Inverter design trade-off for photovoltaic power generation

Photovoltaic (PV) power generation is a very important way of energy conversion. It can convert solar energy into electricity. As the core photovoltaic power generation device, the microinverter design has many challenges. Limitations that need to be considered include high-frequency leakage current generated by parasitic capacitance, the front DC voltage gain of the inverter, and switch components’ number and power consumption. In many cases, these limitations are difficult to solve perfectly simultaneously in a standard H-bridge or three-phase inverters. A front boost converter requires a high DC voltage gain with fewer switches, and Z-source or split-source structures can be applied. But these structures can lead to high leakage current. A traditional half-bridge inverter can solve the problem but with low efficiency. H5 or HERIC full-bridge inverter for PV can eliminate leakage current while more switches are used. A dual-boost common-grounded H-bridge topology can remove the leakage current and achieve a high DC voltage gain with only four switches. However, the power consumption of different switches varies greatly, which may cause an imbalance in heating and damage the device.


Introduction
Compared with fossil fuel, photovoltaic power generation has more advantages, such as simple structure, high efficiency, green and pollution-free, and sustainable development.With the development of technology in recent years, the cost of photovoltaic power generation, including production, installation, and maintenance, continues to decrease.Therefore, photovoltaic power generation will be one of the most promising renewable energy technologies [1].For integrated photovoltaic modules, each contains a microinverter connected to a PV source.This structure's advantage is that it can have continuous output voltage even if some solar panel is shadowed.However, one shortcoming is that the output voltage is low compared with the grid voltage [2].Therefore, a boost circuit is required before the photovoltaic output voltage is connected to the grid.
One way to increase the voltage is to use a transformer on the AC terminal.However, transform efficiency is extremely low, and the device is large and heavy.Therefore, a DC boost circuit at the front end of the inverter will be more popularly considered [3].However, even for a transformerless inverter with a front DC boost circuit, many issues still need to be considered when designing it.A trade-off must be made between the DC voltage gain, the leakage current elimination, the switch count and their power [4].This article describes three important challenges when designing a transformerless inverter with a front DC boost circuit.Section 2 gives some topologies for DC-DC boost circuits.Section 3 explains why leakage current occurs and gives several methods to eliminate it.Section 4 shares the problems of switches, like the number of them and their power consumption of them.Some experimental verifications have been given in sections 2 and 3 to prove the feasibility and limitations of these methods.Section 5 concludes this article.

DC boost circuit and its voltage gain
Because one additional switch is used, the H-bridge inverter has at least five switches.The power efficiency is low because of its higher switching loss.

Z-source/split-source inverter
Generally, boost circuits need switches to control active components' charge or discharge status.However, the inverter circuit already contains four or six switches for H-bridge or three-phase inverters.Therefore, a front-end DC boost circuit without additional switches can be designed by adjusting the modulation of the existing switches.
Figure 3 shows the Z-source three-phase inverter's circuit diagram and modulation method.One diode D is required, and the same six switches are used as a three-phase inverter.It has two additional reference voltages for modulation.When the carrier wave is larger than the highest reference voltage or smaller than the lowest reference voltage, the 000 or 111 null states will be replaced by a shoot-through state and short-circuited.At that time, the inductors will be charged.The voltage gain is determined by the shoot-through time [6]. Because: Then, the voltage gain:   4 gives the circuit diagram and modulation method of the Split-source H-bridge inverter.The inductor charges in 10, 11 and 01 states while only discharges in the 00 states.Then the voltage gain is mainly determined by the discharge time (00 states) [7]. Because: Then, the voltage gain: However, by adding a parasitic capacitor between the negative panel and the ground, there will be very large high-frequency leakage currents, shown in Figure 6, which greatly impact the circuit.Section 3 will focus on the cause and solutions of high-frequency leakage current.

Cause of leakage current
As shown in Figure 7 (a), The main cause of the high frequency of leakage current is the presence of PV panel-to-ground parasitic capacitance [8].  = 0.5  (12) Therefore, once the switching states change in high frequency, the voltage on the parasitic capacitor varies, leading to a high-frequency leakage current.These leakage currents will cause many hazards, including output harmonics, increasing power loss, safety problems, and EMI issues [9].Besides, because the variation of the voltage and leakage currents exist between the PV panel and the ground, people will get an electric shock when touching the solar panel [8].
Unlike ordinary half-bridge inverters, solving high-frequency leakage current for H-bridge inverters and three-phase inverters is more complex.There are some advanced methods, such as introducing auxiliary circuits to absorb the pulse power of parasitic capacitors [9].However, this session mainly focuses on two common methods to eliminate the leakage current when designing inverters.One is to create galvanic isolation by using additional switches; the other is to connect the neutral point of the DC link capacitor with the same ground of the AC output.

Additional switches
Galvanic isolation means isolating the parasitic capacitance from the main circuit using additional switches.It prevents the high-frequency leakage current caused by the sudden voltage change across the parasitic capacitance when the switch states change.The voltage across the parasitic capacitor will remain constant, and no leakage current will be generated.Two familiar methods for galvanic isolation are H5 or HERIC topology.Figure 8 explains how the H5 method works.For H5 full bridge inverter circuit, it contains one extra switch.This switch S5 turns on when in active states (1,0) or (0,1), and everything remains unchanged.Null state (0,0) is not permitted, but the null state (1,1) is permitted, then the switch S5 will be turned off.
Figures 9 (a) and (b) give the H5 and HERIC topology simulation results, respectively.Comparing them with Figure 6, it can be known that the leakage current is well eliminated.However, extra switches may consume more power in a circuit.

Common grounded
Another way to eliminate leakage current is to connect the DC neutral point and the negative terminal of the output to the same ground.It also eliminates voltage changes and reduces leakage currents.Figure 10 gives an advanced topology of a dual-boost H-bridge inverter.It contains four switches, in which the midpoint of the power supply and the output are connected to the same ground, and it can work under the conditions of both single power supply and dual power supply DC input.11, the inverter has four switching states for a dual power supply.In Figure 11(a), L2 is charging, L1 is discharging, and   is charging; In Figure 11(b), L2 is still charging, L1 is still discharging, but   becomes discharging; In Figure 11(c), L1 becomes charging and L2 is discharging; In Figure 11(d), L1 is still charging and L2 is discharging.The circuit can successfully realize a DC boost by charging and discharging the inductor and capacitor.Since the end of the DC input and output terminal are grounded together, the parasitic capacitance-voltage will not change rapidly, so the leakage current is removed.Table 1 gives the charging and discharging status of different active components under different states.
The method of modulation is shown in Figure 12.On the positive half-axis, red is the modulation signal of switch S1, and blue is the modulation signal of switch S3.They have the same common-mode offset, which will determine the gain of the DC boost.The difference between the red and blue signals is sinusoidal, ensuring that the output signal is a sine wave.It is worth noting that the common-mode signals in the positive and negative half-cycle are different.As shown in Figure 12, the DC output signal keeps constant.
This topology has four characteristics.First, it can solve the problem of leakage current; Second, it only uses 4 switches; third, it has a large voltage gain; The fourth point is that it can work in two modes of power supply, including single or dual power supply.But it still has disadvantages.As shown in the simulation results in Figure 13, the power losses of four switches vary greatly due to their special modulation method.

The number of switches
As shown in Figure 14, different inverter structures have a different number of switches.These numbers vary from three to six.Because the switch is not ideal, it will also consume power.Therefore, the more switches, the more energy loss [10].So, reducing the number of switches is important in the design process.Though the number of switches needs to be low, it does not mean the fewer switches, the better the circuit is. Figure 14(b) shows that the three-switch-three-state (TSTS) semiZ-source inverter contains only three switches.However, the voltage gain is low, which is around 2 [11].So, the premise of reducing the number of switches is not to affect the required performance of the circuit.

The power consumption of different switches
As for switch components, another problem to be considered is the power consumption of different switches.Unlike ideal conditions, in practice, the voltage drops in on state are not zero because of the resistance, and it needs delay time for switching on and off [12].So, they consume power and generate heat.The imbalance of switching power consumption will cause two problems.First, it will cause different heat generation.The excessive power of a switch will affect the device packaging and function distortion.Secondly, unequal switching power consumption means that different switches have different lifetimes.
To make the output power of different switches similar or the same, designers should pay attention to two points when designing the circuit.First, the position of switches in a circuit should have a certain symmetry.The switches of half-bridge, H-full-bridge or three-phase inverter are examples of symmetry.However, in H5 full-bridge inverter, switch S5 and the other four switches are not symmetrical.So, its power is difficult to be the same as other switches.The second thing that needs to be considered is the modulation wave.When designing the modulation method, designers should ensure that different switches have similar times and times in one period.To achieve this, the modulation reference wave of different switches should have the same shape and RMS value.

Conclusion
This article explained why a front-end boost circuit is needed, the cause of leakage current, and why the number of switches and power loss should be considered when designing a high-bridge or three-phase inverter for PV power generation.Several PV inverters are introduced, and their advantages and disadvantages are analyzed theoretically.Then, some PV inverters' topologies were simulated using the software PLECS, and their functions and drawbacks were illuminated.The experiments show that Zsource or Split-source inverters can replace traditional DC-DC boost converters without additional switching.But they create a large high-frequency leakage current.One method to eliminate the leakage current is to use additional switches in H5 or HERIC structure.The dual-boost common-grounded Hbridge topology can also remove the leakage current and gain a high DC voltage without additional switches.But it still has the drawback of the imbalance of power consumption of different switches.Therefore, when designing the PV inverter, engineers need to find a trade-off between the voltage gain, the leakage current, and the number and power consumption of switches.In all PV inverters, parasitic capacitors must be considered, and leakage currents must be eliminated.The voltage gain mainly depends on the output voltage of the photovoltaic panel and the standard voltage connected to the grid, and it could not be too low.Without affecting the function, the number of switches should be reduced as much as possible.In the case of unbalanced heat dissipation or considering the device life, the power of different switches should be equal.

2. 1 .
Tradition DC-DC boost converter Figure 1 indicates the structure of the traditional open-loop boost converter.  represent the PV source while   could be replaced by an H-bridge converter.The circuit contains a switch, a diode, an inductor, and a capacitor.The active devices regularly charge and discharge by turning the switch on and off, thus realizing the DC voltage boost.

Figure 1 .Figure 2 .
Figure 1.DC-DC boost converter.The switch status is controlled by Pulse-Width Modulation (PWM).A PWM controller is mainly a comparator.It compares the incoming error voltage to a sawtooth, while the clock signal and a line voltage determine the sawtooth.Figure 2 below shows the diagram of it and how it operates.When the error voltage exceeds the sawtooth, it outputs the signal on, or logical "one".When the error voltage is lower than the sawtooth, it outputs the signal off, or logical "zero" [5].d means the duty cycle, representing the ratio of time to period.Then   =   1 −  (1)

Figure 3 .
Figure 3. Z-source three-phase inverter (a) diagram (b) modulation.Figure4gives the circuit diagram and modulation method of the Split-source H-bridge inverter.The inductor charges in 10, 11 and 01 states while only discharges in the 00 states.Then the voltage gain is mainly determined by the discharge time (00 states)[7].Because:

Figure
Figure 3. Z-source three-phase inverter (a) diagram (b) modulation.Figure4gives the circuit diagram and modulation method of the Split-source H-bridge inverter.The inductor charges in 10, 11 and 01 states while only discharges in the 00 states.Then the voltage gain is mainly determined by the discharge time (00 states)[7].Because:

Figure 4 .Figure 5 .
Figure 4. Split-source H-bridge inverter (a) diagram (b) modulation.2.3.Simulation result of Z-source/split-source inverterSimulation software PLECS was used to verify the boost effect of Z-source and split-source inverters.90 V DC voltage source was used as the power supply.It can be seen from Figure5that both inverter structures boost the input voltage of the DC input voltage.

Figure 7 .
Figure 7. Panel-to-ground parasitic capacitors.Take the H-bridge inverter in Figures7 (b) and (c) as an example; when switch S1 is closed and switch S3 is opened (1,0), then, according to KVL:  = 2  +   (7)   +   = 0 (8)   = 0.5(  −   ) (9) When the switch S1 and switch S3 are both opened (0,0), then:2  +   = 0 (10)   +   = 0 (11) Then:  = 0.5  (12) Therefore, once the switching states change in high frequency, the voltage on the parasitic capacitor varies, leading to a high-frequency leakage current.These leakage currents will cause many hazards, including output harmonics, increasing power loss, safety problems, and EMI issues[9].Besides, because the variation of the voltage and leakage currents exist between the PV panel and the ground, people will get an electric shock when touching the solar panel[8].Unlike ordinary half-bridge inverters, solving high-frequency leakage current for H-bridge inverters and three-phase inverters is more complex.There are some advanced methods, such as introducing auxiliary circuits to absorb the pulse power of parasitic capacitors[9].However, this session mainly focuses on two common methods to eliminate the leakage current when designing inverters.One is to create galvanic isolation by using additional switches; the other is to connect the neutral point of the DC link capacitor with the same ground of the AC output.

Figure 10 .
Figure 10.Dual-boost H-bridge inverter (a) dual source (b) single source.As shown in Figure11, the inverter has four switching states for a dual power supply.In Figure11(a), L2 is charging, L1 is discharging, and   is charging; In Figure11(b), L2 is still charging, L1 is still discharging, but   becomes discharging; In Figure11(c), L1 becomes charging and L2 is discharging; In Figure11(d), L1 is still charging and L2 is discharging.The circuit can successfully realize a DC boost by charging and discharging the inductor and capacitor.Since the end of the DC input and output terminal are grounded together, the parasitic capacitance-voltage will not change rapidly, so the leakage current is removed.Table1gives the charging and discharging status of different active components under different states.

Figure 13 .
Figure 13.Power of four switches.As shown in the simulation results in Figure13, the power losses of four switches vary greatly due to their special modulation method.

Table 1 .
Charging-Discharging status of active components in different states.