RU2794346C1 - Class d amplifier - Google Patents

Abstract

FIELD: amplification; generator technology.

SUBSTANCE: invention can be used in broadband transmission routes with audio frequency range. This result is achieved by the fact that in a class D amplifier containing a master oscillator connected to a sawtooth voltage generator, the outputs of which are connected to the inputs of the first and second comparison circuits, the second inputs of which are connected to the outputs of the first and second operational amplifiers, and the outputs to the inputs of the first and the second pre-amplifiers, the outputs of which are connected to the inputs of the first and second key power amplifiers, the load and the signal conditioner, the third operational amplifier, the third comparison circuit, the third pre-amplifier, the third key power amplifier, as well as three inductive filters, three current sensors, adder and subtractor are additionally introduced, and the sawtooth voltage generator has three channels.

EFFECT: threefold decrease in the required frequency of the PWM conversion while reducing the non-linear distortion of the output signal and reducing the dynamic energy losses.

1 cl, 3 dwg

Description

The invention relates to the field of amplifying and generator technology and can be used in broadband transmission paths of the audio frequency range for broadcasting and underwater communication.

Known key power amplifiers (KUM) [Artym A.D. Class D amplifiers and key generators in radio communications and broadcasting. M.: Communication. 1987], using pulse-width modulation (PWM) and providing highly efficient amplification of wideband audio frequency signals. The implementation of the key amplification method favorably distinguishes PWM amplifiers from linear amplifiers from an energy point of view, especially when operating on a complex load, which includes broadband hydroacoustic emitters.

To amplify alternating low-frequency (LF) signals, as a rule, amplifiers of the ABD class are used, made according to a single-channel circuit with half-bridge terminal stages of the CCM [Artym A.D. …]. Such devices contain a pulse-width converter (PWM), a key power amplifier and a low-pass filter (LPF) made on a choke and a capacitor connected in parallel with the load. The principle of operation of a single-channel amplifier of the ABD class is based on the conversion of an analog signal into a sequence of width-modulated pulses with a switching frequency ƒ ₀ significantly (20-30 times) higher than the upper frequency of the amplified signal, key power amplification with the formation of an alternating pulse voltage (with a level of + E and -E, where E is the amplitude of the pulse voltage) and the allocation of useful low-frequency components of the load excitation voltage at the output of the low-frequency filter.

The key amplification of the pulse signal corresponds to negligible losses in amplifying devices made on high-performance field-effect transistors, which are in the open or closed state. Here, the main energy losses are dynamic in nature and are associated with a change in their state corresponding to the formation of a front or a drop in the pulsed voltage. Thus, the energy losses in the key elements are practically proportional to the switching frequency and can reach a significant relative value (10-15)% with an increase in the switching frequency of high-power CCM stages to (400-600) kHz. It is this PWM frequency that must be provided to amplify broadband signals with an upper frequency F _in = 20 kHz, provided that it is necessary to ensure the uniformity of the frequency response of the output excitation voltage of the complex load with an active power factor of 0.2-0.4, characteristic, in particular, for hydroacoustic emitters.

Significantly reduce (by 2 times) the switching frequency of the switching amplification channels (half-bridge circuits of the KUM) allows the use of bridge circuits of the final stages in class BD amplifiers [Artym A.D. ...], which use two-channel PWMs that form width-modulated pulse sequences that are in-phase by the modulating low-frequency signal and anti-phase by the PWM frequency. At the same time, in the diagonal of the bridge circuit of the KUM, the total impulse pulses, the polarity of which is determined by the polarity of the modulating signal. The resulting frequency with a change in the total pulse voltage is two times higher than the switching frequency of individual channels of the key amplification, which provides a twofold reduction in dynamic energy losses.

However, an increase in the energy efficiency of BD class amplifiers is associated with a significant decrease in the quality indicators of amplified signals, especially signals with a wide dynamic range of change. This circumstance is due to the principle of their operation, according to which the minimum level of modulation corresponds to the minimum duration of the resulting voltage pulses and the almost absence of high-frequency (HF) current components of the LPF inductor. As a result, the distortion of the pulse duration, in particular, due to the turn-on delay of transistors, significantly affects the distortion of low levels of the output voltage, practically limiting the dynamic range of the amplification by the ratio:

where τ _C - the delay time of the front turn on transistors; T is the pulse repetition period with the resulting switching frequency.

Taking into account the required values of the resulting switching frequency ƒ=1/T=(400-600) kHz, even with the minimum required transistor turn-on delays τ _c =150 ns, the gain dynamic range will be limited to 20-26 dB.

As shown in previous studies [Aleksandrov V.A. Analysis of the results of the development of energy-efficient broadband hydroacoustic transmitters for sound underwater communication. // Hydroacoustics. 2017. No. 32 (4). pp. 56-64], in a single-channel ABD class amplifier, distortion is much less and the dynamic range can be more than 30 dB.

At the same time, as noted, the ABD single-channel amplifier fundamentally loses to the BD-class two-channel amplifier in terms of energy efficiency. In this contradiction, a compromise implementation is the transition to a two-channel amplifier class ABD [Pat. RF No. 2188498 prior. 01/21/2001, publ. 08/27/2002 BI No. 24]. The well-known technical solution uses the principle of two-channel PWM in series transformer addition of the pulsed voltages of the switching amplification channels. In this case, the outputs of the KUM channels, in-phase by the low-frequency signal and anti-phase by the PWM, are connected in parallel through the low-pass filter chokes and the blocking capacitor. As a result, in the above analogue, the advantages of an ABD class amplifier with the possibility of using a two-channel PWM are realized. However, the use of this technical solution [US Pat. RF No. 2188498…], under conditions of increasing the natural frequency of the low-pass filter FB leads to a significant (more than 2 times) increase in the upper frequency of the amplified signal, and is also characterized by a significant amplitude of the high-frequency current components of the low-pass filter chokes, the level of which can exceed the maximum amplitude of the low-frequency output current . In addition, the implementation of the well-known two-channel amplifier of the ABD class requires the mandatory use of summing output transformers, the size of which, taking into account a very low cutoff frequency (F _n = 100 ... 300 Hz), determines the weight and size indicators of the amplifying device.

To a large extent eliminate the noted shortcomings of the analogue amplifier [US Pat. RF No. 2188498...] allows the use of a well-known bridge circuit of a key amplifier [US Pat. US2021/0344311 Al. Class D amplifiers. Pub. date Nov.4.2021], which assumes the use of a two-channel PWM with a controlled phase shift between the pulse voltages of the first and second amplification channels. In this case, in the bridge circuit of the class D amplifier, controlled zones of the BD and ABD modes are provided, respectively, for high and low levels of modulating signals.

Known class D amplifier [US Pat. US No. 2021/0344311 A1…], which does not require additional output transformers for summing the impulse voltages of individual channels, is the closest to the proposed technical solution and can be taken as a prototype of the proposed device.

The block diagram of the prototype amplifier is shown in Fig. 1. The prototype amplifier contains a master oscillator 1, a sawtooth voltage generator 2, a signal conditioner 3, two operational amplifiers 4.1 and 4.2, two comparison circuits 5.1 and 5.2, two preamplifiers 6.1 and 6.2, two channels of key power amplifiers 7.1 and 7.2, a source power supply 8 and load 9, which is used as an electrodynamic speaker with a significant inductive component of the input resistance.

In accordance with the principle of operation, the known class D amplifier [US Pat. US2021/0344311 A1…] can operate a two-channel switching amplification circuit in both BD and ABD modes. At the same time, in the ABD mode, especially at low signal levels, a decrease in the nonlinear distortions associated with the “cut-off” of the low-pass filter chokes is achieved, which provides an expansion of the dynamic gain range from 20 dB, typical for the BD mode, to 30-36 dB.

At the same time, the transition to bipolar pulses, typical for the ABD mode, in the prototype amplifier due to the phase shift of the pulse voltages and channels of the key amplification, is associated with a twofold decrease in the resulting switching frequency, which significantly degrades the quality of the output signal due to a significant level of ripple at the initial frequency. ƒ ₀ PWM conversion. Such a decrease in frequency is unacceptable for transmitting paths of underwater communication due to a significant rise in the frequency response at the natural frequency of the low-pass filter, which, as a rule, is set much higher than the upper frequency of the signal being amplified. As a result, the need to significantly increase the frequency of the PWM conversion remains, and, as a result, the energy efficiency characteristics of the transmitting device deteriorate.

The objective of the invention is to increase energy efficiency while expanding the dynamic range of amplification.

The technical result of using the present invention is a threefold decrease in the required PWM conversion frequency while reducing the non-linear distortion of the output signal and reducing dynamic energy losses.

The technical result is achieved by the fact that in a known class D amplifier containing a master oscillator connected to the input of a sawtooth voltage generator, the first and second outputs of which are connected to the first inputs of the first and second comparison circuits, the second inputs of which are connected to the outputs of the first and second operational amplifiers, and the outputs - to the inputs of the first and second preamplifiers, the direct and inverse outputs of which are connected to the first and second inputs of the first and second key power amplifiers, the positive and negative power supply terminals of which are connected to the corresponding outputs of the power source, as well as the load and the signal conditioner by means of that new features have been introduced into its composition, namely: a third operational amplifier, a third comparison circuit, a third preamplifier, a third key power amplifier, as well as three inductive filters, three current sensors, an adder and a subtractor, and a sawtooth voltage generator is connected by a third output to the first input of the third comparison circuit, the second input of which is connected to the output of the third operational amplifier, and the output to the input of the third preamplifier, the direct and inverse outputs of which are connected to the first and second inputs of the third key power amplifier, the positive and negative power supply terminals of which are connected with the corresponding outputs of the power source, and the outputs of the first, second and third key power amplifiers are connected through the first, second and third inductive filters and the first outputs of the first, second and third current sensors, respectively, to the first input of the load, the second input of which is connected to the common output of the power source, wherein the second outputs of the first, second and third current sensors are connected to the first, second and third inputs of the subtractor and the adder connected by the output to the inverse input of the subtractor, the direct input of which is connected to the output of the signal shaper, and the first, second and third outputs, respectively, with the inputs of the first, second and third operational amplifier.

In the proposed three-channel class D amplifier, the implementation of the claimed technical result is provided by a set of newly introduced blocks and connections. An increase in energy efficiency is achieved by reducing dynamic energy losses by reducing the switching frequency ƒ ₀ three times in each of the three channels of key amplification while maintaining the required ratio of the resulting frequency ƒ=3ƒ ₀ to the upper frequency F _in the amplified signal ƒ/F _at =400-600 kHz. In this case, the expansion of the dynamic range is realized by the formation of bipolar pulses of the resulting impulse voltage at a relative level of the input signal m=U _in /U _pm =0...0.33. The solution to the problem of selecting the ABD mode zone is implemented using the parallel connection of three channels of key power amplifiers through inductive filters to the load with equalization of low-frequency output currents, which ensures the reliability of the operation of a class D amplifier with a three-channel PWM with parallel summation of the output power of individual channels of the key amplification.

The essence of the invention is illustrated in Fig. 1-3, which shows block diagrams of the prototype device (Fig. 1) and the proposed device (Fig. 2), as well as timing signal diagrams that explain the operation of the proposed device (Fig. 3).

The block diagram of the proposed class D amplifier (Fig. 2) contains a master oscillator 1, a three-channel sawtooth voltage generator 2, a signal shaper 3, operational amplifiers 4.1, 4.2, 4.3, comparison circuits 5.1, 5.2, 5.3, preamplifiers 6.1, 6.2, 6.3, key power amplifiers 7.1, 7.2, 7.3, inductive filters 10.1, 10.2, 10.3, current sensors 11.1, 11.2, 11.3, adder 12, subtractor 13, power supply 8 and load 9.

To explain the principle of operation of the proposed technical solution in Fig. 3a and fig. 3b, respectively, for large (m≈0.75) and small (m≈0.3) levels of the amplified signal U, the following time diagrams are shown: sawtooth voltages U _p1 , U _p2 , U _p3 , with a uniform time shift T=T ₀ /3 (where T ₀ - the period of switching channels KUM); output voltages V ₁ , V ₂ , V ₃ channels KUM 7.1, 7.2, 7.3; the resulting impulse voltage V, brought through the inductive filters 10.1, 10.2, 10.3 to the load 9.

Figure 3c illustrates the process of equalizing the output currents of the KUM (signals U _t1 , U _t2 , U _t3 current sensors 11.1, 11.2, 11.3) in the process of parallel addition due to the formation of mismatch signals ΔU ₁ , ΔU ₂ , ΔU ₃ .

Parallel addition of impulse voltages V ₁ , V ₂ , V ₃ key power amplifiers 7.1, 7.2, 7.3, despite the high degree of identity of the PWM conversion coefficients, necessarily requires the alignment of low-frequency currents flowing through the inductive filters 10.1, 10.2, 10.3. This alignment is achieved by comparing the output signals of the current sensors U _t1 , U _t2 , U _t3 (Fig. 3c) with the total average value generated by the adder 12:

As a result, the subtractor 13 received mismatch signals ΔU ₁ , ΔU ₂ , ΔU ₃ , which correct the amplified signal U supplied to the inputs of the respective operational amplifiers 4.1, 4.2, 4.3 in the form of corrected values U ₁ , U ₂ , U ₃ :

where k is the coefficient that determines the depth of negative feedback (OS) for the output current of the CCM channels.

The maximum current OS depth can be provided at least 30 dB, which ensures the required current identity of inductive filters 10.1, 10.2, 10.3 (deviation no more than 3%), for the necessary condition for parallel summation of impulse voltages V ₁ , V ₂ , V ₃ .

All structural blocks included in the claimed device are performed according to known rules, and their combined use leads to the achievement of the claimed technical result.

The master oscillator 1 can be implemented according to known schemes for generating pulse signals with a frequency equal to the value of the required resulting frequency of the PWM conversion ƒ=400-600 kHz.

A three-channel sawtooth voltage generator 2 is performed on a three-link ring counter (three D-flip-flops) to the counting inputs of which pulses are received from the output of the master oscillator with a frequency of 2ƒ. As a result, pulse sequences of the meander type with a frequency ƒ ₀ =ƒ/3 are formed at the outputs of individual links. As a result of their integration, sawtooth voltages U _p1 , U _p2 , U _p3 are formed (Fig. 3a, b) uniformly shifted by the time interval τ=1/ ƒ.

As a signal generator 3, it can be used as its own synthesizer of a given modulated low-frequency signal, or as a buffer amplifier and a broadband signal converter from an external device.

To implement operational amplifiers 4.1, 4.2, 4.3, typical 1432UD25BU microcircuits can be used, corresponding to the frequency range of the amplified signals.

As comparison circuits 5.1, 5.2, 5.3, high-speed comparators are used (for example, 1481CA1P), the switching delays of which do not exceed 20 ns, which corresponds to the permissible distortion of the PWM conversion.

Preamplifiers 6.1, 6.2, 6.3 and half-bridge transistors KUM 7.1, 7.2, 7.3 are selected from the technical requirements for the maximum output low-frequency (LF) voltage and load current. For example, for a low-voltage class D amplifier with an output voltage of up to 60 V (p-p) and a rated output power of up to 1000 VA with a maximum total output current of up to 35 A (p-p), the selection of drivers and transistors should correspond to a power supply voltage of 120 V with an allowable voltage of 200 V and permissible output current of at least 20 A. Such requirements, for example, correspond to the CCM on transistors IRFS4227PBF controlled by a floating-point driver IRSZ0124S, or their domestic counterparts. If it is necessary to increase the output voltage for the implementation of the CCM channels, higher-voltage field-effect transistors can be selected, a wide range of which allows the implementation of key amplification circuits with a power supply voltage of up to 1000-1200 V, which corresponds to a typical excitation amplitude level of hydroacoustic emitters of 300-400 V. In this case, KUM channels can be implemented on CREE transistors, for example, type С3М0065100 with a permissible output current of 40A and drivers with optoelectronic isolation of the Si8234 type. Thus, it is possible to ensure the implementation of a key amplification channel with a rated power of up to 4 kVA. Accordingly, the total output power of a three-channel amplifier can reach 10-12 kVA.

Inductive filters 10.1, 10.2, 10.3 ensure the limitation of the amplitude of the high-frequency current and the selection of useful low-frequency components of the total impulse voltage on the load. Inductive filters are made on chokes designed for the rated load current, the inductance L of which is determined by the choice of natural frequency F ₀ of the low-pass filter significantly higher than the upper frequency of the amplified signal -F ₀ ≥ (1.5-2.0) F _in :

where C _n is the load capacity characteristic of hydroacoustic emitters.

In this case, the maximum amplitude of the RF current of each amplification channel, corresponding to the output voltages V ₁ , V ₂ , V ₃ of the meander type, is determined from the relationship:

where E is the amplitude of the impulse voltage.

As a rule, the value of I _{hf max} is set to no more than the maximum low-frequency current of the amplification channel I _{k max} \u003d I _{n max} / 3, which determines the range of filter inductance 10.1, 10.2, 10.3:

Fulfillment of condition (7) at given values of C _n , E, I _{lf max} is provided by the ratio ƒ ₀ /F _in , the value of which can be more than 20…30.

Current sensors 11.1, 11.2, 11.3 are designed to control and equalize the output current of the switching amplification channels 7.1, 7.2, 7.3 and must generate signals U _t1 , U _t2 , U _t3 (Fig. 3c), the voltages of which are proportional to the current of the inductive filters 10.1, 10.2, 10.3. The sensors can be made resistive with galvanic reference to the midpoint of the power supply 8 and additional signal amplification by operational amplifiers.

The adder 12 is designed to form the average value U _t based on the results of the weighted summation of the signals U _t1 , U _t2 , U _t3 (4). With an appropriate level of output voltage of the current sensors 11.1, 11.2, 11.3, the simplest resistive summation circuit can be proposed.

The subtractor 13 must ensure the formation of correction signals to equalize the output currents of the key amplification channels:

When implementing a subtracting device, it is preferable to use operational amplifiers that provide amplification of the signals ΔU ₁ , ΔU ₂ , ΔU ₃ k times to generate the resulting signals U ₁ , U ₂ , U ₃ (4) coming through the operational amplifiers 4.1, 4.2, 4.3 to comparators 5.1, 5.2, 5.3. As a result of the PWM conversion of pulsed voltages V ₁ , V ₂ , V ₃ is implemented taking into account the alignment of the output currents of the switching amplification channels:

The above description of the blocks and the composition of the claimed device confirms the possibility of implementing the proposed technical solution.

The claimed class D amplifier works as follows.

The input signal from the output of the shaper 3 through the subtractor with the necessary correction goes to the direct inputs of the comparison circuits, at the outputs of which, according to the results of comparison with the reference sawtooth voltages U _p1 , U _p2 , U _p3 , PWM signals are formed, as a result of which amplification at the outputs of the KUM 7.1, 7.2, 7.3 impulse voltages V ₁ , V ₂ , V ₃ are formed. In FIG. 3a, b illustrates the formation of pulse-width modulated pulse voltages with the equality U=U ₁ =U ₂ =U ₃ without taking into account the mismatch signals due to very slight differences necessary to equalize the low-frequency currents:

A characteristic feature of the formation of impulse voltages V ₁ , V ₂ , V ₃ is a uniform phase shift at the switching frequency, which corresponds to an equivalent impulse voltage V with a tripled frequency of change applied to the resulting inductive filter formed from inductive filters 10.1, 10.2, 10.3 connected in parallel.

It should be noted that the power supply channels KUM 7.1, 7.2, 7.3 is implemented from the voltages +E and -E of the power supply 8, the middle point of which is connected to the common output of the load. At the same time, alternating voltages corresponding to the ABD mode are formed at the outputs of the CCM. This circumstance is of fundamental importance for reducing distortion and expanding the range of the amplified signal. In FIG. 3b illustrates the formation of a single-cycle sequence of bipolar pulses of equivalent voltage V with tripled resulting switching frequency ƒ=3ƒ ₀ . Thus, the required quality of the output signal is achieved with the possibility of a threefold decrease in the switching frequency of the channels of key amplification with a corresponding decrease in dynamic energy losses. The noted quality of the proposed technical solution is especially important in the implementation of hydroacoustic transmitters that have a pronounced capacitive component of conductivity, which dominates in the high frequency range. Special requirements for the uniformity of the frequency response and phase response of the excitation voltage of such a load must be provided with the necessary frequency ratio ƒ / F _in = (20 ... 30).

The advantages of the proposed solution in comparison with known analogues and the prototype make it possible in these conditions to most fully meet the requirements for energy efficiency and quality indicators of output signals for hydroacoustic broadband transmitters.

Increasing energy efficiency while expanding the dynamic range in the inventive class D amplifier makes it possible to ensure the implementation of the proposed technical solution in hydroacoustic stations of hydroacoustic communication (HS) modes. So, if in the prototype device, when amplifying broadband GS signals with an upper frequency of up to 20 kHz, the switching frequency should be at least 200 kHz, then in the claimed device, it is sufficient to provide a switching frequency of not more than 130 kHz. In this case, the energy loss in the proposed device is reduced by at least 1.5 times, and the dynamic voltage range is increased by at least 20 dB (50 dB compared to 30 dB typical for the prototype).

At present, the enterprise has developed experimental samples of multi-channel key amplifier units, the test results of which confirm the advantages of the proposed three-channel key amplification circuit for use in transmitting equipment for broadband GS modes.

Claims (1)

A class D amplifier containing a master oscillator connected by its output to the input of a sawtooth voltage generator, the first and second outputs of which are connected to the first inputs of the first and second comparison circuits, the second inputs of which are connected to the outputs of the first and second operational amplifiers, and the outputs to the inputs of the first and the second preamplifiers, the direct and inverse outputs of which are connected by the first and second inputs of the first and second key power amplifiers, the positive and negative power supply terminals of which are connected to the corresponding outputs of the power source, as well as the load and the signal conditioner, characterized in that it includes a third an operational amplifier, a third comparison circuit, a third preamplifier, a third switching power amplifier, as well as three inductive filters, three current sensors, an adder and a subtractor, and the sawtooth voltage generator is three-channel and connected by a third output to the first input of the third comparison circuit, the second input which is connected to the output of the third operational amplifier, and the output - to the input of the third pre-amplifier, the direct and inverse outputs of which are connected to the first and second inputs of the third key power amplifier, the positive and negative power supply terminals of which are connected to the corresponding outputs of the power source, and the outputs of the first, the second and third key power amplifiers are connected, respectively, through the first, second and third inductive filters and current sensors connected in series to the first input of the load, the second input of which is connected to the common output of the power source, in turn, the outputs of the first, second and third current sensors connected to the first, second and third inputs of the subtractor and the adder connected by the output to the inverse input of the subtractor, the direct input of which is connected to the output of the signal conditioner, and the first, second and third outputs, respectively, to the inputs of the first, second and third operational amplifiers.

Images