Home >
Power Circuits >Switching Regulated Power Supply > Design and production of switching regulated power supply
Design and production of switching regulated power supply
Source: InternetPublisher:JFET Keywords: switching power supply Updated: 2023/11/17
1. It is required
to design and produce a switching regulated power supply as shown in the figure below.
to design and produce a switching regulated power supply as shown in the figure below.
Requirements: Under resistive load conditions, ① output voltage Uo adjustable range: 30V~36V; ② maximum output current LOmax: 2A; ③ when U2 changes from 15V to 21V, voltage adjustment rate SU≤0.2% (Io=2A); ④Io When changing from 0 to 2A, the load regulation rate S1≤0.5% (U2=18V); ⑤Output noise ripple voltage peak-peak value UOPP≤1V (U2=18V, Uo=36V, Io=2A); ⑥DC-DC conversion The efficiency of the converter is η≥85% (U2=18V, Uo=36V, Io=2A); ⑦ It has over-current protection function, the operating current Io(th)=2.5±0.2A, after the over-current fault is eliminated, the power supply can automatically recover It is in the normal state; ⑧ The output voltage can be set by keyboard and adjusted step by step, with a step value of 1V, and it also has the functions of measuring and digital displaying the output voltage and current; ⑨ The converter (including control circuit) can only be powered by the UIN port , no additional auxiliary power supply is allowed.
2. Overall analysis
First we need to determine the system plan. Among the requirements, the ②③④⑤⑦⑧ have little impact on the overall plan. These indicators are only related to factors such as device selection and manufacturing process, so we focus on the remaining three indicators. First of all, the output voltage Uo has an adjustable range of 30~36V, while the secondary output of the isolation transformer is 15~21V. After rectification and filtering, the maximum is about 27V, which is less than 30V. Obviously, a boost output is required in the entire voltage range. Of course, the question does not limit the form of the rectifier circuit. Another solution is to double the voltage and rectify it first and then filter it, so that the subsequent stage can use a step-down circuit.
First we need to determine the system plan. Among the requirements, the ②③④⑤⑦⑧ have little impact on the overall plan. These indicators are only related to factors such as device selection and manufacturing process, so we focus on the remaining three indicators. First of all, the output voltage Uo has an adjustable range of 30~36V, while the secondary output of the isolation transformer is 15~21V. After rectification and filtering, the maximum is about 27V, which is less than 30V. Obviously, a boost output is required in the entire voltage range. Of course, the question does not limit the form of the rectifier circuit. Another solution is to double the voltage and rectify it first and then filter it, so that the subsequent stage can use a step-down circuit.
Secondly, the overall efficiency of the converter is required to be greater than 85%. For low-power power supplies, this requirement is already relatively high. It can be calculated that at a rated power of 72W and an efficiency of 85%, the loss of the converter cannot exceed 12.7W. To meet this requirement, it is necessary to use as few devices as possible. Whether it is the main power circuit or the control and measurement circuit, it should be kept as simple as possible. The question also requires that the power supply of the control circuit is only drawn from the rectifier output port (UIN), and no additional auxiliary power supply is allowed. This requires a self-made auxiliary power supply, and the efficiency of the auxiliary power supply must not be too low, so linear power supply is not an ideal choice.
From the above analysis, we draw the overall requirements: the main circuit needs to use a boost topology, and the boost amplitude is not large. The circuit structure should be as simple as possible, the number of components should be as small as possible, a self-made auxiliary power supply should be made, and the efficiency should be high. The analysis can also find that there is no isolation requirement for input and output, and the input end is isolated by an isolation transformer, so a circuit topology without electrical isolation for input and output can be selected. Finally, we selected the basic Boost circuit scheme, using Sungyang 16-bit microcontroller as the controller, and FPCA as the drive signal generation. The overall system scheme is shown in the figure below. The 220V AC power is stepped down, rectified, and filtered to obtain a relatively stable DC voltage. The DC voltage is boosted and then filtered by the Boost circuit to obtain a smooth DC output. The output voltage and current are sampled and input into the AD conversion chip, and the voltage is stabilized by the microcontroller PID regulator. And voltage regulation then outputs the command signal to the FPGA and displays it. The FPGA generates a PWM signal and drives the power switch tube through the drive circuit to achieve closed-loop feedback control. When the output current is greater than the protection setting value, an overcurrent protection signal is generated. The overcurrent signal drives the relay to cut off the main circuit and close the drive signal. Then it delays and then tries to power on and perform overcurrent detection. If there is overcurrent, the main circuit is disconnected. , until the circuit returns to normal.
From the above analysis, we draw the overall requirements: the main circuit needs to use a boost topology, and the boost amplitude is not large. The circuit structure should be as simple as possible, the number of components should be as small as possible, a self-made auxiliary power supply should be made, and the efficiency should be high. The analysis can also find that there is no isolation requirement for input and output, and the input end is isolated by an isolation transformer, so a circuit topology without electrical isolation for input and output can be selected. Finally, we selected the basic Boost circuit scheme, using Sungyang 16-bit microcontroller as the controller, and FPCA as the drive signal generation. The overall system scheme is shown in the figure below. The 220V AC power is stepped down, rectified, and filtered to obtain a relatively stable DC voltage. The DC voltage is boosted and then filtered by the Boost circuit to obtain a smooth DC output. The output voltage and current are sampled and input into the AD conversion chip, and the voltage is stabilized by the microcontroller PID regulator. And voltage regulation then outputs the command signal to the FPGA and displays it. The FPGA generates a PWM signal and drives the power switch tube through the drive circuit to achieve closed-loop feedback control. When the output current is greater than the protection setting value, an overcurrent protection signal is generated. The overcurrent signal drives the relay to cut off the main circuit and close the drive signal. Then it delays and then tries to power on and perform overcurrent detection. If there is overcurrent, the main circuit is disconnected. , until the circuit returns to normal.
3. Device selection
First select the circuit switching frequency fs. Because the switching loss is almost proportional to the square of the switching frequency, if the frequency is too high, the loss will increase; but if the frequency is too small, the filter inductor and capacitor will be too large, and the circuit will be prone to audio noise. After comprehensive consideration, fs is selected as 20kHz.
(1) Selection of input inductor and output filter capacitor. First calculate the size of the boost inductor. The rectified output voltage is 19-27V, and the output voltage range is 30-36V. According to the critical current formula Iob=Uo/2Lf8lD (1-D) (2), when D=1/3, the critical current has a maximum value of 1obm =2Uo/27Lfs. To make the inductor current continuous, the minimum load current (the question requires no-load, here is 0.1A) should be greater than Iobm. From this, L≥2Uo/27fsIobm=2*36/27*30*0.1 =1.33mH, take L=2mH. From the output ripple △Uo=DUo/f8RC, R is the load resistance, which can be 15Ω. From this, the size of the filter capacitor can be calculated, where C=4700μF. Since the boost circuit input inductor has a large DC component, an iron core that is not easily saturated should be selected as the inductor core. The winding should be as uniform and compact as possible, otherwise the voltage noise will increase. You can also purchase it directly. Since electrolytic capacitors have large parasitic inductance, the pins should be kept as short as possible when welding, and small-capacity polypropylene capacitors should be connected in parallel, which is very helpful in reducing output voltage spikes.
(2) Selection of switching tubes. The forward voltage endured by switch tube Q when turned off is 36V. Considering a certain peak margin, the forward breakdown voltage of IRF3205 is 55v, and the on-resistance is only 8mΩ, so there will be no breakdown and the conduction loss is very small. The output rectifier diode is a Schottky diode MBR20100 with a small on-resistance, its on-voltage drop is 0.7V, and its reverse breakdown voltage is 100V. The MOSFET driver is a dedicated driver chip IR2110. The driver circuit is shown in the figure below.
First select the circuit switching frequency fs. Because the switching loss is almost proportional to the square of the switching frequency, if the frequency is too high, the loss will increase; but if the frequency is too small, the filter inductor and capacitor will be too large, and the circuit will be prone to audio noise. After comprehensive consideration, fs is selected as 20kHz.
(1) Selection of input inductor and output filter capacitor. First calculate the size of the boost inductor. The rectified output voltage is 19-27V, and the output voltage range is 30-36V. According to the critical current formula Iob=Uo/2Lf8lD (1-D) (2), when D=1/3, the critical current has a maximum value of 1obm =2Uo/27Lfs. To make the inductor current continuous, the minimum load current (the question requires no-load, here is 0.1A) should be greater than Iobm. From this, L≥2Uo/27fsIobm=2*36/27*30*0.1 =1.33mH, take L=2mH. From the output ripple △Uo=DUo/f8RC, R is the load resistance, which can be 15Ω. From this, the size of the filter capacitor can be calculated, where C=4700μF. Since the boost circuit input inductor has a large DC component, an iron core that is not easily saturated should be selected as the inductor core. The winding should be as uniform and compact as possible, otherwise the voltage noise will increase. You can also purchase it directly. Since electrolytic capacitors have large parasitic inductance, the pins should be kept as short as possible when welding, and small-capacity polypropylene capacitors should be connected in parallel, which is very helpful in reducing output voltage spikes.
(2) Selection of switching tubes. The forward voltage endured by switch tube Q when turned off is 36V. Considering a certain peak margin, the forward breakdown voltage of IRF3205 is 55v, and the on-resistance is only 8mΩ, so there will be no breakdown and the conduction loss is very small. The output rectifier diode is a Schottky diode MBR20100 with a small on-resistance, its on-voltage drop is 0.7V, and its reverse breakdown voltage is 100V. The MOSFET driver is a dedicated driver chip IR2110. The driver circuit is shown in the figure below.
(3) Selection of other components. The loss of the measurement control circuit is related to the working voltage of the components. The operational amplifiers used for signal amplification should choose low power supply voltage, Rail-To-Rail type operational amplifiers INA132 and OPA350, which can reduce power consumption.
The power consumption of the microcontroller is related to the CPU clock frequency. Reducing the clock frequency of the microcontroller can also reduce the loss. In this design, the CPU clock of the Sungyang microcontroller is 24.576MHz.
4. Create
the schematic diagram of the main power circuit shown in the figure above, and the auxiliary power circuit diagram shown in the figure below. In the auxiliary power circuit, LM2575, D1, L, and C2 form a Buck circuit. R1 and R2 play the role of feedback regulation. Adjusting R2 can change the output voltage. There are two auxiliary power circuits in this design, which are +5V and +15V. The classic Boost circuit and other voltage and current measurement circuits are relatively simple, and their principles will not be described in detail. Here is a brief explanation of the problems encountered during the production process and their solutions.
The power consumption of the microcontroller is related to the CPU clock frequency. Reducing the clock frequency of the microcontroller can also reduce the loss. In this design, the CPU clock of the Sungyang microcontroller is 24.576MHz.
4. Create
the schematic diagram of the main power circuit shown in the figure above, and the auxiliary power circuit diagram shown in the figure below. In the auxiliary power circuit, LM2575, D1, L, and C2 form a Buck circuit. R1 and R2 play the role of feedback regulation. Adjusting R2 can change the output voltage. There are two auxiliary power circuits in this design, which are +5V and +15V. The classic Boost circuit and other voltage and current measurement circuits are relatively simple, and their principles will not be described in detail. Here is a brief explanation of the problems encountered during the production process and their solutions.
The first problem is that the rectifier bridge (with a current resistance of 10A) is always burned. Analysis shows that the input steady-state current is about SA, which should not damage the rectifier bridge. However, the actual simulation waveform of the current flowing through the rectifier bridge (C=4700μF) is as shown in the figure below. The larger the filter capacitor and the smaller the conduction angle 0 of the diode, the greater the peak current flowing through the diode. Its value can easily be greater than 10A. Later, we connected the inductor L1 in series behind the rectifier bridge. Because the inductor has a certain freewheeling effect, the diode conduction angle becomes larger, thereby reducing the current peak to protect the rectifier bridge. After the improvement, the rectifier bridge no longer burns out. However, the fuse (rated current 10A) is often blown when the machine is turned on. Analysis found that the filter capacitor after the rectifier bridge was in an instantaneous short-circuit state when the machine was turned on, so there was a large inrush current when the machine was turned on. Therefore, we connected NTC (negative temperature coefficient) in series before the rectifier bridge. Thermistor, RV1 in Figure 4), problem is also solved. The principle is that when the power is turned on, the NTC has a low temperature and presents a large resistance, so the power-on current will not be very large. As the circuit is turned on, the NTC heats up and presents a very small resistance, so the voltage drop on the NTC is very small during normal operation. It will affect the normal operation of the circuit.
The second problem we encountered was slow voltage regulation and poor voltage stabilization. At first we thought it was a problem with the software regulator, but after a long inspection we found it was caused by inaccurate voltage measurement. In the circuit of Figure 4, it is obvious that the voltage at both ends of the load is proportional to the voltage between nodes 1 and 2. We have just started to directly measure the voltage between node 2 and ground. On the surface, it seems that the 0.1Ω sampling resistor has little effect, but the current in the circuit When the flowing current is 2A, the voltage drop on the current sampling resistor is 0.2V, and the error is about 0.5%. It can be seen that the error is not small. On the other hand, if this sampling scheme is used, the measurement error will be different due to the different currents in the circuit. The error will show a certain nonlinearity with the voltage change, which will cause trouble in voltage regulation. Therefore, we later switched to a differential method to collect voltage, that is, using a differential op amp to sample between node 1 and node 2. This can greatly reduce the error, and achieved good results after improvement. All aspects of the measurement circuit should be accurate and reliable, and the sampling resistor should also be as accurate and stable as possible. The two sampling circuits shown in the figure below can also reduce the output voltage to one-tenth of the sample, but in the circuit in Figure (b) The accuracy and stability of the amplification factor are higher, that is, the sampling resistor should be made to change as little as possible. Similarly, if the voltage or current signal to be measured needs to be filtered at the input end of the AD conversion, the filter capacitor should not be too large, otherwise it will affect the response time and cause measurement lag, which will naturally make the adjustment inaccurate. Although these problems are simple, they have a great impact. If they can be measured quickly and accurately, the adjustment of the microcontroller will be much smoother.
The third problem is that there is a large spike in the output voltage. This is obviously caused by the high-frequency switching of the switching tube. Especially when it is turned off, due to the parasitic inductance in the circuit, the instantaneous current cut off will cause an impulse voltage to appear at both ends of the inductor. . Our solution is to add a buffer circuit to the switch tube on the one hand to improve the turn-off performance. The basic principle is as shown below. When the switch tube Q is turned off, part of the current of the original circuit passes through the fast recovery diode 1) to charge the capacitor C, so that both switches The terminal voltage E rises slowly, and the current decrease rate in the circuit also slows down. A simple buffer circuit can eliminate the need for diode D. The specific parameter design of RCD is more complicated. You can refer to relevant switching power supply books when designing. On the other hand, polypropylene capacitors with good high-frequency characteristics and small parasitic inductance are connected in parallel at both ends of the output filter electrolytic capacitor. Multiple parallel connections have better effects, but the leads must be kept as short as possible. At the same time, in order to reduce the line inductance, for the power main circuit, the wiring should be as short as possible and the wire diameter should be slightly thicker.
Then there is a characteristic of the Boost circuit itself - it cannot operate in an open circuit. However, the meaning of the question obviously requires that the power supply can be open circuit. Because when the load is open circuit, the input inductor continues to store and release energy periodically as usual, and the energy is not consumed by the load. The capacitor voltage will continue to rise, that is, the excess energy will be stored between the capacitor plates, which will soon lead to capacitor breakdown. . One solution is to add a dummy load, that is, when it detects that the power supply is in an no-load state, it automatically puts in a light load. This load has a large resistance value, which can maintain the output voltage at a given value and reduce its own power loss. smaller.
The above are the problems and some solutions encountered in the production of this switching power supply. After taking these measures and careful debugging, good results were achieved.
The above are the problems and some solutions encountered in the production of this switching power supply. After taking these measures and careful debugging, good results were achieved.
Latest Power Circuits Circuits
- Cleverly use the electronic ballast of waste energy-saving lamps as power supply transformer
- Using Lead-acid Battery as Power Source for Walkman
- Driving circuit of switching power supply field effect tube
- Brief Analysis of the Working Principle of AC Voltage Stabilizer Circuit
- Transformerless AC-DC constant current LED driver circuit
- Offline 8w LED Flyback Power Supply with PFC using NCP1014
- Charging control circuit made by solar energy
- Adjustable voltage power socket
- UC3842 Electric Vehicle Charger
- Car Audio Power Supply
Popular Circuits
- Power supply design experience summary
- Switching power supply constant current charger circuit diagram
- Uninterruptible power supply circuit diagram composed of MIC5158
- Small power switching power supply circuit diagram
- Practical circuit of switching power supply for dot matrix printer using M51978P-FP chip
- Switching power supply circuit diagram using TDA4605-2
- Oscilloscope test switching power supply example
- Power supply noise filter circuit for use in amplifier switching power supplies
- Basic circuit diagram of switching power supply AC input and rectification and filtering
- Lixin LCD switching power supply partial circuit diagram
京公网安备 11010802033920号