Microinverters provide an effective solution to solar-energy harvesting by providing power conversion at the individual panel level. The emergence of highly integrated MCUs offers an attractive approach to microinverter design, providing an option that reduces the cost of complexity which limited widespread adoption of microinverters in the past. Today, designers can build highly efficient microinverter designs using available MCUs from semiconductor manufacturers including Freescale Semiconductor, Infineon Technologies, Microchip Technology, Spansion, and Texas Instruments, among others.
Solar-energy-harvesting systems have continued to evolve away from traditional centralized solutions (Figure 1). Unlike systems based on a single central inverter or even multiple string inverters, microinverters convert power from a single panel. In turn, the AC power generated by microinverters on each panel is combined on the output to the load.
Figure 1: Solar-energy-harvesting methods continue to evolve from central-inverter systems (left) to more distributed string-inverter systems (middle), and finally microinverters (right) built into individual panels. (Courtesy of Freescale Semiconductor)
By generating AC power on each solar panel, the microinverter approach reduces or eliminates costs associated with expensive cabling, cooling, and other facility requirements associated with central- or even string-inverter systems. At the facility level, losses in overall conversion efficiency due to differences in illumination, shading, dirt, or panel age are greatly reduced. Although the use of microinverters increases the cost of an individual panel, the solar installation overall typically has lower costs and higher-conversion efficiency.
For all their apparent advantages, microinverters have been slow to emerge as cost-effective alternatives to traditional inverter systems. In the past, the complex functionality of solar-inverter designs increased the cost of these systems enough to offset the downside of centralized-solar-energy solutions. Indeed, the need to respond to changing environmental conditions to maximize solar-energy conversion requires sophisticated system designs that were difficult to achieve at the cost and effectiveness required for use in individual solar panels.
In an ideal situation, a solar panel produces its maximum power output at a specific point on its IV curve dictated by the characteristics of the environment and the panel itself (Figure 2, left). In practice, this maximum-power point (MPP) can be elusive, shifting to different points on the power curve as shadows pass across the panel from clouds or debris collects on the panel (Figure 2, right). Sophisticated solar-energy systems use maximum-power-point-tracking (MPPT) methods to modify the operating voltage of the panel to ensure that the panel produces its maximum power output despite changing conditions.
Figure 2: Power output from a 72-cell 180 W solar panel in full sun (left) and in partial shade (right) reaches a maximum at a specific point on the power curve, but finding that maximum power point can represent a significant challenge. (Courtesy of Linear Technology)
Popular MPPT methods such as Perturb & Observe (P&O) use the simple expedient of periodically adjusting the operating voltage of the panel in incremental steps, seeking any increase or decrease in panel-operating voltage that can boost panel-power output. If an incremental increase in operating voltage resulted in lower-power output, the P&O algorithm would incrementally decrease operating voltage at the next adjustment step – and repeat this process until an incremental change in operating voltage in either direction would result in lower panel-power output. In practice, however, the appearance of local maxima in the power-output curve (see Figure 2, right) typically requires a more sophisticated approach than simple step increments.
In the past, building an effective MPPT system was a complex undertaking that could quickly increase costs and lengthen schedules as engineers dealt with corner cases such as local maxima or other factors. Today, engineers can find a wide variety of available devices able to provide complete solutions with a minimum of additional components. In fact, integrated MCUs provide on-chip functionality able to address the measurement and analysis requirements, often requiring only complementary analog circuitry for voltage and current sensing on the input and power conditioning on the output (Figure 3). MCU manufacturers typically provide associated software libraries that include ready-to-use MPPT algorithms, further simplifying the design process for microinverter designers.
Figure 3: Integrated MCUs such as the Microchip PIC16F690 provide a complete set of features and on-chip peripherals needed for digital control of a microinverter. (Courtesy of Microchip Technology)
For an MPPT implementation, suitable MCUs such as those in the Microchip Technology PIC 16F series, Spansion FM3 MB9B520M series, and Infineon Technologies XMC4000 series combine a processor core and memory with a full complement of analog peripherals. At a minimum, these MCUs provide an analog-digital converter (ADC) for measuring panel voltage and current, a voltage reference and analog comparator for accurate analog processing, and the pulse-width modulation (PWM) output needed in the voltage-converter stage.
For example, devices in the Microchip Technology PIC 16F MCU series combine an 8-bit CMOS MCU core with Flash memory and the required set of analog peripherals. The PIC16F690
MCU integrates a 12-channel 10-bit ADC, two analog comparators, programmable on-chip voltage reference, and Capture/Compare/PWM, providing 16-bit capture with resolution down to 12.5 ns and 16-bit compare with resolution down to 200 ns. Other members of the PIC 16F MCU series extend these capabilities with additional peripherals needed to provide further enhancements in a microinverter design. For example, the Microchip PIC16F913
MCU augments the base peripheral set with an LCD control module, enabling designers to provide user feedback directly on the panel.
Freescale Semiconductor’s MC56F82xx digital-signal controller (DSC) family provides a 32-bit DSP core and on-chip peripherals targeted for microinverter designs. Among its features, the MC56F82xx family is based on a 60 MIPS 32-bit 56800E core. The family’s Harvard-style architecture combines three execution units operating in parallel, allowing as many as six operations per instruction cycle. At the same time, these devices provide an MCU-style programming model and optimized instruction set. Among their peripherals, these devices include two 12-bit 8-channel ADCs, three analog comparators with integrated ADC, and Freescale’s Enhanced FlexPWM (eFlexPWM) with a broad array of control, edge-placement, and triggering capabilities.
Texas Instruments addresses the need for real-time performance with its line of C2000 MCUs comprising the 32-bit Piccolo MCU, 32-bit Delfino multicore MCU, 32-bit fixed-point DSC, and 16-bit DSP MCU series. Designed specifically for real-time applications, the C2000 MCU family provides a wide range of price/performance points in highly integrated devices able to quickly acquire analog data, perform required calculations, and adjust PWM output within a single clock cycle.
Beyond the need for an MCU with a necessary complement of on-chip peripherals and real-time performance, engineers developing high-voltage applications in solar-energy harvesting can also find themselves facing safety requirements, such as the IEC 61508 SIL-3 safety standard. To address this emerging requirement, Texas Instruments offers its Hercules RM family of safety MCUs. Built around a pair of ARM Cortex-R4F cores operating in lockstep, these devices combine ADCs, PWMs with features designed to continually monitor their own operation and provide near-instant fault detection without an impact on performance.
Placed on each individual solar panel, microinverters can offer highly efficient solar-energy conversion and reduce costs. In the past, however, the need for sophisticated designs able to extract maximum power from solar arrays precluded wide use of this decentralized approach. Today, engineers can take advantage of a broad array of MCUs that integrate the peripherals required to perform efficient energy harvesting. As a result, engineers can quickly implement cost-effective microinverter designs based on MCUs capable of meeting a wide range of application performance requirements and functional capabilities.
For more information on the parts discussed in this article, use the links provided to access product pages on the Hotenda website.