502030 battery

　　Redundant structure and current sharing technology of parallel power supply of power modules

　　With the development of power electronics technology, various electronic devices have higher and higher requirements for power supply and current. However, affected by the properties of the semiconductor power devices and magnetic materials that make up the power module, The output parameters (such as voltage, current, power) of a single switching power supply module often cannot meet the requirements. If multiple power modules are used to supply power in parallel, as shown in Figure 1, it can not only provide the required current, but also form an N+m redundant structure, which improves the stability of the system and kills two birds with one stone.

　　However, when the power modules are running in parallel, due to the dispersion of the parameters of each module, the output current cannot be exactly the same, resulting in some modules being overloaded and some modules being too light. This will reduce the stability of the system and bring serious consequences to our production and life, and the life of the power module itself will also be greatly shortened. Foreign data show that the service life of electronic components when the working environment temperature exceeds 50°C is 1/6 of that at normal temperature (25°C). Therefore, evenly distributing the output current of each parallel power supply module is a problem that must be solved to improve the stability of the parallel power supply system.

　　This article starts from the topology of the current sharing circuit and introduces several traditional parallel current sharing schemes. Other current sharing schemes (such as the automatic current sharing method based on thermal stress) will not be discussed for the time being. Each current sharing method mentioned in the article is introduced in detail, and combined with a simple circuit diagram, its working principle, advantages and disadvantages are described [1][2][3][4]. In the last part of the article, a brief outlook on the development of parallel current sharing is given.

　　Benefits of 2N+m redundant structure

　　Using N+m redundant structure operation can improve system stability.

　　N+m redundant structure means that N+m power modules supply power to the system together. Here N represents the number of power modules during normal operation, and m represents the number of redundant modules. The larger the value of m, the higher the reliability of system operation, but the system cost will also increase accordingly. Under normal working conditions, it is powered by N modules. When one or some of the modules fail, they will withdraw from power supply and be replaced by one or all of the m modules, thereby ensuring the continuity and stability of the entire system.

　　Taking a system with an output current of 100A as an example to illustrate the benefits of redundant structure operation, only three working modes of 1+1, 2+1, and 3+1 are discussed here, as shown in Figure 2. The working condition of each power module is determined by the closing condition of Kn.

　　If a 1+1 redundant structure is adopted, two power modules with an output current of 100A are used to supply power in parallel. Under normal circumstances, only one module works. When it fails and stops working, another module starts working, and the system can still run normally.

　　If a 2+1 redundant structure is adopted, three power modules with an output current of 50A are used to supply power in parallel. Normally, only two modules work. When one of them fails and stops working, the other module starts working, and the system can still run normally.

　　If a 3+1 redundant structure is adopted, that is, four power supply modules with an output current of 33A are used to supply power in parallel. Under normal circumstances, only three modules work. When one of them fails and stops working, the other module starts working, and the system can still work. normal operation.

　　Comparing the above three working methods, the 2+1 method is the best. This is because half of the power in the 1+1 method is idle, while the 3+1 method uses too many components, is too costly, and is not economical.

　　3 Several Traditional Parallel Current Sharing Schemes

　　3.1 Drooping method

　　The full name of the droop method is the external characteristic droop method, also called the slope control method. In a parallel power module system, each power module works independently. Each module determines its output current based on its external characteristics and voltage parameter values. In the droop method, the current feedback signal is mainly used to adjust the output impedance of each module, that is, to adjust the slope of Vo=f(Io), thereby adjusting the output current. Its working principle diagram is shown in Figure 3.

　　Ri is the sampling resistor of the output current Io of any parallel module. After current amplification, the current feedback voltage signal Vi is generated. Vf is the output voltage feedback. Vr is the sum of Vi and Vf. Vg is the control reference voltage (5V). Ve is the error voltage. . When the output current Io of a certain module is too large, the combined voltage and current feedback signal Vr = Vi + Vf increases. After comparing with Vg, Ve decreases, and Ve is fed back to the control part of the power module to increase the output voltage Vo of the module. decreases, Io decreases, that is, the external characteristics of Vo=f(Io) decrease. Each module adjusts its own output current, so that the parallel current sharing of each module can be achieved.

　　The advantage of this method is that it is simple, does not require a special current balancing device, and is an open-loop control. The disadvantage is that the adjustment accuracy is not high, and each module must be adjusted individually. If the parallel-connected modules have different powers, current imbalance between modules may easily occur.

　　3.2 Master-slave power supply method

　　The master-slave power supply method uses one of the multiple power modules connected in parallel as the master module, and the other modules follow the master module to work. The specific working process is: compare the operating current of the main module with the output feedback signal, and feed the difference signal back to the control circuit of each power module (including the main module and the slave module), thereby adjusting the output current of each module.

　　As shown in Figure 4, assume that module 1 is the main module, the sampling voltage of its output current is V1, and the sampling voltage of the output current of other modules is Vn. When the output current of a certain module is too large, the corresponding Vn increases. Compared with V1, the obtained Ven decreases and is fed back to the control circuit of the module to reduce its output current, thereby achieving current sharing.

　　The advantage of the master-slave module method is that no special control circuit is required. Its disadvantage is that communication between each module is required, and the wiring is relatively complicated; its biggest disadvantage is that once the main module fails, the entire power system will collapse, so it cannot be used in a redundant structure.

　　3.3 Automatic current sharing method and maximum current method

　　The automatic current sharing method is also called the single-wire method. Its working principle is to connect each power module to a current sharing bus through a current sensor and a sampling resistor. As shown in Figure 5, when the output reaches current sharing, the output current I1 is zero. On the contrary, due to the current I1 flowing through the resistor R, a voltage Uab is generated at both ends of the resistor R. This voltage passes through the output voltage Uc of the amplifier A, and the ΔU after comparing it with the reference voltage Ur is fed back to the control part of the power module, thus Adjust the output current to ultimately achieve current sharing.

　　The advantage of the automatic current sharing method is that the circuit is simple and easy to implement. The disadvantage is that if a module is short-circuited to the current sharing bus, the system cannot share current, and current limiting of a single module may also cause system instability.

　　If the resistor in Figure 5 is replaced by a diode, the positive terminal of the diode is connected to a and the negative terminal is connected to b. In this way, among the N power modules connected in parallel, only the current of the module with the largest output current can turn on the diode connected to it, so the current sharing bus voltage is equal to the output voltage of the module, and the other modules are on the current sharing bus. The voltage is used as the reference to adjust their respective output currents to achieve current sharing. If the sampling resistor is simply replaced by a diode, the current sharing accuracy of the master module will be reduced due to the forward voltage drop of the diode itself, while the slave module will not be affected. This can be replaced by the buffer shown in Figure 6 to improve the current sharing accuracy.

　　Using this current sharing method, the N power modules participating in current sharing are based on the one with the largest output current. The maximum current module is random. This current sharing method is also called the "democratic current sharing method". Since the maximum current sharing unit works in the master control state and other units work in the controlled state, this method is also called the "automatic master-slave current sharing method".

　　The UC3907 series of integrated current sharing control chips developed by the American company Unitrode adopt this working method.

　　The UC3907 chip allows multiple power modules connected in parallel to bear a part of the total load current, and the load currents they bear are equal. By monitoring the current of each module, the current balancing bus determines which parallel module has the highest output current, and sets it as the main module, and then adjusts the output current of other modules according to the current of the main module to achieve current sharing.

　　3.4 External controller method

　　The external controller method is to add an external module specifically for parallel current sharing control in addition to each parallel power supply module, as shown in Figure 7.

　　The output current of each module is sampled, converted into a voltage signal, and compared with the given voltage Vcc. The resulting difference is input to the control part of each power module, so that the parallel current sharing of the output current of each module can be achieved.

　　This working method requires an additional special controller, which increases investment, and the controller and power modules need to be connected in multiple ways. The wiring is complicated, but the current sharing effect is very good, and the output current of each module is basically equal. 4 Current status and future prospects of the development of power supply parallel current sharing technology

　　The most commonly used parallel current sharing technology is the master-slave control method, and the UC3907 series of chips developed by the American company Unitrode based on the maximum current method has been widely used due to its simple structure and powerful functions. Its detailed parameters and working process.

　　Due to the rapid development of microcontroller and DSP technology, some people use them to control the current sharing of parallel power modules, and the effect is very good. However, due to the high cost of the chip and the insufficient accuracy of its own A/D and D/A, if you want to obtain ideal parameters, you must add special A/D and D/A chips, so it has not been widely used.

Read recommendations:

LR20

What factors should be considered when using high-temperature batteries?12v deep cycle marine batter

48v 10kwh energy storage solar system.Can chargers for lead-acid batteries and lithium batteries be

601435 polymer battery

12V23A battery