Solar Inverter Architecture Components Revealed (with Pictures)

　　Starting from the DC input of the solar panel, through the DC-AC conversion process, to the AC output delivered to the grid, what characteristics this design needs to achieve to meet the various safety and other performance standards, as well as the strict requirements of power companies for the signals entering their grid.

　　We take a typical solar inverter SMA "Sunny Boy" as an example to see the main components and parts selection in the design, including EMI suppression capacitors from Vishay to TMS320F2812 DSP from TI. The design places special emphasis on isolation and protection, and wisely uses optical isolation MOSFET gate drivers such as Avago's HCPL-316J and HCPL-312J.

　　Photovoltaic power systems include multiple components, such as photovoltaic panels responsible for converting sunlight into electrical energy, mechanical and electrical connections and mounting, and solar inverters. Solar inverters are key to transmitting solar-generated electricity to the grid. Figure 1 shows a basic, but very complete, photovoltaic system block diagram.

Figure 1: Complete photovoltaic system block diagram

　　What is a solar photovoltaic inverter?

　　The main function of the inverter is to convert the variable DC voltage from the sun stored in the photovoltaic panel or battery into a specific AC voltage and frequency used by the power supply equipment, and then feed it back to the grid. Of course, this AC output varies from region to region. In North America, it is 60Hz/115VAC, while in most parts of Europe, it is 50Hz/230VAC.

　　SMA Solar Technology AG, headquartered in Germany, has developed the "Sunny Boy" series of solar inverters. The inverter motherboard shown in Figure 2 is used in the transformerless versions of the "Sunny Boy" 3000TL, 4000TL and 5000TL, which are suitable for 3kW, 4kW and 4.6kW AC output power systems (@230V, 50Hz) respectively.

　　The inverter mainboard uses multi-string technology, and with two independent DC converters, it makes highly complex generator configurations easy to implement. The input section is shown in the lower left quadrant of Figure 2. Both DC inputs use Vishay's EMI suppression capacitor #339MKP as part of the filter, which also includes a DC common mode filter inductor wound on a common magnetic core and a 15μF boost converter smoothing capacitor #MKPC4AE series, as shown in the lower left quadrant of Figure 2.

　　Still on the DC input side, two relays are used to monitor the insulation resistance in pure IT AC systems in accordance with IEC 61557-8. See the upper left quadrant of Figure 2. What needs to be measured is the insulation resistance between the system line and the system ground. When it drops below an adjustable threshold, the output relay switches to the fault state. With the help of these relays, a superimposed DC measurement signal can be used to perform the measurement function. From the superimposed DC measurement voltage and its resultant current, the value of the insulation resistance of the measured system can be calculated. Note the Hall effect current sensor in Figure 2.

　　One of the most impressive features on this SMA inverter motherboard is the use of extremely high-quality active and passive components, which enhances the reliability and performance of this inverter design.

Figure 2: SMA's "Sunny Boy" series solar inverter motherboard

　　Maximum Power Point (MPP)

　　The first DC function encountered in this signal chain is the maximum power point (MPP).

　　The inverter works to compensate for environmental conditions that affect power output. For example, the output voltage and current of a PV panel are extremely sensitive to temperature changes and the light intensity per cell area (see "irradiance"). The cell output voltage is inversely proportional to the cell temperature, and the cell current is proportional to the irradiance. This variation, along with other key parameters, causes the optimum inverter voltage/current operating point to shift significantly. The inverter solves this problem by using closed-loop control to maintain operation at the maximum power point (where the product of voltage and current is at a maximum). SMA uses OptiTrac Global Peak maximum power point tracker. With this added capability, OptiTrac, a proven work tracker management system, can find and use the optimal operating point to achieve good output yield even when the PV plant is partially shaded. The TI DSP controller is the brains of maximum power point tracking (MPPT).

　　The most common method for determining the MPP is to perturb the panel’s operating voltage with a controller and observe the output during each MPPT cycle. This method continuously oscillates around the MPP over a large range to avoid localized, misleading changes in the power curve due to cloud cover or other conditions. This perturbation-and-observe method is inefficient because it oscillates away from the MPP during each cycle. The incremental inductance method is an alternative that addresses the problem of the power curve’s derivative being zero, defined as the peak, and then stabilizes to a resolved voltage level. However, while this method avoids the inefficiencies caused by oscillations, it creates other inefficiencies because it stabilizes at a local peak rather than the MPP. Combining the two methods maintains the level of the incremental inductance method while periodically scanning over a larger range to avoid selecting local peaks. This method is the most efficient, but it also places very high demands on the controller’s performance.

　　Figure 3 shows how the determination of MPP varies with different conditions.

Figure 3: MPP under various weather, time of day, and panel thermal conditions (Courtesy of TI)

　　A capacitor is usually used to store the energy that must be stored and extracted by the inverter. This capacitor is usually located on the PV bus and must be large enough to control the bus ripple voltage. Otherwise, the ripple may impair the accuracy of the MPPT.

　　Electrolytic capacitors are well suited for controlling ripple because they have low equivalent series resistance (ESR) and high capacitance. Along the top edge of the PCB in Figure 2 is the smoothing capacitor bank.

　　Boost DC-DC Step-Up Converter

　　The next device is the stepper DC-DC converter, which is responsible for stepping up the DC input to the switching MOSFET bridge so that the inverter can efficiently generate a 230V, 50Hz AC sine wave for transmission to the grid. This DC-DC step-up converter, along with the H5 switching bridge, is contained in a separate power module that is attached to the back of the inverter motherboard and can be well cooled to the base. This module will be installed in the middle of the upper part of the motherboard in Figure 2 during the final assembly.

　　Figure 4 shows the basic DC/AC conversion circuit or inverter in a typical transformerless configuration system. In it: the DC/DC converter steps up or down the input PV voltage, adjusting its output to achieve maximum efficiency in the DC/AC conversion stage; capacitors provide further voltage buffering; the IGBTs or MOSFETs in the H4 bridge generate AC voltage using a switching frequency in the range of 20kHz; the coil smoothly switches the AC voltage into a sinusoidal signal for generating AC output at the grid frequency. Transformerless Inverter 　　Technology

　　The idea of ​​transformerless switching technology existed long before the photovoltaic market developed. Equipment engineers knew that a pair of field-effect transistors worked most efficiently when they were either fully on or fully off, because no current was flowing through them and they produced no power losses. Therefore, the theoretical efficiency of amplifying an ideal square wave could reach 100%. The signal was modulated by a higher-frequency square wave, and the result was pulse-width modulation (PWM), a circuit known as a Class D circuit. In this way, it was possible to achieve efficient DC-DC conversion, or switching from DC to AC. For solar inverters, this technology was not suitable in the past due to the high cost of MOSFET and IGBT devices. However, as these devices continue to become cheaper and faster, the technology is becoming more cost-effective than analog switching with a large number of copper and iron devices. This technology can also be used to build electric vehicles.

　　Transformerless inverters have been available in Europe for several years now, and SMA received UL certification to sell such products in the United States in August 2010. This certification applies to SMA's transformerless inverters Sunny Boy 8000TL-US, Sunny Boy 9000TL-US and Sunny Boy 10000TL-US, and guarantees compliance with the UL 1741 standard for photovoltaic and battery-powered inverters, which for the first time covers the regulatory requirements for transformerless inverters. Compared with devices with galvanic isolation, transformerless inverters are much simpler; and, with advanced switching circuits, transformerless inverters can provide a wider operating voltage range than traditional inverters.

Figure 4: Transformerless DC/AC conversion circuit - inverter (Courtesy of TI)

　　The negative consequence of not having galvanic isolation is that a ground fault can occur, damaging the inverter and causing sparks. In the case of a transformer, if the secondary circuit is shorted, all current will flow through the primary circuit and will be stopped (hopefully) by the thermal cutout once the transformer overheats. In the case of no transformer, if there is no protection circuit or the protection circuit fails, fails to detect the ground fault or trips, the high power MOSFET or IGBT will fail immediately in a catastrophic manner. Fortunately, such events are very rare and all such inverters must comply with UL 1741 requirements to have ground fault protection. Regardless, this function must be maintained to ensure that the backfeed current in the event of an undetected ground fault is taken into account when sizing the combiner and isolation fuses.

　　If accurate and simple calculations can be performed, transformerless inverters have few disadvantages and offer many advantages.

　　However, PV inverters have many other key functions.

　　PV inverters also provide a grid disconnect function to prevent the PV system from supplying power to a disconnected facility; that is, the inverter remains online during a grid disconnect or when power is transmitted through an unreliable connection, which can cause the PV system to feed back to the local equipment transformer, generating thousands of volts at the equipment end and endangering workers. Safety standards specifications IEEE 1547 and UL 1741 require that all grid-connected inverters must disconnect when the AC line voltage or frequency is not within specified limits, or completely shut down when the grid is no longer present. When reconnecting, the inverter will not transmit power until the inverter detects the rated equipment voltage and frequency for more than 5 minutes. As shown in the upper right quadrant of Figure 2, four LF-G relays rated at 22A and 250VAC are used.

　　But this is not the inverter's final duty. In addition to the above tasks, the inverter supports manual and automatic input/output disconnection during operation, EMI /RFI conducted and radiated interference suppression, ground fault interruption, PC compatible communication interface ("Sunny Boy" series has Bluetooth function), and many other functions. The inverter is installed in a rugged enclosure to work at full power outdoors for more than 25 years! A typical single-phase photovoltaic inverter like the SMA motherboard uses a digital power controller, DSP, and a pair of high-side/low-side gate drivers to drive a full-bridge PWM converter. This inverter and many high-performance inverter applications use the full H-bridge topology because it has the highest power carrying capacity of any switch mode topology. SMA uses H5 technology. The fifth power semiconductor device between the output capacitor and the H-bridge prevents the excitation oscillation loss of the charge and significantly reduces the power loss again. Compared with the classic inverter bridge circuit (H4 topology), H5 has achieved significant improvement, with a maximum conversion efficiency of 98%. To prevent power fluctuations from the PV generator, the architecture disconnects the DC side from the AC side during the inverter's idle operation.

　　Compared with the H4 bridge in Figure 4, the H5 topology shown in Figure 5 only requires one more switch device. Switching devices T5, T2, and T4 operate at a high frequency of about 20kHz, and T1 and T3 operate at the grid frequency of 50 Hz. During the idle operation, T5 is open to disconnect the DC and AC sides. The forward current idle operation path passes through the reverse diodes of T1 and T3, and the reverse current forms a loop through the diodes of T3 and T1.

Figure 5: SMA's H5 bridge topology

　　The PWM voltage switching action forms a discrete noisy 50Hz current waveform at the full-bridge output. The high-frequency noise components are filtered out and a medium-low amplitude 50Hz sine wave is generated. The H-bridge operates using an asymmetric unipolar modulation method. The high side of the asymmetric H-bridge should be driven by a 50Hz half-wave depending on the polarity of the main line; while the corresponding low side is PWM modulated to form a sine wave. The right side of the inverter motherboard shown in Figure 2 is the AC output filter section with EMI suppression capacitors. The output sine wave filter containing large inductors will also be fixed to this area on the motherboard to achieve the AC filter.

　　There are many design trade-offs to be made when designing a PV inverter, and making the wrong trade-offs can lead to headaches for the designer. For example, the need for PV systems to operate reliably and at full output for at least 25 years at a competitive price forces designers to make cost/reliability trade-offs. PV systems require highly efficient inverters because more efficient inverters generate less heat and last longer than their less efficient counterparts, and they save money for PV system manufacturers and users. SMA has done an excellent job in this regard.

　　Control Architecture

　　The "brain" of the inverter is its controller, typically a digital power controller (DPC), or in this case a digital signal processor (DSP). DSP-based controllers, such as the TI TMS320F2812 used in this design, provide the advanced computing performance and programmable flexibility required for real-time signal processing in solar inverters. Highly integrated digital signal controllers help inverter manufacturers launch more efficient and cost-effective products that can meet the rapidly growing demand for solar applications in the coming years.

　　The inverter's control processor must address a number of real-time processing challenges in order to effectively execute the precise algorithms required for efficient DC/AC conversion and circuit protection. Although MPPT and battery charging control only require near real-time response, they also involve advanced processing algorithms. Digital signal processors that combine high-performance DSPs with integrated control peripherals provide an excellent solution for real-time control of the DC/AC conversion bridge, MPPT, and protection circuits in solar inverters.

　　DSP controllers inherently support fast mathematical calculations in real-time control algorithms. Integrated peripherals such as analog-to-digital converters (ADCs) and pulse-width modulation (PWMs) can directly sense inputs and control power IGBTs or MOSFETs, saving system space and cost. On-chip flash memory facilitates programming and data acquisition, and communication ports enable networking of devices such as meters and other inverters to simplify design. The high efficiency of DSP controllers in solar inverters has been proven by design, cutting conversion efficiency losses by more than 50% while significantly reducing cost. Typically, controller firmware is implemented in a state machine format to achieve the highest execution efficiency of non-blocking (pass-through) code, which prevents execution from accidentally entering an infinite loop. Firmware execution is hierarchical, with the highest priority functions generally being serviced more frequently than lower-level functions. In PV inverters, isolated feedback loop compensation and power switch modulation are usually the highest priority, followed by critical protection functions to support safety standards, and finally efficiency control or maximum power point (MPP). The remaining firmware tasks are mostly related to optimizing operation at the current operating point, monitoring system operation, and supporting system communications.

　　Integrated functionality keeps the system operating cost-effective. TI's TMS320F2812 controller has an ultra-fast 12-bit ADC that supports up to 16 input channels for current and voltage sensing to achieve a regular sine wave. For safety, the ADC can also provide current sensing in residual current protection devices (RCDs).

　　12 individually controlled enhanced PWM (EPWM) channels provide variable duty cycle for high-speed switching in the converter bridge and battery charging circuits. Each EPWM has its own timer and phase register to program the phase delay, and all EPWMs can be synchronized to drive multiple stages at the same frequency. Multiple timers give multiple frequencies, and fast interrupt management supports additional control tasks. Multiple standard communication ports, including CAN bus, provide simple interfaces to other components and systems.

　　isolation

Figure 6: Alternative energy systems require an isolated connection (red) between the high-voltage power circuits and the controllers that manage the power flow. (Courtesy of Avago)

　　In the center of the SMA inverter motherboard, we find five Avago isolated gate drivers. See Figure 2.

　　The two isolated MOSFET drivers that control the switching of T1 and T3 at a line frequency of 50Hz are Avago's HCPL-316J, 2.5A gate drive optocouplers with integrated (VCE) desaturation detection and fault status feedback. The other three isolated MOSFET drivers that control the switching of T2, T4, and T5 at higher frequencies are Avago's HCPL-312J, 2.5A output current MOSFET gate drive optocouplers. The H5 configuration is shown in Figure 5.

　　Especially in transformerless inverter designs, optocouplers provide reinforced isolation and fail-safe protection in the event of a fault condition.

　　Why is reactive power control important in PV inverters?

　　The "Sunny Boy" 3000TL/4000TL/5000TL inverter with reactive power control is now available on the market.

　　Reactive power occurs any time energy is transferred via alternating current. Reactive power is growing in importance to solar engineers and PV system operators for both larger and smaller systems. The most important reality is this: Reactive power is not a problem at all. In fact, it is a solution to some problems.

　　Starting from July 1, 2010, photovoltaic systems fed into the grid at medium voltage in Germany must provide reactive power to the grid. This has been mentioned in the 2008 edition of the medium voltage equipment access specification of the German Federal Energy and Water Association (BDEW). For low-voltage grids, more stringent requirements are under discussion.

　　How does reactive power develop?

　　For direct current, the equation is simple: power is the product of voltage and current. However, for alternating current, things are a little more complicated because the magnitude and direction of both the current and voltage change regularly. See Figure 7.

Figure 7: In addition to the received PV active power, the required reactive power is also generated in the inverter. The geometric sum of these two is the apparent power; it is decisive for the inverter design. (Courtesy of SMA) 　　In the mains grid, both have a sinusoidal trajectory with a frequency of 50 or 60 Hz. As long as the current and voltage are "in phase", i.e. move at the same pace, the product of these two oscillating factors will also be an oscillating output with a positive average value - pure active power (Figure 8a).

Figure 8a: When there is no phase shift, the product of current i and voltage u is an oscillating, but always positive output - pure active power (Courtesy of SMA)

　　However, once the sinusoidal trajectories of current and voltage are shifted in opposition to each other, their product will be an output in which positive and negative signals alternate. In the extreme case, current and voltage are phase-shifted over a quarter cycle: the current always reaches its maximum intensity when the voltage is zero—and vice versa. The result is pure reactive power, with positive and negative signals completely canceling each other (Figure 8b).

Figure 8b: At a 90° phase shift between current i and voltage u, an alternating positive and negative output with zero average value is produced – pure reactive power (courtesy of SMA)

　　This phase shift usually occurs in both directions. Phase shift occurs when a coil and a capacitor are placed in an AC circuit - this is usually the case: all engines and transformers have coils (for inductive shift); capacitors (for capacitive shift) are also common.

　　Multicore cables act like capacitors and high-voltage overhead cables can be considered as extremely long coils. Therefore, a certain degree of phase shift (i.e. reactive power) is unavoidable in AC grids. The measured parameter of the phase shift is the shift factor cos(φ) which has a value between 0 and 1. It can be used to easily convert to output values. The unit of reactive power is called var (VAR) instead of watt (see formula 1)

Formula 1: Calculate reactive power using the Pythagorean theorem for a right triangle. (Courtesy of SMA)

　　What is the impact of reactive power in the grid?

　　In reality, only active power is usable power. It can power a machine, make a light bulb glow, or turn on an electric heater. Reactive power is different: it cannot be used and therefore cannot power any electronic equipment. It simply moves back and forth in the grid, acting as an additional load. In addition, all the cables, switches, transformers, and other parts also need to take reactive power into account.

　　This means that they need to be designed for apparent power, which is the geometric sum of real power and reactive power. Resistive losses during energy conduction occur on top of the apparent power; therefore additional reactive power results in greater conduction losses.

　　Looking ahead

　　Photovoltaic systems are a relatively new technology in the field of power generation. Like other emerging technologies, photovoltaic systems will change rapidly as the technology matures. Therefore, photovoltaic systems will undoubtedly continue to change to meet market demands for higher capacity, lower costs, and higher reliability. In time, photovoltaic inverters will expand their capabilities, and designers will require more integrated, application-specific component-level parts. As these demonstrate, photovoltaic systems will become widely popular and eventually become a viable mainstream facility, significantly reducing our dependence on fossil fuels.