3 branch. After 400 nanoseconds, the driver circuit 2 sends a high level, turning on the main switch Q2. After another 100 nanoseconds, the driver circuit 3 sends a low level, turning off the auxiliary switch Q3. When the MCU calculates the turn-on time, the driver circuit 2 sends a low level, turning off Q2.

[0075] Optionally, the current of the freewheeling diode DQ1 can be detected. When Q3 is turned on and DQ1 is forcibly reverse-biased, if the current of DQ1 is zero, the main switch Q2 can be turned on with an appropriate delay. See Figure 10.

[0076] During the negative half-cycle of the AC power supply, the main switch is Q1, the freewheeling diode is Q2, and the auxiliary diode is Q4. The control method for each switch is as follows: When the driver circuit 1 sends a low level, 100 nanoseconds after the main switch Q1 turns off (this delay should be determined based on the characteristics of the driver and the main switch, and should be as short as possible while ensuring that Q2 turns on at zero voltage), the driver circuit 2 sends a high level to turn on the freewheeling diode Q2. Since the freewheeling diode DQ2 conducts first, Q2 turns on at zero voltage. The turn-on time of Q2 can be set according to the turn-on time of Q1 calculated by the loop and the set dead time. After this time, the driver circuit 2 sends a low level to turn off the freewheeling diode Q2. Since the freewheeling diode DQ2 is still conducting, Q2 turns off at zero voltage. After Q2 goes low for 60 nanoseconds, the driver circuit 4 outputs a high level, Q4 turns on, DQ2 is forcibly reverse-biased and cut off, and the freewheeling current is transferred to the D5 branch. 400 nanoseconds later, the driver circuit 1 outputs a high level, turning on the main switch Q1. 100 nanoseconds later, the driver circuit 4 outputs a low level, turning off the auxiliary switch Q4. When the turn-on time of Q1 calculated by the MCU expires, the driver circuit 1 outputs a low level, turning off Q1.

[0077] Optionally, the current of the freewheeling diode DQ2 can be detected. When Q4 is turned on and DQ2 is forcibly reverse-biased, if the current of DQ2 is zero, the main switch Q1 can be turned on with an appropriate delay, as shown in Figure 10.

[0078] The above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit the application of this technical solution to totem-pole bridgeless PFC topology. Those skilled in the art should understand that modifications or equivalent substitutions to the technical solution of the present invention without departing from the core of the technical solution should be covered within the scope of the claims of the present invention. For example, in a single-phase bidirectional grid-connected inverter, as shown in Figure 11, D1 and D2 are replaced with ordinary silicon MOSFETs Q5 and Q6, referred to as low-frequency arms. During AC-DC conversion, it is a totem-pole bridgeless PFC circuit. Q5 and Q6 have the same function as D1 and D2, providing a low-frequency path, but the use of MOSFETs reduces conduction losses. During DC-AC conversion, in the positive half-cycle of AC, Q6 is on and Q5 is off. During each high-frequency operating cycle, before the freewheeling phase of L1, Q2 is on, discharging the AC voltage source. This phase accounts for the majority of the freewheeling phase. After the freewheeling phase of L1, Q2 turns off at zero voltage. After a 60-nanosecond delay, Q4 turns on, forcibly reverse-biasing DQ2, transferring the freewheeling current to the D5 branch. 400 nanoseconds after Q4 turns on, Q1 turns on, ending the freewheeling and beginning to charge L1 and the AC voltage source branch. This also avoids the problem of long inversion recovery time associated with ordinary silicon MOSFETs. The situation is similar in the negative half-cycle of AC and will not be elaborated further.

[0079] Optionally, D3, Q1, T1 and D5, Q2, T2 can be connected as shown in Figure 9 to achieve the function of the present invention. However, since the secondary winding of the auxiliary transformer is connected in series with the drain of the main switching transistors Q1 and Q2, the losses of the auxiliary transformer, especially the copper losses, are greatly increased. The power capacity of the auxiliary transformer must be increased. This method is not the optimal solution, but since its core is also to use the transformer to step down the voltage and force the reverse bias of the body diode to transfer the freewheeling current, it should also be within the scope of the claims of the present invention.

Claims

1. A continuous-mode totem-pole bridgeless PFC converter, characterized in that, The circuit includes an AC voltage source (AC), diodes D1 and D2, a boost inductor L1, main switches Q1 and Q2, a driver circuit 1, a driver circuit 2, an AC positive half-cycle control circuit, an AC negative half-cycle control circuit, and an output capacitor C1. The AC positive half-cycle control circuit consists of a driver circuit 3, an auxiliary switch Q3, a diode D4, a resistor R2, a capacitor C2, an auxiliary transformer T1, and an auxiliary ultrafast diode D3. The AC negative half-cycle control circuit consists of a driver circuit 4, an auxiliary switch Q4, a diode D6, a resistor R3, a capacitor C3, an auxiliary transformer T2, and an auxiliary ultrafast diode D5. The first terminal of the boost inductor L1 is connected to the live wire of the AC voltage source, and the second terminal of the boost inductor L1 is connected to the midpoint of the bridge arm formed by the main switches Q1 and Q2 connected in series. The drain of the upper transistor in the bridge arm is connected to the positive terminal of the output capacitor C1, and the source of the lower transistor in the bridge arm is connected to the negative terminal of the output capacitor C1. The auxiliary transformer T1... The secondary non-same-name terminal of the auxiliary transformer T1 is connected to the source of the main switching transistor Q1, and the same-name terminal is connected to the positive terminal of the auxiliary ultrafast diode D3. The negative terminal of diode D3 is connected to the drain of the main switching transistor Q1. The same-name terminal of the primary winding of the auxiliary transformer T1 is connected to the positive terminal of capacitor C1. The non-same-name terminal of the primary winding of the auxiliary transformer T1 is connected to the drain of the auxiliary MOSFET Q3. The source of the auxiliary MOSFET Q3 is connected to the negative terminal of capacitor C1. The gate of the auxiliary MOSFET Q3 is connected to the drive circuit. R2 and C2 are connected in parallel, one end of which is connected to the same-name terminal of the primary winding of the auxiliary transformer T1, and the other end is connected to the negative terminal of the ultrafast diode D4. The positive terminal of D4 is connected to the non-same-name terminal of the primary winding of the auxiliary transformer T1, forming a magnetic reset circuit. The non-same-name terminal of the secondary winding of the auxiliary transformer T2 is connected to the negative terminal of the auxiliary ultrafast diode D5. The same-name terminal of the secondary winding of the auxiliary transformer T2 is connected to the drain of the main switching transistor Q2. The auxiliary ultrafast diode D5... The positive terminal is connected to the source of Q2. The positive terminal of the auxiliary ultrafast diode D5 is connected to the source of the main switch Q2. The primary terminal of the auxiliary transformer T2 is connected to the positive terminal of capacitor C1. The secondary terminal of the auxiliary transformer T2 is connected to the drain of the auxiliary MOSFET Q4. The source of the auxiliary MOSFET Q4 is connected to the negative terminal of capacitor C1. The gate of the auxiliary MOSFET Q4 is connected to the drive circuit. R3 and C3 are connected in parallel, with one end connected to the primary terminal of the auxiliary transformer T2 and the other end connected to the negative terminal of the ultrafast diode D6. The positive terminal of D6 is connected to the secondary terminal of the auxiliary transformer T2, forming a magnetic reset circuit.

2. The continuous-mode totem-pole bridgeless PFC converter according to claim 1, characterized in that, The main switch is a standard silicon MOS transistor.

3. The continuous-mode totem-pole bridgeless PFC converter according to claim 2, characterized in that, It also includes at least four PWM timers with synchronization functions to accurately determine the switching timing of each switching transistor.

4. The continuous-mode totem-pole bridgeless PFC converter according to claim 3, characterized in that, It also includes two current detectors for detecting the current in the body diode when the freewheeling diode is reverse biased.

5. A control method for a continuous-mode totem-pole bridgeless PFC converter, applied to the continuous-mode totem-pole bridgeless PFC converter according to any one of claims 1-4, characterized in that, The control method includes: during the positive half-cycle of the AC power supply, after the main switch Q2 is turned off, the freewheeling diode Q1 is turned on 100 nanoseconds later. The body diode of the freewheeling diode conducts first due to the current from the boost inductor L1, so Q1 is turned on at zero voltage. When the conduction time of the freewheeling diode Q1, as calculated according to the loop, ends, Q1 is turned off. Because the current from the boost inductor L1 has not reached zero, the body diode of Q1 remains conducting, so Q1 is turned off at zero voltage. 60 nanoseconds after Q1 is turned off, the auxiliary switch Q3 is turned on. The secondary winding of the auxiliary transformer T1 generates a voltage that forcibly reverse-biases the body diode of Q1, causing the current from the inductor L1 to transfer from the body diode of Q1 to the branch of the auxiliary ultrafast diode D3. After Q3 is turned on for 400 nanoseconds, the main switch Q2 is turned on. After Q2 is turned on for 100 nanoseconds, Q3 is turned off. The main switch Q2 is turned on according to the loop calculation. When the conduction time ends, Q2 is turned off. During the negative half-cycle of the AC power supply, after the main switch Q1 is turned off, the freewheeling diode Q2 is turned on 100 nanoseconds later. The body diode of the freewheeling diode conducts first due to the current of the boost inductor L1, and Q2 is turned on at zero voltage. When the conduction time of the freewheeling diode Q2, as calculated according to the loop, ends, Q2 is turned off. Because the current of the boost inductor L1 has not reached zero, the body diode of Q2 is still conducting, and Q2 is turned off at zero voltage. 60 nanoseconds after Q2 is turned off, the auxiliary transistor Q4 is turned on. The secondary side of the auxiliary transformer T2 generates a voltage, which forcibly reverse-biases the body diode of Q2, causing the current of inductor L1 to transfer from the body diode of Q2 to the branch of the auxiliary ultrafast diode D5. After Q4 is turned on for 400 nanoseconds, the main switch Q1 is turned on. After Q1 is turned on for 100 nanoseconds, Q4 is turned off. When the conduction time of the main switch Q1, as calculated according to the loop, ends, Q1 is turned off.

Images