Solar Inverter Optimization

The process starts with simulation tools that enable you to look at the coupled interaction between the control software that runs on the inverter and the analog circuitry of the switching power system such as the power transistors, capacitors, inductors, transmission lines, and so on. The goal is for designers to be able to see the interaction between the embedded system they are designing and the rest of the power grid on day one when they start writing the control software.

“What we’re seeing is a strong interest in field programmable gate arrays (FPGAs) for control needs across the board for smart grid applications,” said MacCleery. “Any technology that is grid-tied needs to be out in the field for a very long time, and the FPGA provides a way to rewire the hardware down at the silicon level. That’s a big comfort not just for the designers, but also for the customer writing the check and deploying the new equipment en masse.”

Since a device can be deployed on the grid for years, the use of FPGAs makes it possible to come back later and literally rewire the internal circuitry of the system. That’s a big competitive advantage for companies when customers want to know how their systems are going to keep up with the new communication protocols and control algorithms, and take advantage of updates to comply with new standards that get passed a year from now. Rather than having to grandfather in old equipment and deal with all sorts of intercompatibility issues that weren’t discovered when it was designed, reconfigurable hardware can literally evolve digitally as the grid itself evolves to become more digital.

So in terms of providing an improved tool chain, the FPGA is really being designed as the heart of new developments for smart grid and clean renewable energy systems because reconfigurability is such a big advantage. But another reason they are increasingly popular is that they are actually now hybrid devices with embedded DSPs inside.

“The Spartan-6 LX45 FPGA from Xilinx offers 58 DSP cores inside the device, so rather than buying a multicore DSP, you can use an FPGA with an array of DSPs inside that are distributed throughout the fabric itself,” MacCleery said.

“Particularly for switching power applications like grid-tied power converters, that combination of programmable logic and DSP computing is appealing. For specific functions such as pulse-width modulation (PWM) algorithms, the combination of precisely controlled timing logic with fast digital signal processing provides an ability to reduce the EMI noise coming out of the inverter, and simultaneously extend the lifetime of components that are the common failure modes such as the power transistors by reducing the temperature fluctuations and the capacitors by reducing current ripple.”

The technology can help stabilize the temperature of the power transistors, usually IGBTs that are used in these inverters, for example. Similar to Flash memory, every time you heat and cool these devices, it takes a tick off their lifetime. So if you can implement a control scheme that uses those DSP cores to simulate the IGBT junction temperature based on the pulse width modulation that’s occurring and then regulates that temperature to be as constant as possible, then you have just extended the life of those IGBTs.

Another important development is aimed at the deployment of inverters because there are so many new designs going into the field. The growth rate of inverters and power converters is spectacular because they are used by so many renewable and alternative energy applications such as wind turbines and grid storage systems. There is also a class of equipment that regulates the power on distribution and transmission networks called FACTS or Flexible AC Transmission Systems. These systems use the same types of technologies that are used in solar inverters to convert power to AC and put it on the grid.