Research into solar cells has succeeded in improving the efficiency of the process of converting photons into electrical energy. Laboratory cells have demonstrated efficiencies beyond 20% and scientists are confident they can push this figure much further. However, the efficiency of the cell itself is only part of the story.
The metric that truly matters is the number of kWhr a year that a panel will deliver in real world conditions. In addition to the cells, the electronics tucked away underneath each panel play an increasingly crucial role in that equation. Efficiency tests assume a constant level of light but, even under ideal conditions, this never happens. Only during the middle of the day will the light falling on a fixed module be at the optimum level and angle. For the rest of the time, the energy output from the module will be lower. Without the right electronics and system architecture, the output can be much lower than expected, even before factors such as shading are taken into account. Shade falling on less than 3% of the area of a solar installation can reduce its output efficiency by more than 15%, according to tests carried out by the US National Renewable Energy Laboratory. This is because shade on one part of a panel can reduce the output of subsequent cells within the panel if they are on the same circuit. Each photovoltaic (PV) module has a characteristic current-voltage (IV) curve that depends on temperature and incident light. A typical silicon-based module might generate a high voltage, but very low current on a cold, drab winter day. As the light level increases, the voltage will drop slightly but current will rise dramatically until it reaches an asymptote. Beyond that, current might rise slightly further but, at this point, voltage starts to fall dramatically. As a result, the 'knee' of the curve presents the optimum operating point for the module. Temperature also affects the peak output efficiency of a module: high temperatures cause the output voltage to drop. As a result, even when they should be at peak efficiency during periods of intense sunlight, PV panels can sufferfrom drops in conversion efficiency if the electronic circuitry does not compensate for the voltage drop. There are two ways to prevent efficiency losses. As heat is a problem for PV modules, the design should focus on channelling heat away from the cells. The second angle is to ensure that conversion to grid-compatible electricity is conducted at a suitable voltage level and only in modules that are, at any given point in time, going to be net contributors of electrical energy. To deal with the heat problem directly, power transistors and diodes developed for the PV market focus on the efficient transfer of heat energy. An example is Microsemi's SPx family of low-profile power transistor modules. These power modules incorporate copper baseplates to conduct heat away. It is also important to ensure the power transistors themselves do not generate heat unnecessarily. For example, Microsemi's developed the MOS 8 family of IGBTs to minimise conduction losses, so they do not contribute to self heating that might be passed on to the solar cells. To ensure that only productive cells are included in the array, modules typically incorporate bypass diodes to switch out cells that might consume, instead of generating, electricity. Ideally, a bypass diode has a low forward-bias voltage drop to allow current from upstream panels to flow easily to the inverter which converts the energy to grid compatible electricity. The LX2400 solar bypass device, for example, uses Microsemi's CoolRUN technology to cut resistance that causes forward voltage drop and therefore minimises heat generation. The voltage drop of the LX2400 is just 50mV at 10A. Furthermore, the temperature rise for such a high current is only 10°C – preventing efficiency-sapping heat increases near productive cells. For productive cells, the key to efficiency in PV module design is a flexible conversion scheme that responds to the modules' change in IV properties with environmental conditions. The aim is to perform maximum power point tracking (MPPT) – the power point being the voltage and current at which optimal transfer of energy can take place. Traditionally, architectures have deployed a single MPPT engine in a common inverter. While this is simple, the conditions will not be optimal for all panels in an array. Depending on shading or dirt build up, each panel has its own characteristic maximum power point at any point that will not be reflected by the conditions set by the common MPPT engine. Deploying an inverter within each module allows a finer level of granularity in terms of MPPT. This means more circuitry within each panel, but the architecture can ensure a faster payback, particularly when shading from nearby buildings is an issue. By offering finer control over power conversion, it becomes more practical to install modules on surfaces that might experience some level of shading because these modules will not affect the optimum power conversion of those rarely shaded. Inverter designers can choose from a variety of MPPT algorithms – each has advantages and drawbacks. One technique is 'perturb and observe', in which voltage or current are changed and the change in power output noted. If the output increases, the perturbation moves in the correct direction. If not, the next perturbation will be in the opposite direction. However, this technique is susceptible to oscillations if lighting conditions change quickly or the steps are too large, particularly at low levels of irradiance. An alternative approach, incremental conductance, also uses small changes to feed data to a control loop that calculates the direction in which the power point should be moved. It generally provides a better guide on the direction in which the maximum power point is moving, but is more computationally intensive and thus may increase hardware costs. Typically, the heart of the solar inverter is a programmable microcontroller (mcu) running a number of control loops that implement the switching algorithms and MPPT functions. The mcu's programmability makes it possible for PV system manufacturers to differentiate their offerings – changes to the MPPT algorithm can deliver efficiency improvements that look small on paper, but which make a big difference to payback over the operational life of the module. Field programmable gate array (FPGA) technology provides a way to assist the mcu when running more complex MPPT algorithms and to improve response times – it even makes it possible to provide field upgrades as MPPT technology improves. The FPGA can be configured to act as a coprocessor for the core mcu, not just offloading repetitive calculations, but also accelerating them and reducing the time between power-point updates. The SmartFusion customisable SoC makes it possible to combine a high-performance microcontroller element – using the ARM Cortex-M3 processor core – with FPGA logic. Performance can be further optimised through the use of the analogue compute engine (ACE). This combines a sample-sequencing engine (SSE) with a post-processing engine (PPE) to offload the job of reading analogue inputs from the Cortex-M3. The SSE captures data from the analogue inputs, passing them to the PPE, which can perform functions such as low-pass filtering to remove noise, and transform the data into a format convenient for the processor. A combination of solar-oriented power components and system-logic devices is responding to the demand for solar energy – and the greater levels of efficiency needed to ensure faster payback as subsidies and incentives such as feed-in tariffs reduce over time. Semiconductor technology is ensuring that the high efficiency of modern solar cells does not go to waste.