Maximising efficiency from renewable energy sources
Digital signal controllers can deliver significant efficiency benefits for power inverters in green energy installations.
The need for alternative forms of power generation are well established and with a current focus on renewable sources, solar panels and wind turbines are receiving considerable attention. As an example, there are now commercial companies offering free panels with installation and maintenance contracts for domestic and commercial users, which provide the property owner with power when they need it, while feeding surplus power back in to the national grid.
This form of energy generation is, unfortunately, susceptible to inefficiencies. Photo voltaic (PV) cells offer limited conversion efficiency under the best conditions, but maximising their potential also requires significant system intelligence, in order to provide the best return on investment.
Using complex algorithms, it is possible to achieve high levels of output power even under less than ideal conditions and the power conversion stage is key in this.
Typically a solar panel installation will use a number of PV cells connected in series and parallel, along with a number of DC/AC inverters to convert the PV output to a mains-rated voltage for use within domestic or commercial premises, or to feed back in to the national grid. Some installations may also use batteries to store energy during good weather conditions, to provide energy during periods of cloud cover or even through the night. In these installations a number of DC/DC converters will also be used.
Maximising the efficiency of the PV output is crucial and is typically controlled by the duty cycle of the pulse width modulation (PWM) signal controlling the DC/DC and/or DC/AC converters. Achieving the most efficient power conversion requires close monitoring of the power (product of the voltage and current outputs) from the PV cell. For this a microprocessor is used, along with a number of ADCs. The processor must be capable of calculating the PWM signal to achieve optimal conversion conditions, predominantly using a technique known as Maximum Power Point Tracking (MPPT).
Due to the constantly variable conditions, the MPPT process is cyclic; it must be performed reiteratively to compensate for changes in output caused by temperature, cloud cover, the time of day and even seasonal variations. The objective is to continuously tune the inverters, through PWM control, to match the peak voltage and peak current outputs of the cell(s). This involves measuring and converting the voltage and current outputs and passing their values to the MPPT algorithm, calculating the optimal control and feeding that back in the form of a PWM control signal. While this can be achieved using several discrete devices, a digital signal controller such as Texas Instruments’ TMS320C200 family of DSP controllers has the right combination of peripherals, processing power and architectural features to perform this complex process in real time in a single chip, thereby delivering the maximum possible efficiency for a PV installation.
The Maximum Peak Power algorithm is the most commonly used solution for maximising power output, but is inherently variable. Another solution sometimes employed is the incremental inductance algorithm, which solves the derivative of the voltage/current product (power) curve for 0, which is by definition a peak. However this method is prone to errors caused by settling on ‘false’ peaks. A more elegant solution is to employ a combination of the two and while this is the most efficient, it also requires the most processing power.
TI’s TMS320F28x digital signal controller is ideally suited to this task; it uses a 32bit device that can deliver up to 150 MIPS performance and integrates a range of peripherals that enable a highly efficient solar panel inverter. Its use has already been reported to have cut conversion inefficiency by up to 50% in some installations.
Article by Helga Stevenson - written for DesignSpark