Solar Photovoltaic Inverter Topology and Design Ideas [Graphic]

When designing traditional power electronic devices, we usually measure their cost-effectiveness by the price per kilowatt. However, when designing photovoltaic inverters, the pursuit of maximum power is only secondary, and maximizing European efficiency is the most important. Because for photovoltaic inverters, not only can the increase in maximum output power be converted into economic benefits, but the improvement in European efficiency can also be converted into economic benefits, and it is more obvious [1]. The definition of European efficiency is different from what we usually call average efficiency or maximum efficiency. It fully considers the changes in sunlight intensity and more accurately describes the performance of photovoltaic inverters. European efficiency is obtained by adding up the efficiencies under different load conditions according to different proportions, among which the efficiency under half load accounts for the largest component (see Figure 1).

Figure 1. European efficiency calculation ratio

Therefore, in order to improve the European efficiency of photovoltaic inverters, it is not enough to reduce the loss at rated load, but it is necessary to improve the efficiency under different load conditions at the same time. European efficiency is a new parameter, which is mainly proposed for photovoltaic inverters. Since the intensity of sunlight is different at different times, photovoltaic inverters do not actually work at rated power all the time, but more often work at light loads. Therefore, the efficiency of photovoltaic inverters cannot be measured entirely by the efficiency at rated power. Therefore, Europeans have come up with a new parameter - European efficiency. The calculation method of European efficiency is shown in Table 1.

The economic benefits brought about by the improvement of European efficiency can also be easily calculated. For example, taking a 3kW rated photovoltaic inverter as an example, according to current market cost estimates, the installation cost of photovoltaic power generation is about 4,000 euros per kilowatt [2]. This means that every 1% improvement in the European efficiency of the photovoltaic inverter can save 120 euros (the current cost of photovoltaic power generation is about 4,000 euros per kilowatt, or 4 euros per watt, including solar cells and photovoltaic inverters. For a 3kW power generation device, if the inverter efficiency is improved by 1%, that is, 30W more power is generated, then the cost can be saved by 4×30=120 euros). The economic benefits brought about by improving the European efficiency of photovoltaic inverters are obvious, and the pursuit of higher European efficiency at all costs has become the current trend of photovoltaic inverter development.

2 Selection of power devices

In the design of general inverters, IGBT is the most used device, considering the cost-effectiveness factor. Because of the nonlinear characteristics of the IGBT conduction voltage drop, the IGBT conduction voltage drop will not increase significantly with the increase of current. This ensures that the inverter can still maintain low loss and high efficiency under maximum load. However, for photovoltaic inverters, this characteristic of IGBT has become a disadvantage. Because the European efficiency is mainly related to the efficiency of the inverter under different light load conditions. Under light load, the conduction voltage drop of IGBT will not drop significantly, which reduces the European efficiency of the inverter. On the contrary, the conduction voltage drop of MOSFET is linear, and it has a lower conduction voltage drop under light load conditions. In addition, considering its excellent dynamic characteristics and high-frequency working ability, MOSFET has become the first choice for photovoltaic inverters. In addition, considering the huge economic returns after improving European efficiency, the latest relatively expensive devices, such as SIC diodes, are also being used more and more in the design of photovoltaic inverters. SIC Schottky diodes can significantly reduce the conduction loss of the switch tube and reduce electromagnetic interference.

3 Design goals of PV inverters

For transformerless photovoltaic inverters, its main design goals are:

(1) Track the maximum power point of the solar cell input voltage to obtain the maximum input power;

(2) Pursuing the maximum efficiency of photovoltaic inverters;

(3) Low electromagnetic interference.

In order to obtain the maximum input power, the circuit must have the function of automatically adjusting the input voltage according to different sunlight conditions. The maximum power point is generally around 70% of the open-loop voltage. Of course, this is also related to the characteristics of the specific photovoltaic cells used. The typical circuit is implemented through a boost circuit. Then the DC power is inverted into a grid-connected sinusoidal AC power through an inverter.

4 Introduction to single-phase transformerless photovoltaic inverter topology

The choice of topology is related to the rated output power of the photovoltaic inverter. For photovoltaic inverters below 4kw, a topology with a DC bus not exceeding 500v and single-phase output is usually selected.

This function (see Figure 2) can be realized by the following schematic diagram (see Figure 3).

Figure 2 Functional diagram of a single-phase transformerless photovoltaic inverter

Figure 3 Schematic diagram of single-phase transformerless photovoltaic inverter

The boost circuit achieves maximum power point tracking by adjusting the input voltage. The H-bridge inverter converts DC power into sinusoidal AC power and injects it into the grid. The IGBT of the upper half bridge acts as a polarity controller and operates at 50 Hz, thereby reducing the total loss and the output electromagnetic interference of the inverter. The IGBT or MOSFET of the lower half bridge performs PWM high-frequency switching. In order to minimize the size of the boost inductor and the output filter, the switching frequency is required to be as high as possible, such as 16 kHz.

4.1 Advantages of single-phase transformerless photovoltaic inverter

We recommend using power modules to design PV inverters because integrating all the components in the topology of Figure 3 into one module can provide the following advantages:

(1) Easy to install and reliable;

(2) Short R&D and design cycles allow products to be brought to market more quickly;

(3) Better electrical performance.

4.2 Indicators that must be achieved for module design

For the design of the module, we must ensure:

(1) Low inductance design of DC bus loop

To achieve this goal, we must reduce the parasitic inductance inside and outside the module at the same time. To reduce the parasitic inductance inside the module, we must optimize the binding wires, pin layout and internal routing inside the module. To reduce the parasitic inductance outside the module, we must ensure that the positive and negative ends of the DC bus of the boost circuit and the inverter bridge circuit are as close as possible while meeting the safety spacing.

(2) Configure dedicated drive pins for fast switching transistors

During the switching process of the switch tube, the parasitic inductance of the binding wire will cause the driving voltage to decrease. This will lead to an increase in switching losses and even oscillation of the switching waveform. Inside the module, by configuring a dedicated driving pin for each switch tube (directly drawn from the chip), it can be ensured that no large current flows in the driving loop, thereby ensuring the stability and reliability of the driving loop. This solution can only be achieved by power modules at present, and single-tube IGBTs cannot do it yet.

Figure 4 shows Vincotech's latest dedicated module for photovoltaic inverters, FlowSol-BI (P896-E01), which integrates the advantages mentioned above.

Figure 4 Flowsol-bi boost circuit and full-bridge inverter circuit

4.3 Technical parameters

(1) The boost circuit consists of a MOSFET (600V/45Mω) and a SIC diode;

(2) The bypass diode is mainly used to bypass the boost circuit when the input exceeds the rated load, thereby improving the overall efficiency of the inverter;

(3) The upper half of the H-bridge circuit consists of a 75A/600V IGBT and a SIC diode, and the lower half consists of a MOSFET (600V/45Mω);

(4) Integrated temperature detection resistor.

5 Efficiency calculation of flowsol0-bi, a dedicated module for single-phase transformerless photovoltaic inverter

Here we mainly consider the loss of power semiconductors, and the loss of other passive components, such as boost inductors and output filter inductors, is not included.

Based on the relevant parameters of this circuit, the simulation results are as follows:

condition

●pin=2kw;

●fpwm = 16khz;

●v pv -nominal = 300v;

●vdc = 400v.

According to the simulation results of Figures 5 and 6, it can be seen that the efficiency of the module hardly decreases with the decrease of load. The total European efficiency of the module (boost+inverter) can reach 98.8%. Even with the loss of passive components, the total efficiency of the photovoltaic inverter can still reach 98%. The dotted line in Figure 6 shows the efficiency change of the inverter using conventional power devices. It can be clearly seen that at low load, the inverter efficiency drops rapidly.

Figure 5 Boost circuit efficiency simulation results ee = 99.6%

Figure 6 FlowSOL-BI inverter circuit efficiency simulation results -EE = 99.2% Standard IGBT full bridge -EE = 97.2% (dashed line)

6 Introduction to three-phase transformerless photovoltaic inverter topology

High-power photovoltaic inverters require more photovoltaic battery groups and three-phase inverter outputs (see Figure 7), and the maximum DC bus voltage will reach 1000V.

Figure 7 Functional diagram of three-phase transformerless photovoltaic inverter

The standard application here is to use a three-phase full-bridge circuit. Considering that the DC bus voltage will reach 1000V, the switching device must use 1200V. As we know, the switching speed of 1200V power devices will be much slower than that of 600V devices, which will increase losses and affect efficiency. For this application, a better alternative is to use a center point clamped (NPC=neutral point clamped) topology (see Figure 8). In this way, 600V devices can be used to replace 1200V devices.

Figure 8 Schematic diagram of three-phase transformerless NPC photovoltaic inverter

In order to minimize the parasitic inductance in the loop, it is best to integrate the symmetrical dual boost circuit and the NPC inverter bridge into one module.

(1) Technical parameters of dual boost module (see Figure 9)

Figure 9 FlowSOL-NPB—Symmetrical dual boost circuit

●The dual boost circuits are both composed of MOSFET (600V/45Mω) and SIC diode;

●The bypass diode is mainly used to bypass the boost circuit when the input exceeds the rated load, thereby improving the overall efficiency of the inverter;

●The module has an integrated temperature detection resistor.

(2) Technical parameters of npc inverter bridge module (see Figure 10)

Figure 10 Flowsol-NPi-NPC inverter bridge

●The intermediate switching link is composed of 75A/600V IGBT and fast recovery diode;

●The upper and lower high-frequency switching links are composed of MOSFET (600V/45Mω);

●The center point clamping diode is composed of a sic diode;

●The module has an integrated temperature detection resistor.

For this topology, the design requirements for the module are basically similar to the single-phase inverter module mentioned above. The only thing that requires extra attention is that whether it is a dual boost circuit or an NPC inverter bridge, a low inductance design must be ensured between DC+, DC- and the center point.

With these two modules, it is easy to design a higher power output photovoltaic inverter. For example, using two dual boost circuits in parallel and a three-phase NPC inverter bridge can get a high-efficiency 10kw photovoltaic inverter. Moreover, the pin design of these two modules fully considers the needs of parallel connection, and parallel use is very convenient. Figure 11 is a layout diagram of a dual boost module in parallel and a three-phase NPC inverter output module.

Figure 11 Layout of dual boost modules in parallel and three-phase NPC inverter output module

For PV inverters with a DC bus voltage of 1000V, the NPC topology inverter is the most efficient on the market. Figure 12 compares the efficiency of the NPC module (MOSFET + IGBT) and the 1200V IGBT half-bridge module.

Figure 12 Comparison of NPC inverter bridge output efficiency (solid line) and half-bridge inverter efficiency (dashed line)

According to the simulation results, the efficiency of the NPC inverter can reach 99.2%, while the efficiency of the latter is only 96.4%. The advantages of the NPC topology are obvious.

7 Introduction to the design ideas of the next generation photovoltaic inverter topology

At present, the hybrid H-bridge (MOSFET+IGBT) topology has achieved a high efficiency level. The next generation of photovoltaic inverters will focus on improving the following performance:

(1) Further improvement of efficiency;

(2) Reactive power compensation;

(3) Efficient bidirectional conversion mode.

7.1 Single-phase PV Inverter Topology

For single-phase photovoltaic inverters, we first discuss how to further improve the efficiency of the hybrid H-bridge topology (see Figure 13).

In Figure 13, the switching frequency of the upper bridge arm IGBT is generally set to the grid frequency (e.g. 50 Hz), while the lower bridge arm MOSFET operates at a higher switching frequency, such as 16 kHz, to achieve a sinusoidal output. Simulations show that this inverter topology can achieve an efficiency of 99.2% at a rated power output of 2 kW. Due to the slow speed of the built-in diode of the MOSFET, the MOSFET cannot be used in the upper bridge arm.

Figure 13 Development of PV Inverters - Hybrid

Since the IGBT of the upper bridge arm operates at a switching frequency of 50 Hz, it is not actually necessary to filter this branch. Therefore, by optimizing the circuit topology, the emitter open circuit topology shown in Figure 14 can be obtained. The advantage of this topology is that only the branch through which the high-frequency current passes has a filter inductor, thereby reducing the loss of the output filter circuit.

Figure 14 Improved transformerless upper bridge arm open emitter topology

Currently, Vincotech already has a standard open emitter IGBT module product, model number is Flowsol0-Bi Open E (P896-E02), as shown in Figure 15.

Figure 15 flowsol0-bi-open e (p896-e02)

Technical parameters:

(1) The boost circuit is composed of MOSFET (600V/45Mω) and SIC diode;

(2) The bypass diode is mainly used to bypass the boost circuit when the input exceeds the rated load, thereby improving the overall efficiency of the inverter;

(3) The upper arm of the H-bridge uses IGBT (600V/75A) and SIC diode, and the lower arm uses MOSFET (600V/45Mω);

(4) The module has an internal integrated temperature detection resistor.

Next, let's analyze the open-emitter topology shown in Figure 14. When the MOSFET in the lower bridge arm is working, the diode in anti-parallel with the IGBT in the upper bridge arm does not work due to the effect of the filter inductor. In this way, the MOSFET can also be used in the upper bridge arm to improve the efficiency of the inverter when the load is light. The simulation results show that at a rated power output of 2kW, the European efficiency of this photovoltaic inverter can be increased by 0.2%, so that the efficiency reaches 99.4%. In actual applications, this topology will improve the efficiency more, because the simulation results are obtained under the assumption that the chip junction temperature is 125℃, but because the MOSFET is large in size and the photovoltaic inverter often works under light load, the junction temperature of the MOSFET chip is much lower than 125℃, so the on-resistance RDS-ON of the MOSFET in actual operation will be lower than the value in simulation, and the loss will be correspondingly smaller.

How to solve the problem of reactive power? The only way to handle reactive power with this circuit topology is to use FRED-FETs, but the on-resistance rds-on of these devices is usually very high. Another disadvantage is that their reverse recovery characteristics are poor, which affects the performance during reactive power compensation and bidirectional conversion. However, in some special applications, if reactive power must be used to measure line impedance or protect certain components, the topology shown in Figure 16 will meet the above requirements.

The topology shown in Figure 16 allows pure reactive loads, which can improve reactive compensation for the grid and also meet bidirectional power flow, such as efficient battery charging. If SIC Schottky diodes are applied, this circuit topology will be able to achieve higher efficiency levels. Table 2 shows the European efficiency of different topologies at 2kW rated power.

Figure 16 Full MOSFET topology for reactive loads

7.2 Three-phase photovoltaic inverter topology

Similar improvements can be made to the three-phase photovoltaic inverter of the NPC topology.

Taking one phase as an example, at a rated output of 2 kW, the three-level inverter (see Figure 17) can achieve 99.2% efficiency (see Table 2). With a slight modification, this topology can achieve reactive power flow.

Figure 17 Three-level inverter

After adding a 1200V diode between the output and the DC bus, this topology (see Figure 18) can output reactive power. It can also be used as a high-efficiency bidirectional inverter to achieve reverse energy conversion. In order to reduce losses, it is recommended to use SIC diodes for D3 and D4.

Figure 18 NPC topology inverter capable of achieving reactive power output

However, since the price of 1200V SIC is too high, the topology shown in Figure 19 would be a better choice.

Figure 19 NPC topology inverter that can achieve reactive power output (two SIC diodes and four SI diodes are added)

This topology uses only two 600V SIC diodes (D4, D6). Fast SI diodes are used for D3 and D5, and small SI diodes are used for D7 and D8 to prevent the SIC diodes from being damaged by overvoltage.

Is it possible to use MOSFETs here? The answer is yes, provided that the body diode of the MOSFET is bypassed. This can be achieved by separating the output terminals of the upper and lower half bridges and equipping them with their own filter inductors.

The circuit topology of Figure 20 can improve efficiency at light loads.

Figure 20 NPC topology inverter using MOSFET to achieve reactive power output

Figure 21 is a comparison of the efficiency of the all-MOSFET solution and the hybrid solution at a rated power of 2kW.

Figure 21 Efficiency comparison of all-MOSFET solution and hybrid solution at rated power of 2kW

Its European efficiency can be increased from 99.2% to 99.4%. Reactive power is achieved by the 1200V fast diode path. When selecting diodes, it is recommended to use SIC diodes, which can achieve higher efficiency during reverse conversion. Or as shown in Figure 22, D4 and D6 use 600V SIC diodes, and the other four use fast recovery SI diodes.

Figure 22 NPC inverter topology using 2 SIC diodes, 4 SI diodes and separate outputs

8 Conclusion

These new topologies enable the inverter to achieve higher efficiency levels. Even at an output power of 0.4kw, we can still achieve the highest efficiency, which also allows the system capacity to be further increased by paralleling modules. At this point, the return on investment can be easily calculated, which also shows the important role of efficiency level in photovoltaic inverter applications.

The improvement of reactive power output also makes this topology have the following characteristics and wider application:

(1) Line reactive power compensation;

(2) Efficient battery charging for applications in backup power systems, electric vehicles, and hybrid electric vehicles;

(3) High efficiency and high speed motor drive.