Advanced Electronnics for Solar Power Conversion

Solar Power Opening up Advances for Control and Conversion Modules

The dropping price of photovoltaic solar panels along with the development of improved ways to extract their optimal power is opening up a world of opportunities for embedded electronics as the control moves closer to the panel.


  • Page 1 of 1
    Bookmark and Share

Article Media

There are increasingly clear indications that we may soon be looking at a surge in renewable energy, particularly in solar power. Interestingly, this does not appear to be happening as a result of government regulation such as the recent carbon limits announced by the EPA. Rather, it seems to be a result of cold, hard numbers representing the costs of solar vs. those of coal-fired plants, with solar poised to drop below the price of coal. In fact, Barclay’s Bank recently downgraded the bonds of several utilities based on their continued use of coal. Germany is already setting an example with a huge expansion of solar energy as well as wind power.

The expansion of solar photovoltaic energy could have huge implications for the semiconductor industry as well as for the developers of embedded controllers and the build-out of the Smart Grid. The rather simple reason is that while solar panels convert sunlight into energy and present a voltage at their terminals, is doesn’t end there. The power produced by the panels must be controlled, combined with the output of other panels and converted into AC power that can be used to run homes, offices and facilities, placed on the grid and/or stored for later use.

There are relentless efforts underway to improve the efficiency of solar panels themselves to produce more power per surface area. There are also intense efforts to improve the efficiency of the control and conversion systems that ultimately deliver the power to the user. Basically, that involves taking the DC output of the panel, converting it to a target DC voltage and then sending it to an inverter to produce AC.

One of the main reasons for the DC/DC conversion is that the output voltage varies among panels so that it is necessary to get a single common DC voltage to feed into the inverter. One of the earlier and most common approaches has been to connect groups of panels serially in several “strings” using special diodes (which cause losses) to allow for the different string voltages and simply feed them into a single large inverter. This had been the traditional approach taken by large arrays at solar farms and utility plants. The DC from the strings must still be converted to a fixed DC in the inverter before being converted to AC. This is not really optimal because it is converting the output of all the panels and some panels may not be delivering the peak power they may be capable of (Figure 1).

Figure 1
The “traditional” central topology for a solar array involves feeding the output of several strings into a single, often large, central converter that converts the panel outputs to a single DC voltage that is then fed into a DC/AC inverter (Courtesy Texas Instruments).

The trend for controller/inverters is moving toward implementing them on smaller groups of panels within an array, and further toward implementing them on each individual panel. One variation on the above approach it to do DC/DC conversion at each panel using microconverters while connecting the panels in parallel and feeding their output into a central DC/AC inverter. This increases efficiency somewhat but still involves a single large inverter, which may be fine for a central utility plant but is less attractive for systems placed on residential rooftops or commercial buildings.

Upending the System

This last little item, which we can collectively refer to as “rooftop solar,” is inherently a disruptive technology. Taking power generation out of a centralized location with its distribution and billing systems and dispersing it at user locations has the potential to radically change the assumptions long associated with local monopolies (regulated though they may be) like utility companies. While relatively few users are likely to be completely off grid, the vast majority will increasingly have the opportunity to use rooftop solar to supplement the power they take from the grid by having systems that push surplus power onto the grid and remove only what they need when they need it. Another potential alternative is to add local storage to such a system so that very little if any interaction with the grid will be needed. With the cost of solar installations falling about 10% per year, we can look for some pretty rapid acceleration that will not be significantly affected by added fees and/or political resistance. Nor will it be significantly accelerated by government regulation. Market economics can be expected to act as the main driver.

Texas Instruments has produced a number of solar development kits to help engineers design and evaluate controllers, converters and inverters for solar power applications. The most recent such kit, based around TI’s C2000 Piccolo TMS320F28035 microcontroller, supports the most recent and advanced approach to solar panel power conversion and control, the micro-inverter (Figure 2). It also appears to indicate TI’s recognition of the extremely high market potential for solar inverter technology and the semiconductor products that will support it.

Figure 2
The Texas Instruments C2000 micro-inverter development kit supports the latest trend toward the implementation of solar panel arrays.

The micro-inverter is used at the output of each individual panel and integrates DC/DC conversion along with DC/AC inversion plus some control and optimization functions to be described later. According to TI MCU application engineer Manish Bbhardwaj, the most frequently used approach for rooftop solar until recently has been the string topology in which 10 to 12 panels are connected in serial and then handled by one converter/inverter. Several such strings can then be connected in parallel. Now, however, the trend is moving toward the micro-inverter, which performs DC/DC conversion and DC/AC inversion at each panel (Figure 3).

Figure 3
The micro-inverter topology implements DC/DC conversions along with DC/AC at each panel. The modules also implement AC phase lock among panels and MPPT for optimal power output under variable conditions (Courtesy Texas Instruments).

There are several immediate advantages for residential users and the companies that install rooftop systems. First, if one panel in the array is shaded or damaged, it does not affect the output of the other panels in the array. Second, the system is easy to expand. Simply adding one or more panels with their own micro-inverters adds power to the systems without having to worry about upgrading a central converter to handle the additional current.

While micro-inverters are used for grid-tie systems in which power is put on and taken off the grid as needed, the micro-inverter is not involved in the “bookkeeping.” The ultimate charge or credit from the electric utility is handled by the smart meters that are rapidly replacing the old-style electric meters that required a human meter reader. Smart meters, in conjunction with wireless and power line communications, are in many locations already capable of handling grid-tie solar panel systems.

Let’s Get Even More Efficient

As noted earlier, the actual output of which a solar panel is capable can vary according to such things as temperature, partial shading or even damage. Their maximum output also varies according to time of day. It is obviously important, therefore, to be able to extract the maximum power available from each panel throughout these changing conditions where that maximum obtainable power will vary. The micro-inverter’s internal controller has two main responsibilities. First, in doing the DC/AC conversion, it also must lock the AC output to the frequency of the grid. Second, it must see to it that the panel produces its maximum available power from moment to moment via a process called maximum peak power tracking (MPPT).

The real-time processors—in this case, the TI C2000 series—use A/D converters for input and pulse width modulators (PWMs) for output to adjust the DC/DC converter and the DC/AC inverter by changing the PWM duty cycle according to the load. Thus the DC/DC conversion happens at the optimal power output. The AC output voltage is the same as that of all the other panels, but each panel then delivers the maximum power of which it is capable in a process that is continuously updated.

Two things appear to be converging: the dropping price of solar panels along with their constantly improving efficiency, and the advancing capabilities of the electronic components and modules that can produce optimal useable electrical power in a vastly distributed world of grid-tied systems, systems with increasing storage capabilities, and a growing but much smaller number of off-grid systems. The opportunities for even more improved control and conversion technologies do indeed seem bright.

Texas Instruments
Dallas, TX
(972) 995-2011