SPWM regulates voltage in DC-AC inverter designs

Article By : Alberto Leibovich and Pablo Leibovich, Dialog Semiconductor

SPWM is widely used in power electronics to regulate voltage in DC-AC inverter designs.

Renewable energy continues to be a massive trend around the world. As methods for capturing wind, solar, and other forms of renewable energy have developed, the cost and efficiency of renewable energy systems have become more appealing to both companies and consumers. In fact, in 2016, the global monetary investment in renewable energy fell to the lowest number in years, while simultaneously the record was broken for the largest number of renewable energy installations in a given year.

Amongst the components used for renewable energy resources, the inverter stands out as a uniquely critical system component. As most renewable energy is generated in DC, an inverter plays a key role in converting DC into AC for effective integration into existing power grids. In hybrid power systems—that combine different renewable resources—and micro-grid systems, the use of inverters is essential.

Renewable energy inverters also play a critical role in industrial applications where single-phase and three-phase motors and other rotary machines are used. Variable frequency and voltage—obtained from an inverter—embodies the principle of autonomous control in this type of equipment.

To implement the power conversion, DC-AC inverters usually apply the pulse width modulation (PWM) technique. It’s a useful technique wherein switches like power MOSFETs are controlled with pulses of variable widths. In order to obtain an automatic control and regulation, the PWM technique is used to maintain the AC voltage output of the inverter and its frequency at the nominal value independent of the output load.

Many studies and technologies have been developed to obtain a better response from the inverter control system. Conventional inverters change the output voltage according to the changes in the load. To reduce the sensitivity of the output voltage to load changes, PWM-based inverters regulate the output voltage by changing the width of the pulses generated at a comparatively high frequency. As a result, the output voltage depends on the switching frequency and pulse width, which varies according to the value of the load connected at the output. With this type of regulation, the inverters provide a constant nominal voltage and a frequency independent of the output power.

Several methods of generating PWM have been studied. The efficiency parameters of an inverter, such as switching losses and harmonic reduction, are the main factors considered in any modulation technique evaluation. So, sinusoidal PWM (SPWM) is widely used in power electronics as the modulation method for inverter designs.

SPWM inverter concept

A three-phase wave bridge inverter is the most commonly-used inverter topology in industrial applications. To simplify the concept, a single-phase version is analyzed. The single-phase design includes switching transistors or IGBTs on each arm of the H-bridge with antiparallel freewheeling diodes to discharge when the switch is turned off. Its schematic is shown in Figure 1.

H-bridge circuit shows building blocks of an inverterFigure 1 This basic H-bridge circuit shows the key building blocks of an inverter. Source: Dialog Semiconductor

The transistors—usually power MOSFETs—are identified as S1, S2, S3, and S4. The switches are alternated such that both transistors of the same arm are not conducting or opened simultaneously, thereby preventing a short circuit.

To generate the alternating current in the load, transistors operate in a pair: S1 and S2 conducting and S3 and S4 opened or vice versa. Table 1 shows the different switching stages and the applied voltage to the load.

table outlines the logic of switchesTable 1 This outline shows the logic of switches. Source: Dialog Semiconductor

A square wave inverter, also known as basic inverter, is operated by two square waves in the opposite phase and with a frequency equal to the desired frequency at the output. One of the waves is applied to S1 and S2 and the other waveform to S3 and S4. Figure 2 shows the PWM control signals and the obtained voltage in the load if this type of inverter is used.

graphs of PWM control signals and output voltageFigure 2 PWM control signals and output voltage look like this if a basic inverter is used. Source: Dialog Semiconductor

The PWM technique is based on the generation of constant amplitude pulses and modulation of the pulse duration by varying the duty cycle. The reference or modulated signal is the desired signal output—sinusoidal in the case of voltage waveforms at the output of an inverter—and the carrier signal must have a frequency much greater than the modulated one. This is the main reason for the usage of sinusoidal PWM or SPWM as the modulation method for PWM inverters.

cover image for the Power Electroincs News Sept 2020 ebookDownload the Power Electronics News eBook on how GaN- and SiC-based devices are key to addressing a primary hurdle for the expansion of renewable energy.


SPWM characteristics

SPWM is based on constant amplitude pulses with different duty cycles for each period. The width of pulses is obtained by modulation of a carrier to obtain the desired output voltage and to reduce its harmonic content.

The carrier signal of SPWM is usually a triangular wave with a high frequency, generally in several kHz. The modulation signal of SPWM is a sinusoidal waveform with a frequency equal to the desired output voltage frequency, generally 50 Hz or 60 Hz.

In Figure 3, a simplified schematic of sinusoidal modulation is shown. The switching signal is generated by comparing the sinusoidal waveform and the triangular carrier waveform. The comparator output is high when the sinusoidal voltage is greater than the triangular voltage. The output pulses of the comparator are used as the gate pulses of the H-bridge presented in the previous section.

schematic showing how SPWM generates the switching signalFigure 3 A simplified schematic shows how SPWM generates the switching signal. Source: Dialog Semiconductor

To obtain better results, the frequency ratio between the triangular and the sinusoidal waveforms must be an integer N = fC/fS, where fC is the carrier frequency or the triangular waveform and fS is the modulation frequency or the sine waveform. With this condition, the number of voltage pulses per half-cycle results in N/2. This effect can be seen in Figure 4, where the triangular, sine, and the PWM output waveforms are shown.

graphs of frequency ratio between the triangular and the sinusoidal waveformsFigure 4 The frequency ratio between the triangular and the sinusoidal waveforms must be an integer. Source: Dialog Semiconductor

The modulation process of the duty cycle is designed for modulation index m equal to or less than one. If m is higher, there will be periods of the triangle signal in which there is no intersection of the carrier and the modulation signal will exist. The effect on the output signal is shown in Figure 5. However, it’s important to note that a certain amount of over-modulation is sometimes used to obtain a higher AC voltage amplitude.

graphs showing over-modulationFigure 5 A certain amount of over-modulation is sometimes required to raise AC voltage amplitude. Source: Dialog Semiconductor

When SPWM is analyzed in terms of waveform quality, the harmonics must be considered. The SPWM generates different harmonics of several orders in the voltage waveform. However, the dominant ones are of order N and N±2 whereas N is defined as fC/fS. If over-modulation is considered, the output voltage has a higher harmonics content as a trade-off in generating a higher voltage. By varying the sinusoidal voltage, the output voltage can be regulated.

These concepts of triangular waveforms with fixed amplitude and frequency and sinusoidal waveforms with a fixed frequency equal to the output frequency and variable amplitude are the basis of the SPWM generator implemented in this article. The SPWM generator is shown in the block diagram in Figure 6.

SPWM generator block diagramFigure 6 This block diagram shows how the SPWM generator is designed and implemented. Source: Dialog Semiconductor

A high-frequency triangular waveform is necessary to generate the SPWM signals. This task is implemented with finite state machines (FSM), counters, and D-type flip flops, and is referenced as HF triangle generator in the above diagram. The generator is based on the AN-CM-265 programmable limits PWM, generating a PWM with a triangular variation of the duty cycle. As a triangular waveform is required, a low-pass filter is applied to eliminate the very high frequency of the square wave.

This triangular waveform is compared with an external low-voltage 50 or 60 Hz sine waveform with GreenPAK analog comparators. With this comparison, the sinusoidal modulation of the PWM is implemented as described in the previous section. Finally, an inverter is used to generate the complementary signals for the SPWM outputs.

schematic showing the output of the SPWM generatorFigure 7 The output of the SPWM generator is connected to an H-bridge. Source: Dialog Semiconductor

As seen in Figure 7, the SPWM output and its complementary signal are connected to each transistor of the same leg. The output of the H-bridge contains an LC-filter, so the high-frequency component of the SPWM is filtered, and finally, the sinusoidal waveform of 50 or 60 Hz is applied to the load.

Implementing SPWM generator

The implementation of the SPWM generator is based on the SLG46826V, a configurable mixed-signal IC (CMIC) that contains FSM digital counters, high-speed analog comparators, and high-frequency oscillators. That allows the CMIC to be used for generating the required triangular waveform and the sinusoidal modulation.

As mentioned earlier, the high-frequency triangular waveform generator is based on the AN-CM-265 programmable limits PWM. The implementation of the generator in the GreenPAK Designer software can be seen in Figure 8.

triangle waveform generator circuit shown in the GreenPAK Designer softwareFigure 8. The implementation of the triangle waveform generator is shown here in the GreenPAK Designer software. Source: Dialog Semiconductor

The generator uses the internal 25 MHz oscillator, configured for an output frequency of 12.5 MHz. This clock, combined with macrocells CNT1 and CNT2, generates the corresponding square waveform with the desired duty cycle.

With this configuration, the triangular waveform has a period of 1 ms, so a 1-kHz triangular waveform is obtained. In this design, a 50-Hz SPWM inverter is implemented but can be modified for 60 Hz or other frequencies by changing the period of the triangular waveform.

The high-frequency PWM with triangular variation is connected to PIN 16, where the corresponding, external low-pass filter is connected. This filter is based on a first-order RC filter, with a 1.5 kΩ resistor and a 10 nF capacitor, so the cut-off frequency of the filter is 10.6 kHz.

The output of the filter, shown in Figure 6, is connected to the high-speed analog comparator ACMP0H. The configuration of ACMP0H is shown in Figure 9.

ACMP0H screenshot of the configuration of the high-speed analog comparatorFigure 9 This configuration of the high-speed analog comparator shows odd output control. Source: Dialog Semiconductor

This block is used to compare the voltage between PIN 20 and PIN 3. For best performance, the hysteresis and the bandwidth limit options must be disabled. Thus, a low-voltage sinusoidal waveform generator must be connected to PIN 3, so the sinusoidal PWM modulation is obtained (Figure 10).

block diagram of the modulatorFigure 10 This is the block diagram of the modulator that facilitates the sinusoidal PWM modulation. Source: Dialog Semiconductor

To generate the complementary signals for PWM output, the 3-L1 look-up table is configured as a logical inverter. Finally, PWM outputs are connected to PIN 5 and PIN 6. As PIN 8 and PIN 9 are connected to the I2C module of the GreenPAK chip, it’s necessary to connect them to VDD using a pull-up resistor. The entire SPWM implementation diagram is shown in Figure 11.

SPWM generator implementation circuitFigure 11 Here is a complete view of the SPWM generator implementation. Source: Dialog Semiconductor

Testing SPWM implementation

To test the implementation, the entire system was assembled and analyzed with an oscilloscope. The 50-Hz sine wave signal was generated with a function signal generator, with a corresponding amplitude and offset such that it can be connected directly to the SLG46826V CMIC. The entire system can be seen in Figure 12.

annotated photo of the SPWM generator test setupFigure 12 A view of the entire setup used to test the SPWM generator implementation. Source: Dialog Semiconductor

This article has demonstrated the implementation of an SPWM generator, one of the most widely used methods for implementing power inverters commonly used in applications such as motor controls and renewable energy, including each step of SPWM generation and how it can be connected and filtered at the output.

This article was originally published on EDN.

You can download the original design file, and see how the whole application was built, here.

Related articles:

Leave a comment