How do we achieve current equalization? When two modules are used in parallel, it is common for one module to have a high output current and the other module to have a low output current. In this case, if we can reduce the voltage of the module with a high output current or increase the voltage of the module with a low output current, we can adjust the current of each module in the direction of current sharing.

There are generally several ways to achieve the current sharing function of the module:

(1) Output impedance method (droop method)

This method achieves approximate current sharing by adjusting the output impedance of the module. As shown in the figure below, when used alone, if the output current of the module increases, the current signal is amplified by the current detection resistor R1 and superimposed on the negative feedback input terminal of the voltage loop. After comparison with the reference voltage Vr, the output signal is fed back through the loop, causing the output voltage to decrease and the module output current to decrease.

When using modules in parallel, assuming that the output current of Module 1 is high, the current signal through negative feedback causes the output voltage of Module 1 to decrease. At this time, the output current of Module 1 decreases, so the output current of Module 2 will increase. The final two modules achieve current sharing.

And this current sharing method, as the current increases, the output voltage will decrease, which is not suitable for backend devices with voltage accuracy requirements, and the current sharing accuracy is relatively low.

(2) Master-slave setting method

This method involves manually setting up a main module, and all other modules are assigned current as a reference by the main module, as shown in the following figure:

Each power module in the figure is a dual loop control system. In this control system, the engineer sets module l as the main module and makes it operate under voltage control, while the other modules are set as secondary modules and operate under current type control.

Ur is the reference voltage of the main module, and Uf is the output voltage feedback signal. After passing through the voltage error amplifier, the error voltage Ue is obtained, which serves as the current reference for the main module. After comparison with Ui1 (which reflects the magnitude of the main module current), the control voltage Uc1 is generated to control the operation of the modulator and driver. The main module current will be modulated according to the current reference Ue.

When modules are used in parallel, the voltage error amplifiers of each slave module are connected in the form of followers. The voltage error Ue of the main module is input to each follower, and the output of the follower is Ue, which becomes the current reference for each slave module. Therefore, the current of each slave module is modulated according to Ue value, which is basically consistent with the current of the main module, thus achieving current sharing between modules.

This method can effectively ensure the stable operation of the product without problems such as poor current distribution characteristics. However, this current sharing method requires communication connections between the master and slave modules, and if the master module fails, the entire power system cannot function properly. Therefore, the stability of the master module determines the reliability of the entire system, so it can only be used for current sharing and is not suitable for forming a redundant parallel system.

(3) Average current method

This current sharing method requires the output terminals of the current amplifiers of each module to be connected in parallel to a common bus through a resistor R with the same resistance value, called the ShareBus, as shown in the following figure:

The above figure shows the circuit schematic diagram of each single module in the parallel module automatically sharing current according to the average current. From the above figure, it can be seen that the input of the voltage amplifier is Vr 'and the feedback circuit Vr. Vr' is the sum of the reference voltage Vr and the current sharing control voltage VC. It is compared and amplified with Vf to generate Ve (error voltage amplification), which controls the PWM and driver to adjust the output of the power stage.

When modules are used in parallel, the outputs of all modules are amplified through current sampling and connected to a common current sharing bus. According to Kirchhoff's law, the sum of the currents flowing into the busbar from all branches is 0, indicating that:

At this point, Vb reflects the average current of all modules, and the difference between Vi and Vb is the current sharing error. After passing through the error amplifier, the control voltage Vc is obtained. When Vb=Vi, it indicates that each module has equal current; When Vb ≠ Vi, it indicates that the current distribution is uneven. The error control signal between Vi of the uneven module and the current sharing bus Vb is added to Vr and input to the voltage amplifier to output the error signal Ve, thereby adjusting the module output to achieve current sharing for each module.

This method can achieve current sharing accurately, but when the current sharing bus is grounded or a module fails. It will cause a decrease in the voltage of the module, leading to output faults.

(4) Busbar peak current method

The peak current method requires parallel connection of the output terminals of each module's current amplifier, connected to the same current sharing bus (ShareBus) through the same diode D, as shown in the following figure:

When modules are used in parallel, the current sharing bus Vb is equal to Vi Vdf (diode voltage drop). That is to say, the current sharing bus voltage at this time is positively correlated with the maximum Vb in each module. This current sharing method can automatically turn the module with the highest load current into the main module, while the other modules control the circuit to bring the output current closer to the main module.

This method can ensure that the system can still function normally when one module fails, and the other modules will choose another main module. However, due to the difference of one diode voltage drop between the current sampling voltage of the current sharing bus and the module, the other modules cannot maintain the same current as the main module, which has a certain impact on the accuracy of current sharing.

3、 Summary
According to different market application environments and product demand positioning, different current sharing schemes can be selected. At present, the common current sharing method in the market is the bus peak current method: for example, the LMF1000 series of Jinshengyang, through the bus peak current sharing method, not only can achieve high current sharing accuracy, but also can achieve parallel redundancy function, to ensure that a module failure will not have a significant impact on the back-end, greatly improving the reliability and stability of the system.

With the exploration and research of market demand, Jinshengyang has always been committed to power technology innovation, providing customers with higher quality and more stable one-stop power solutions.