Overview

The photovoltaic market requires DC/DC converters that adjust the solar input voltage to the DC-link or battery level and DC/AC converters (inverters) to deliver the solar energy to the public grid.

There are a few key architecture types of solar inverters in the market:

  • Single-phase string inverters for residential areas with a few kVA. See figure 1.
  • Three-phase multi-string inverters for residential, commercial and utility-scale systems. Most consist of an input DC/DC boost converter with one or more Maximum Power Point (MPP) trackers and an inverter in the output stage. Sub-100 to 200kVA levels are typical. See figure 2.
  • Microinverters consist of a single solar panel with one single inverter. There is no high voltage DC power here as in string inverters. Reliability is better here since if one or two microinverters fail, the remaining ones will continue to produce power. See figure 1.
  • DC power optimizers fall in between string inverters and microinverters with typical voltage of 400V. See figure 1.
Figure 1

Figure 1: Central (or string) inverters, microinverters and power optimizers. (Courtesy of solartribune.com)

Figure 2: Three-phase, multi-string inverter (Schematic courtesy of Semikron)

In April 2019, the United States passed a significant milestone. In that 30-day period, renewable energy output surpassed that of coal generation in the U.S. for the first time in history. On Business Insider, an expert states, ‘The inverter breakthrough has solar bulls making what I call “hockey stick” graphs, projections of rising growth over several years. Costs for solar power have fallen by 90% in the last decade, but by 2022 they could fall another 34%.’

Inverter design options: Silicon vs Wide Bandgap

It’s all about Size, Weight, and Power (SWaP)

WBG semiconductor materials enable smaller, faster, more reliable power electronic components at higher efficiency than their Si-based counterparts. These capabilities lead to reduced weight, volume, and life-cycle costs in a wide range of power applications.

WBG has lower intrinsic leakage currents and higher operating temperatures than Si devices. This is due to higher bandgaps (Eg) in WBG devices than in Si.

WBG has higher breakdown voltage and lower power loss than Si due to the 10x smaller WBG drift region. This allows for a lower Rds(on) for a given area with WBG semi’s vs Si.

WBG operates at faster speeds and lower switching losses than Si, enabling design architectures with smaller magnetics and heat sinks thus reducing size, weight and total BOM cost.

WBG Advantages in Solar Inverters

The conversion from DC to AC power, in solar energy designs, is done using Inverters, which are expected to be extremely efficient (over 97%) and to last for a very long time (in some cases over 25 years). To achieve these performance parameters, Silicon Carbide (SiC) and Gallium Nitride (GaN) WBG power transistors need to be employed.

Designing solar inverters with WBG devices will improve architectures with elimination of up to 90% of power losses in a DC-to-AC power inverter.

Solar inverters can be located in harsh, high ambient temperature environments. WBG’s higher operating temperatures can enable higher reliability of these systems. Solar inverters employing WBG devices will also be lighter in weight due to smaller magnetics and thermal requirements. Lighter, more compact inverter design allows for easier lifting and maneuverability during installation and potential repair. Figure 1 shows a size comparison of a 50kW Si IGBT- based design vs SiC-based design. Typically 60% smaller in size and 10x lighter in weight as shown in Figure 3, below.

Figure 3

Figure 3: Silicon carbide enables a smaller, lighter, higher power density and lower overall system cost solar PV inverter compared to silicon.
(Image courtesy of Wolfspeed)

Power module or discrete components?

The topology and components have to be carefully selected to strike the right balance between cost and performance. Component costs can be a concern either way, but if we factor in manufacturing costs, production yield, power density, Improved reliability and thermal performance into the equation, the use of power modules can be an appealing proposition.

A power module can offer a validated and specified solution, while a discrete design enables more customization to the application.

Figure 4: Wolfspeed SiC half-bridge module illustrating compact layout configuration

Power modules can potentially integrate more functionality, sometimes in a smaller overall footprint. Modules allow for multiple die to be placed close together, in a compact layout configuration, to maximize performance and space. Additional components like SiC SBD diodes, gate resistors and capacitors can also be included. Multiple standard and custom topologies are available including half-bridge, full-bridge, six-pack and 3-level just to name a few. See Figure 4, above, showing the inside of a half-bridge power module using 5 parallel SiC in each switch position.

Discrete designs are typically less expensive in lower power applications where fewer discrete parts are used in parallel to meet current requirements. As power increases more discrete devices are needed and assembly costs and potential lower reliability increase. Discrete designs can also allow for a more flexible board layout around fixed component areas in an inverter design.

Conclusion

When solar power was in its infancy, inverters tended towards centralization with capacities in excess of 100 kW. In more modern times this trend has changed with operators preferring to use strings of sub-100 kW inverters. In all cases, the architecture is similar to a DC/DC boost converter to increase the voltage from the PV panel and a DC/AC inverter that generates an AC voltage at the correct frequency for the local grid (50 Hz / 60 Hz). The system also adds protection circuitry and sophisticated monitoring/control to ensure maximum efficiency.

While the topology chosen for the inverter will have an impact on the efficiency, the primary semiconductor switching devices (WBG and Si MOSFETs, IGBTs and diodes) are critical in achieving the efficiency needed for today’s solar power applications. In the early days, silicon (Si) had been the primary material used and, through years of incremental innovation, this technology has reached a point where very little further enhancement is possible.

Semiconductor manufacturers have been exploring other materials to build future switching devices from. Wide bandgap (WBG) materials including gallium nitride (GaN) and silicon carbide (SiC) have risen because of their properties which are ideally suited for developing efficient semiconductor devices.

WBG materials have inherently lower on-resistance than Si-based devices, reducing static losses when conducting continuously. As switching frequencies rise to reduce the size of magnetic components, WBG technology further improves efficiency as the gate charge is reduced compared to silicon, reducing dynamic losses as well.