Analysis and Research on Bidirectional Synchronous Demodulation Control Circuit of Differential Frequency High-frequency Chain

    Abstract: Discusses the working principle of the differential frequency high-frequency chain bidirectional synchronous demodulation control circuit, and analyzes the control method of the bidirectional synchronous demodulation circuit under pure resistive load and inductive load. The experimental results show that the control method has a simple structure, reliable performance and low cost.

    Keywords: difference frequency, high frequency chain, two-way synchronous demodulation

1 Introduction

    In the application field of small and medium-power inverter power supplies, especially in various types of special variable frequency power supplies and UPS, people have put forward increasingly higher requirements for their electrical performance, work efficiency and other indicators to meet the needs of specific occasions. However, due to There is likely to be a low-frequency isolation transformer in the device, which greatly limits the improvement of the power density index of the inverter power supply. In order to overcome the influence of low-frequency transformers, various high-frequency inverter power supplies have been developed and designed in recent years, breaking through the major obstacle of the weight and volume index of low-frequency transformers. However, most of the existing high-frequency chain inverters are unidirectional voltage source high-frequency chains [ 2], using one-way transmission. Although the bidirectional voltage source high-frequency chain inverter solves the problem of bidirectional power transmission [1][3][5][6], it has problems such as complex control circuits because it is often controlled by a single-chip microcomputer.

    From the perspective of analog control, this paper proposes a bidirectional synchronous demodulation control circuit for a bidirectional voltage source inverter.

2 Working principle of difference frequency high frequency chain inverter circuit

    Figure 1 is an isolated push-pull high-frequency chain inverter circuit based on the difference frequency principle. The basic working principle of this circuit is to use two sets of high-frequency push-pull inverters to obtain the frequencies of (fs+f0) and (fs-f0) The switching frequencies of the two sets of high-frequency inverters (fs is the carrier frequency, f0 is the output fundamental frequency), after isolation by the high-frequency transformer, form a high-frequency sinusoidal voltage on the secondary side of the high-frequency transformer, and then obtain a high-frequency sinusoidal voltage based on the difference frequency principle. The difference frequency voltage waveform of the bidirectional voltage source characteristics is shown in Figure 2. The waveform obtained after bidirectional synchronous demodulation of this difference frequency voltage waveform is shown in Figure 3.

3 Working principle of bidirectional synchronous demodulation circuit

    The bidirectional synchronous demodulation circuit is shown in Figure 4, in which S1, S2, S3, and S4 form a bidirectional voltage and current switch, and L, C and high-frequency transformer windings N1 and N2 form a bidirectional full-wave demodulation circuit.

    The principle of demodulation can be derived from the following formula:

    It is known that the mathematical expression of the modulated signal, that is, the difference frequency voltage is:

    u s =U m sin[2π(f s +f 0 )t]－U m sin[2π(f s－f 0 )t]

    =2U m cos2πf s t·sin2πf 0 t (1)

    In the formula: U m - high-frequency sinusoidal voltage amplitude;

    f s + f 0 , f s -f 0 ——the operating frequency of the two push-pull circuits;

    f s ——carrier frequency;

    f 0 ——modulation frequency.

    According to the working principle of the product detector, the modulated signal is multiplied by the carrier signal to get:

    u=u a (t)uS(t)=U2 m cosω0t(1+cos2ω s t) (2)

    The modulated signal can then be restored through a low-pass filter.

    Multiplier and adder circuits can be formed by utilizing the nonlinear characteristics of diodes and the balancing circuit of diodes, as shown in Figure 5.

    Assuming that the static characteristic curve of the diode is nonlinear, it can be expanded with a power series as:

    id=I o＋b 1 u s＋b 2 u s 2＋…… (3)

    In the formula: Io——diode bias operating point current;

    us——the differential frequency voltage after input modulation;

    b1, b2——expansion coefficient (constant).

    Suppose the nonlinear current flowing through the diode is:

    i 1 =a 0＋a 1 (u 2＋u 1 )＋a 2 (u 2＋u 1 )2＋……

    i 2 =a 0＋a 1 (u 1－u 2 )＋a 2 (u 1－u 2 )2＋……

    i 1 -i 2 =2a 0 u 2 +4a 1 u 2 u 1 +...... (4)

    In the formula: u 1 ——carrier voltage;

    u 2 ——Voltage after modulation.

    If a diode is considered a switching device in a power circuit, then

    i∝(i 1 -i 2 )=2a 1 u 2     (5)

    It can be seen from equation (5) that if the u1 signal in Figure 5 is changed into a difference frequency modulation signal, the modulated signal can be obtained as the output:

    u O (t)∝2a 1 u 2 Z 0

    The output signal in Figure 5 is only a half-wave voltage. Through the combination of controllable switching power devices, the output full-wave voltage in Figure 4 can be obtained, and the waveform is shown in Figure 3.

    Due to the different properties of Z0, the two-way full-wave demodulation control mode is different.

4. Control mode of bidirectional synchronous demodulation circuit

    Due to the limitation of the reverse withstand voltage of existing power switching devices, bidirectional synchronous demodulation circuits are generally composed of two fully controlled switching tubes in reverse series. Depending on the nature of the load, different control methods are used to realize the bidirectional flow of current. Thus demodulation is achieved.

4.1 Demodulation control of purely resistive load

    When the load is a purely resistive load, the phase difference between the output current and the output voltage is φ=0, that is, the output current and the output voltage are in the same phase. As shown in Figure 6, when the difference frequency output voltage is zero, the current in the filter inductor is also zero. At this time, the bidirectional synchronous demodulation switch is reversed and turned on by controlling the drive signal, and the current of the switch tube is theoretically zero. , that is, all switching tubes are turned on with zero voltage and zero current, and the control is the simplest, with only two stages: forward demodulation and reverse demodulation.

    (1) Forward demodulation stage

    In interval I, at time t0, when the difference frequency output voltage is zero, the forward demodulation switches S1 and S3 are controlled to be turned on, and the reverse demodulation switches S2 and S4 are turned off. At this time, the current in the filter inductor is zero, the bidirectional synchronous demodulation circuit works in the forward demodulation stage, and S1 and S3 work in the soft switching mode.

    (2) Reverse demodulation stage

    In interval II, at time t1, when the difference frequency output voltage is zero, the current in the output inductor is also zero, and the reverse demodulation switches S2 and S4 are controlled to be turned on, and the forward demodulation switches S1 and S3 are turned off. At this time, the bidirectional synchronous demodulation circuit works in the reverse demodulation stage, and S2 and S4 work in the soft switching mode.

    The output voltage, output current and driving waveforms of each switch tube are shown in Figure 6.

4.2 Demodulation control of inductive load

    When the load is inductive, because the inductor current phase lags, the phase difference between its output voltage and output current is φ>0. That is, when the output voltage is zero, its output current is not zero. At this time, the current in the inductor does not reverse direction. If At this time, when the bidirectional synchronous demodulation switch is turned on, the switch will bear a large forward or reverse peak current. Therefore, before the load current changes from zero to positive, the forward demodulation switch must be turned on, and the reverse demodulation switch must be turned on. The switch tube should be turned off, and the current in the inductor will feed energy back to the power supply; before the load current changes from zero to negative, the reverse demodulation switch tube must be turned on, and the forward demodulation switch tube should be turned off, and the current in the inductor will Energy is fed back to the power source. Its two-way synchronous demodulation process can be divided into four stages, see Figure 7.

    (1) Forward demodulation stage

    In the interval I, t0 moment, when the difference frequency output current is zero, the forward demodulation switches S1 and S3 are controlled to be turned on, the reverse demodulation switches S2 and S4 are turned off, and the forward demodulation switches S1 and S3 are always controlled. When a trigger signal is added, the switching tube works in a non-controlled demodulation mode, and S1 and S3 work in soft switching mode.

    (2) Positive energy feedback stage

    In interval II, at time t1, when the difference frequency output voltage is zero, because the output current iout>0, the forward demodulation switches S1 and S3 are controlled to be on and off at high frequency, and the reverse demodulation switches S2 and S4 are turned off, UA When <0 (see Figure 1), S1 is turned on, when UB<0, S3 is turned on, and S1 and S3 are turned on alternately.

    (3) Reverse demodulation stage

    In interval III, at time t2, when the difference frequency output current is zero, the reverse demodulation switches S2 and S4 are controlled to be turned on, the forward demodulation switches S1 and S3 are turned off, and the reverse demodulation switches S2 and S4 are always controlled. Adding a trigger signal, the switch tube works in a non-controlled demodulation mode, and S2 and S4 work in soft switching mode.

    (4) Reverse energy feedback stage

    In interval IV, moment t3, when the difference frequency output voltage is zero, because the output current iout<0, the reverse demodulation switches S2 and S4 are controlled to be on and off at high frequency, and the forward demodulation switches S1 and S3 are turned off, UA >0, S2 is turned on, when UB>0, S4 is turned on, and S2 and S4 are turned on alternately.

    It can be seen that a positive half-wave output voltage can be obtained in intervals I and IV, and a negative half-wave output voltage can be obtained in intervals II and III.

    Its bidirectional synchronous demodulation output voltage, output current, filtered output voltage waveform and driving waveform of each switch tube are shown in Figure 7.

4.3 Demodulation control of capacitive load

    When the load is capacitive, because the capacitor voltage phase angle lags, the phase difference between its output voltage and output current φ<0, that is, when the output current is zero, its output voltage is not zero. If the bidirectional synchronous demodulation switch is turned on at this time tube, the switch tube will bear a large forward or reverse peak current. Therefore, when the load current changes from zero to positive, the reverse demodulation switch tube must be turned off, and the forward demodulation switch tube should be turned on at high frequency. The current in the inductor feeds energy back to the power supply; when the load current changes from zero to negative, the reverse demodulation switch must be turned on, the forward demodulation switch must be turned off, and the current in the inductor feeds energy back to the power supply. The driving waveform of the switch tube under capacitive load is roughly the same as the driving waveform for inductive load, except that the control phase is different.

5 Implementation of bidirectional synchronous demodulation control circuit

    Based on the above analysis, the control circuit of the bidirectional synchronous demodulation circuit under pure resistive load and inductive load is designed. The control circuit and basic structure are shown in Figure 8.

    In the experiment, the driving signal of the bidirectional synchronous demodulation circuit is sampled from the output voltage and output current. After passing through the zero-crossing comparator and the phase comparator, the pulse waveforms in the II and IV intervals in Figure 7a, b, c and d are taken out respectively, and then compared with The high-frequency pulse square waves output by the two 3525 pass through the AND gate circuit to form a high-frequency pulse wave (see the high-frequency pulses in the II and IV intervals in Figure 7a, b, c, d) and the 400Hz square wave signal generated by the 555 passes through the OR The gate circuit forms the drive waveform in Figure 7. According to the voltage relationship between points A and B of the main circuit, an analog switch is used to drive the bidirectional synchronous demodulation switch tubes S1, S2, S3, and S4.

    It should be noted that when uO=0, iO≠0, during the PWM operation of 3525, switch tubes S1, S2, S3, and S4 are all turned on at the same time, but because the inductor current is not commutated, S2, S4, or S1, S3 can't really turn on.

6 Experimental results

    Figures 9 to 12 are the experimental waveforms of an aviation power supply we developed. In the experiment, the operating frequencies of the two push-pull inverters are 20kHz and 19.2kHz respectively, the input voltage is 27V, and the output voltage is a sinusoidal voltage of 400Hz, 115V. The bidirectional synchronous demodulation circuit uses ordinary MOS tubes, and the experimental results are basically consistent with the theoretical analysis.