High-voltage amplifier control method, device and equipment

Abstract

The application is suitable for the field of power electronics, and provides a high-voltage amplifier control method, device and equipment. The method generates a dynamic power supply voltage rail, and extracts an error signal of the target output. The error signal is decomposed into at least two frequency components, and is processed respectively by using differentiated amplification strategies. The processed components are combined into a compensation signal, which is superimposed with the power supply voltage rail, so as to generate a final output signal with high fidelity. The application improves the efficiency by using dynamic power supply, and at the same time, the residual error signal is divided into frequency and compensated by using differentiated strategies, so that various distortions and noises are effectively eliminated, and the unification of high efficiency and high precision is realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power electronics, and particularly relates to a high-voltage amplifier control method, device and equipment. BACKGROUND

[0002] As an important electronic device, a high-voltage amplifier is widely used in occasions requiring high-voltage driving, such as industrial testing, scientific research and medical equipment. In order to ensure that no clipping distortion is generated when the peak value of an output signal is guaranteed, a power tube in a traditional high-performance linear amplifier, such as a class A or AB amplifier, usually needs to be powered by a fixed high-voltage DC power supply much higher than the maximum output voltage. The disadvantage of this power supply mode is that when the instantaneous amplitude of the output signal is low, the power tube will bear a large voltage drop, resulting in a large amount of energy being wasted in the form of heat, so that the energy efficiency of the amplifier is very low, and a serious heating problem is caused.

[0003] In order to improve the efficiency, some dynamic power supply technologies, such as envelope tracking technology, are proposed in the prior art. The core idea is to enable the supply voltage of the amplifier to be dynamically adjusted according to the change of the output signal envelope, so as to maintain a margin slightly higher than the voltage required by the current output signal, thereby significantly reducing the average voltage drop and power loss of the power tube. However, this kind of technology also exposes new problems in actual application. The switching power supply used to generate dynamic voltage itself introduces a large output ripple and switching noise, and its tracking of the signal envelope also has certain delay and non-linear error. These residual errors introduced by the dynamic power supply process will be directly coupled or fed to the final amplifier output, seriously affecting the output accuracy and signal fidelity of the amplifier, and making it difficult to meet the needs of high-precision applications. Therefore, how to improve the efficiency of the high-voltage amplifier while effectively solving the precision deterioration problem introduced by the dynamic power supply is a technical problem to be solved in the field. SUMMARY

[0004] The embodiments of the application provide a high-voltage amplifier control method, device and equipment, which aims to solve the technical problem that the efficiency and precision of the high-voltage amplifier are difficult to be considered in the prior art, and can effectively suppress the noise, ripple and distortion errors introduced by the dynamic power supply process while improving the efficiency by using the dynamic power supply technology, so as to realize the unity of high efficiency and high precision.

[0005] In a first aspect, the embodiments of the application provide a high-voltage amplifier control method, comprising:

[0006] In response to an input signal, a dynamically changing supply voltage rail is generated; wherein a voltage value of the supply voltage rail and a target output voltage value corresponding to the input signal maintain a preset voltage difference relationship;

[0007] generating an error signal representing a difference between a target output voltage value corresponding to the input signal and the supply voltage rail;

[0008] decomposing the error signal into at least two frequency components according to a preset frequency band;

[0009] processing different frequency components of the at least two frequency components by using an amplification strategy corresponding to the different frequency components to generate a plurality of processed signal components;

[0010] combining the plurality of processed signal components into a compensation signal, and superimposing the compensation signal on the supply voltage rail to generate a final amplifier output signal.

[0011] In a possible implementation manner of the first aspect, the generating a dynamically changing supply voltage rail in response to the input signal comprises:

[0012] determining a target output voltage value corresponding to the input signal by using a feedforward prediction manner based on an amplitude, a frequency and a change rate of the input signal;

[0013] generating the supply voltage rail according to the target output voltage value.

[0014] In a possible implementation manner of the first aspect, the generating the supply voltage rail according to the target output voltage value comprises:

[0015] coarsely adjusting the target output voltage value by using a switching modulator to generate a main voltage;

[0016] finely adjusting the main voltage by using a linear voltage stabilizer to output the supply voltage rail.

[0017] In a possible implementation manner of the first aspect, the method further comprises:

[0018] collecting the final amplifier output signal;

[0019] adjusting a voltage of the supply voltage rail by using a feedback control loop, so that a preset voltage difference relationship is maintained between the voltage of the supply voltage rail and a voltage of the final amplifier output signal.

[0020] In a possible implementation manner of the first aspect, the decomposing the error signal into at least two frequency components according to a preset frequency band comprises:

[0021] decomposing the error signal into a high-frequency component and a low-frequency component by using a digital filter or an analog filter set.

[0022] In a possible implementation manner of the first aspect, the processing of different frequency components in the at least two frequency components by using corresponding amplification strategies to generate a plurality of processed signal components comprises:

[0023] The high-frequency component is processed by using incremental accumulation modulation and switch amplification.

[0024] The low-frequency component is processed by using linear amplification.

[0025] In a possible implementation manner of the first aspect, the decomposing the error signal into at least two frequency components according to a preset frequency band comprises:

[0026] The error signal is decomposed into a low-frequency component, a medium-frequency component and a high-frequency component.

[0027] In a possible implementation manner of the first aspect, the combining the plurality of processed signal components into a compensation signal comprises:

[0028] The plurality of processed signal components are fused by an LC synthesis network to form the compensation signal.

[0029] In the second aspect, an embodiment of the present application provides a high-voltage amplifier control device, comprising:

[0030] A supply voltage rail generation module is configured to generate a dynamically changing supply voltage rail in response to an input signal, wherein a voltage value of the supply voltage rail and a target output voltage value corresponding to the input signal maintain a preset pressure difference relationship.

[0031] An error signal generation module is configured to generate an error signal representing a difference between the target output voltage value corresponding to the input signal and the supply voltage rail.

[0032] A decomposition module is configured to decompose the error signal into at least two frequency components according to a preset frequency band.

[0033] An amplification processing module is configured to process different frequency components in the at least two frequency components by using corresponding amplification strategies to generate a plurality of processed signal components.

[0034] A synthesis module is configured to combine the plurality of processed signal components into a compensation signal, and superimpose the compensation signal and the supply voltage rail to generate a final amplifier output signal.

[0035] In a third aspect, the embodiments of the present application provide a control device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the high-voltage amplifier control method of any one of the first aspect when executing the computer program.

[0036] The embodiments of the present application generate a dynamically changing power supply voltage rail, so that the voltage is always only slightly higher than the actual required output voltage, greatly reducing the voltage drop and power loss on the power tube of the amplifier, significantly improving the overall energy efficiency, and solving the problem of low efficiency of the traditional linear amplifier. In addition, the residual error signal after dynamic power supply is frequency band decomposed, and the optimal differential amplification strategy is adopted for the error characteristics of different frequency bands (such as high-frequency switching noise and low-frequency linear distortion) for accurate compensation and correction. This fine processing method can effectively eliminate various distortions and noises, thereby realizing high efficiency while ensuring high fidelity and low noise of the output signal, solving the problem that the prior art cannot balance efficiency and accuracy. In addition, the dynamic power supply scheme and the error processing scheme of the present application are deeply coupled and work cooperatively, the former is responsible for the main energy supply and solves the efficiency problem, and the latter is specifically responsible for eliminating the imperfections of the former to guarantee the final output accuracy, and the combination of the two realizes the synergistic gain effect beyond simple technical combination. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0038] Figure 1 A flowchart of a high-voltage amplifier control method provided by an embodiment of the present application;

[0039] Figure 2 A structural diagram of a high-voltage amplifier control device provided by an embodiment of the present application;

[0040] Figure 3 A structural diagram of a dynamic power supply unit in another embodiment of the present application.

[0041] Wherein, the main figure mark explanation is as follows: 100 - dynamic power supply unit; 110 - voltage predictor; 120 - switch mode linear composite power supply; 200 - error processing unit; 210 - error signal generation module; 220 - decomposition module; 230 - amplification processing module; 231 - high frequency processing channel; 232 - low frequency processing channel; 300 - synthesis output unit; 310 - synthesis module; 320 - superposition module; S101 - generating a dynamically changing power supply voltage rail step; S102 - generating an error signal step; S103 - decomposing the error signal step; S104 - differentiating amplification processing step; S105 - synthesizing compensation and superimposing output step. DETAILED DESCRIPTION

[0042] In the following description, for the purposes of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments described. However, it will be apparent to those skilled in the art that the application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the application with unnecessary detail.

[0043] It should be understood that the term "comprises" when used in this specification and the appended claims, specifies the presence of stated features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0044] It should also be understood that the term "and / or" when used in this specification and the appended claims, means any one or more of the associated listed items can be present, and includes multiples of any one or more of the associated listed items.

[0045] As used in this specification and the appended claims, the term "if" can be interpreted as meaning "when" or "once" or "in response to a determination" or "in response to detecting," depending on the context. Similarly, the phrase "if determined" or "if detected" can be interpreted as meaning "once determined" or "in response to a determination" or "once detected" or "in response to detecting," depending on the context.

[0046] In addition, in the description of the application and the appended claims, the terms "first", "second", "third", etc. are used only to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0047] Reference within the specification of this application to "one embodiment" or "some embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in additional embodiments," and so on, in various places in the specification are not necessarily all referring to the same embodiment, unless otherwise specifically specified. The terms "comprises," "comprising," "includes," "including," "has," "having," and the like are meant to be open-ended and non-limiting.

[0048] For the purpose of making the technical solutions and advantages of the present application more clear, the technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Those skilled in the art should know that the technical solutions of the present application can have various modifications and equivalent replacements. For example, the specific filter type, amplifier category, parameter value or digital implementation platform (such as FPGA or DSP) described in the embodiments are all exemplary, and can be adjusted according to specific performance requirements, cost budget and technical conditions. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of the present application.

[0049] Embodiment 1

[0050] The present embodiment provides a high-voltage amplifier control method and device based on multi-dimensional state space prediction and frequency compensation. The core of the technical solution is to construct a double-loop nested architecture that deeply couples the energy loop and the precision loop and works cooperatively. This architecture fundamentally decouples the two tasks of high-power transmission and high-precision signal shaping, which are mutually restrictive, aiming to achieve high efficiency and high precision of the high-voltage amplifier when driving complex loads (such as capacitive or inductive loads) at the same time. The application scenarios of the technical solution are wide, and it is especially suitable for fields with strict requirements on power efficiency and signal fidelity, such as arbitrary waveform generators (AWG), medical ultrasonic probe driving, semiconductor automatic test equipment (ATE), laser modulator driving, and high-precision piezoelectric servo control, etc.

[0051] Reference Figure 1 which is a whole flowchart of a high-voltage amplifier control method provided by the present embodiment. The method mainly includes the following steps:

[0052] Step S101: In response to an input signal, a dynamically changing supply voltage rail that can closely envelope the waveform of the input signal is generated. This step constitutes the energy loop of the system, and its main goal is to provide most of the energy required by the signal with as high efficiency as possible.

[0053] Step S102: generating an error signal that precisely characterizes the instantaneous difference between the ideal target output voltage value corresponding to the input signal and the dynamic supply voltage rail. This error signal aggregates all the imperfections of the energy loop, including noise, ripple, and tracking error.

[0054] Step S103: decomposing the error signal into at least one low-frequency component and one high-frequency component according to a pre-defined frequency band. This step is a diagnosis of the error, decomposing it into components with different physical origins and characteristics.

[0055] Step S104: parallel processing the low-frequency component and the high-frequency component using differentiated amplification strategies that match their characteristics to generate multiple processed signal components. This step matches the optimal processing mode for error components of different nature.

[0056] Step S105: distortionless combining the multiple processed signal components into a unified compensation signal and superimposing the compensation signal with the supply voltage rail to correct its error, thereby generating the final high-fidelity, high-voltage amplifier output signal. This step constitutes the precision loop of the system, restoring the perfect form of the signal through active error cancellation.

[0057] Further, please refer to Figure 2 which is a structural schematic diagram of a high-voltage amplifier control device for implementing the above method. The device mainly includes a dynamic supply unit 100, an error processing unit 200, and a synthesis output unit 300. The technical solutions of the present embodiment will be described in detail below in combination with Figure 1 and Figure 2 .

[0058] In a specific application scenario, an external signal source provides an input signal Vin that needs to be amplified by a high-voltage. The input signal Vin can be any waveform, such as a sine wave, a square wave, a triangular wave, or a more complex modulated signal generated by an arbitrary waveform generator (AWG). The input signal Vin is sent into the control device provided by the present embodiment and is divided into two paths, serving as the reference input of the dynamic supply unit 100 and the error processing unit 200, respectively.

[0059] First, in step S101, a dynamically changing supply voltage rail is generated. This step is achieved by Figure 2The shown dynamic power supply unit 100 is implemented, which plays the role of energy ring in the architecture, and its main goal is to efficiently generate a coarse adjustment power supply voltage that can closely follow the trend of signal changes. Specifically, the input signal Vin is first sent to the voltage predictor 110 in the dynamic power supply unit 100. The voltage predictor 110 is a key module in this embodiment, and its function is to predict the target voltage value required by the amplifier to output the signal without distortion based on real-time analysis of the input signal Vin after a very short time Δt in the future. This prediction is proactive and preemptive, which overcomes the inherent tracking lag problem caused by system loop delay in traditional envelope tracking technology. Especially when the signal slope changes sharply (such as the edges of a square wave or the transient part of a complex waveform), it can avoid signal top or bottom clipping caused by insufficient power supply, or huge instantaneous power waste caused by excessive power supply.

[0060] Unlike the way in the prior art, which only predicts based on the current amplitude or simple first-order differential of the signal, the voltage predictor 110 in this embodiment adopts a voltage prediction model based on a multi-dimensional state space. This model not only considers the current value (position) and rate of change (speed) of the signal, but also innovatively introduces the acceleration (i.e. second-order differential) of the signal and the temperature of the key devices in the system as prediction variables, thereby constructing a more complete and accurate system state description and achieving more accurate prediction of future voltage demand. The prediction model can be realized by real-time digital calculation in a high-performance digital signal processor (DSP) or field programmable gate array (FPGA). The predicted power supply rail voltage can be described by the following formula:

[0061] ;

[0062] wherein, represents the current time, represents the predicted future time; is the instantaneous amplitude of the input signal at the current time; is the first-order differential of the input signal, representing the instantaneous rate of change or slope of the signal; is the second-order differential of the input signal, representing the curvature or acceleration of the signal. It can be understood that the introduction of the second-order differential term is an important improvement of this embodiment, which enables the voltage predictor 110 to predict the inflection point of the signal waveform. For example, at the peak or trough of a sine wave, the first-order differential is zero, and the traditional prediction model will misjudge that the signal tends to be stable, but the absolute value of the second-order differential is maximum at this time. This term can help the system to prepare for deceleration or reverse acceleration in advance, thereby greatly improving the ability to follow rapid changes in the signal and effectively reducing overshoot and lag. The coefficients and is an optimized look-ahead coefficient obtained by system identification or offline calibration according to the dynamic response characteristics of the whole amplifier system (such as loop delay of switching power supply, turn-on time of power tube, etc.), which is used to weight the first and second order differential terms. is a nonlinear compensation function whose input is the real-time temperature T of the power device (such as switching tube or linear regulating tube) monitored by a temperature sensor (such as NTC thermistor attached near the power device or integrated temperature sensor). This function is used to compensate the device parameter drift (such as the change of on-resistance , threshold voltage of MOSFET) caused by temperature change, so as to offset the influence of thermal effect on the stability of supply voltage, which is particularly important under long-time high-power working conditions. This function can be implemented as a lookup table (LUT) or a polynomial function. is a preset fixed voltage margin, which can be set to 3 to 5 volts for example. Its role is to ensure that the supply voltage rail Vrail generated by the dynamic power supply unit 100 is always higher than the signal voltage to be finally output by a safe margin, leaving necessary voltage headroom for the subsequent error compensation loop and the operation of linear amplification stage, preventing it from entering the saturation region. The selection of this margin is a trade-off: too small may cause clipping during transient response, and too large will reduce system efficiency.

[0063] The predicted supply rail voltage is calculated by the voltage predictor 110, which is then sent as a control instruction to the switching-linear composite power supply 120. The switching-linear composite power supply 120 is a composite power supply structure that takes into account both efficiency and accuracy. Its design idea is to let the two different types of power supply each play to their strengths and make up for each other's weaknesses. In this embodiment, it includes a front-end high-frequency step-down switching regulator (for example, a synchronous Buck converter working at several hundred kHz) and a rear-stage series high-speed linear regulator (for example, a low-dropout regulator LDO). The high-frequency switching regulator converts a high-voltage DC bus voltage to a coarse main voltage quickly according to the instruction, completing the main power transfer. Due to the working characteristics of the switching regulator, this main voltage will contain certain switching ripple and high-frequency noise. In order to eliminate these disturbances, the main voltage is sent to the rear-stage linear regulator. The linear regulator takes the main voltage as input, uses it as its own power supply, and performs fine adjustment and filtering on it, taking advantage of its excellent power supply rejection ratio (PSRR) characteristics to filter out most of the switching ripple and transient noise, and finally outputs a relatively pure voltage value that accurately follows The power supply voltage rail Vrail changes dynamically according to the command. Through this composite structure of switching coarse adjustment and linear fine adjustment, the switching-linear composite power supply 120 utilizes the high efficiency of switching power supplies (typically >90%) and achieves low output noise and fast transient response with the help of linear power supplies, laying the foundation for the efficient operation of the entire system.

[0064] Next, step S102 is executed to generate an error signal. This step is performed by... Figure 2 The error signal generation module 210 in the error processing unit 200 shown completes this process. The error signal generation module 210 receives two inputs: one is the original input signal Vin (or a signal scaled proportionally by a precision attenuator, representing the ideal target output voltage), and the other is the actual supply voltage rail Vrail generated by the dynamic power supply unit 100. Internally, the error signal generation module 210 can be a high-precision, high common-mode rejection ratio (CMRR) differential amplifier circuit that calculates the difference between these two signals to generate the error signal Verror. (Where k is the system's preset total gain). This error signal Verror comprehensively reflects all residual deviations between the energy loop and the ideal target after dynamic voltage tracking. These deviations are a collection of various errors, including but not limited to: the output ripple of the switching linear composite power supply 120 itself, high-frequency noise, and other factors. The error is caused by instruction tracking delay and nonlinear distortion, as well as prediction errors due to imperfections in the prediction model itself. Since the dynamic power supply unit 100 has completed the supply of the main energy to the signal, the amplitude of this error signal Verror is usually much smaller than the amplitude of the main signal Vin, and its energy is also greatly reduced, which creates favorable conditions for subsequent low-power, high-precision processing.

[0065] Subsequently, in step S103, the error signal is decomposed. The error signal Verror is output from the error signal generation module 210 and sent to the decomposition module 220. The role of the decomposition module 220 is to decompose the error signal Verror containing a plurality of complex components according to a predetermined frequency band, so as to be subsequently processed in a targeted manner. In an embodiment of the present application, we decompose the error signal into two frequency components: a high-frequency component and a low-frequency component. The decomposition operation can be completed in the digital domain. For example, the error signal Verror can be first sampled and digitized by a high-speed, high-resolution analog-to-digital converter (ADC), and then sent to a digital filter bank integrated in the same FPGA as the voltage predictor 110. In order to ensure that the decomposition process does not introduce additional phase distortion, thereby affecting the accurate synthesis of the subsequent signal, the embodiment preferably uses a linear phase finite impulse response (FIR) filter bank. The filter bank can be composed of a low-pass FIR filter and a high-pass FIR filter, and the cutoff frequency (or crossover frequency) can be set according to the system characteristics and error spectrum analysis. For example, a typical dividing point is 20 kilohertz, and this selection is based on the analysis of the error sources: error components below 20 kilohertz are divided into low-frequency components, mainly including linear distortion, harmonic distortion of the signal, and slow drift caused by temperature changes; error components above 20 kilohertz are divided into high-frequency components, mainly including switching noise, high-frequency ripple introduced by the switching linear composite power supply 120, and errors caused by rapid transients of the signal.

[0066] Then, in step S104, the decomposed frequency components are differentially amplified. This step is completed by the amplification processing module 230, which contains dedicated processing channels for different frequency components inside, embodying the core idea of the present application of divide and conquer. The high-frequency component and the low-frequency component output by the decomposition module 220 are sent to the high-frequency processing channel 231 and the low-frequency processing channel 232, respectively.

[0067] In the high-frequency processing channel 231, an efficient amplification strategy is adopted for the high-frequency error component. Specifically, the high-frequency component is first sent to a high-order delta-sigma modulator, for example, a seventh-order delta-sigma modulator. ​​Modulation techniques are able to quantize a signal with a high oversampling ratio, and utilize its noise shaping property to push the noise energy generated in the quantization process to a high frequency region far above the signal bandwidth, thus achieving a very high signal-to-noise ratio within the signal band. After modulation, the high frequency error component is converted into a high speed single-bit or multi-bit digital stream (i.e. a pulse density modulated or pulse width modulated signal). This digital stream is then used to directly drive a high efficiency class-D switching power bridge. Class-D amplifiers are well known for their theoretical efficiency close to 100%, and are very suitable for amplifying this pulse encoded signal. Accordingly, by utilizing the high efficiency property of class-D amplifiers to handle the high frequency error component containing a large amount of switching noise and high frequency dynamics, the precise amplification and compensation of the high frequency error can be accomplished without introducing significant additional power consumption.

[0068] In the low frequency processing channel 232, a high precision amplification strategy is employed for the low frequency error component. The low frequency component mainly contains linear distortion and harmonic distortion components which have a great impact on the fidelity of the final output signal. Therefore, this component is sent to an AB class analog amplifier with very high open loop gain, very low noise figure and excellent linearity for amplification. It can be understood that since the AB class amplifier only needs to handle a low frequency error signal with small amplitude and low frequency, and it is working in a far from full power output state, its static and dynamic power consumption is very low, and its impact on the overall efficiency is minimal. At the same time, the high linearity inherent in the AB class amplifier (especially the elimination of crossover distortion in class-B amplifiers) can ensure that the low frequency error component does not generate new distortion during amplification, thus providing a high fidelity guarantee for subsequent precise compensation.

[0069] Through the above differential processing, the present application adopts an efficient way (class-D amplification) to handle the high frequency noise which may have large energy but has low requirements on linearity, and adopts a high precision way (AB class linear amplification) to handle the low frequency distortion which has very high requirements on linearity but has small energy, thus realizing the optimal allocation of system resources and avoiding the need for a single amplifier to simultaneously meet the contradictory requirements of efficiency and precision.

[0070] Finally, in step S105, the compensation signal is synthesized and superimposed with the supply voltage rail to generate the final output. This step is accomplished by the synthesis output unit 300. The two processed signal components output by the high frequency processing channel 231 and the low frequency processing channel 232 are sent to the synthesis module 310 in the synthesis output unit 300. As an optional implementation, in order to avoid introducing new distortion due to load effect and impedance mismatch in the signal synthesis process, the present embodiment employs a floating current mode synthesis bridge circuit. Specifically, the amplified signal from the high frequency processing channel 231 (which can need to pass through a simple LC low pass filter to filter out the ultrahigh frequency modulation carrier) and the amplified signal from the low frequency processing channel 232 are first converted into current signals by respective voltage controlled current sources (VCCS). The two current signals are injected onto a common low impedance synthesis node (e.g. the virtual ground of an operational amplifier), where they are summed up without distortion to form a total compensation current. This compensation current is then converted back into a voltage signal by a transimpedance amplifier (TIA), thereby generating the final compensation signal Vcomp. This compensation signal Vcomp precisely replicates the amplified version of the original error signal Verror, and its polarity is opposite to the original error.

[0071] The compensation signal Vcomp is then sent to the superimposition module 320 together with the supply voltage rail Vrail from the dynamic power supply unit 100. The superimposition module 320 can be a high speed, high voltage slew rate adder circuit, or in some implementations, the compensation signal can be directly coupled to the supply voltage rail through a fast power amplifier in series or through a transformer. The superimposition module 320 superimposes the compensation signal Vcomp with the supply voltage rail Vrail. Since Vcomp is an exact inverse compensation of the original error , this superimposition operation is equivalent to actively subtracting all the error components from the coarsely regulated supply voltage rail Vrail. Finally, the output of the superimposition module 320 generates the final high voltage amplifier output signal Vout. This output signal Vout retains the high efficiency brought by the dynamic power supply, and at the same time eliminates various distortions and noises through the fine compensation of the error loop, thereby achieving extremely high signal fidelity, with a theoretical output of , by properly setting the compensation loop gain G, it is possible to unlimitedly approach the ideal .

[0072] In summary, the embodiment realizes high-efficiency dynamic power supply energy ring through voltage prediction and switch linear composite power supply, and realizes accurate compensation of errors through fine frequency division processing and optimal differential amplification strategy of residual error signals. The overall efficiency can reach more than 85%, and the total harmonic distortion and noise (THD+N) of the output signal can be lower than -100dB, thereby successfully solving the problem that the efficiency and accuracy are difficult to balance in the prior art.

[0073] Embodiment 2

[0074] As an optional implementation manner, the embodiment is a variant of embodiment 1, and aims to show the possibility of realizing error signal decomposition and processing in the pure analog domain. This scheme provides an effective alternative implementation for some ultra-wideband application scenarios (such as high-speed probe driving, laser modulation, etc.) that are inconvenient to use digital processors or have extreme requirements for system delay. The overall architecture of the embodiment is similar to that shown in Figure 2 The main difference between the embodiment and embodiment 1 is the specific implementation manner of the decomposition module 220 in the error processing unit 200.

[0075] In embodiment 1, the decomposition module 220 adopts a digital manner and is realized through a digital filter group in the FPGA, which includes the sampling delay of the ADC and the pipeline delay of the digital processing. In the embodiment, the analog error signal Verror generated by the error signal generation module 210 is no longer converted by the analog-to-digital converter, but is directly sent to a pure analog filter circuit for decomposition.

[0076] Specifically, the decomposition module 220 can be realized by a fourth-order state variable filter. Compared with traditional analog filters such as Butterworth or Chebyshev, the state variable filter has unique advantages. A standard state variable filter circuit can provide accurate high-pass, low-pass and band-pass filter results at different output ends simultaneously using only a single input, and the output signals have a strict and fixed phase relationship (for example, the low-pass and high-pass outputs differ by 180 degrees), which greatly facilitates subsequent signal synthesis and avoids complex phase calibration networks.

[0077] In the embodiment, the high-pass and low-pass outputs are used. By carefully designing the element parameters (such as precise resistance and capacitance values) of the filter, the same crossover frequency as the digital filter in embodiment 1 can be set, for example, 20 kHz. After the error signal Verror is input, the high-pass output of the state variable filter obtains the high-frequency error component, and the low-pass output obtains the low-frequency error component.

[0078] It is noted that a key advantage of the state variable filter used in this embodiment is that by adjusting the feedback network (Q value) inside, a smooth power complementary crossover characteristic (e.g. a Linkwitz-Riley filter response) can be achieved around the crossover frequency. That is, around the crossover frequency, the attenuation of the high-pass filter and the low-pass filter are complementary to each other, so that the sum of the power of the two output components remains constant throughout the frequency band. This ensures that the high and low frequency components can perfectly reconstruct the original error signal Verror after being superimposed in the subsequent synthesis module 310, fundamentally avoiding the crossover distortion caused by the phase nonlinearity and amplitude drop of the traditional analog filter around the crossover frequency.

[0079] After the error signal is decomposed into analog high and low frequency components, the subsequent processing process remains the same as in Embodiment 1. The high frequency component is sent to the high frequency processing channel 231, which can include an analog input modulator and a class-D amplifier. The low frequency component is sent to the AB class linear amplifier in the low frequency processing channel 232 for high-fidelity amplification. Finally, the two amplified analog signal components are synthesized in the synthesis module 310 (e.g. through a current-mode synthesis bridge) to generate the compensation signal Vcomp, which is superimposed with the supply voltage rail Vrail and output.

[0080] This embodiment demonstrates the flexibility of the technical solution of the present application by implementing error decomposition in the analog domain. The advantage of this scheme is that it completely avoids the quantization noise, harmonic distortion and processing delay that may be introduced by the analog-to-digital converter and digital-to-analog converter. Therefore, in some ultra-high frequency applications that require extremely high transient response speed and extremely wide compensation bandwidth, the analog processing scheme of this embodiment may exhibit superior performance and wider closed-loop bandwidth than the digital scheme.

[0081] Embodiment 3

[0082] This embodiment provides another alternative implementation of generating a dynamic supply voltage rail as a parallel scheme to the feedforward prediction scheme in Embodiment 1. This scheme uses feedback control based on the final output signal to dynamically adjust the supply voltage rail, rather than feedforward prediction based on the input signal. This scheme has better robustness and adaptability when the input signal contains large noise or its derivative characteristics are difficult to obtain, or when the system model is difficult to accurately identify.

[0083] Referring to Figure 3 , which is a structural schematic diagram of the dynamic supply unit used in this embodiment. The dynamic supply unit mainly includes a feedback controller and a switching linear composite power supply. The hardware structure of the switching linear composite power supply can be the same as that of the switching linear composite power supply 120 in Embodiment 1.

[0084] Unlike the embodiment 1, this embodiment does not rely on complex differential operation and prediction on the input signal Vin. The starting point of its control logic is to directly monitor the final amplifier output signal Vout and ensure that the supply voltage rail Vrail is always accurately floating above Vout by a fixed margin .

[0085] The specific workflow is as follows: the final amplifier output signal Vout is sampled and, together with the actual supply voltage rail Vrail generated by the switched linear hybrid power supply, is input as a feedback signal into the feedback controller. The core of the feedback controller is a hybrid proportional-integral (PI) control law with output feedforward.

[0086] It can be understood that simple feedback control (such as traditional PI or PID controllers) is often limited in response speed and tracking accuracy when tracking high-speed changing signals due to inherent processing delay and phase lag of the integral element of the system, and it is difficult to maintain a constant voltage margin when the signal changes dramatically, which may result in excessive margin (loss of efficiency) or too small margin (generation of clipping distortion).

[0087] To overcome this defect, the feedback controller of this embodiment adopts a hybrid control law, which generates a control signal that follows the following formula:

[0088] ;

[0089] In this formula, the feedback error is defined as the difference between the target supply rail voltage and the actual supply rail voltage, i.e. where is the preset voltage margin (equivalent to in the embodiment 1). The part of the formula is a standard PI controller, which is responsible for adjusting the error . Its main role is to eliminate steady-state error and ensure that can accurately converge to the target value when the signal changes slowly or stably, ensuring the static accuracy of the system.

[0090] A key improvement of this scheme is the introduction of the output feedforward term . This term directly feeds forward the final output signal Vout after weighting by a proportional coefficient and is superimposed on the output of the PI controller. This design enables the supply system to respond immediately and quickly to changes in the output voltage without waiting for the error and accumulation. When the output signal Vout rises suddenly, the feedforward term will immediately increase the control signal , driving the switch-mode linear hybrid power supply to rapidly boost its output voltage, and vice versa. The PI control loop then makes fine corrections on the basis of this rapid coarse adjustment provided by the feedforward term, to eliminate the remaining small errors. This control strategy can be analogized as an experienced driver: the feedforward term is like the driver's steering action in advance according to the road conditions in the distance, and the PI feedback term is like the driver's fine adjustment according to the actual position of the vehicle in the lane, and the combination of the two achieves rapid and smooth driving.

[0091] This hybrid control method combining the rapid response of feedforward and the accurate adjustment of feedback integrates the speed advantage of feedforward control and the accuracy advantage of feedback control, so that the dynamic power supply unit can achieve rapid, stable and accurate dynamic following of the final output signal.

[0092] The subsequent error processing flow of this embodiment (i.e. the working of the error processing unit 200 and the synthetic output unit 300) is exactly the same as that of embodiment 1 or embodiment 2. By adopting this feedback type dynamic power supply scheme, high efficiency and high accuracy can also be achieved. The advantage of this scheme is that it does not need to perform complex differentiation operation on the input signal, so it is not sensitive to the noise that may exist in the input signal, the overall robustness of the system is stronger, and the design and debugging are relatively simple, thus showing multiple possibilities for implementing the technical scheme of the present application.

[0093] Embodiment 4

[0094] This embodiment is a further extension and optimization of embodiment 1, aiming to show the potential of the technical scheme of the present application in pursuing extreme performance. This embodiment decomposes the error signal into three or more frequency bands, and performs more fine frequency division processing and compensation on the error, to obtain lower distortion and noise than the double-band decomposition scheme. This scheme reflects the scalability of the architecture of the present application, which can be customized according to the extreme requirements of the application on performance.

[0095] In this embodiment, the overall architecture of the high-voltage amplifier control device can still refer to Figure 2 , but the decomposition module 220 and the amplification processing module 230 inside the error processing unit 200 are designed to be more complex and fine.

[0096] In particular, the decomposition module 220 decomposes the error signal Verror into three frequency bands: low frequency components, mid frequency components, and high frequency components. This multi-band decomposition is based on deep insights into the physical sources of the error. This can be achieved by a filter bank composed of band-pass filters and high / low-pass filters, either in the digital domain or in the analog domain. For example, two cut-off frequencies can be set, such as 1 kHz and 100 kHz, to divide the error spectrum into three regions, each corresponding to a different physical source of error:

[0097] 1) Low frequency components (frequency < 1 kHz): The errors in this band mainly correspond to the 1 / f noise (i.e. flicker noise) of the system and slow thermal drifts caused by environmental temperature, device aging, etc. These ultra-low frequency errors are crucial for the DC accuracy and long-term stability of the signal.

[0098] 2) Mid frequency components (1 kHz < frequency < 100 kHz): This band is usually where the main harmonic distortion components of the signal are concentrated. The second, third, and higher harmonics generated by the non-linearity of the amplifier mostly fall within this range, directly affecting the fidelity of the signal (i.e. the THD index).

[0099] 3) High frequency components (frequency > 100 kHz): The errors in this band mainly correspond to the high-frequency switching noise and ripple introduced by the switching power supply in the dynamic power supply unit, as well as the transient intermodulation distortion (TIM) caused by the slew rate limitation of the power tube when processing fast signal transients, etc.

[0100] Correspondingly, the amplification and processing module 230 also sets up three parallel, respectively optimized processing channels to differentially amplify the three error components:

[0101] 1) For low frequency components (< 1 kHz): An ultra-high precision class-A amplifier is used for processing. In order to maximize the suppression of 1 / f noise and DC drift, the amplifier can integrate Chopper-Stabilized technology or Auto-Zeroing technology. These technologies effectively remove the low-frequency noise and offset voltage of the amplifier itself from the signal path by modulating the low-frequency signal to a higher carrier frequency for amplification, and then demodulating it back to the original frequency, which can reduce the equivalent input offset voltage drift to the nanovolt (nV) level, providing the system with extremely high DC accuracy and long-term stability.

[0102] 2) For mid frequency components (1 kHz-100 kHz): A high-speed, high-slew-rate class-AB amplifier designed specifically for audio or high-fidelity applications is used for processing. The design focus of this amplifier is to provide the best linearity and lowest harmonic distortion in this core frequency band to ensure the most accurate compensation for the main distortion components of the signal.

[0103] 3) For high frequency components (>100kHz): the same processing strategy as in embodiment 1 is adopted, with a high-efficiency modulator and class-D switching amplifier, focusing on the highest efficiency to suppress high-frequency noise and ripple, completing the cleaning work in the high-frequency band.

[0104] After the fine amplification processing through the three special channels, the three processed signal components are finally fused without distortion in the synthesis module 310 (for example, through a three-input current-mode synthesis bridge) to form a unified compensation signal Vcomp.

[0105] By performing more fine frequency band division on the error signal and configuring an optimal amplification strategy for each frequency band-specific error source (such as 1 / f noise, harmonic distortion, switching noise), this embodiment can achieve a lower overall distortion and noise level than the dual-band decomposition scheme, achieving extreme signal fidelity. This scheme is particularly suitable for application fields with the most stringent requirements for signal quality, such as metrology standard sources, high-precision physical experiment equipment, semiconductor parameter testing systems, etc., fully demonstrating the flexibility and scalability of the technical solution of the present application.

[0106] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Various modifications and changes can be made by those skilled in the art. For example, the decomposition of the error signal can be extended to N frequency bands, each using the optimal amplification technique; the prediction model or feedback control law of the dynamic power supply unit can also use more complex algorithms, such as Kalman filter-based prediction or adaptive fuzzy control, etc. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included within the scope of protection of the present application.

[0107] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included within the scope of protection of the present application.​

Claims

1. A high-voltage amplifier control method, characterized by, The method comprises: generating a dynamically changing power supply voltage rail in response to an input signal; wherein a preset voltage difference relationship is maintained between the voltage value of the power supply voltage rail and the target output voltage value corresponding to the input signal; generating an error signal representing the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail; decomposing the error signal into at least two frequency components according to a preset frequency band; processing different frequency components in the at least two frequency components using corresponding amplification strategies to generate a plurality of processed signal components; combining the plurality of processed signal components into a compensation signal, and superimposing the compensation signal with the power supply voltage rail to generate a final amplifier output signal; wherein the generating of the dynamically changing power supply voltage rail in response to the input signal comprises: using a voltage predictor to perform real-time analysis on the input signal, predicting a target output voltage value required by the amplifier to output the amplifier output signal without distortion, and generating the power supply voltage rail according to the target output voltage value; the generating of the error signal representing the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail comprises: calculating the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail based on a differential amplification circuit to generate the error signal.

2. The high-voltage amplifier control method of claim 1, wherein, the generating of the dynamically changing power supply voltage rail in response to the input signal comprises: determining the target output voltage value corresponding to the input signal using a feedforward prediction method based on the amplitude, frequency and rate of change of the input signal; generating the power supply voltage rail according to the target output voltage value.

3. The high-voltage amplifier control method of claim 2, wherein, the generating of the power supply voltage rail according to the target output voltage value comprises: coarsely adjusting the target output voltage value using a switching modulator to generate a main voltage; finely adjusting the main voltage using a linear voltage stabilizer to output the power supply voltage rail.

4. The high-voltage amplifier control method of claim 1, wherein, The method further comprises: collecting the final amplifier output signal; adjusting the voltage of the power supply voltage rail through a feedback control loop, so that the voltage value of the power supply voltage rail and the voltage value of the final amplifier output signal maintain the preset voltage difference relationship.

5. The high-voltage amplifier control method of claim 1, wherein, the decomposing of the error signal into at least two frequency components according to a preset frequency band comprises: decomposing the error signal into a high frequency component and a low frequency component through a digital filter or an analog filter bank.

6. The high-voltage amplifier control method of claim 5, wherein, the processing of different frequency components in the at least two frequency components using corresponding amplification strategies to generate a plurality of processed signal components comprises: processing the high frequency component using incremental cumulative modulation and switching amplification; processing the low frequency component using linear amplification.

7. The high-voltage amplifier control method of claim 1, wherein, the decomposing of the error signal into at least two frequency components according to a preset frequency band comprises: decomposing the error signal into a low frequency component, a medium frequency component and a high frequency component.

8. The high-voltage amplifier control method of claim 1, wherein, the combining of the plurality of processed signal components into a compensation signal comprises: fusing the plurality of processed signal components through an LC synthesis network to form the compensation signal.

9. A high-voltage amplifier control device, characterized by comprising: The application relates to a high-voltage amplifier control method, comprising the following steps: a power supply voltage rail generating module is used to generate a dynamically changing power supply voltage rail in response to an input signal; wherein a preset pressure difference relationship is maintained between the voltage value of the power supply voltage rail and the target output voltage value corresponding to the input signal; an error signal generating module is used to generate an error signal representing the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail; a decomposition module is used to decompose the error signal into at least two frequency components according to a preset frequency band; an amplification processing module is used to process different frequency components in the at least two frequency components by using corresponding amplification strategies to generate a plurality of processed signal components; a synthesis module is used to synthesize the plurality of processed signal components into a compensation signal, and the compensation signal is superimposed with the power supply voltage rail to generate a final amplifier output signal; wherein the step of generating a dynamically changing power supply voltage rail in response to an input signal comprises the following steps: a voltage predictor is used to analyze the input signal in real time, to predict the target output voltage value required by the amplifier to output the amplifier output signal without distortion, and to generate the power supply voltage rail according to the target output voltage value; the step of generating an error signal representing the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail comprises the following steps: a differential amplification circuit inside the error signal generating module is used to calculate the difference between the target output voltage value corresponding to the input signal and the power supply voltage rail, and to generate the error signal.

10. A control device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the high-voltage amplifier control method according to any one of claims 1 to 8.

Images