Switching power amplifier of digital control scheme

Power amplifiers in the audio amplifier, launch systems, servo systems, sonar, vibration testing, and many areas have been widely used. The traditional linear amplifier power amplifier, its efficiency is low (40% 1 60%), and bulky, so applications are restricted. To address the shortcomings of conventional power amplifiers, switches, power amplifier came into being.

At home and abroad in high-power (5 kW or more) amplifier design, in order to meet the power requirements of commonly used IGBT-based full-bridge inverter topology. In contrast to MOSFET power devices for the high power amplifier system is designed only a few, and its development of control methods can not be a good solution to the system are between the flow control module and the capacitor mid-point potential control problems . Therefore an urgent need to develop to M0SFET based high power amplifier systems to reliably improve the amplifier system performance. This paper presents a modular system for high-power amplifiers used in inverter unit, and introduces elements of topology and digital control theory, experimental results show its good performance.

A main circuit topology

Traditional two-level full-bridge inverter topology used in high power amplifier system, due to the breakdown voltage limit, difficult to use high frequency MOSFET, it can not effectively improve system performance. Draws on existing research, we use the literature is the midpoint of the five-level diode clamped inverter topology ("Five-Level NPC Inverter", the following abbreviated as "FNI") as a basis for the power unit. Figure l shows the circuit for the FNI.

Switching power amplifier of digital control scheme

The basis of the structure of this FNI - NPC inverter topology, was first used by Nable-and others made in 1981. With the traditional two-level converter, compared with the following advantages: high-power systems Zhong, Direct Series will power Qi Jian Wu Xu Shi Yong Er plus auxiliary circuits; devices down to DC voltage Ji Xian Ce voltage of the half, Shi device of Xuanqubiande flexible; output waveform harmonic components relative to the two-level converter greatly reduced, thereby reducing the burden of filtering part; load voltage ripple decreases, suppress electromagnetic interference.

2 Comparison of Control and Improvement

2.l control scheme has been introduced

Literature, Lau WH, etc. The advantage of the development of control programs is to increase the equivalent switching frequency of the output module, suppress output harmonics; disadvantage is that the system through the PWM input signal is modulated, still can not use as a drive signal is required to continue more tedious calculations, it is not a good use of the existing digital signal processing chip. See the detailed analysis of the program literature.

2.2 The principle of the improved control scheme

Improved control methods in the literature first carrier frequency doubled to 2fc, and adjust its bias, then compared to PWM, as shown in Figure 2, the signal shall be modulated drive signals. And control the left and right leg (Legl, Leg2) the carrier phase the same, there is no documentation required by the phase control method, its advantage is to avoid the system to adjust the switching frequency is also a need to adjust the phase difference at the same time help the system adjust the DC side capacitor mid-point potential.

Comparing Figure 2 and documentation can be found, switch the drive signal is the same, so the output waveform must be the same. Improved control strategy can easily be applied to digital signal processing chips, while retaining the advantages of document control mode. For example, DSP (TMS320LF2407) chip PWM modulation can be full-time event management module directly to the completion of EVA and EVB, this approach greatly reduces control of the realization of the difficulty.

Switching power amplifier of digital control scheme

Improved control strategy has also had shortcomings that do not provide solutions to the mid-point potential of DC capacitor imbalance problem. According to the experimental results can be found that due to the inherent resistance characteristics of circuit elements caused by the asymmetry of the mid-point potential of the static capacitance error can not be ignored. Figure 9 (f) for the 400 V DC power supply for the case when the midpoint potential, 13.2 V can be found in the static error.

2.3 midpoint potential hazards and solutions unbalanced

Document analysis system on the DC side of the mid-point potential drift THD of the output shown in Figure 3. Figure 3 The k value: the number of switching power amplifier control scheme, represents the midpoint of the degree of imbalance. In other industrial applications, the output waveform distortion due to less demanding, the midpoint of the appropriate drift is allowed. However, in systems such as the power amplifier requires high quality output waveform application, the midpoint of the imbalance can be one important reason for the output distortion. In order to overcome the imbalance caused by the mid-point decline in the output waveform quality, we control the way in improving balance control by adding the midpoint of the program flow chart shown in Figure 4, the midpoint of balance control program block diagram shown in Figure 5. Mid-point of balance control principle is, the beginning of each switching cycle, the first DC bus voltage sampled Kc1 and Vc2 (see Figure 1), then Vc1 and Vc2. Computing the difference between doing PI. If the PI results of operations is positive, then and POSREF (system can tolerate more than Vc2 Vc1 maximum) compared to the output of the comparator if found positive, which means that the more severe the mid-point of the drift. And then detected the magnitude of reference signal Vs, if the amplitude Vs is negative (as when the O and the performance of the amplitude of Vs by comparing the converters, the output is positive), the carrier frequency will increase as 2fc; the other hand, if the voltage difference the result of computing the value of PI is less than NECREF, and the amplitude reference signal Vs is positive, it will increase the carrier frequency 2fc; other cases, the carrier frequency fc to maintain the same. Figure 5 Switch module function, if the middle of the module on the left foot of the input signal is positive, then the right side of the module output with the module input signals of the left most offal; the middle of the left foot if the module input signal is zero, then the right side of the module output most on the left foot with the module input signal line.

Carrier frequency is confirmed. Vs function value will be loaded to the DSP chip event manager module compare unit, ready to carry out PWM modulation with the carrier. Load upon completion of the reset interrupt.

This is the essence of the mid-point control mode by adjusting the carrier frequency to change the midpoint of the current flow. By comparing before and after the midpoint of the current carrier frequency doubling the flow simulation, we can know: If the reference signal frequency fs Ks as a reference, the carrier frequency doubling before the midpoint of the flow of current ineu alternate week period (see Figure 6) , transform the frequency 2fc; carrier frequency doubling, the midpoint of the flow of current ineu only change once a week during the period (see Figure 7), which transform the frequencies 2fs. And because the latter mid-point of the flow of current reference signal with the amplitude of Ks, so in deciding whether to double the carrier frequency before, need to detect the amplitude of the positive and negative Vs.

When the carrier frequency doubling the output waveform with the original agreement. In the DSF (TMS320LF2407) chip, carrier frequency only when the carrier amplitude is O can be changed; it carrier-free phase leg can control the carrier frequency is about the same time change has no impact on the output waveform.

3 Experimental Validation and Results

This juice had a single module based multi-level circuit of the experimental model, the specific circuit parameters and specifications are as follows:

Full output power of 1 kW:

Output frequency 2 kHz;

DC input voltage 400 V:

Basic switching frequency 100 kHz.

Switch drive signal provided by the DSP, PWM modulation drive signals are generated within the DSP to complete. Figures 9 and 10 respectively before and after using the midpoint of balance control output waveform and compare the mid-point potential.

Figure 9 (e) and Figure 9 (f) below, using the midpoint of balance control, the DC bus voltage static error 3.2V; used before the mid-point of balance control, DC bus voltage static error 13.2V.

Switching power amplifier of digital control scheme for switching power amplifier of the digital control scheme

4 Conclusion

This paper analyzes the switching power amplifier topology and digital control program. In the control design, the introduction of a mid-point for the five-level diode clamped inverter topology PWM control technology, it can increase output and reduce the equivalent switching frequency DC-side of the midpoint potential of the drift and improve the system output waveform quality.

Man in the FNI power module can be used in parallel staggered manner and to improve the system's total power output waveform level number, then it will expand the system to achieve the required power level, but also reduces the output distortion of the system, detailed analysis, see literature.

Of course there are many problems to be solved, such as multi-level inverter dead time compensation issues, and multi-module Q of all current issues, these will serve as the focus of future research work.