TW498663B - Low distortion digital power amplification method and system therefor - Google Patents

Abstract

The invention relates to a low distortion digital power amplification method and system there for, in which primarily a modulation unit or an adjustment unit provide a comparison signal to input a control signal in accordance with the comparison signal and a feedback signal of a power switching unit so that the adjustment can have corresponding adjustment concerning each pulse width and phase of the inputted pulse width modulation (PWM) signal based on the control signal to generate an adjustment signal of the input power switching unit, making the center position of each pulse and the center position of the corresponding pulse for the inputted PWM signal have an expected phase difference and achieve the objective of eliminating phase noise and limiting system noise.

Description

Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 498663 A7 _ B7 V. Description of the Invention (1) The present invention relates to a low-distortion power amplification method and system, and particularly to a method that can eliminate asymmetrical signals due to phase shift Low-distortion digital power amplification method of generated phase noise and internal noise of the system and a system thereof. According to the general, digital power amplifiers mainly use a power switching circuit to amplify the input signal. At present, there are two types of power switching circuits, which are unilateral (using "outside magic switching circuits and bilateral (H-bridge) switching circuits, and of which The switching action is based on the transistor as the main switching element. If it is assumed that the general transistor system is an ideal element, the pulse width modulation (PWM) signal of the digital power amplifier is input through the switching element of the power switching circuit. 'It should be amplified very realistically, and then through a low-pass filter and wave filter to remove high-frequency components, a complete low-frequency amplified signal should be obtained at the output of the digital power amplifier. However, in a real circuit It is not possible to generate such a complete signal, because the power supply of a general digital power amplifier generates ripples, and the transistor used as its switching element is not an ideal element, and when the switching element is turned ON alternately, 0FF is switched to When the signal is amplified, the switching timing of its ON and OFF cannot be fully matched, and various factors such as load effect will be in the A lot of noise is generated in the bit power amplifier, so that the input PWM signal cannot get a complete amplified signal after being amplified by the digital power amplifier, which causes serious distortion of the output signal. Therefore, in order to eliminate the noise caused by the above factors, Xi It is known that a pulse edge delay (pulse edge delay) technology is proposed. Please refer to the first figure, which is a digital power amplifier disclosed in International Publication No. W098 / 44626 and International Application No. PCT / DK / 00133. Outline Circuit Page 3 This paper size is applicable to China National Standard (CNS) A4 (210 X 297 mm) (Please read the precautions on the back before filling this page) -------- Order I! I- -. 498663 Printed page A7 of the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 5. Description of the invention (2 Block diagram, the digital power amplification 1. ^, t 15 1 mainly includes a correction unit n, 'a power switching unit 12 、 和 一 ERROR # _ First is the difference processing early stage> 13, where the correction unit 11 is connected to the power cutting system ― Qian 丨 tool change early 7G 12 and the error processing unit 13 respectively, and the correction unit 11 is Temporary士 # 输入 is input to the pulse modulation signal Vi, and according to the error processing unit 13, ^ v, which is one of the error signals Ve, and the pulse wave modulation signal Vi is relative to each pulse edge M ^ •豕 The relative delay is to output a corrected pulse oil after adjusting the width of each pulse wave. The pulse signal Vc of the magic wave is adjusted to the power cut β 奂 unit 1 2 so that the power is passed through the power switching unit. After proper amplification, an output signal Vo ′ is generated and is switched by the power switch. Wu Wuwu 2 returns the output signal V0 to the error processing guide 13 and binds the input signal Vi to the input signal Vi. After the comparison process, a new error signal Ve is generated when the error is 5 and X is received -ve is transported to the modified early element n to adjust the pulse width of the next input pulse modulation signal Vi. Therefore, the conventional digital power amplifier * is crying. Bing m〆qian double generation 1 uses the pulse wave edge delay method 'according to the feedback error signal Ve, suitable for winter, models private x 1b crossing the field to adjust the input pulse wave modulation signal vi The width of each pulse wave can achieve the relative suppression of the internal noise of the digital power amplifier, so as to obtain a more complete output signal waveform. However, although the above-mentioned knowledge has the effect of eliminating the noise caused by the chelation value and the negative month of the digital power amplifier, in the process of eliminating such noise, it also generates another phase noise. K Please refer to the second figure, which is the correction of the pulse wave modulation signal V i waveform of the knife state 1 when the digital power is entered and the output is adjusted by the correction unit n outline 单元 ^ 1 The waveform of the signal Vc can be seen from the figure, and the distance between the center position of each pulse wave of the signal Vc and the center of the input signal Vi—the relative position of the pulse wave (phase difference △ 1, △ 2 and △ 3 w) are different. This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm). M -------- ^ --------- (Please read the precautions on the back before filling this page) 498663 A7

V. Explanation of the invention (3) After the printed signal Vc is printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs, it will provide a low phase noise reduction method, a low power amplification system, true amplification signal rate amplification method, and a control method based on the comparison. There is a pre-system noise rate amplification system switching unit between the modification and input of the modulation signal switching level, and the input signal is transferred: the processing unit correction unit is connected to the output unit of the switching unit; the repair, but these are not Therefore, the noise of the power amplifying side is effective. The complete amplifying edge of the power amplifying side is as follows, including the following steps: the power switching control signal, the width of the signal and the phase of the modified signal. Each, by this means, a modulation unit element, in which the width modulation signal and the above-mentioned modulation modulation unit input signals generate a same phase output signal Vo and cause the output signal of the main method of the present invention and its system. . The other method of the present invention and the system signal thereof achieve the low loss of the present invention: providing a ratio of a level output correspondingly adjusting the input bit to generate a phase of each pulse relative to the pulse wave to eliminate the phase. The modulation unit number alone, and the comparison signal of the unit and power output control signal will be distorted by the spectrum display number.

In order to achieve the purpose of eliminating the noise, the digital power output signal with low distortion is output and the signal is given. The pulse width produced is the center noise of the center position of the input power wave and the true digital power is suppressed. A power is a digital one that can compare the signal switching unit and the signal and power into page 5 of the correction sheet, causing the digital distortion of the phase miscellaneous and the digital distortion of the system's internal distortion can output more effective. ’Mainly include a signal ring and a signal; according to the positive signal of each pulse, the phase difference effect of the pulse width is adjusted. 'Mainly includes and a correction unit is replaced by a pulse 70' are connected separately to feedback the positive unit according to one, which is II .-------- r — MW -------- Order --------- Line-#-(Please read the precautions on the back before filling this page) This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) 498663 A7 B7 5 4. Description of the invention (4) For the pulse width modulation signal input of the aforementioned modulation unit, and according to the control signal of the processing unit, the corresponding adjustment of each pulse width modulation signal • the width and phase of the pulse wave To generate a correction signal to input the power switching unit, so that there is an expected phase difference between the center position of each pulse wave of the correction signal and the center position of each relative pulse wave of the pulse width modulation signal, thereby achieving The effect of eliminating phase noise and system noise. In order to have a more detailed and clear understanding of the purpose, effects and structural features of the present invention, the preferred embodiment is described below in conjunction with the drawings. First, please refer to the third and tenth drawings, which are the first embodiment of the present invention. The schematic flow chart of the low-distortion digital power amplification method of the preferred embodiment and the schematic diagram of the internal circuit of one of the power switching stages 33 in the digital power amplifier. And in order to eliminate the phase noise caused by the inconsistent phase offset mentioned above, the method is to adjust the phase and width of each pulse of a pulse width modulation signal appropriately, so that the power through the digital power amplifier is switched. Level 3 3 After amplification, 'a low-distortion output signal can be obtained; please refer to the third figure' This method includes the following steps: 'First, a comparison signal Vm is provided, and the comparison signal Vm and a digital The feedback signal _Vf output from the power switching stage 33 of the power amplifier generates a control signal Vc; — Then, according to the control signal Vc, each pulse of the pulse width modulation signal PWM of the input digital power amplifier is adjusted correspondingly. Width and Phase 'to generate a modified signal Ve input to the power switching stage 3 3 such that the center position of each pulse wave of the modified signal Ve and the center of each relative pulse wave of the input pulse width modulation signal PWM There is an expectation between the positions on page 6. This paper size applies to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) ------ I ---- II · II I --- 1 Order ·! I!-Line (Please read the notes on the back before filling out this page) Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 498663 A7 ____ B7 V. Phase difference of the invention description (5). Among them, in this embodiment, the pulse width modulation signal PWM is divided into two signals CH-R and CH-L, and the correction signal Ve inputted into the power switching stage 33 is divided into R-U, R-d, L a u and L — d four signals, then the feedback signal Vf output by the power switching stage 33, as shown in the ninth figure, 'is actually the correction signal V e amplified by the power switching stage 3 3 The signals Vol, Vo2, Vo3, and Vo4 are subtracted, that is, the difference between Vo2 and Vo4 (or Vo 1 and Vo 3) is subtracted, and then divided by a proportionality constant κ. The comparison signal Vm, as shown in the third figure, can be the difference signal obtained by subtracting the two signals of CH — R and CH — L of the pulse width modulation signal PWM according to different requirements of the system design. , Or the difference signal obtained by subtracting the R — U and L — u (or R — d and L_d) of the correction signal Ve. Then the control signal Vc can be obtained by subtracting the comparison signal vm from the feedback signal vf, but because the feedback signal Vf amplified by the power switching stage 33 contains complex noise, the comparison signal Vm and the feedback The difference signal obtained by subtracting the signal Vf is actually a wideband analog signal containing a lot of noise. Therefore, the difference signal must be processed through a low-pass filtering and sample-and-hold step. A control signal Vc containing only low-frequency components is obtained, and the control signal Vc can generate a relative displacement amount for each pulse wave of the input pulse width modulation signal PWM as a pulse wave number of 1, 2, 3, etc.) . In addition, in this embodiment, a table test signal Vmax is provided, which can generate a relative reference displacement Amax. Therefore, the present invention is based on the control signal Vc, each time the pulse width modulation signal PWM is adjusted. The relative displacement of a pulse wave is △ max- △ n / 2 distance (n: = pulse wave number. Page 7 This paper applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) --- r --- ----------------- Order --------- ^ --Awl (Please read the notes on the back before filling in this page} 498663 Shi Hongzheng Employees ’Consumer Cooperation Du printed A7 V. Description of the Invention (6) 1,2,3 ···) 'Then adjust each pulse wave correspondingly, such as the pulse wave number 1,2,3 ··· ) Width, the center position of each pulse wave of the resulting modified signal and the center position of each relative pulse wave of the input pulse width modulation signal pwM differ by an expected phase difference, and the The expected phase difference is not affected by other noise in the system and the error during implementation, as follows: • Δ max-Δ η / 2 + Δ η / 2 = Δ max The above equation can be known, although Enter it The value of △ ❿ to be adjusted for each pulse of the wave width modulation signal is different, but each pulse is shifted by △ 7. And the width adjustment “after” will cause the correction signal to be generated. The phase difference of Ve relative to the pulse width modulation signal is almost the same (that is, equivalent to Δπ ^ χ), so that the output signal M per pulse is relative to each pulse of the input signal (pulse width modulation signal PWM). The phase difference of the wave system is equivalent to the phase difference of △ max, and the SNR (signal-to-noise ratio) of this I ~ system, so the turtle not only suppresses the internal noise of the digital power amplifier, but also eliminates the possibility of The phase noise makes the output signal V. The distortion is reduced to the maximum. In addition, it is worthwhile / thinking that, for the pulse width modulation, the present invention can adjust the displacement and width of the PWM signal in the order of adjustment, that is, the pulse width modulation signal can be adjusted first. After adjusting the width of each pulse wave of pwM, then adjust its displacement distance, or phase shift and width adjustment are performed at the same time. The suppression effect of the phase noise obtained is the same as the method of adjusting the displacement and adjusting the width described above. . The paper size on page 8 applies to China National Standard (CNS) A4 (210 X 297 cm) --------------------- 1 Order * II --- --I (Please read the notes on the back before filling this page) 498663 A7 -----_ B7 V. Description of the invention (7) (Please read the notes on the back before filling this page) In addition 'In this embodiment The provided reference signal Vmax is set according to the maximum displacement distance of the pulse wave to be adjusted by the system, which may be a relative signal of the maximum displacement distance in the adjusted pulse wave, or a signal that can generate a value greater than the maximum displacement distance. Reference signal for reference distance. Next, a detailed description of a low-distortion digital power amplifier system implementing one of the above methods will be described below. First, please refer to the fourth figure, which is a circuit block diagram of a preferred embodiment of the low distortion digital power amplifier system of the present invention, and in this embodiment, 'is a low distortion digital power amplifier system based on H-bridge architecture. 3 as an illustrative example, where the system 3 mainly includes a modulation unit 3 1, a correction unit 32, a power switching unit 33, and a processing unit 34; among them, the employee co-operative cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs prints the modulation unit 31, which is connected to the above-mentioned correction unit 32, and is used to convert an input signal (may be an analog signal or a digital signal) into a pulse width modulation (PWM) signal 30, and then input the correction unit 32; and in this embodiment In the modulation unit 31, the PWM signal 30 is divided into two signals CH-R and CH-L, and the modulation method is shown in the fifth figure. The modulation signal is an input signal, such as a pulse-coded modulation signal. PCM, divided into PCM and -PCM (that is, PCM multiplied by -1) two signals, and then after intermodulation with a triangle wave ratio car, you can get two different pwM modulation signals: CH-R and CH-L. These are respectively input into the correction unit 32. The processing unit 34 is connected to the modulation unit 31, the power switching unit 33 and the correction unit, respectively. It includes two subtractors 341, 342, a low-pass filter, a wave filter 3 4 3, and a sample-and-hold circuit. 3 4 4; Of which, the paper size on page 9 applies to the Chinese National Standard (CNS) A4 (210 X 297 mm). Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs. 498663 A7 B7 5. Explanation of the invention (8) Subtraction The subtracter 341 takes two pulse wave modulation signals CH — R and CH — L input from the modulation unit 31 and subtracts them to obtain a comparison signal Vm. • The second subtractor 342 is used for the comparison signal Vm. And a subtraction of a feedback signal Vf output by the power switching unit 33 to obtain an error signal 35, but because the feedback signal Vf is a difference signal amplified by the power switching unit 3 3 'It contains complex noise, so the error signal 35 obtained by subtracting the comparison signal—Vm and the feedback signal Vf—is actually _ A wideband analog signal that contains a lot of noise, so the error signal needs to be 35 into the low-pass chirped wave filter 343 After the signal, and then through the corresponding sample and hold circuit 344 generates the control pulse, to obtain a low frequency component of a control signal Vc of, and fed to the correction unit 32. In addition, it is worth mentioning here that when this digital amplification system is designed to amplify lower frequency signals and low frequency signals are required, the low-pass wideband filter 343 must have a very low cut-off frequency to filter In addition to high-frequency signals, only lower-frequency signals are left (for example, audio, it needs about 22kHz), because if the low-pass signal and the bandwidth of the waveband are very wide, the sampling frequency must be very high. There is a 1 ias (artificial frequency), but in practice, the design cost of the low-pass filter 3 4 3 with dry and low cut-off frequency is relatively high. Therefore, in order to reduce the cost of designing a low-pass filter, in this embodiment, the low-pass filter 343 and the sample-and-hold circuit 344 of the production unit 34 can be a low-pass filter 343 with a higher cut-off frequency. Replaced in series by a supersampling circuit 345 and a decimator 346, as shown in the sixth figure, to form an equivalent low-pass filter circuit with a very low cut-off frequency. The method is to input low Pass filter 343 'The error signal 35 filters out a part of page 10 This paper size applies to China National Standard (CNS) A4 specification (210 X 297 mm) ----- I ---------- II --- Order ------- I · Line (Please read the notes on the back before filling out this page) Printed by the Employees' Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 498663 A7 B7 V. Invention Description (9) High Frequency After the components, they are sent to the oversampling circuit 345, and the error signal 35 · is sampled at, for example, μ times the sampling speed ', and then passed through a decimator 346 at 1 / M times (M = 1, 2, 3._ · Positive integer) by sampling the error signal 35, the control signal Vc with a low frequency component can be obtained, and because the oversampling Path 345 is a discrete-time signal processing system (discrete signal processing) device, which has the ability to compensate for the overall frequency response, it is also desired to control the spectral response of the system. The correction unit 32, as shown in the fourth figure, adjusts the width and phase of each pulse wave of the input pulse width modulation signal 30 according to the control signal Vc ′ input by the processing unit 34 to generate a correction. The signal Ve is input to the power switching unit 33. Please refer to the seventh figure, which is a schematic diagram of the internal structure of the correction unit 32. The correction unit 32 mainly includes two phase shift circuits 321 and two corresponding strings for the input of the pulse width modulation signals CH-R and CH_L, respectively. The pulse width adjustment circuit 322 after the two phase shift circuits 321 and a reference signal generator 324 providing a reference signal Vmax, and the reference signal Vmax can relatively generate a reference correction amount Δmax. Among them, when the aforementioned control signal Vc is input, because its signal is positive or negative, it will first pass a comparison circuit 323 to convert the control signal Vc to a positive value, and then multiply it by a gain value related to the system's correction coefficient K c (Such as 3000), and then multiply it with a basic quantity (2e · 9) to generate an appropriately amplified adjustment signal Vc ', and the adjustment signal Vc' can be relative to the input PWM signal 30 (CH-R and CH-L Each pulse wave of) generates a relative correction amount ΔnCn. Pulse number 1, 2, 3 · ·). Then, the adjustment signal Vc 'is divided into two paths, one is multiplied by 1/2 by a multiplier 325, and then inputted into a subtractor 326 and referenced on page 11. This paper standard applies the China National Standard (CNS) A4 specification (210 X 297 mm) --1 -------- * ---------- Order --------- line 丨 (Please read the notes on the back before filling in this Page) 498663 'A7 B7 V. Description of the invention (10) Subtract the Vmax signal from the test signal to obtain a phase shift control signal Vmax-l / 2Vc. Each pulse wave with relative pulse wave modulation signal 30 generates a relative displacement. △ max-1 / 2 △ n (n = pulse wave number 1, 2, 3, ···), as shown in the eighth figure, each pulse of the pulse wave modulation signal 30 of the input correction unit 32 will be changed. Waves (n = 1, 2, 3 ...) correspond to displacements △ max-1 / 2 △ n (n = 1, 2, 3 ...) 'respectively to obtain the second waveform as shown in the eighth figure; at the same time, The adjustment signal V c '-is input to the pulse wave width adjustment circuit 322 from another path to cause each pulse wave (n = 1, 2, 3 ...) of the PWM signal 30, output by the phase shift circuit 321. The width is relatively adjusted by a correction amount A n (n = 1, 2, 3 ...), and the result is as shown in the eighth figure. Three waveform (i.e., the correction signal V e). To further explain the operation principle of the above-mentioned phase shift and width adjustment, please refer to the ninth figure. Assume that (a) is a pulse width modulation signal input to the correction unit 32 within the period T s. As shown in figure (b), a ramp wave generator will modulate the rising edge and falling edge of the signal according to the pulse width to generate a rising ramp wave with the same angle respectively, thereby 'as shown in (b) and Figure (C), when the control signal Vc of the input correction unit 32 is zero, the rising ramp wave correspondingly generates a △ max phase shift amount, so that the pulse width modulation signal moves backward by a △ max displacement. As shown in Figures (b) and (d), when the control signal Vc of the input correction unit 32 is a positive value, the rising ramp wave of J1 will generate a △ 1 phase shift amount, so that The pulse width modulation signal is delayed backwards by △ max- △ 1 phase shift amount, and at the same time the front and rear edges are pulled outward by the width of Δι; otherwise, as shown in (b) and (e), when the input correction unit When the control signal V c of 3 2 is a negative value, a rising ramp wave correspondingly generates a Δ2 phase shift amount, and The pulse width modulation signal is delayed backwards △ max-one page 12 The paper size is suitable for China National Standard (CNS) A4 (210 X 297 mm) ------------- installation- ------ Order --------- Line (Please read the precautions on the back before filling this page) 498663 A7 B7 V. Description of the invention (11) △ 2 phase shift amount, and make it at the same time The leading and trailing edges are retracted inwardly by a width of Δ2, so that the displacement and width adjustment of the pulse width modulation signal are completed. Therefore, it can be seen from the eighth figure that the input pulse wave modulation signal 3 passes through the phase shift circuit 321 and the width adjustment circuit 322 of the correction unit 32. The phase and width of the pulse wave are appropriately adjusted according to the control signal V c Then, without being disturbed by other noise of the system, there is an equivalent △ between the center position of each pulse wave of the correction signal Ve generated and the center position of each relative pulse wave of the PWM signal 30. The expected phase difference of max, and the expected phase difference (Δmax) does not affect the system's SNR (signal-to-noise ratio) when the factors of the system's immediate correction are not considered. The power switching unit 3 3 amplifies the above-mentioned correction signal Ve to generate an output signal Vo to drive the load, and generates a feedback signal Vf of the input processing unit 34. Please refer to the tenth figure, which is a schematic diagram of the internal circuit of the power switching unit 33. In this embodiment, the correction signal Ve output by the correction unit 32 is divided into R-u, R_d, L-u, and L_d. The four signals are respectively input to the power switching unit 33, and after being appropriately amplified through the switching elements 332, 333, 3 34, and 3 35, respectively, an amplified signal (Vol, Vol, Vo3, and Vo4) is output, and the two sets of differential circuits are passed through 336, 3 37 and a low-pass filter 338 perform differential low-pass filtering to generate an output signal Vo to drive a load 331, and the feedback signal vf is the amplified signal Vo2 and Vo4 are subtracted and divided by one The difference signal obtained by the proportionality constant K (appropriately reducing the voltage value of the amplified signal). Therefore, by inputting the feedback signal Vf into the processing unit 34, the comparison signal Vm is subtracted from the input terminal and a control signal Vc is generated to be returned to page 13. This paper standard applies the Chinese National Standard (CNS) A4 specification (210 X 297) (Please read the notes on the back before filling out this page) Order --------- Line · Printed by the Consumer Consumption Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs 498663 A7 B7 V. Invention Description (12 ) In the correction unit 32, the pulse phase and width of the input PWM signal 30 are adjusted so that the generated correction signal Ve input power switching unit 33 performs signal amplification to minimize the distortion of the output signal Vo. Therefore, through the above, The pulse wave phase and width adjustment action can indeed eliminate the phase noise and internal noise generated by the input signal 30 due to phase shift, so that the digital power amplifier system 3 can output the high-fax amplified signal.-Therefore According to the above-mentioned low-distortion digital power large system 3, one can obtain a low-frequency equivalent linear module as shown in the eleventh figure, where: X is the low level of the PWM signal of the correction unit 32 of the input system 3. Component signal; nl is noise generated by mismatch, non-ideal, approximate assumption and phase asymmetry of internal components of digital power amplifier system 3; n2 is power switching unit 33 of digital power amplifier system 3 Noise caused by factors such as non-linearity, non-idealities, load effects, and ripples on the power supply; _ y is the amplified signal output by the power switching unit 33; and let x, = xe-jwA / 2 s Kc, = Kc e-jwA / 2. And Kp x, + {H (w) Kcx,-[yH (w) Kc,] / K} Kp + z = y where z = nlKp + n2 is determined by The effect module can find STF = y / x | w = 0 = [Kp + H (w) KcKp] e (-jWA / 2) / [l + H (w) Kc, Kp / K] NTF = y / w | x = 0 = i / [i + H (w) Kc5Kp / K] Page 14 This paper size is applicable to China National Standard (CNS) A4 (21〇χ 297 mm) -------- ---- I ------- 1 Order --- ----- I (Please read the notes on the back before filling this page) 498663 A7 B7 V. Description of the invention (13) Among them, (Please (Read the precautions on the back before filling in this page) △ / 2 is the center position of the correction signal and the turn-in pulse width modulation signal in each Ts period The time difference.

Kc is an important reference coefficient when system 3 adjusts the pulse width of the pulse width modulation signal. It is a correction obtained by modifying the signal Ve to reduce the pulse width modulation signal PWM and dividing it by the control signal Vc. The coefficient 'and the derivation of the Kc value are as shown in the example in Fig. 12, where it is assumed that the graph (a) is the control signal Vc' output after the processing unit 34 samples once per second and if only its low frequency components are considered ( For example, the audio frequency is about 22kHz), the control signal Vc can be equivalent to the waveform drawn in Figure (b), and the compound's expression of the waveform can be obtained as: [a · eJ “+ a · e” 01] ri · [Sir ^ TTfrO / 7rf7: i] =? (2a) r! [Sin (7r fr x) / π fr x] ........ (1) The result of the pulse width modulation signal PWM after the correction signal Ve is subtracted from the △ / 2 shift, such as the waveform shown in Figure (c), can be obtained from the waveform. Its compound-leaf expression is: 4C · r 2 [sin (7r fr 2) / 7Γ fz * 2] ............ (2) can be expressed by the above formula (1) (2) Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs

Kc = {4C * r2 [sin (7rfr2) / ^ fr2]} / {(a + b) [siiUTrfrJ / TrfrJ} (because 1 * 2 is proportional to a, and when f is a low frequency, Kc is a constant value. )

Kp is the amplification gain of the power switching unit 33; and K is a scale constant in the power switching unit 33; The STF is a signal conversion function. The higher the STF, the digital power is put on page 15. This paper is applicable to the paper scale. § National Standard (CNS) A4 Specification (21 × 297 mm) 498663 A7

(Large amplifier plus) • The rate of coverage and construction and repair, and the large signal of the output of the 10 system 3 are larger; the NTF is a noise conversion function, the lower the NTF, 〇α, non-digit power generation to suppress noise The better the communication ability; the H (w) is the overall conversion function of the processing unit; therefore '4d can be obtained from the above STF and NTF formulas, + / when the K value increases

At the same time, the STF of the system will increase, but the NTF will also increase. In order to reduce the NTF of the system, you can increase the H (w) Kc, Kp or reduce the number of nl hands. Then, please refer to the thirteenth figure, which is another preferred embodiment of the low-distortion digital power amplification system of the present invention. The digital power amplification system 4 includes a modulation unit 4, a correction unit 42, and a power switch. The unit 43 is a processing unit 44; and the connection relationship between these units and the internal circuit structure are the same as those in the above embodiment, so they will not be repeated here. The difference between the above embodiments of this embodiment is mainly that the comparison signal is generated by the positive unit 42, and in this embodiment, the correction signal Ve output by the correction unit 42 is also divided into R_u, R_d, and L. —U and L — d four signals, the comparison signal Vm is obtained by subtracting R_u and L-u (or) R-d and L-d; therefore, a digital power amplification system 4 can also be obtained as Another equivalent low-frequency linear module shown in the fourth figure, where y = n2 + β Κρ and # = nl + xe, jwA / 2 + [// -y / K] H (w) Kc e.jwA / 2 Let x '= xe · * w △ / 2

Kc ^ Kc e-* wA / 2 can be found on page 16 of this module. The paper size of this module applies to China National Standard (CNS) A4 (210 X 297 mm) --- I! ----- -Install i ---- I 1 order II--II--line (please read the precautions on the back before filling this page) 498663 Printed by the Consumers ’Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention ( 15) STF = y / x '| nl h2 = 0 = Kp / [l- H (w) Kc, + Kp H (w) Kc, / K] N2TF = y / n2 | x' and nl = 0 = [ l- H (w) Kc,] / [1- H (w) Kc, + KpH (w) Kc, / K] NJF = y / nl | x 'and n2 = 0 = Kp / [l- H (w ) Kc, + Kp H (w) Kc, / K] Among them, the definition of Kc, Kp, K, STF and NTF is the same as the above module, and to further explain, the above two feedback modes, which The setting of the K value is related to the Kp value, so a rough estimate of the Kp value is needed. Generally, the approximate value of Kp can be obtained by estimating the DC voltage value of the power supply. In addition, it is worth noting that, since the phase shift circuit 321 and the width delay circuit 322 of the correction unit 32 or 42 use a control signal Vc to control a delay circuit to generate a time delay line (de 1 ay 1 ine), and then The delay line is used to phase-shift and adjust the width of the pulse wave; but because the delay circuit is generally affected by temperature changes, the delay time of the delay line corresponding to the control signal Vc will drift with temperature change, so that The unstable delay time causes distortion of the system output and affects the Kc value of the system. Therefore, 'to improve this problem, a compensation circuit of one degree / m degree can be set in the correction unit 32 or 42', which uses a very high oscillation frequency oscillator to control the delay line corresponding to the control signal Vc at different temperatures. The pulse generated by the delay value (such as the output of the AND gate of the pulse width adjustment circuit 3 22 in the seventh figure) is sampled to measure this pulse width, and after D / A conversion, Obtain the delay line slope and intrinsic delay value C of the control signal Vc at different temperatures (that is, the initial delay value when the control signal Vc is 0). For example, as shown in the fifteenth figure, input the two control signals A and B respectively. Measured from the width of the delay line and converted by D / A, the corresponding page 17 of this paper is applicable to the Chinese National Standard (CNS) A4 specification (21〇X 297 mm) II --------- — # -------- Order --------- line · < Please read the notes on the back before filling this page) 498663 A7 B7 V. Description of the invention (16) See the value dl And d2, you can calculate the delay line slope and c value, and then-the delay line slope and the essential delay value C and EC - comparing the difference obtained from one of the insult recognize Cloth 4 c and the reference value to compensate for slope changes in the delay line, the delay line can be reduced by the degree of influence of temperature. In another implementation method of the 2 delay line, the C value may be zero or non-zero, and when the C value is not zero, relative compensation must be made. Again and again, one point that needs to be discussed here is, please refer to the illustration in Figure 16. When the s control signal Vc is synchronized with the pulse clock of the PWM signal (PWM clock), if the pulse modulation signal is The pulse wave pl is adjusted by the control pulse wave a of the control signal Ve, so that the pulse width after the displacement exceeds the period Ts fe and extends to the period ts of the next pulse wave p2, it will receive the next control pulse wave. The adjustment of b resulted in a false correction signal. Therefore, in order to solve this unsatisfactory situation, the clock of the pulse modulation signal can be slightly delayed to the right, so that the pulse of the control signal is delayed to the right, as shown in the seventeenth figure, the controlled signal Vc can be controlled. Each pulse ^ of the adjusted modified signal Ve is within the range of its period Ts, and the range that can be corrected is increased, and the maximum distance that can be delayed to the right by No. I Vc is the modulation of the pulse width modulation signal One half of the modulation index. For example, in the example in Figure 17, the modulation coefficient of the pulse wave p 1 is 1/2, which means that the pulse width J occupies 1/2 of the entire period TS, then The maximum distance that the control signal Vc can be delayed to the right is 1/4 of the period Ts, which is the three values in the figure. This ensures that each pulse of the modified signal Ve adjusted after the control signal V c is located in its period Ts range. Therefore, when the above-mentioned control signal vc is delayed by 5 distances to the right, 'Kc.e-jwA / 2 ^ in the linear module shown in Figures ^-and 14] Page 18 This paper applies Chinese national standards (CNS) A4 specification (210 X 297 mm) ^ · I I ----- t --- I ----- line (Please read the precautions on the back before filling this page) 498663 A7 B7 Ministry of Economic Affairs The Intellectual Property Bureau employee consumer cooperative printed item 5. Item (17) of the invention description, because Kc A is set to a constant, so it must be changed to Kc w a] ', which is equivalent to shifting the control signal Vc to the right by a δ distance. Moreover, it is worth mentioning here that the present invention can not only be applied to the above-mentioned preferred embodiment of the digital power amplifier system based on H-bridge, it can also be applied to a unilateral digital power amplifier based on ge The main difference between the system and its H-bridge is that the PWM signal modulated by its modulation unit is a single channel (that is, it is not divided into two signals of ch_r and-L, so its correction unit only needs a phase shift circuit. And a pulse width adjustment switch, and the correction signal generated by the correction unit is only divided into two signals input to its power switching unit, so the switching (switching) element of its power switching unit only needs to be provided on one side. And the aforementioned The two linear modules can also be applied to the digital power amplifier of the half_bddge architecture, and its Kc value can also be derived from the same explanation as above. In summary, the low distortion digital power amplification method and low distortion digital power of the present invention The amplification system, through feedback of the control signal Vc to the correction unit 32, appropriately adjusts the phase and width of each pulse wave of the input pulse width modulation signal, which can surely achieve suppression of system noise and In addition to the effects of phase noise caused by uneven phase shift, etc., but the 'mentioned one' is only a preferred embodiment of the present invention, so the equivalent structural changes for the application of the present specification and the scope of patent applications It should be included in the patent scope of the present invention. Brief description of the drawing: The figure is a schematic circuit block diagram of a digital power amplifier. Page 19 This paper is applicable to China's standard specifications (21G x 297 cm) ) ---------------------- Order --------- line ^ 1 ^ · (Please read the notes on the back before filling this page ) 498663 A7 B7 V. Description of the invention (18) The digital signal of the proportional wave power wave is learned from the meridian and the output signal, and the wave is integrated. The figure of the second pulse of the pulse wave is larger than the streamlined method of the French method. The true power loss is low. This figure is shown in the figure. Missing picture, low block, bright side, and the first line of the main line of the electric system. Fill the squares in the metatonic tone. Road maps are shown in Figure 4 of the series. Figure 5th in the 5th schedule and the schedule wide effect is in Yuandan S at the middle of the squares. 〇 Figure six block square road ^ ¾ r ^ Jr Yuan Λ in the mouth of Yuandan is repairing the square block square road | ^ & The elementary signal of the positive wave repair is shown in Figure 4. The block diagram of the square wave width of the radio wave pulse pattern is four o'clock and the whole system is the eighth phase. 3 yuan 〇 Zhengli Xiu originally showed the figure of the 9th wide and the first parrot of the movement system --- I ----------. II (Please read the precautions on the back before filling this page > Order: • Line · Sri Qi, I ’m talking about the members !. Figure 4 of the square circuit diagram of the block diagram of the power consumption in the yuan conversion and cutting rate power consumption within the printing unit. The first path τρβτ part and other I come out, and the square path T ^ nr in the block square response. The system's graphic mode has the same ten-line frequency and low efficiency. The other system has a large power rate. K True picture guide block is pushed down. Yiming Road shows that the power generation display is mainly the second and third examples of the main map. The value of the neutron example of the tenth best figure is better than the figure. Page 20 The Chinese standard is applicable to this paper size. (CNS) A4 specification (210 X 297 mm) H-youuj V. Description of the invention (19) The fourteenth figure corresponds to the circuit block diagram of the tenth = one figure and an equivalent low-frequency linear module is obtained. The fifteenth figure is the digital power amplifier system of the low-generation # a low-distortion digital power amplifier system of the present invention. Please read the notes on the back first and then fill out this page to illustrate the example of the compensation circuit for degree change. The sixteenth figure shows the unsatisfactory situation that occurs when controlling the field service signal Vc to synchronize with the pulse clock (PWM clock). The seventeenth figure shows that when the pulse of the control signal is delayed to the right, the reference number in the figure: When the distance is reached, you can solve the unsatisfactory situation of the + 丄 circle * 丁 _ listen to Fu Le X /. 3, 4 Digital power amplifier system 31, 41 Modulation unit 32, 42 Correction unit 33, 43 Power switching (stage) unit 34, 44 Processing unit 35 Error signal 30, PWM signal 321 Phase shift circuit 322 Pulse width adjustment circuit 323 Comparison circuit 324 Reference signal generator 325, 326 Multiplier 331 Load 3 3 2 ~ 3 3 5 Switching element 336, 337 Differential circuit 338 Low-pass filter 341, 342 Subtractor 343, 343 'Low-pass filter 344 Sample and hold Circuit 345 Oversampling circuit 346 Divider Vm Comparison signal Vf Feedback signal Vc Control signal PWM Pulse width modulation signal Ve Correction signal CH—R, CH _L PWM signal Printed by the Consumer Cooperative of Intellectual Property Bureau of the Ministry of Economic Affairs Page 21 This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) 498663 A7 B7 V. Description of the invention (20) R—u, R_d, L_u, L_d Vol, Vo2, Vo3, Vo4 A, B control signals p 1, p 2 pulse wave Ts pulse wave period correction signal amplification signal dl, d2 width value a, b control pulse wave ------------- install -------- order- -------- Line (Please read the notes on the back before filling this page) Qi Zou Zhihui Property Office employees consumer cooperatives printed on page 22 Paper scale applicable to Chinese National Standard (CNS) A4 size (210 X 297 mm)

Claims (1)

498663 A8B8C8D8 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 6. Patent application scope1. A low-distortion digital power amplification method includes the following steps: Provide a comparison signal and switch the stage based on the comparison signal and by a power The output feedback signal generates a control signal; according to the control signal, the width and phase of each pulse of the input pulse width modulation signal are adjusted correspondingly to generate a correction signal input to the power switching stage, so that the An expected phase difference is generated between the center position of each pulse wave of the correction signal and the center position of each relative pulse wave of the input pulse width modulation signal. 2. The method according to item 1 of the scope of the patent application, wherein a reference signal is provided, which can generate a reference correction amount △ max, and according to the control signal, each of the pulse width modulation signals can be machined. The pulse wave generates a relative correction amount Δ η, so that each pulse wave of the pulse width modulation signal is shifted by A max-A η / 2 distance, and each pulse wave width is adjusted by a Δη width. There is an expected phase difference between the center position of each pulse of the correction signal and the center position of each relative pulse of the pulse width modulation signal, and the expected phase difference is equivalent to the reference correction amount A max . 3. The method according to item 1 of the scope of patent application, wherein the comparison signal is the input pulse width modulation signal. 4. The method according to item 1 of the scope of patent application, wherein the comparison signal is the correction signal. 5. The method according to item 1 of the scope of patent application, wherein the pulse width modulation signal is divided into two signals of CH-R and CH-L, and the comparison signal is composed of the CH-R signal and CH- L signal is subtracted. — Page 23 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) — · -------- 一 —-------- tr ------ -• Line (please read the precautions on the back before filling this page) 498663 A8B8C8D8 6. Application for patents and patents 6. According to No. 1 of the scope of patent application, it is divided into Ru, Rd, and Lu signals are R-u signals Subtracted from L-u. 7. —A kind of low-distortion digital power amplifier, a processing unit, and a method of power description, in which four signal signals of Ld or Rd signal, the correction signal, and the comparison number and Ld signal ------ -------- Installation -------- Order --------- line (please read the precautions on the back before filling this page) In the system: adjust the signal of the early signal and the output signal to adjust the pulse 8 · —type, — Medium: Modulate the modulation signal The processing and correction power are cut into the repair and correction input, the pulse width is 1 Correct the mid-low distortion correction single unit of each pulse wave, and produce ATI as early as 70. The unit is connected to the positive unit of the unit, and the digital position of the center position of the signal input wave is adjusted according to the degree. It includes a modulation unit change unit and a correction unit, the input signal of which is converted into a pulse width and a comparison signal; a feedback signal is input with the modulation unit according to the modulation unit, and a control signal is generated respectively. To the output; A single signal enters the center of the work center and has a comparison between the power output and the power switching output. The control signal of the pulse unit input of the modulation unit is the width rate switching unit of each pulse wave. The expected large phase system includes the wave width modulation number and the corresponding harmonic phase to obtain the phase differences of the correction signal. A modulation unit replacement unit and a processing unit, the modulation unit converts an input signal into a pulse width signal; page 24 This paper uses the Chinese National Standard (CNS) A4 specification (210 X 297) (Mm) 498663 A8B8C8D8 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 6. The scope of patent application This processing unit is connected to the above correction unit and power switching unit respectively to compare the signal and the output of the power switching unit according to one of the output of the correction unit One of the feedback signals generates a control signal to be input into the correction unit; the correction unit is connected to the modulation unit and the power switching unit respectively for the input of the pulse width modulation signal, and according to the input of the processing unit The control signal correspondingly adjusts the width and phase of each pulse wave of the pulse width modulation signal to generate a correction signal input to the power switching unit, so that the center position and pulse width of each pulse wave of the correction signal There is an expected phase difference between the center positions of the relative pulses of the modulation signal. 9. The low-distortion digital power amplifier system according to item 7 or 8 of the scope of the patent application, wherein the correction unit provides a reference signal, and the reference signal can relatively generate a reference correction amount Δmax, and the control signal can be A relative correction amount Δ η is generated with respect to each pulse wave of the pulse width modulation signal, so that each pulse wave of the pulse width modulation signal is displaced by a distance Δ max- Δ η / 2, and each After the pulse width is adjusted by a Δ η width, there is an expected phase difference between the center position of each pulse wave of the generated correction signal and the center position of each relative pulse wave of the pulse width modulation signal, and the The expected phase difference is equivalent to the reference correction amount Amax. I 0. The low-distortion digital power amplifier system according to item 7 of the scope of the patent application, wherein the comparison signal is the pulse width modulation signal. II · Digital Power Amplification System with Low Distortion as described in item 7 of the scope of patent application Page 25 I —ί — — — — — — ΙΓΙ —Aw — — — — — — — — — — — — — — — — — I — I ^ w— (Please read the precautions on the back before filling out this page) This button is applicable to Zhongguanjia Standard (CNS) A4 specifications ⑽ X 297 ~ 498663 A8B8C8D8 6. Application for patent scope (Please read the precautions on the back before (Fill in this page) system, where the pulse width modulation signal is divided into two signals: CH-R and CH-L, and the processing unit includes a first subtractor, a second subtractor, and a low-pass filter. And a sample-and-hold circuit, wherein 'the first subtractor is used for subtracting the above-mentioned CH-R and CH-L signals' to obtain the comparison signal, and the second subtractor is used for the comparison signal and the power The feedback signal of the switching unit is subtracted to obtain an error signal, and after the error signal is processed by the low-pass filter and the φ sample and hold circuit, the control signal is obtained. 12. The low-distortion digital power amplifier system according to item 8 of the scope of the patent application, wherein the comparison signal is the correction signal. Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs 1 3 · According to the low-distortion digital power amplifier system described in item 8 of the scope of patent application, the correction signal is divided into four signals: Ru, Rd, Lu, and Ld, and the The processing unit includes a first subtractor, a second subtractor, a low-pass filter, and a sample-and-hold circuit. The first subtractor is used for the Ru signal and the Lu signal (or Rd and Ld signals). Subtract to obtain the comparison signal, and the _ second subtractor is used to subtract the comparison signal and the feedback signal of the power switching unit to obtain an error signal and pass the error signal through the low pass After processing by the filter and the sample-and-hold circuit, the control signal is obtained. 14. The low-distortion digital power amplifier system according to item 7 or 8 of the scope of the patent application, wherein the processing unit includes a first subtractor, a second subtractor, a low-pass filter, and a sample-and-hold circuit. And the control signal is the comparison signal and the feedback signal through the second subtractor-page 26. The paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) 498663 A8B8C8D8 A signal generated after subtraction is processed by the low-pass filter and the sample-and-hold circuit. 15 · According to the low-distortion digital power system described in item 14 of the scope of the patent application, the low-pass filter and sample-and-hold circuit can be combined with a low-pass filter with a higher cut-off frequency and an over-sampling circuit. And a decimator. 16. According to the low-distortion digital power amplification system described in item 7 or 8 of the scope of the patent application, a temperature compensation circuit is further provided in the correction unit, which can respond to different temperature changes and correspondingly compensate for the temperature change inside the correction unit. The phenomenon of unstable time delay occurs. (Please read the precautions on the back before filling this page) tr ------—— line! Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs Page 27 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) -I 1 I n an n 1 · <