CN100486111C - Class d amplifier - Google Patents

Class d amplifier Download PDF

Info

Publication number
CN100486111C
CN100486111C CNB2005101317185A CN200510131718A CN100486111C CN 100486111 C CN100486111 C CN 100486111C CN B2005101317185 A CNB2005101317185 A CN B2005101317185A CN 200510131718 A CN200510131718 A CN 200510131718A CN 100486111 C CN100486111 C CN 100486111C
Authority
CN
China
Prior art keywords
signal
waveform
pulse
output
distortion
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CNB2005101317185A
Other languages
Chinese (zh)
Other versions
CN1790898A (en
Inventor
田浦贤一
辻雅之
石田雅之
仲田刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Publication of CN1790898A publication Critical patent/CN1790898A/en
Application granted granted Critical
Publication of CN100486111C publication Critical patent/CN100486111C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Abstract

A pulse modulated signal (ei) output from a pulse modulator ( 1 ) and a feedback signal (ef) containing distortion caused by a power switch ( 3 ) are integrated in a first integrator ( 21 ) and a second integrator ( 24 ), respectively, and are input to input terminals of a comparator ( 25 or 29 ), respectively, so that a correction signal (Vc) is generated.

Description

The D class A amplifier A
The application be that September 30, application number in 2003 are 03164888.6 the applying date, denomination of invention divides an application for the application of " D class A amplifier A ".
Technical field
The present invention relates to a kind of D class A amplifier A, and relate in particular to a kind of D class A amplifier A with correcting circuit.
Background technology
The D class is amplified usually by by the audio signal of high frequency, low-power consumption is carried out power amplification so that the method for equipment miniaturization is sampled.In a kind of well-known D class A amplifier A, a digital audio signal directly is converted to pulse width modulating signal and is offered mains switch.This mains switch generally comprises the switching device shifter of a switching device shifter that links to each other with constant voltage source and a ground connection (or cathode power).
Also well-known a kind of method is modulated the rounding error that quantizer caused again that reduces by needing PWM (pulse width modulation) conversion by delta sigma exactly, and this can obtain High-Accuracy PWM (pulse width modulation) signal.This pwm signal is from having the output of high-precision mains switch, therefore can be from the D class A amplifier A outputting high quality audio signal (and JP2001-292040) referring to JP11-261347 (1999).
, in fact above-mentioned method can cause such problem: use the desirable constant voltage source that is the mains switch power supply to be difficult in general realize that zero expends, and the consumed power of constant voltage source can increase, this will cause the forfeiture of the intrinsic advantage of D class A amplifier A.Consider these problems, though be not all, adopted the mode of compromise under a variety of situations, this mode only suppresses the alternating compenent of audio frequency, but this also can be owing to the LC filter becomes problem.
Further, according to above-mentioned method, each power transfer device that constitutes mains switch all has one and is used to the finite delay time of opening and shutting.Therefore, open for one in the switching device shifter of very difficult usually switching device shifter that will link to each other and ground connection, and at one time another is shut with constant voltage source.Like this, just need be provided with one therein device almost by the dead time between another device is opened after being shut fully.
Power-supply fluctuation meeting above-mentioned directly occurs with the form of the amplitude fluctuations of mains switch output signal, and this will cause the distortion of amplifier output audio signal.
Further, because the distortion that the distortion of the mains switch output signal that is caused also can cause the amplifier output audio signal is set dead time.
As solution to the problems described above, well-known is exactly corrective system (referring to national translation publication 2001-51739, for example Fig. 3-8) according to common technology. and this technology commonly used will be below describes with reference to the accompanying drawing of display structure.
Figure 32 is the block diagram with D class A amplifier A of conventional corrective system.
In Figure 32, pulse modulator 100, correcting unit 102, mains switch 103 and LPF (low pass filter) 104 are cascaded mutually.Error processor 101 and correcting unit 102 and mains switch 103 parallel being connected between node N100 and the N101, and its output is linked to each other with correcting unit 102.
Have in the D class A amplifier A of the corrective system that disposes as shown in the figure at this, pulse modulator 100 generates a binary pulse modulation signal Vr by an audio signal is modulated.
This mains switch 103 switches between constant voltage source and ground by the value of foundation correction signal Vc and carries out power amplification, wherein this correction signal Vc is a binary pulse signal by correcting unit 102 transmission, and makes power supply to the electric that links to each other with amplifier output.Here, this mains switch 103 comprises a factor (being called distortion factor here) that causes the audio signal distortion, for example is provided with the dead time of power-supply fluctuation and switching device shifter operation.
This error processor 101 detects the distortion of mains switch 103 output signals, and especially detect error among the feedback signal Vs of mains switch 103 outputs, thereby generate and export an error signal Ve corresponding to this error with reference to the pulse-modulated signal Vr of pulse modulator 100 output.
This correcting unit 102 changes its width by the error signal Ve according to error processor 101 output to be proofreaied and correct from the pulse-modulated signal Vr of pulse modulator 100 inputs, controls thus so that reduce error signal Ve from error processor 101.
The internal structure of this correcting unit 102 will be described below.
Figure 33 is the internal structure block diagram of correcting unit 102. and in Figure 33, "-" of integrator 200, amplitude limiter 201 and comparator 202 end is cascaded. and "+" of this comparator 202 end links to each other with the output of error processor 101 and its output links to each other with the input of mains switch 103.The input of this integrator 200 links to each other with the output of pulse modulator 100.
With reference to Figure 34 the operation of various piece in the correcting unit 102 is described below, wherein shows the signal waveform of correcting unit 102 at each point.
In Figure 34, reference marker 210 expressions input to the waveform of the pulse-modulated signal Vr of integrator 200,211 expressions input to the trapezoidal wave of the input signal Vi of comparator 202 "-" end, and it is to obtain from the pulse-modulated signal that is converted in by integrator 200 and amplitude limiter 201.By the effect of integrator 200, the trailing edge of trapezoidal wave 211 and rising edge tilt to have a certain degree.Effect by amplitude limiter 201 with the amplitude limitation of trapezoidal wave 211 within certain scope.
Reference marker 212 and 213 all represents from error processor 101 output and inputs to the waveform of the error signal Ve of comparator 202 "+" end, and 214 and 215 all represent to generate in comparator 202 by input signal Vi and the error signal Ve that inputs to comparator 202 are compared and the waveform of the correction signal Vc of output therefrom.
Here, this waveform 212 and 213 is to obtain from the error signal Ve that has different value each other.Waveform 214 is to obtain among the correction signal Vc that generates from comparator 202 according to waveform 212, and waveform 215 is to obtain among the correction signal Vc that generates from comparator 202 according to waveform 213.
As can be seen from Figure 34, when error signal Ve has high potential (under the situation of waveform 212), comparator 202 in the correcting unit 102 can generate the correction signal Vc with broad pulse width (just waveform 214), and it is opposite, when error signal Ve has electronegative potential (under the situation of waveform 213), comparator 202 can generate the correction signal Vc with narrow pulse width (just waveform 215).
Therefore, from as the pulse-modulated signal Vr of pulse modulator 100 input of benchmark and from the feedback signal Vs of mains switch 103 inputs in the process of generated error signal Ve, this error processor 101 is set to make it to generate an error signal Ve who reduces as waveform 213 on current potential under more than or equal to the situation as the pulse duration of the pulse-modulated signal Vr of benchmark in the pulse duration of the feedback signal Vs that contains error, and generates an error signal Ve who raises as waveform 212 on current potential under the pulse duration of the feedback signal Vs that contains error is less than or equal to situation as the pulse duration of the pulse-modulated signal Vr of benchmark.
The use of D class A amplifier A with corrective system of said structure can reduce the error with respect to the pulse-modulated signal Vr that is used as benchmark from the feedback signal Vs of mains switch 103 outputs automatically.
Like this, just can automatically be corrected owing to the distorted signals that causes is set the dead time in fluctuation in the supply voltage and the mains switch 103, this has just prevented the appearance of the distortion phenomenon of amplifier output audio signal.
, the problem below can occurring by the above-mentioned corrective system that feedback in the disclosed D class A amplifier A is proofreaied and correct in national translation publication 2001-51739.
At first, in order to improve the effect of correction, need the signal Vi that will input to comparator 202 "-" end to be converted to high-precision trapezoidal wave signal., the shortcoming that generates a high accuracy trapezoidal wave signal is exactly the circuit structure that needs a complexity of comparing with the circuit shown in Figure 33.
Secondly, pulse-modulated signal Vr and the feedback signal Vs that inputs to error processor 101 are pulse signals. be difficult to generated error signal Ve from this pulse signal usually, and remaining pulse can't be removed thoroughly also in error signal Ve.These remaining pulse meetings cause being difficult to obtain sufficient calibration result.
Under the situation that the residual impulse part is thoroughly removed, the operation of this circuit is restricted.That is to say that when segment pulse in the nonlinear area of correcting unit 102 distortion took place, error signal Ve also distortion can take place, this just can't carry out suitable correction.Therefore, optimal situation is exactly not comprise segment pulse among the error signal Ve that generates in error processor 101, and this part can reflect the difference between pulse-modulated signal Vr and the feedback signal Vs low frequency part.
In fact, the rotation of the phase place of feedback signal Vs can't be stablized cycling in the error processor 101, and this is difficult to error processor 101 is filtered with regard to making, therefore makes segment pulse be attenuated.On the other hand,, also need error signal Ve is fully amplified and proofreaies and correct in order to obtain sufficient feedback effects, contradiction be that this can amplify remaining segment pulse simultaneously.
In front under the condition of being mentioned, owing to exist residual impulse therefore to be difficult to obtain sufficient calibration result (reducing the distortion of audio signal).
For above-mentioned known structure, can be by pwm signal being reflected the high-quality audio signal of output to obtain the High-Accuracy PWM signal and to export as amplifier to the high-precision power switch.
, offer that the fluctuation meeting of supply voltage of mains switch is very disadvantageous to cause output signal generation distortion.If offer mains switch with what a certain magnitude of voltage continued by a constant voltage circuit, then the distortion of output signal will reduce,, this mains switch will the sizable power of loss, and be used for providing the power loss of the constant voltage circuit of a certain magnitude of voltage also can increase to mains switch, this another problem that will cause is exactly to carry out power amplification to audio signal efficient, low-power consumption.
Summary of the invention
One object of the present invention just provides a kind of D class A amplifier A with simple circuit structure, this amplifier can reduce the distortion of high accuracy audio signal, just, a kind of class A amplifier A of D efficiently, the distortion that its amplifier with routine is compared the output signal that the fluctuation owing to the supply voltage that offers mains switch causes can reduce greatly, even and supply voltage in a sizable scope, change and also any problem can not occur.
According to a first aspect of the invention, provide a kind of D class A amplifier A, comprise
A pulse modulator is with generating a pulse-modulated signal;
A correcting circuit is used for reference to described pulse-modulated signal to proofreading and correct by the feedback signal of feedback input; And
A mains switch is used for generating a voltage signal according to the correction signal from described correcting circuit output, wherein
Described feedback signal generates according to described voltage signal, and
Described correcting circuit comprises:
A first integrator is used for carrying out integration according to described pulse-modulated signal;
A second integral device is used for carrying out integration according to described feedback signal; And
A comparator is used for first integral signal of exporting from described first integrator and the second integral signal of exporting from described second integral device are compared, and generates described correction signal according to comparative result.
This pulse-modulated signal and feedback signal are converted into the integrated signal with the low frequency part that is reinforced respectively in first and second integrators.This comparator compares integrated signal, generates correction signal and it is exported to correcting circuit as feedback signal.Therefore, the distortion that appears at the audio signal in the mains switch can be proofreaied and correct by a pulse signal that is removed (low frequency part that just is reinforced). and this can prevent that circuit operation is subjected to the restriction of residual impulse part, just, can avoid this to enter the correcting circuit nonlinear area and error signal distortion that the residual impulse part of distortion caused takes place and the interference of correct correction.And, also can in correcting circuit, directly generate correction signal according to the error between pulse-modulated signal and the feedback signal, this can make circuit structure simplify on the whole.
According to a second aspect of the invention, provide a kind of D class A amplifier A, comprising:
A pulse modulator is with generating a pulse-modulated signal;
A correcting circuit is used for reference to described pulse-modulated signal to proofreading and correct by the feedback signal of feedback input; And
A mains switch is used for generating a voltage signal according to the correction signal from described correcting circuit output, wherein
Described feedback signal generates according to described voltage signal, and
Described correcting circuit comprises:
A first integrator is used for carrying out integration according to described pulse-modulated signal;
A second integral device is used for carrying out integration according to described feedback signal;
One first subtracter is used for obtaining from the first integral signal of described first integrator output and from the difference between the second integral signal of described second integral device output;
A third integral device is used for first differential signal of exporting from described first subtracter is carried out integration;
A reverser is used for the third integral signal of exporting from described third integral device is carried out oppositely; And
A comparator is used for comparing to described first differential signal with by the reverse described third integral signal of described reverser, and generates described correction signal according to comparative result.
The low frequency part of first differential signal (distortion of audio signal just) can also be strengthened in the third integral device and then be undertaken oppositely by reverser, therefore this comparator can provide first differential signal part that distortion caused by audio signal, and generates the correction signal that the distortion of audio signal is further strengthened.This D class A amplifier A with first aspect is just compared and can be carried out more high-precision correction.
According to a third aspect of the invention we, provide a kind of D class A amplifier A, comprising:
A mains switch, being used for the response pulse duration modulation signal carries out on/off switch to the power supply that supply voltage is provided;
A correcting circuit is used for proofreading and correct being transfused to the pulse duration of the pulse width modulating signal of described mains switch according to the amplitude of the feedback signal that generates from the output of described mains switch; And
An arithmetic element is used for regulating being transfused to the amplitude of the feedback signal of described correcting circuit according to described supply voltage value.
The efficient of this D class A amplifier A is very high, the distortion of the output signal that fluctuation caused that wherein offers the supply voltage of mains switch is compared with the D class A amplifier A of routine and is greatly reduced, even and supply voltage fluctuates in a quite wide scope, the output level of the audio signal when not having distortion phenomenon to occur also can reduce hardly.
According to a forth aspect of the invention, provide a kind of D class A amplifier A, comprising:
A mains switch is used for according to pulse width modulating signal the power supply that supply voltage is provided being carried out on/off switch; And
A correcting circuit is used for proofreading and correct being transfused to the pulse duration of the described pulse width modulating signal of described mains switch according to the amplitude of the feedback signal that generates from the output of described mains switch, wherein
Wherein said correcting circuit comprises:
A first integrator is used for described pulse width modulating signal is carried out integration;
A second integral device, the difference between the reference voltage that is used for generating to described feedback signal with according to the dc component of described supply voltage is carried out integration; And
A comparator is used for the output of more described first and second integrators, and
The output of described comparator is transfused to described mains switch.
The efficient of this D class A amplifier A is very high, the distortion of the output signal that fluctuation caused that wherein offers the supply voltage of mains switch is compared with the D class A amplifier A of routine and is greatly reduced, even and supply voltage fluctuates in a quite wide scope, the output level of the audio signal when not having distortion phenomenon to occur also can reduce hardly.
According to a fifth aspect of the invention, provide a kind of D class A amplifier A, comprising:
A mains switch is used for according to pulse width modulating signal the power supply that supply voltage is provided being carried out on/off switch;
A correcting circuit is used for proofreading and correct being transfused to the pulse duration of the described pulse width modulating signal of mains switch according to the amplitude of described mains switch output signal;
A level reference signal generator is used for generating a level reference signal from described supply voltage; And
A level adjustment circuit is used for regulating being transfused to the amplitude of the described pulse width modulating signal of described correcting circuit according to described level reference signal value.
The efficient of this D class A amplifier A is very high, the distortion of the output signal that fluctuation caused that wherein offers the supply voltage of mains switch is compared with the D class A amplifier A of routine and is greatly reduced, even and supply voltage fluctuates in a quite wide scope, the use that it also can be without a doubt.
According to a sixth aspect of the invention, provide a kind of D class A amplifier A, comprising:
A pulse modulator is used for the pulse duration of input signal is modulated to export a pulse width modulating signal;
A mains switch is used for according to described pulse width modulating signal the power supply that supply voltage is provided being carried out on/off switch;
A correcting circuit is used for proofreading and correct being transfused to the pulse duration of the described pulse width modulating signal of described mains switch according to the amplitude of described mains switch output signal;
A level reference signal generator is used for generating a level reference signal from described supply voltage;
A modulation index control-signals generator is used for generating a modulation index control signal from described supply voltage;
A level adjustment circuit is used for regulating being transfused to the amplitude of the described pulse width modulating signal of described correcting circuit according to described level reference signal value; And
A modulation index regulating circuit is used for according to described modulation index control signal value the modulation index in the described pulse modulator being regulated.
The efficient of this D class A amplifier A is very high, the distortion of the output signal that fluctuation caused that wherein offers the supply voltage of mains switch is compared with the D class A amplifier A of routine and is greatly reduced, even and supply voltage fluctuates in a quite wide scope, the use that it also can be without a doubt.。
These and other purpose of the present invention, feature, aspect and advantage will be along with below with reference to becoming more obvious as the figure detailed description of the present invention.
Description of drawings
Fig. 1 is the circuit structure block diagram according to D class A amplifier A of the present invention;
Fig. 2 is the block diagram of correcting circuit internal structure according to a first advantageous embodiment of the invention;
Fig. 3 is when not having distortion to take place, the steady-state signal waveform of each point in the correcting circuit according to a first advantageous embodiment of the invention;
Fig. 4 A-4B shows first type distortion;
Fig. 5 A-5B shows second type distortion;
Fig. 6 A-6B shows the distortion of the third type;
Fig. 7 A-7B shows the 4th type distortion;
Fig. 8 is when the distortion that occurs first type, according to the steady-state signal waveform of each point in the correcting circuit of first preferred embodiment;
Fig. 9 is when the distortion that occurs second type, according to the steady-state signal waveform of each point in the correcting circuit of first preferred embodiment;
Figure 10 is for when the distortion that the third type occurs, according to the steady-state signal waveform of each point in the correcting circuit of first preferred embodiment;
Figure 11 is when the distortion that occurs the 4th type, according to the steady-state signal waveform of each point in the correcting circuit of first preferred embodiment;
Figure 12 is the particular circuit configurations figure according to the correcting circuit of first preferred embodiment;
Figure 13 is the block diagram according to the correcting circuit of second preferred embodiment of the invention;
Figure 14 is when not having distortion phenomenon to take place, the steady-state signal waveform of each point in the correcting circuit according to a second, preferred embodiment of the present invention;
Figure 15 is when the distortion that occurs first type, according to the steady-state signal waveform of each point in the correcting circuit of second preferred embodiment;
Figure 16 is when the distortion that occurs second type, according to the steady-state signal waveform of each point in the correcting circuit of second preferred embodiment;
Figure 17 is for when the distortion that the third type occurs, according to the steady-state signal waveform of each point in the correcting circuit of second preferred embodiment;
Figure 18 is when the distortion that occurs the 4th type, according to the steady-state signal waveform of each point in the correcting circuit of second preferred embodiment;
Figure 19 is the particular circuit configurations figure according to the correcting circuit of second preferred embodiment;
Figure 20 is the measuring circuit structure chart that is used for estimating the calibration result of being realized by correcting circuit according to the present invention;
Figure 21 has drawn the curve chart of indication by the measurement data of the calibration result of correcting circuit realization according to the present invention;
Figure 22 is the block diagram according to the D class A amplifier A of third preferred embodiment of the invention;
Figure 23 is the output waveform according to the D class A amplifier A of third preferred embodiment of the invention;
Figure 24 is the block diagram according to the D class A amplifier A of four preferred embodiment of the invention;
Figure 25 is the circuit diagram according to the correcting circuit of the D class A amplifier A of the 4th preferred embodiment;
Figure 26 is the block diagram according to the D class A amplifier A of fifth preferred embodiment of the invention;
Figure 27 is according to the signal waveform in the various piece of the correcting circuit of the D class A amplifier A of the 5th preferred embodiment;
Figure 28 is the cut-away view according to the level adjustment circuit of the D class A amplifier A of the 5th preferred embodiment;
Figure 29 is the block diagram according to the D class A amplifier A of sixth preferred embodiment of the invention;
Figure 30 is the cut-away view according to the level reference signal generator of the D class A amplifier A of the 7th preferred embodiment;
Figure 31 is the curve chart of drafting according to the input-output characteristic of the level reference signal generator of the D class A amplifier A of the 7th preferred embodiment;
Figure 32 is the block diagram according to the D class A amplifier A of routine techniques;
Figure 33 is the internal structure block diagram according to the correcting unit of routine techniques;
Figure 34 is the correction schematic diagram according to routine techniques.
Embodiment
First preferred embodiment
Being included in can be to carrying out integration from the pulse-modulated signal of pulse modulator output and from the feedback signal of mains switch output according to the correcting circuit in the D class A amplifier A of this embodiment, adopting pulse-modulated signal in comparator is that benchmark compares integrated signal, be used for to output signal of mains switch output corresponding to comparative result, and the distorted signals that occurs in the last calibration power supply switch.
Fig. 1 is according to the circuit structure block diagram that contains the D class A amplifier A of correcting circuit of the present invention.
D class A amplifier A shown in Fig. 1 comprises 2, one mains switches 3 of 1, one correcting circuit of a pulse modulator and a low pass filter (LPF) 4, and they are series connection mutually each other.Signal is sent to LPF4 from pulse modulator 1.A feedback circuit 5 is provided between node N1 and correcting circuit 2, and this just makes that the part of output signal is fed to correcting circuit 2 from mains switch 3.
This correcting circuit 2 comprises two inputs, and one of them is from pulse modulator 1 received signal, and another is receiving feedback signals from feedback circuit 5.
In the structure of mentioning in front, the operation of each parts will be described below.
This pulse modulator 1 is converted to pulse-modulated signal with numeral or simulated audio signal, for example binary pulse bandwidth modulation signals or binary pulse digital modulation signals, and export this pulse-modulated signal.
This correcting circuit 2 receives this pulse-modulated signal and return next feedback signal from feedback circuit 5, and in mains switch 3, use pulse-modulated signal the distortion factor that is included in the feedback signal to be proofreaied and correct, and export the pulse-modulated signal after this correction as benchmark.
This mains switch 3 is made of the switching device shifter of a switching device shifter that links to each other with power supply and ground connection (or link to each other with cathode power), carry out power amplification by between power supply and ground connection, switching, make that power supply can be to an electric that links to each other with amplifier output according to the pulse-modulated signal that is used for the correcting distortion factor from correcting circuit 2 outputs.
This LPF4 is from from removing HFS in the power amplification signal the mains switch 3, thereby exports the audio signal of a demodulation.
At last, the amplitude of the power amplification signal of 5 pairs of mains switches of this feedback circuit, 3 outputs is carried out level adjustment, and the signal after the level adjustment is offered another input of correcting circuit 2.When the signal amplitude in the mains switch 3 distortion do not occur, the level adjustment of signal amplitude above-mentioned will make with pulse modulator 1 in the amplitude of pulse-modulated signal of output have the fixing decay gain of amplitude attenuation one of power amplification signal in the mains switch 3 of same level.
Fig. 2 is the block diagram according to the correcting circuit internal structure that is used to proofread and correct in the D class A amplifier A of the present invention.
To the structure of the correcting circuit shown in Fig. 22 be described below.
The output of subtracter 20 links to each other with the input of first integrator 21, an input of the output of first integrator 21 and comparator 25, and just "+" input links to each other.
The output of subtracter 23 links to each other with the input of second integral device 24, the other end of the output of second integral device 24 and comparator 25, and just "-" end links to each other.
And the output of first integrator 21 also links to each other with the input of gain controller 22 with node N2 between the comparator 25 by first integrator 21. the output of this gain controller 22 is also continuous with "-" end of subtracter 20,23 respectively by node N3.
"+" end of subtracter 20 links to each other with the output of pulse modulator 1, and "+" end of subtracter 23 links to each other with the output of feedback circuit 5.The output of comparator 25 links to each other with the input of mains switch 3.
In the structure of mentioning in the above, subtracter 20, first integrator 21 and gain controller 22 have constituted one provides degenerative integrating circuit by gain controller 22. in this integrating circuit, this subtracter 20 has obtained difference between pulse-modulated signal of exporting and the signal of exporting from gain controller 22 from pulse modulator 1, thereby generate first differential signal, and carry out integration by 21 pairs of these first differential signals of first integrator.Just, first integrator 21 is carried out integration according to pulse-modulated signal to first differential signal, in order to strengthen the low frequency part in the pulse-modulated signal, suitably restrain low-frequency gain by the negative feedback of gain controller 22 simultaneously, this just can prevent to surpass from the integrated signal in the first integrator 21 working range of circuit.
Further, the integrating circuit that is made of subtracter 23 and second integral device 24 also deducts the output signal of gain controller 22 from the feedback signal of feedback circuit 5, thereby generates one second differential signal, and this second differential signal is carried out integration.That is to say, second integral device 24 carries out integration according to feedback signal to second differential signal, in order to strengthen the low frequency part in the feedback signal, subtracter 23 also deducts the output signal of gain controller 22 from the feedback signal of feedback circuit 5 simultaneously, thereby the minimizing low frequency part, this just can prevent that the integrated signal of second integral device 24 from surpassing the working range of circuit.
The waveform of 25 pairs of integrated signals of output from first integrator 21 and second integral device 24 of this comparator compares, and exports a comparative result as correction signal Vc, and this signal is a binary pulse signal.
Hereinafter, the operation of using arithmetic expression to comparator 25 is described.
At first, with the pulse-modulated signal of ei indicating impulse modulator 1 output, represent the integrated signal of output from first integrator 21 with eo1, this integrated signal eo1 can be expressed as:
Wherein G1 is the gain constant of first integrator 21, and Gf is that the gain of gain controller 22 is normal
eo1＝G1∫(ei-Gf·eo1)dt
＝G1∫ei·dt-Gf·G1∫eo1·dt ...(1)
Number.
Further, represent the feedback signal that feedback circuit 5 is exported with ef, represent the integrated signal that second integral device 24 is exported with eo2, this integrated signal eo2 can be expressed as:
eo2＝G2∫(ef-Gf·eo1)dt
＝G2∫ef·dt-Gf·G2∫eo1·dt ...(2)
Wherein G2 is the gain constant of second integral device 24.
In expression formula (1) and (2), the low frequency part of integrated signal eo1 and eo2 has all suitably been reduced by second (the comprising gain constant Gf's) on right side.
According to the output calibration signal Vc of numerical value (eo1-eo2) generation comparator 25, as binary pulse " H " or " L " (just " 0 " or " 1 ").
For example, (eo1-eo2) is timing when numerical value, and output calibration signal Vc " H " is as binary pulse signal from comparator 25.Then, this correction signal Vc " H " makes the signal of a same phase be fed by mains switch 3 and feedback circuit 5. like this, feedback signal ef just is changed to " H " level and increases the value of integrated signal eo2.Therefore, numerical value (eo1-eo2) will reduce and restrain to being similar to zero with stable state.
On the other hand, when numerical value (eo1-eo2) when negative, output calibration signal Vc " L " is as binary pulse signal from comparator 25.Then, this correction signal Vc " L " makes the signal of a same phase be fed by mains switch 3 and feedback circuit 5.Like this, feedback signal ef just is changed to " L " level and reduces the value of integrated signal eo2.Therefore, numerical value (eo1-eo2) will increase and restrain to being similar to zero with stable state.
That is to say the following expression that from expression formula (1) and (2), obtains:
G1∫ei·dt-G2∫ef·dt+Gf·(G2-G1)∫eo1·dt ...(3)
Be similar to zero.
Here, if constant G1 that needn't be equal to each other and G2 approximately equal and be expressed as constant G all, expression formula (3) can be by approximate being expressed as:
G(∫ei?dt-∫efdt)＝0 ...(4)
This shows that the correcting circuit 2 with said structure is mainly used in makes the low frequency part of input signal ei and feedback signal ef equate.Like this, just correction signal Vc can be generated from comparator 25, in order to the difference between the low frequency part that reduces pulse-modulated signal ei and feedback signal ef, the just distortion in the audio signal.
As mentioned above, can be by proofreading and correct the distortion that reduces in the audio signal.From top description as can be seen, even constant G1 has different values each other with G2, distortion also can be reduced in the scope of Gf (G2-G1) ∫ eo1 dt.,, then just can reduce distortion (correcting distortion) with higher precision if constant G1 has identical value with G2.
Next, will the signal waveform state of each point in the correcting circuit 2 be described.
At first, will the situation that not have distortion phenomenon to occur in the mains switch 3 be described.Fig. 3 just shows the steady-state signal waveform of each point in the correcting circuit 2 in this case.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Fig. 3, the integrated waveform of the integrated signal eo1 that reference marker 30 expression output and be input to the impulse waveform of the pulse-modulated signal ei of correcting circuit 2 from pulse modulator 1,31 expressions generate in first integrator 21 according to waveform 30.
Reference marker 34 expression output and be transfused to impulse waveform from feedback circuit 5, and the integrated waveform of the integrated signal eo2 that in second integral device 24, generates according to waveform 34 of 32 expressions to the feedback signal ef of correcting circuit 2.
The impulse waveform of the correction signal Vc that reference marker 33 expression generates in comparator 25 according to the difference of integrated waveform 31 and 32 is as binary pulse " H " or " L " (just " 0 " or " 1 ").Particularly, when waveform 31 is higher than waveform 32, production burst " H " (just " 1 "), and when waveform 31 is lower than waveform 32, production burst " L " (just " 0 ").
The amplitude range of impulse waveform 30 should be similar to 0 and Vsig between.If from the voltage of constant voltage source supply power switch 3 is that fixed attenuation gain Vpow and the feedback circuit 5 is 1/K, then the amplitude of the impulse waveform of feedback circuit 5 output should be similar to 0 and Vpow/K between, and feedback circuit 5 is arranged so that the amplitude of impulse waveform 34 equals the amplitude (Vpow/K=Vsig just) of the impulse waveform 30 of output from pulse modulator 1.
Be in as shown in Figure 3 state in impulse waveform 30 and 34, and under the situation of on the basis that acts on approximate Vsig/2 of gain controller 22 etc. first integrator 21 and second integral device 24 being operated, will generate integrated waveform 31 and 32 as shown in Figure 3, and the correction signal Vc of comparator 25 output will become and is similar to impulse waveform 33.Here, under stable state, impulse waveform 34 has δ time of delay that is mainly caused by mains switch 3 with respect to the impulse waveform 33 of comparator 25 outputs.
That is to say, as shown in Figure 3, when correcting circuit 2 is usually operated at can not cause mains switch 3 stable state of distortion to occur the time, the impulse waveform 34 of feedback signal ef just becomes the waveform that similarly has δ time of delay with the impulse waveform 30 of pulse-modulated signal ei, and this is equal to each other with regard to the low frequency part that means impulse waveform 30 and 34 and audio signal is not normally transmitted distortion can be taken place.
In fact comprise the wave distortion that mainly causes among the feedback signal ef by mains switch 3.This distortion makes the waveform generation deformation of feedback signal ef, and this just makes the low frequency part of feedback signal ef with the low frequency part difference of pulse-modulated signal ei.
Therefore, now just causing by mains switch 3 under the situation of the wave distortion of four kinds of patterns shown in Fig. 4 A-7B, the waveform in the correcting circuit 2 that is in stable state is being described in the variation that each point takes place according to the general correct operation of the correcting circuit 2 of the D class A amplifier A of present embodiment.In Fig. 4 A-7B, the trunnion axis express time, vertical axis is represented magnitude of voltage. these four kinds of wave distortions or the combination between them can be represented any distortion that in fact occurs.
The situation that Fig. 4 A and 4B represent is: the pulse duration that the fall delay of mains switch 3 makes mains switch 3 outputs is than the big te1 of pulse duration (being called first type distortion later on) of the correction signal Vc that will input to mains switch 3.
Fig. 4 A shows the impulse waveform of the correction signal Vc that exported from comparator 25 before distortion takes place, and Fig. 4 B shows under the situation that first type of distortion takes place, after the correction signal Vc shown in Fig. 4 A is transfused to mains switch 3, the impulse waveform of mains switch 3 output signals.
The situation that Fig. 5 A and 5B represent is: the pulse duration that the rising delay of mains switch 3 makes mains switch 3 outputs is than the little te2 of pulse duration (being called second type distortion later on) of the correction signal Vc that will input to mains switch 3.
Fig. 5 A shows the impulse waveform of the correction signal Vc that exported from comparator 25 before distortion takes place, and Fig. 5 B shows under the situation that second type of distortion takes place, after the correction signal Vc shown in Fig. 5 A is transfused to mains switch 3, the impulse waveform of mains switch 3 output signals.
The situation that Fig. 6 A and 6B represent is: the fluctuation of supply voltage makes the pulse duration of mains switch 3 outputs offer the big Δ V1 of reference voltage value Vpow (being called the distortion of the third type later on) of mains switch 3 than constant voltage source in the mains switch 3.
Fig. 6 A shows the impulse waveform of the correction signal Vc that exported from comparator 25 before distortion takes place, and Fig. 6 B shows under the situation that the third type distortion takes place, after as shown in Figure 6A correction signal Vc is transfused to mains switch 3, the impulse waveform of mains switch 3 output signals.
The situation that Fig. 7 A and 7B represent is: the fluctuation of supply voltage makes the pulse duration of mains switch 3 outputs offer the little Δ V2 of reference voltage value Vpow (being called the 4th type distortion later on) of mains switch 3 than constant voltage source in the mains switch 3.
Fig. 7 A shows the impulse waveform of the correction signal Vc that exported from comparator 25 before distortion takes place, and Fig. 7 B shows under the situation that the 4th type of distortion takes place, after the correction signal Vc shown in Fig. 7 A is transfused to mains switch 3, the impulse waveform of mains switch 3 output signals.
The time of delay that reference symbol δ presentation graphs 4A generates in the mains switch 3 of 7B.
First and the distortion situation of the third type under, distortion appears at the increase of low frequency part as signal level.Under second and the 4th type distortion situation, distortion appears at the reduction of low frequency part as signal level.
Below just in front under the situation of the wave distortion of described first to fourth kind of pattern, the waveform in the correcting circuit 2 that is in stable state is described in the variation that each point takes place.
At first, first type distortion is described.Fig. 8 shows by proofreading and correct the waveform of the each point of the correcting circuit 2 that is placed in stable state when the distortion that occurs first type.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Fig. 8, therefore identical with shown in Fig. 3 of impulse waveform 30 and integrated waveform 31 dispensed the explanation of those repetitions here.
Reference marker 34a represents to export and input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, be placed in stable state by the aligning step of being carried out by the D class A amplifier A of this embodiment when it first type of distortion occur in mains switch 3, reference marker 32a is illustrated in the integrated waveform of the integrated signal eo2 that generates according to impulse waveform 34a in the second integral device 24.
Reference marker 33a represents that the impulse waveform of the correction signal Vc that generates according to the difference between integrated waveform 31 and the 32a is as binary pulse " H " or " L " (that is, " 0 " or " 1 ") in comparator 25.Particularly, when waveform 31 is higher than waveform 32a, production burst " H " (perhaps " 1 "), and when waveform 31 is lower than waveform 32a, production burst " L " (perhaps " 0 ").
The amplitude range of impulse waveform 30 should be similar to 0 and Vsig between.Amplitude distortion can not occur in first type of distortion.Therefore, if from the voltage of constant voltage source supply power switch 3 is that fixed attenuation gain Vpow and the feedback circuit 5 is 1/K, then the amplitude of the impulse waveform 34a of feedback circuit 5 output should be similar to 0 and Vpow/K between, and equal the amplitude (Vpow/K=Vsig just) of the impulse waveform 30 of output from pulse modulator 1 by the effect of feedback circuit 5.
When in correcting circuit 2, carrying out normal correction, equal pulse area corresponding to the pulse area of the impulse waveform 34a of a frequency period in the stable state corresponding to the impulse waveform 30 of a frequency period.
Be in as shown in Figure 8 state at impulse waveform 30 and 34a, and under the situation of on the basis that acts on approximate Vsig/2 of gain controller 22 etc. first integrator 21 and second integral device 24 being operated, will form as shown in Figure 8 integrated waveform 31 and 32a.
In first type distortion of mains switch 3 (Fig. 4 A, 4B), the mean value of integrated waveform 32a will be higher than the mean value of the integrated waveform 32 that generates when distortion not occurring as shown in Figure 3.Like this, integrated waveform 31 just is shorter than time cycle under as shown in Figure 3 the situation that distortion do not occur greater than time cycle of integrated waveform 32a.
Therefore, the correction signal Vc that is exported by comparator 25 according to the difference between integrated waveform 31 and the 32a just has impulse waveform 33a.
As shown in Figure 8, respond first type distortion, routine effect by comparator 25, the pulse duration of impulse waveform 33a is than the little te1 of pulse duration of impulse waveform 34a (that is to say that the time cycle that impulse waveform 33a is in " H " level is shorter than the time cycle under as shown in Figure 3 the situation that distortion do not occur).
Therefore, even the impulse waveform 33a of pulse duration is fed and is transfused to once more to mains switch 3 as shown in Figure 8, this can make and occur first type distortion in the mains switch 3, and cause pulse duration to increase te1, and the pulse duration of this impulse waveform 34a also can equal the pulse duration of waveform 30. and Fig. 8 also shows δ time of delay that is mainly caused by mains switch 3 and occurs between impulse waveform 33a and 34a.
As mentioned above, in mains switch 3, occur under the situation of first type of distortion, the D class A amplifier A of this embodiment is this distortion of first type of response, the pulse duration of the impulse waveform 33a that comparator 25 can be exported is arranged to the little te1 of pulse duration than impulse waveform 30, first type distortion in the calibration power supply switch 3 thus, with so that the pulse duration of impulse waveform 34a is approximately equal to the pulse duration of impulse waveform 30 under the stable state, just, make pulse area corresponding to the impulse waveform 34a of a frequency period be approximately equal to pulse area corresponding to the impulse waveform 30 of a frequency period. Here it is will make the low frequency part of feedback signal ef and pulse-modulated signal ei be similar to identical, this means between these signals does not have error, has promptly proofreaied and correct the distortion of audio signal.
Next, second type distortion is described.Fig. 9 shows by proofreading and correct the waveform of the each point of the correcting circuit 2 that is placed in stable state when the distortion that occurs second type.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Fig. 9, therefore identical with shown in Fig. 8 of impulse waveform 30 and integrated waveform 31 dispensed the explanation of those repetitions here.
Reference marker 34b represents to export and input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, be placed in stable state by the aligning step of being carried out by the D class A amplifier A of this embodiment when it second type of distortion occur in mains switch 3, reference marker 32b is illustrated in the integrated waveform of the integrated signal eo2 that generates according to impulse waveform 34b in the second integral device 24.
Reference marker 33b represents that the impulse waveform of the correction signal Vc that generates according to the difference between integrated waveform 31 and the 32b is as binary pulse " H " or " L " (that is, " 0 " or " 1 ") in comparator 25.Particularly, when waveform 31 is higher than waveform 32b, production burst " H " (perhaps " 1 "), and when waveform 31 is lower than waveform 32b, production burst " L " (perhaps " 0 ").
The amplitude range of impulse waveform 30 should be similar to 0 and Vsig between.Amplitude distortion can not occur in second type of distortion. therefore, if from the voltage of constant voltage source supply power switch 3 is that fixed attenuation gain Vpow and the feedback circuit 5 is 1/K, then the amplitude of the impulse waveform 34b of feedback circuit 5 output should be similar to 0 and Vpow/K between, and equal the amplitude (Vpow/K=Vsig just) of the impulse waveform 30 of output from pulse modulator 1 by the effect of feedback circuit 5.
When in correcting circuit 2, carrying out normal correction, equal pulse area corresponding to the pulse area of the impulse waveform 34b of a frequency period in the stable state corresponding to the impulse waveform 30 of a frequency period.
Be in as shown in Figure 9 state at impulse waveform 30 and 34b, and under the situation of on the basis that acts on approximate Vsig/2 of gain controller 22 etc. first integrator 21 and second integral device 24 being operated, will form as shown in Figure 9 integrated waveform 31 and 32b.
In second type distortion of mains switch 3 (Fig. 5 A, 5B), the mean value of the integrated waveform 32 that the mean value of integrated waveform 32b will generate when distortion not occurring as shown in Figure 3.Like this, integrated waveform 31 just is shorter than time cycle under as shown in Figure 3 the situation that distortion do not occur greater than time cycle of integrated waveform 32b.
Therefore, the correction signal Vc that is exported by comparator 25 according to the difference between integrated waveform 31 and the 32b just has impulse waveform 33b.
As shown in Figure 9, respond second type distortion, routine effect by comparator 25, the pulse duration of impulse waveform 33b is than the big te2 of pulse duration of impulse waveform 34b (that is to say that the time cycle that impulse waveform 33b is in " H " level is longer than the time cycle under as shown in Figure 3 the situation that distortion do not occur).
Therefore, when the impulse waveform 33b of pulse duration as shown in Figure 9 is fed and is transfused to mains switch 3 once more, this can make and occur second type distortion in the mains switch 3, and causing pulse duration to increase te2, the pulse duration of this impulse waveform 34b also can equal the pulse duration of waveform 30.Fig. 9 also shows δ time of delay that is mainly caused by mains switch 3 and occurs between impulse waveform 33b and 34b.
As mentioned above, in mains switch 3, occur under the situation of second type of distortion, the D class A amplifier A of this embodiment is this distortion of second type of response, the pulse duration of the impulse waveform 33b that comparator 25 can be exported is arranged to the big te2 of pulse duration than impulse waveform 30, second type distortion in the calibration power supply switch 3 thus, with so that the pulse duration of impulse waveform 34b is approximately equal to the pulse duration of impulse waveform 30 under the stable state, just, make pulse area corresponding to the impulse waveform 34b of a frequency period be approximately equal to pulse area corresponding to the impulse waveform 30 of a frequency period. Here it is will make the low frequency part of feedback signal ef and pulse-modulated signal ei be similar to identical, this means between these signals does not have error, has promptly proofreaied and correct the distortion of audio signal.
Next, the distortion to the third type describes.Figure 10 shows by be placed in the waveform of each point of the correcting circuit 2 of stable state when the distortion that the third type occurs by correction.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 10, therefore identical with shown in Fig. 8 of impulse waveform 30 and integrated waveform 31 dispensed the explanation of those repetitions here.
Reference marker 34c represents to export and input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, be placed in stable state by the aligning step of being carried out by the D class A amplifier A of this embodiment when it the third type distortion occurs in mains switch 3, reference marker 32c is illustrated in the integrated waveform of the integrated signal eo2 that generates according to impulse waveform 34c in the second integral device 24.
Reference marker 33c represents that the impulse waveform of the correction signal Vc that generates according to the difference between integrated waveform 31 and the 32c is as binary pulse " H " or " L " (i.e. " 0 " or " 1 ") in comparator 25.Particularly, when waveform 31 is higher than waveform 32c, production burst " H " (perhaps " 1 "), and when waveform 31 is lower than waveform 32c, production burst " L " (perhaps " 0 ").
The amplitude range of impulse waveform 30 should be similar to 0 and Vsig between. in the third type distortion, amplitude is with respect to the distortion of the reference voltage value Vpow appearance+Δ V1 that is provided to mains switch 3 by constant voltage voltage. therefore, if the fixed attenuation in the feedback circuit 5 gain is 1/K, then the amplitude of the impulse waveform 34c of feedback circuit 5 outputs should be similar to 0 and (Vpow+ Δ V1)/K between.Like this, the amplitude of impulse waveform 34c just is not equal to but the big Δ V1/K of amplitude of ratio impulse waveform 30 of output from pulse modulator 1.
When in correcting circuit 2, carrying out normal correction, equal pulse area corresponding to the pulse area of the impulse waveform 34c of a frequency period in the stable state corresponding to the impulse waveform 30 of a frequency period.
Be in as shown in Figure 10 state at impulse waveform 30 and 34c, and under the situation of on the basis that acts on approximate Vsig/2 of gain controller 22 etc. first integrator 21 and second integral device 24 being operated, will form as shown in Figure 10 integrated waveform 31 and 32c.
When the distortion of the third type occurring in mains switch 3 (Fig. 6 A, 6B), the mean value of integrated waveform 32c will be higher than the mean value of the integrated waveform 32 that generates when distortion not occurring as shown in Figure 3.Like this, integrated waveform 31 just is shorter than time cycle under as shown in Figure 3 the situation that distortion do not occur greater than time cycle of integrated waveform 32c during.
Therefore, the correction signal Vc that is exported by comparator 25 according to the difference between integrated waveform 31 and the 32c just has impulse waveform 33c.
As shown in figure 10, respond the distortion of the third type, effect by comparator 25, the pulse duration of impulse waveform 33c is than the pulse duration of impulse waveform 30 little (be amplitude increase Δ V1) (that is to say that the time cycle that impulse waveform 33c is in " H " level is shorter than the time cycle under as shown in Figure 3 the situation that distortion do not occur).
Therefore, even when the impulse waveform 33c of pulse duration as shown in figure 10 is fed and is transfused to mains switch 3 once more, this can make the distortion that occurs the third type in the mains switch 3, and cause amplitude to be compared increasing Δ V1 with the pulse duration normal value that remains unchanged, be approximately equal to pulse area corresponding to the pulse area of the impulse waveform 34c of a frequency period corresponding to the impulse waveform 30 of a frequency period.Figure 10 also shows δ time of delay that is mainly caused by mains switch 3 and occurs between impulse waveform 33c and 34c.
As mentioned above, in mains switch 3, occur under the situation of the third type distortion, the D class A amplifier A of this embodiment responds the distortion of this third type, the pulse duration of the impulse waveform 33c that comparator 25 can be exported is arranged to the pulse duration little (being that amplitude increases Δ V1) than impulse waveform 30, the distortion of the third type in the calibration power supply switch 3 thus is with being approximately equal to pulse area corresponding to the impulse waveform 30 of a frequency period so that be in the pulse area corresponding to the impulse waveform 34c of a frequency period of stable state.Here it is, and will to make that the low frequency part of feedback signal ef and pulse-modulated signal ei is similar to identical, and this means between these signals does not have error, has promptly proofreaied and correct the distortion of audio signal.
In this case, the increase of integrated waveform 32c on amplitude reduced the pulse duration of the correction signal Vc of comparator 25 outputs effectively.
Next, the 4th type distortion being described. Figure 11 shows by be placed in the waveform of each point of the correcting circuit 2 of stable state when the distortion that occurs the 4th type by correction.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 11, therefore identical with shown in Fig. 8 of impulse waveform 30 and integrated waveform 31 dispensed the explanation of those repetitions here.
Reference marker 34d represents to export and input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, be placed in stable state by the correction of being carried out by the D class A amplifier A of this embodiment when it the 4th type of distortion occur in mains switch 3, reference marker 32d is illustrated in the integrated waveform of the integrated signal eo2 that generates according to impulse waveform 34d in the second integral device 24.
Reference marker 33d represents that the impulse waveform of the correction signal Vc that generates according to the difference between integrated waveform 31 and the 32d is as binary pulse " H " or " L " (i.e. " 0 " or " 1 ") in comparator 25.Particularly, when waveform 31 is higher than waveform 32d, production burst " H " (perhaps " 1 "), and when waveform 31 is lower than waveform 32d, production burst " L " (perhaps " 0 ").
The amplitude range of impulse waveform 30 should be similar to 0 and Vsig between.In the 4th type of distortion, amplitude is with respect to the distortion of the reference voltage value Vpow appearance-Δ V2 that is provided to mains switch 3 by constant voltage voltage. therefore, if the fixed attenuation in the feedback circuit 5 gain is 1/K, then the amplitude of the impulse waveform 34d of feedback circuit 5 outputs should 0 and (Vpow-Δ V2)/K between.Like this, the amplitude of impulse waveform 34d just is not equal to but the little Δ V2/K of amplitude of ratio impulse waveform 30 of output from pulse modulator 1.
When in correcting circuit 2, carrying out normal correction, equal pulse area corresponding to the pulse area of the impulse waveform 34d of a frequency period in the stable state corresponding to the impulse waveform 30 of a frequency period.
Be in as shown in Figure 11 state at impulse waveform 30 and 34d, and under the situation of on the basis that acts on approximate Vsig/2 of gain controller 22 etc. first integrator 21 and second integral device 24 being operated, will form as shown in Figure 11 integrated waveform 31 and 32d.
In the 4th type distortion of mains switch 3 (Fig. 7 A, 7B), the mean value of integrated waveform 32d will be lower than the mean value of the integrated waveform 32 that generates when distortion not occurring as shown in Figure 3.Like this, integrated waveform 31 just is longer than time cycle under as shown in Figure 3 the situation that distortion do not occur greater than the time cycle of integrated waveform 32d.
Therefore, the correction signal Vc that is exported by comparator 25 according to the difference between integrated waveform 31 and the 32d just has impulse waveform 33d.
As shown in figure 11, respond the 4th type distortion, effect by comparator 25, the pulse duration of impulse waveform 33d is than the pulse duration of impulse waveform 30 big (being that amplitude reduces Δ V2) (that is to say that the time cycle that impulse waveform 33d is in " H " level is longer than the time cycle under as shown in Figure 3 the situation that distortion do not occur).
Therefore, even when the impulse waveform 33d of pulse duration as shown in figure 11 is fed and is transfused to mains switch 3 once more, can make and occur the 4th type distortion in the mains switch 3, and cause amplitude to be compared with the pulse duration normal value that remains unchanged having reduced Δ V2, be approximately equal to pulse area corresponding to the pulse area of the impulse waveform 34d of a frequency period corresponding to the impulse waveform 30 of a frequency period.Figure 11 also shows δ time of delay that is mainly caused by mains switch 3 and occurs between impulse waveform 33d and 34d.
As mentioned above, in mains switch 3, occur under the situation of the 4th type of distortion, the D class A amplifier A of this embodiment responds the 4th type distortion, the pulse duration of the impulse waveform 33d that comparator 25 can be exported is arranged to the pulse duration big (being that amplitude reduces Δ V2) than impulse waveform 30, the 4th type distortion in the calibration power supply switch 3 thus is with being approximately equal to pulse area corresponding to the impulse waveform 30 of a frequency period so that be in the pulse area corresponding to the impulse waveform 34d of a frequency period of stable state.Here it is, and will to make that the low frequency part of feedback signal ef and pulse-modulated signal ei is similar to identical, and this means between these signals does not have error, has promptly proofreaied and correct the distortion of audio signal.
In this case, the increase of integrated waveform 32d on amplitude increased the pulse duration of the correction signal Vc of comparator 25 outputs effectively.
More than send pulse-modulated signal ei at correcting circuit 2 and be illustrated based on the process that feedback is carried out correction simultaneously to its output according to current embodiment.
As mentioned above, in (or correction) audio signal that has as illustrated in fig. 1 and 2 that the D class A amplifier A of structure can reduce mainly by the caused by operations distortion of mains switch 3.
In other words, this first integrator 21 is carried out integration based on pulse-modulated signal ei to signal, second integral device 24 carries out integration based on feedback signal ef to signal, comparator 25 compares integrated signal with reference to pulse-modulated signal ei, and according to pulse signal of comparative result generation, be used for to mains switch 3 output pulse signals, thus the factor that distortion caused that mainly causes in can correcting audio signals of this correcting circuit 2 by mains switch 3.
And in the present embodiment, first and second integrators 21 and 24 generate the signal (just, the signal that its segment pulse has been removed) that its HFS has been removed, and generate a correction signal by comparator 25 according to it.This just can prevent that as in routine techniques the work of circuit is subjected to the restriction of residual impulse part.That is to say, can prevent that this residual impulse that enters the nonlinear area of correcting circuit 2 and distortion takes place from partly causing the distortion of error signal and disturbing correct correction.
Routine techniques requires to form high-precision trapezoidal wave, and requires to generate an error signal and generate a correction signal according to this error signal in correcting unit 102 in error processor 101, and this has just caused complicated structure., in the present invention, do not need to form this trapezoidal wave and directly generate correction signal in correcting circuit 2 and do not need the generated error signal, this just makes circuit structure simplify on the whole.
And, the correcting circuit of present embodiment comprises a gain controller 22, be used for controlling gain from the integrated waveform of first integrator 21, subtracter 20 deducts the output signal (difference between the picked up signal) of gain controller 22 from pulse-modulated signal ei, subtracter 23 deducts the output signal (difference between the picked up signal) of gain controller 22 from feedback signal ef, and respectively the differential signal that subtracter generates is carried out integration by first and second integrators 21 and 24, therefore can suitably control the low-frequency gain of the integrated signal that generates in integrator 21 and 24, this just can prevent that integrated signal from surpassing the working range of circuit.
Though as shown in Figure 2, the output signal of first integrator 21 is transfused to gain controller 22, and gain controller 22 jointly provides input signal to subtracter 20 and 23, the output signal of second integral device 24 also can be transfused to gain controller 22. alternatively, also can adopt this structure: comprising two gain controllers 22, and the output signal of first integrator 21 is transfused to a gain controller 22 and is transfused to subtracter 20 from the signal of a gain controller 22 outputs, and the output signal of second integral device 24 is transfused to another gain controller 22 and is transfused to subtracter 23 from the signal of another gain controller 22 outputs simultaneously.
; in structure shown in Figure 2; the output signal that the output signal of first integrator 21 is transfused to gain controller 22 and gain controller 22 jointly is input to subtracter 20 and 23; this structure just can prevent that the gain of second integral device 24 is attenuated, and also just can carry out correct operation with higher precision.
And D class A amplifier A of the present invention comprises feedback circuit 5, is used for making the amplitude attenuation of mains switch 3 output signals and the feedback signal ef that output will be input to correcting circuit 2.Like this, just can be in feedback circuit 5 with the amplitude attenuation of amplifying signal in the mains switch 3 for the situation that distortion in mains switch 3, do not occur under from pulse modulator 1 level of the amplitude of the pulse-modulated signal ei of output identical, the compare operation in the correcting circuit 2 does not appear can simplifying under the situation of distortion simultaneously in mains switch 3.Therefore, can realize having the D class A amplifier A of this calibration function easily.
The concrete example of first preferred embodiment
In this first preferred embodiment, one of correcting circuit structure more specifically has been shown among Figure 12.Structure among Figure 12 will illustrate below.
In Figure 12, the output of pulse modulator 1 as shown in fig. 1 links to each other with an end of resistance 50, and the other end of resistance 50 links to each other with reverse ("-") input of operational amplifier 56 with N7 by node N4.And the output of this operational amplifier 56 provides negative feedback at node N5 bifurcated by electric capacity 54 and node N5, links to each other with "-" input and the node N6 of comparator 58 simultaneously.
One end of resistance 52 and an end of resistance 51 link together by node N6.The other end of resistance 51 links to each other with reverse ("-") input of operational amplifier 56 by node N7.The other end of resistance 52 links to each other with reverse ("-") input of operational amplifier 57 with N9 by node N8.
The output of operational amplifier 57 provides negative feedback at node N10 bifurcated by electric capacity 55 and node N10, links to each other with non-return ("+") input of comparator 58 simultaneously.The output of feedback circuit 5 as shown in Figure 1 links to each other with reverse ("-") input of operational amplifier 57 with node N8 by resistance 53.
Non-return ("+") input of operational amplifier 56 and operational amplifier 57 links to each other with suitable fixed potential respectively.The input of the same mains switch 3 as shown in Figure 1 of the output of comparator 58 links to each other.
In correcting circuit 2 with said structure, to be combined in a reverse input end that is input to operational amplifier 56 through the pulse-modulated signal ei of resistance 50 input and through the output signal of the operational amplifier 56 of resistance 51 inputs, just be equivalent to the operation of subtracter 20 among Fig. 2.That is to say that relatively reverse input end carries out oppositely just being equivalent to obtain pulse-modulated signal ei in subtracter 20 to the output signal of operational amplifier 56 and from the difference between the signal of gain controller 22.
And the signal that inputs to the reverse input end of operational amplifier 56 gathers the function that electric charge just is equivalent to the first integrator 21 shown in Fig. 2 in electric capacity 54.Also have, the resistivity between the adjusting resistance 50 and 51 just is equivalent to the function of the gain controller 22 shown in Fig. 2.
On the other hand, will be combined in the operation that a reverse input end that is input to operational amplifier 56 just is equivalent to subtracter 23 among Fig. 2 through the feedback signal ef of resistance 53 input and through the output signal of the operational amplifier 56 of resistance 52 inputs.That is to say that relatively reverse input end carries out oppositely just being equivalent to obtain feedback signal ef in subtracter 23 to the output signal of operational amplifier 57 and from the difference between the signal of gain controller 22.
And the signal that inputs to the reverse input end of operational amplifier 57 gathers the function that electric charge just is equivalent to the second integral device 24 shown in Fig. 2 in electric capacity 55.Also have, the resistivity between the resistance 52 and 53 just plays the effect of the gain controller 22 shown in Fig. 2.
Resistivity Gf1 between the resistance 50 and 51 equals the resistivity Gf2 between resistance 52 and 53
In addition, comparator 58 is equivalent to the comparator 25 shown in Fig. 2.Because the output of operational amplifier 56 and 57 is reversed first and second integrators 21 shown in Fig. 2 and 24 output respectively, the output of operational amplifier 56 is continuous with "-" input of comparator 58 and output operational amplifier 57 is continuous with "+" input of comparator 58, having constituted and the reverse structures of interconnection of comparator shown in Figure 2 25, so the output of comparator 58 is identical with the phase place of the output of comparator 25.
Recited above is exactly concrete demonstrative structure according to the correcting circuit 2 of this first preferred embodiment.
In the superincumbent instantiation, though the resistivity Gf1 between resistance 50 and 51 equates with resistivity Gf2 between resistance 52 and 53, even these resistivity are some difference each other, to the distortion that occurs in the mains switch 3 proofread and correct also can only paired pulses modulation signal ei and work period of feedback signal ef in fluctuation carry out., be equal to each other by making resistivity, do not make that the waveform of pulse-modulated signal ei and feedback signal ef is identical when just can distortion not occur in mains switch 3, this is simple correction the and do not need complicated structure very just.
Further, the value that resistivity Gf2 between resistivity Gf1 between the resistance 50 and 51 and resistance 52 and 53 is set to differ from one another, and regulate resistivity Gf1, Gf2, the gain constant G1 that is used for first integrator 21 is used for the gain constant G2 of second integral device 24 and the function that just is equivalent to feedback circuit 5 by the voltage Vpow that constant voltage source offers mains switch 3.Like this, just can dispense feedback circuit.
Second preferred embodiment
Combine according to the correcting circuit of this embodiment D class A amplifier A of describing in first preferred embodiment as shown in Figure 1 that can coexist.This correcting circuit is not to be directly used in the signal of comparison from first integrator and second integral device, but be used for obtaining from the difference between the signal of first integrator and second integral device at subtracter, and then to comparing by the signal that differential signal is carried out integration and then in reverser, carry out oppositely obtaining from the output signal (differential signal) of subtracter and in the third integral device, thereby, generate a correction signal.
Internal structure block diagram according to present embodiment correcting circuit 2 is shown in Figure 13.Structure shown in Figure 13 will specify hereinafter.In Figure 13, subtracter 20, first integrator 21, gain controller 22, the parts of describing in subtracter 23 and second integral device 24 and first preferred embodiment as shown in Figure 2 have identical structure, and carry out identical functions and operation, dispense identical description here.
In Figure 13, first integrator 21 has the output that is connected to "+" input of subtracter 26 through node N2, and second integral device 24 has the output of "-" input that is connected to described subtracter 26.In addition, the output of subtracter 26 is at node N11 bifurcated, is connected respectively to "+" input of comparator 29 and "-" input that is connected to comparator 29 through third integral device 27 and a reverser 28.
Identical with first preferred embodiment, the output of comparator 29 is connected to the input of mains switch 3.
In the present embodiment, as shown in figure 13, directly do not compare with the integrated waveform of exporting from second integral device 24 from the integrated waveform of first integrator 21 outputs, but in subtracter 26, ask poor between to integrated waveform, then to the signal (differential signal) of output from subtracter 26 with in reverser 28, its reverse signal that obtains is compared then by in third integral device 27, this differential signal being carried out integration, thereby in comparator 29, produce correction signal Vc.
Here, comprise difference between the low frequency part of the low frequency part of pulse-modulated signal ei and feedback signal ef from the differential signal of subtracter 26.Difference between the low frequency part represents to be included in the error among the feedback signal ef, i.e. distortion in the output audio signal, and this distortion is corrected in comparator 29.
Clearly, if for example fixed reference current potential (Vsig/2) is connected to "-" input of comparator 29, then described correct operation is identical with correct operation among first embodiment.
Yet, in the present embodiment, this fixed reference current potential (Vsig/2) is free of attachment to "-" input of comparator 29, but third integral device 27 and reverser 28 are connected to "-" input of comparator 29, described third integral device 27 is used for the differential signal of the subtracter 26 that comes self-reference Vsig/2 work is carried out integration, and described reverser 28 is used for described differential signal is carried out oppositely.Except above-mentioned calibration result, it also provides following effect.
That is to say, the difference (being the distortion in the output audio signal) that is included between the low frequency part of pulse-modulated signal ei in the differential signal that subtracter 26 exported and feedback signal ef further is integrated in third integral device 27 and is reinforced, in reverser 28, be reversed afterwards, thereby the distortion that is reinforced in the output audio signal is added in comparator 29 in the differential signal of subtracter 26 outputs.Therefore, can in comparator 29, produce correction signal Vc, have the distortion of strengthening more than in first preferred embodiment corresponding to the distortion in the output audio signal.Obtained like this than the higher calibration result of correcting circuit described in first preferred embodiment.
Then, will the waveform state of each point in the correcting circuit 2 be described.
The situation of any distortion at first, explanation is not taken place in mains switch 3.Figure 14 shows in this case each waveform when stable state in the correcting circuit 2.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 14, reference marker 40 expression is from pulse modulator 1 output and be input to the impulse waveform of the pulse-modulated signal ei of correcting circuit 2, the integrated waveform of the integrated signal eo1 that 41 expressions produce in first integrator 21 according to impulse waveform 40.
Reference marker 44 expression is input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, the integrated waveform of the integrated signal eo2 that 42 expressions produce in second integral device 24 according to impulse waveform 44.
Reference marker 45 is illustrated in the subtracter 26 by deducting the waveform of the signal that integrated waveform 42 generates from integrated waveform 41,46 expressions by in third integral device 27 to waveform 45 integrations and the integrated waveforms that in reverser 28, waveform 45 oppositely obtained.
Reference marker 43 expressions are represented as binary pulse " H " or " L " (i.e. " 0 " or " 1 ") according to the impulse waveform of the correction signal Vc that the difference between waveform 45 and 46 produces in comparator 29.Particularly, when waveform 45 is higher than waveform 46, produce pulse " H " (or " 1 "), and when waveform 45 is lower than waveform 46, produce pulse " L " (or " 0 ").
Impulse waveform 40 should have the amplitude range that is similar between zero-sum Vsig.The voltage of supposing to offer from constant voltage source mains switch 3 is Vpow, and the gain of the fixed attenuation in the feedback circuit 5 is 1/K, then have amplitude range between approximate zero and Vpow/K, and feedback circuit 5 is set makes the amplitude of impulse waveform 44 equal from the amplitude (being Vpow/K=Vsig) of the impulse waveform 40 of pulse modulator 1 output from the impulse waveform 44 of feedback circuit 5 output.
Be in state shown in Figure 14 and first integrator 21 and second integral device 24 under the situation of operating by gain controller 22 and similar device on the basis of approximate Vsig/2 in impulse waveform 40 and 44, form integrated waveform 41 and 42 as shown in figure 14.If subtracter 26 and third integral device 27 are worked, then form waveform 45 and 46 as shown in figure 14 on the basis of approximate Vsig/2.
Therefore, the correction signal Vc from comparator 29 outputs has impulse waveform 43.Here, under stable state, impulse waveform 44 has δ time of delay with respect to the impulse waveform 43 from comparator 29 outputs, is mainly caused by mains switch 3 this time of delay.
That is to say, Figure 14 shows correcting circuit 2 and can not cause under the normal operation of any distortion at mains switch 3, as the impulse waveform 44 of feedback signal ef have with as the identical waveform of the impulse waveform 40 of pulse-modulated signal ei. the low frequency part that this means impulse waveform 40 and 44 is equal to each other, and can be normal and without any the transmit audio signals of distortion.
In fact feedback signal ef comprises the wave distortion that is mainly caused by mains switch 3.This distortion makes the waveform distortion of feedback signal ef, causes the low frequency part of feedback signal ef to be different from the low frequency part of pulse-modulated signal ei.
Following interpretation in the four kind modes of Fig. 4 A to the wave distortion that causes by mains switch 3 shown in the 7B, when these four kinds of distortion modes take place, carry out a normal correct operation according to the correcting circuit of the D class A amplifier A of present embodiment. the combination of these four kinds of modes or these modes can be represented any distortion of actual generation.
To how change when these four kinds of distortion modes of generation the waveform of each point in the correcting circuit 2 below and describe in stable state.
At first, first kind of distortion mode will be described.Figure 15 shows under the situation of first type distortion the waveform of each point that is placed in the correcting circuit 2 of stable state by correction.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 15, impulse waveform 40 and integrated waveform 41 are with shown in Figure 14 identical, in this no longer repeat specification.
When reference marker 44a is illustrated in the distortion of first type of mains switch 3 generation, export and be input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, it carry out to proofread and correct the integrated waveform that places stable state, 42a to represent the integrated signal eo2 that produces according to impulse waveform 44a in second integral device 24 by the D class A amplifier A in the present embodiment.
Reference marker 45a is illustrated in the subtracter 26 by deducting the waveform of the signal that integrated waveform 42a generates from integrated waveform 41,46a represent by in third integral device 27 to waveform 45a integration and the integrated waveform that in reverser 28, waveform 45a oppositely formed.
Reference marker 43a represents that the impulse waveform of the correction signal Vc that produces according to the difference between integrated waveform 45a and the 46a is as binary pulse " H " or " L " (i.e. " 0 " or " 1 ") in comparator 29.Particularly, when waveform 45a is higher than waveform 46a, produce pulse " H " (or " 1 "), and when waveform 45a is lower than waveform 46a, produce pulse " L " (or " 0 ").
Impulse waveform 40 should have the amplitude range that is similar between zero-sum Vsig.Any distortion can not take place in the amplitude in first type distortion.Therefore, if offering the voltage of mains switch 3 from constant voltage source is Vpow, and the gain of the fixed attenuation in the feedback circuit 5 is 1/K, then have amplitude range between approximate zero and Vpow/K, and the effect by feedback circuit 5 becomes and equals amplitude (being Vpow/K=Vsig) from the impulse waveform 40 of pulse modulator 1 output from the impulse waveform 44a of feedback circuit 5 output.
When the normal timing of carrying out in correcting circuit 2, becoming corresponding to the pulse area of the impulse waveform 44a of a frequency period of stable state equals pulse area corresponding to the impulse waveform 40 of a frequency period.
In impulse waveform 40 and 44a is situation shown in Figure 15, and first and second integrators 21 and 24 are when operating by gain controller 22 and similar device on the basis of approximate Vsig/2, and integrated waveform 41 and 42a form as shown in figure 15.
(Fig. 4 A, 4B), the mean value of integrated waveform 42a is greater than the mean value of the integrated waveform 42 that produces when any distortion not taking place as shown in figure 14 in first type distortion of mains switch 3.
This has shortened the integrated waveform 45a that produces according to the difference between integrated waveform 41 and the 42a and has surpassed the used time cycle of reference level Vsig/2 in subtracter 26. and on the other hand, third integral device 27 and reverser 28 on average surpass reference level Vsig/2 according to the integrated waveform 46a that waveform 45a forms.
Therefore, the correction signal Vc that exports from comparator 29 based on the difference between waveform 45a and the 46a has impulse waveform 43a.In the present embodiment, correction signal Vc can proofread and correct the distortion that is reinforced in the output audio signal.
As shown in figure 15, respond first type distortion, normal running by comparator 29, formation is than the impulse waveform 43a pulse duration of the little te1 of pulse duration of impulse waveform 44a (that is, impulse waveform 43a time cycle of being in " H " level is compared with the situation that any distortion does not take place as shown in figure 14 and has been shortened).
Therefore, even the impulse waveform 43a of pulse duration is fed and imports once more mains switch 3 as shown in figure 15, make and at mains switch 3 first type distortion takes place, it makes pulse duration increase te1, and the pulse duration of impulse waveform 44a becomes and equals the pulse duration of waveform 40.Figure 15 also shows and δ time of delay occurred between impulse waveform 43a and 44a, and this time of delay, δ was mainly caused by mains switch 3.
As described, when first type distortion takes place in mains switch 3, the D class A amplifier A of present embodiment responds this distortion of first type, setting is from the pulse duration of the impulse waveform 43a of the comparator 29 output pulse duration te1 less than impulse waveform 40, thereby first type distortion in the calibration power supply switch 3 so that the pulse duration of impulse waveform 44a in the stable state pulse duration of impulse waveform 40 no better than, promptly make corresponding to the pulse area of the impulse waveform 44a of a frequency period no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period.This makes the low frequency part of feedback signal ef and the low frequency part of pulse-modulated signal ei almost be equal to each other, and this means between these signals that without any error, the distortion in the audio signal just is corrected.
Then, with the distortion that illustrates second type.Figure 16 shows when the distortion that takes place second type, by carrying out the waveform of each point of proofreading and correct the correcting circuit 2 that is placed in stable state.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 16, impulse waveform 40 and integrated waveform 41 no longer describe in detail at this with shown in Figure 15 identical.
When reference marker 44b is illustrated in the distortion of second type of mains switch 3 generation, export and be input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, it carry out to proofread and correct the integrated waveform that places stable state, 42b to represent the integrated signal eo2 that produces according to impulse waveform 44b in second integral device 24 by the D class A amplifier A among the current embodiment.
Reference marker 45b is illustrated in the subtracter 26 by deducting the waveform of the signal that integrated waveform 42b generates from integrated waveform 41,46b represent by in third integral device 27 to waveform 45b integration and the integrated waveform that in reverser 28, waveform 45b oppositely formed.
Reference marker 43b represents the impulse waveform of the correction signal Vc that produces according to the difference between integrated waveform 45b and the 46b in comparator 29, as binary pulse " H " or " L " (i.e. " 0 " or " 1 ").Particularly, when waveform 45b is higher than waveform 46b, produce pulse " H " (or " 1 "), and when waveform 45b is lower than waveform 46b, produce pulse " L " (or " 0 ").
Impulse waveform 40 should have the amplitude range that is similar between zero-sum Vsig.Any distortion can not take place in the amplitude in second type distortion.Therefore, if offering the voltage of mains switch 3 from constant voltage source is Vpow, and the gain of the fixed attenuation in the feedback circuit 5 is 1/K, then have amplitude range between approximate zero and Vpow/K, and the effect by feedback circuit 5 becomes and equals amplitude (being Vpow/K=Vsig) from the impulse waveform 40 of pulse modulator 1 output from the impulse waveform 44b of feedback circuit 5 output.
When the normal timing of carrying out in correcting circuit 2, becoming corresponding to the pulse area of the impulse waveform 44b of a frequency period of stable state equals pulse area corresponding to the impulse waveform 40 of a frequency period.
In impulse waveform 40 and 44b is situation shown in Figure 16, and first and second integrators 21 and 24 form as shown in figure 16 integrated waveform 41 and 42b when operating by gain controller 22 and similar device on the basis of approximate Vsig/2.
(Fig. 5 A, 5B), the mean value of integrated waveform 42b is greater than the mean value of the integrated waveform 42 that produces when any distortion not taking place as shown in figure 14 in second type distortion of mains switch 3.
Therefore, the integrated waveform 45b that produces in subtracter 26 according to the difference between integrated waveform 41 and the 42b was extended above the used time cycle of reference level Vsig/2. on the other hand, on average be lower than reference level Vsig/2 according to the integrated waveform 46b that waveform 45b forms by third integral device 27 and reverser 28.
Therefore, the correction signal Vc that exports from comparator 29 based on the difference between waveform 45b and the 46b has impulse waveform 43b.In current embodiment, correction signal Vc can proofread and correct the distortion that is reinforced in the output audio signal.
As shown in figure 16, respond second type distortion, by the routine effect of comparator 29, the pulse duration of impulse waveform 43b is formed the big te2 of pulse duration than impulse waveform 44b (being to compare with the situation that takes place without any distortion as shown in figure 14 time cycle that impulse waveform 43b is in " H " level to be extended).
Therefore, even when the impulse waveform 43b of pulse duration as shown in figure 16 is fed and imports mains switch 3 once more, make and at mains switch 3 second type distortion takes place, it makes pulse duration reduce te2, and the pulse duration of impulse waveform 44b becomes and equals the pulse duration of waveform 40.Figure 16 also shows and δ time of delay occurred between impulse waveform 43b and 44b, and this time of delay, δ was mainly caused by mains switch 3.
As described, when second type distortion takes place in mains switch 3, the distortion that the D class A amplifier A response of present embodiment is second type, setting is from the pulse duration of the impulse waveform 43b of the comparator 29 output pulse duration te2 greater than impulse waveform 40, thereby second type distortion in the calibration power supply switch 3 so that the pulse duration of impulse waveform 44b in the stable state pulse duration of impulse waveform 40 no better than, promptly make corresponding to the pulse area of the impulse waveform 44b of a frequency period no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period.This makes the low frequency part of feedback signal ef and the low frequency part of pulse-modulated signal ei almost be equal to each other, and this means between these signals that without any error, the distortion in the audio signal just is corrected.
Then, will the distortion of the third type be described.Figure 17 shows when the distortion that the third type takes place the waveform of each point that is placed in the correcting circuit 2 of stable state by correction.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 17, impulse waveform 40 and integrated waveform 41 are with shown in Figure 15 identical, and the repetitive description thereof will be omitted at this.
When reference marker 44c is illustrated in the distortion of mains switch 3 the third types of generation, export and be input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, it carry out to proofread and correct the integrated waveform that places stable state, 42c to represent the integrated signal eo2 that produces according to impulse waveform 44c in second integral device 24 by the D class A amplifier A among the current embodiment.
Reference marker 45c is illustrated in the subtracter 26 by deducting the waveform of the signal that integrated waveform 42c generates from integrated waveform 41,46c represent by in third integral device 27 to waveform 45c integration and the integrated waveform that in reverser 28, waveform 45c oppositely formed.
Reference marker 43c represents that according to the difference between integrated waveform 45c and 46c, the impulse waveform of the correction signal Vc that produces is as binary pulse " H " or " L " (i.e. " 0 " or " 1 ") in comparator 29.Particularly, when waveform 45c is higher than waveform 46c, produce pulse " H " (or " 1 "), and when waveform 45c is lower than waveform 46c, produce pulse " L " (or " 0 ").
Impulse waveform 40 should have the amplitude range that is similar between zero-sum Vsig.In the distortion of the third type, distortion occur in respect to the reference voltage level Vpow that offers mains switch 3 from constant voltage source+amplitude of Δ V1. therefore, if the gain of the fixed attenuation in the feedback circuit 5 is 1/K, be similar between zero-sum (Vpow+ Δ V1)/K from the amplitude of the impulse waveform 44c of feedback circuit 5 output, thereby be not equal to but greater than amplitude Δ V1/K from the impulse waveform 40 of pulse modulator 1 output.
When correcting circuit 2 was carried out normal corrections, becoming corresponding to the pulse area of the impulse waveform 44c of a frequency period of stable state equaled pulse area corresponding to the impulse waveform 40 of a frequency period.
In impulse waveform 40 and 44c is situation shown in Figure 17, and first and second integrators 21 and 24 form integrated waveform 41 and 42c. when operating by the action of gain controller 22 and similar device as shown in figure 17 on the basis of approximate Vsig/2
(Fig. 6 A, 6B), the mean value of integrated waveform 42c is greater than the mean value of the integrated waveform 42 that produces when distortion not taking place as shown in figure 14 in the distortion of the third type of mains switch 3.
Therefore, according to the difference between integrated waveform 41 and the 42c, the integrated waveform 45c that produces in subtracter 26 was shortened above the used time cycle of reference level Vsig/2.On the other hand, on average surpass reference level Vsig/2 by third integral device 27 and reverser 28 according to the integrated waveform 46c that waveform 45c forms.
Therefore, the correction signal Vc that exports from comparator 29 based on the difference between waveform 45c and the 46c has impulse waveform 43c.In the present embodiment, correction signal Vc can proofread and correct the distortion that is reinforced in the output audio signal.
As shown in figure 17, respond the distortion of the third type, normal effect by normal comparator 29, impulse waveform 43c pulse duration is formed less than the pulse duration of pulse duration waveform 40 (be amplitude increase Δ V1) (just, impulse waveform 43c time cycle of being in " H " level is compared with the situation that takes place without any distortion as shown in figure 14 and has been shortened).
Therefore, even when the impulse waveform 43c of pulse duration as shown in figure 17 is fed and imports mains switch 3 once more, make the distortion that the third types take place at mains switch 3, it makes amplitude compare with normal value has increased Δ V1, pulse duration remains unchanged, and becomes no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period corresponding to the pulse area of the impulse waveform 44c of a frequency period.Figure 17 also shows and δ time of delay occurred between impulse waveform 43c and 44c, and this time of delay, δ was mainly caused by mains switch 3.
As described, when the distortion of the third type takes place in mains switch 3, the D class A amplifier A of present embodiment responds the distortion of the third type, setting is from the pulse duration of the impulse waveform 43c of the comparator 29 output pulse duration (being that amplitude increases Δ V1) less than impulse waveform 40, thus the distortion of the third type in the calibration power supply switch 3 so that corresponding to the pulse area of the impulse waveform 44c of a frequency period no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period.This makes the low frequency part of feedback signal ef and the low frequency part of pulse-modulated signal ei almost be equal to each other, and this means between these signals that without any error, the distortion in the audio signal just is corrected.
In this case, the amplitude that increases integrated waveform 42c causes that the rising edge of waveform 45c has a reverse caster, and this pulse duration for the correction signal Vc that reduces comparator 29 outputs is effective.
Then, with the distortion that illustrates the 4th type.Figure 18 shows when the distortion that takes place the 4th type the waveform of each point that is placed in the correcting circuit 2 of stable state by correction.The trunnion axis express time, vertical axis is represented magnitude of voltage.
In Figure 18, impulse waveform 40 and integrated waveform 41 have omitted the explanation that repeats with shown in Figure 15 identical at this.
When reference marker 44d is illustrated in the distortion of mains switch 3 the third types of generation, export and be input to the impulse waveform of the feedback signal ef of correcting circuit 2 from feedback circuit 5, it carry out to be proofreaied and correct by the D class A amplifier A among the current embodiment and places stable state, and 42d represents the integrated waveform of the integrated signal eo2 that produces according to impulse waveform 44d in second integral device 24.
Reference marker 45d is illustrated in the subtracter 26 by deducting the waveform of the signal that integrated waveform 42d generates from integrated waveform 41,46d represent by in third integral device 27 to waveform 45d integration and the integrated waveform that in reverser 28, waveform 45d oppositely formed.
Reference marker 43d represents the impulse waveform according to the correction signal Vc that produces in the difference between integrated waveform 45d and the 46d in comparator 29, as binary pulse " H " or " L " (i.e. " 0 " or " 1 ").Particularly, when waveform 45d is higher than waveform 46d, produce pulse " H " (or " 1 "), and when waveform 45d is lower than waveform 46d, produce pulse " L " (or " 0 ").
Impulse waveform 40 should have the amplitude range that is similar between zero-sum Vsig.In the 4th type distortion, distortion occur in respect to the reference voltage level Vpow that offers mains switch 3 from constant voltage source-amplitude of Δ V2.Therefore, if the gain of the fixed attenuation in the feedback circuit 5 is 1/K, have the amplitude range that is similar between zero-sum (Vpow-Δ V2)/K from the impulse waveform 44d of feedback circuit 5 output. therefore, the amplitude of impulse waveform 44d is not equal to but less than the amplitude Δ V2/K of the impulse waveform 40 of pulse modulator 1 output.
When correcting circuit 2 was carried out normal corrections, becoming corresponding to the pulse area of the impulse waveform 44d of a frequency period of stable state equaled pulse area corresponding to the impulse waveform 40 of a frequency period.
In impulse waveform 40 and 44d is situation shown in Figure 180, and first and second integrators 21 and 24 during the effect operation by gain controller 22 and similar device, form as shown in figure 18 integrated waveform 41 and 42d on the basis of approximate Vsig/2.
(Fig. 7 A, 7B), the mean value of integrated waveform 42d is less than the mean value of the integrated waveform 42 that produces when any distortion not taking place as shown in figure 14 in the 4th type distortion of mains switch 3.
Therefore, the integrated waveform 45d that produces in subtracter 26 according to the difference between integrated waveform 41 and the 42d was extended above the used time cycle of reference level Vsig/2.On the other hand, on average be lower than reference level Vsig/2 by third integral device 27 and reverser 28 according to the integrated waveform 46d that waveform 45d forms.
Therefore, the correction signal Vc that exports from comparator 29 based on the difference between waveform 45d and the 46d has impulse waveform 43d.In the present embodiment, correction signal Vc can proofread and correct the distortion that is reinforced in the output audio signal.
As shown in figure 18, respond the 4th type distortion, by the normal effect of comparator 29, impulse waveform 43d pulse duration is formed greater than the pulse duration of impulse waveform 40 (being that amplitude reduces Δ V2) (just impulse waveform 43d time cycle of being in " H " level is compared with the situation that takes place without any distortion as shown in figure 14 and has been extended).
Therefore, even when the impulse waveform 43d of pulse duration as shown in figure 18 is fed and imports mains switch 3 once more, make and the 4th type distortion takes place at mains switch 3, it makes amplitude compare with normal value to have reduced Δ V2, pulse duration remains unchanged, and becomes no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period corresponding to the pulse area of the impulse waveform 44d of a frequency period.Figure 18 also shows and δ time of delay occurred between impulse waveform 43d and 44d, and this time of delay, δ was mainly caused by mains switch 3.
As described, when the 4th type distortion takes place in mains switch 3, the D class A amplifier A of present embodiment responds the 4th type distortion, setting is from the pulse duration of the impulse waveform 43d of the comparator 29 output pulse duration (being that amplitude reduces Δ V2) greater than impulse waveform 40, thus the 4th type distortion in the calibration power supply switch 3 so that corresponding to the pulse area of the impulse waveform 44d of a frequency period no better than corresponding to the pulse area of the impulse waveform 40 of a frequency period.This makes the low frequency part of feedback signal ef and the low frequency part of pulse-modulated signal ei almost be equal to each other, and this means between these signals that without any error, the distortion in the audio signal just is corrected.
In this case, the amplitude that increases integrated waveform 42d causes that the rising edge of waveform 45d has an anacline, and this pulse duration for the correction signal Vc that increases comparator 29 outputs is effective.
More than send pulse-modulated signal ei at correcting circuit 2 and be illustrated based on the process that feedback is carried out correction simultaneously to its output according to present embodiment.
As described, the D class A amplifier A with structure shown in Fig. 1 and 13 can reduce (or correction) main distortion that is caused by the operation of mains switch 3 in audio signal.
In addition, in the D class A amplifier A according to present embodiment, first and second integrators 21 and 24 integrated waveform are not directly inputted in the comparator, but carry out subtraction in subtracter 26, and export resulting subtraction signal.Thereafter, this subtraction signal be used for comparison by in third integral device 27, this subtraction signal integration and the signal that in reverser 28 this signals reverse obtained all being imported in the comparator 29, thereby strengthen the distortion in the audio signal.Compare with first preferred embodiment, this can further improve the calibration result that reduces the distortion in (or correction) audio signal.
The concrete example of second preferred embodiment
Below will be to more specifically describe the structure of correcting circuit 2 according to current embodiment shown in Figure 19.
The structure of the level before operational amplifier 56 and 57 the output comprises the subtracter 20 and 23 shown in Figure 13 block diagram, first integrator 21, second integral device 24 and gain controller 22 with shown in Figure 12 identical.The concrete annexation of each circuit and function are illustrated in the concrete example of first preferred embodiment, and the repetitive description thereof will be omitted at this.
Correcting circuit 2 below with reference to current concrete example shown in Figure 19, operational amplifier 56 has the output that is connected to reverse ("-") input of operational amplifier 66 through node N5, resistance 60 and node N12, and operational amplifier 57 has the output that is connected to non-return ("+") input of operational amplifier 66 through node N10, resistance 62 and node N13 simultaneously.
Operational amplifier 66 provides negative feedback by resistance 61, and has the non-inverting input that is connected to a suitable fixed potential through node N13 and resistance 63.
The output of operational amplifier 66 is at node N14 bifurcated, is connected respectively to non-return ("+") input of comparator 68 and reverse ("-") input that is connected to operational amplifier 67 through resistance 64 and node N15.Operational amplifier 67 has non-return ("+") input that is connected to a suitable fixed potential.
Operational amplifier 67 provides negative feedback through node N16, electric capacity 65 and node N15.
Operational amplifier 67 has the output that is connected to reverse ("-") input of comparator 68 through node N16.Comparator 68 has the output of the input that is connected to mains switch shown in Figure 13.
In the correcting circuit 2 of configuration as implied above, resistance 60 to 63 and operational amplifier 66 constitute subtracter 26 shown in Figure 13.
Yet, be reversed first and second integrators 21 shown in Figure 13 and 24 output signal respectively from the signal of operational amplifier 56 and 57 outputs. therefore, the output of operational amplifier 56 is connected to the reverse input end of operational amplifier 66, the output of operational amplifier 57 is connected to the non-inverting input of operational amplifier 66, thereby makes the signal of operational amplifier shown in Figure 13 66 and subtracter 26 outputs produce same phase.
Resistance 64, electric capacity 65 and operational amplifier 67 constitute third integral device 27 and reverser 28 shown in Figure 13, and the signal that has exporting from operational amplifier 66 carries out integration and reverse function.
As described, illustrative circuitry shown in Figure 19 constitutes according to correcting circuit 2 of the present invention.
Measurement data
Figure 20 is explanation is used to measure the measuring circuit of actual calibration result under the situation of using correcting circuit shown in Figure 12 (the concrete example of first preferred embodiment) or Figure 19 (the concrete example of second preferred embodiment) a block diagram.
The measurement of using measuring circuit shown in Figure 20 has 50% duty ratio by input after the interference of the signal generator 301 that PWM modulation circuit 300 is provided square wave is carried out to correcting circuit 2 and the feedback input end of exporting to correcting circuit 2 of importing correcting circuit 2.
By relatively square wave being directly inputted to PWM modulation circuit 300 not under the observed result of the AC voltmeter 303 of the output level of decoding by the impulse waveform input signal that obtains through the correcting circuit 2 of LPF302 and the state that suppresses by correcting circuit 2 in the interference that provides by the signal generator in the PWM modulation circuit 300 in the AC voltmeter 303 of the output level of decoding of impulse waveform input signal the result of observation estimate.
Estimation is the result be plotted in the chart of Figure 21. in the chart of Figure 21, vertical axis represents to disturb the residue level, trunnion axis is represented interfering frequency. here, measurement data A uses the shown in Figure 12 correcting circuit identical with correcting circuit 2 to obtain, and measurement data B uses the shown in Figure 19 correcting circuit acquisition identical with correcting circuit 2.
Clearly, the correcting circuit of describing in first preferred embodiment 2 is carried out calibration result preferably from the chart of Figure 21, and the correcting circuit of describing in second preferred embodiment 2 is carried out better calibration result.
Although output stage is pressed single-ended description in the above description, the present invention can also be applied to has two so-called BTL (balun transformer) structures that are used for the output stage of 180 ° of output audio signals of mutual out-phase.Just, use correcting circuit each output stage in the BTL structure in addition according to the present invention, can obtain the effect of correcting distortion.
The 3rd preferred embodiment
Figure 22 is the structured flowchart according to the D class A amplifier A of the 3rd preferred embodiment of the present invention.The D class A amplifier A comprises pulse modulator 1, correcting circuit 2, mains switch 3, LPF4, feedback circuit 5, loud speaker 406, the first constant-voltage control circuits 407, the second constant-voltage control circuits 408, dc output reference signal generator 410 and arithmetic element 411.Provide power source voltage Vcc by a power terminal 409 to the D class A amplifier A from an external source.
Pulse modulator 1 produces a pulse-modulated signal, and mains switch 3 is carried out switching according to the correction signal of proofreading and correct between power supply and ground in correcting circuit 2, make the power supply supply company receive the load of the output of D class A amplifier A (loud speaker 406).This handover operation by as the switching device shifter of the configuration of description in first preferred embodiment carry out.
LPF4 deletion HFS from the signal of mains switch 3 output will offer the audio signal of loud speaker 406 with demodulation, thereby reproduce sound. and feedback circuit 5 makes the suitable level and provide it to correcting circuit 2 of amplitude attenuation to of the signal of mains switch 3 outputs.
Pulse modulator 1 comprises the delta sigma modulating device 1a that is used for the delta sigma modulation digital audio signal and is used to change the pulse-modulated signal conversion equipment of delta sigma audio signal modulation to pulse-modulated signal.
Dc output reference signal generator 410 comprises LPF401 and level regulator 402.Arithmetic element 411 comprises subtracter 403 and adder 404.
Mainly first constant-voltage control circuit 407 that is made of a logical circuit makes the power source voltage Vcc that provides by power terminal 409 from external source be stabilized in a certain value, and provides this stable power voltage vcc to pulse modulator 1.
Second constant-voltage control circuit 408 also mainly is made of a logical circuit, makes the power source voltage Vcc that provides by power terminal 409 from external source be stabilized in a certain value, and provides this stable power voltage vcc to correcting circuit 2.
Although be directly connected to terminal 409 in Figure 22, in fact mains switch 3 generally is connected to terminal 409 by the low pass filter that is made of an inductance and an electric capacity.Yet this low pass filter will be provided by the high frequency noise that is included in the power source voltage Vcc that provides by terminal 409, and it is different from constant- voltage control circuit 407 and 408, can't produce the effect of the voltage fluctuation that suppresses the low frequency part in the audio band.
This be because be used to make offer need powerful relatively mains switch 3 the stable constant-voltage control circuit of voltage can bring unfavorable, promptly in constant-voltage control circuit bigger power loss can take place, and will increase the cost that this constant-voltage control circuit is installed.
In current embodiment, the correcting circuit 2 that is used for the correction feedback signal replaces constant-voltage control circuit to use, and has structure same as shown in Figure 2, and the repetitive description thereof will be omitted at this.In addition, the signal waveform of each point is with shown in Figure 3 identical in correcting circuit 2, and the repetitive description thereof will be omitted at this.
D class A amplifier A shown in Figure 22 is carried out a series of feedback operation, wherein based on the pulse-modulated signal ei that exports from pulse modulator 1 and from the relatively generation one correction signal Vc between the feedback signal ef of mains switch 3 outputs, and be input in the correcting circuit 2 by a reponse system (feedback circuit 5 and arithmetic element 411), correction signal Vc becomes feedback signal ef after by mains switch 3 and reponse system.
Waveform shown in Figure 3 has much at one amplitude in the impulse waveform 34 of the impulse waveform 30 of pulse-modulated signal ei and feedback signal ef, and has δ and do not take place to obtain under the situation of wave distortion in mains switch 3 time of delay.In addition, the impulse waveform 30 of correction signal Vc is similar each other with impulse waveform 33.
Yet, therefore the amplitude that surpasses a predetermined value and impulse waveform 34 when the power source voltage Vcc that offers mains switch 3 by terminal 409 becomes amplitude than impulse waveform 30 when big, increase from the signal level of second integral device 24 outputs, make integrated waveform 32 and Fig. 3 move up by comparison.
In this case, as shown in figure 10, in the time cycle of integrated waveform 31 above integrated waveform 32c, i.e. the time cycle that the output of comparator 25 is in " H " level is shortened.Shown in Figure 10 as reference, this has reduced the pulse duration of the impulse waveform of correction signal Vc, and make pulse area corresponding to the impulse waveform 34c of a frequency period equal pulse area corresponding to the impulse waveform 30 of a frequency period. with this process, the increase of the supply voltage that offers mains switch 3 carried out proofread and correct.
On the contrary, therefore when the amplitude that is lower than a predetermined value and impulse waveform 34 when the power source voltage Vcc that offers mains switch 3 by terminal 409 becomes less than the amplitude of impulse waveform 30, the output signal level of second integral device 24 reduces, and makes that comparing integrated waveform 32 with Fig. 3 moves down.
In this case, as shown in figure 11, in the time cycle of integrated waveform 31 above integrated waveform 32d, i.e. the time cycle that the output of comparator 25 is in " H " level is extended.As with reference to shown in Figure 11, this has increased the pulse duration of the impulse waveform of correction signal Vc, and makes pulse area corresponding to the impulse waveform 34d of a frequency period equal pulse area corresponding to the impulse waveform 30 of a frequency period.With this process, reducing to carry out and proofread and correct the supply voltage that offers mains switch 3.
As described, correcting circuit 2 is output calibration signal Vc when carrying out correction (impulse correction) according to feedback signal ef paired pulses modulation signal ei.
The particular circuit configurations of correcting circuit 2 as shown in figure 12.
In the above description, be input to the pulse-modulated signal ei of correcting circuit 2 and suitably all comprising a dc component after the decay from pulse modulator 1 from mains switch 3 output and the feedback signal ef that is input to correcting circuit 2.
Correcting circuit 2 is carried out the correction that comprises the dc component.Reason will illustrate below.
Although comprise a dc component, pulse-modulated signal ei is a binary pulse signal basically, and is illustrated in the scheduled voltage of each point in the circuit with a level among two level " H " and " L ".Therefore, use the conventional method that adopts analog circuit for example to block the dc component by an electric capacity, it is very difficult that a dc operating point is set separately.
In addition, correcting circuit 2 comprises the integrator with high dc gain, and it is very real therefore making circuit operation point stable through integrator by the dc feedback.
When power source voltage Vcc generation minor variations, the correction of the above-mentioned dc of comprising component can not go wrong.Yet, when power source voltage Vcc greatly changes, will go wrong.
For example, even the instrument of configuration automobile usually needs power source voltage Vcc to change also running without error between 11V and 16V.Therefore, when the design centre is set to 13.2V, in scope approximately ± 20%, change also work without error even need guarantee supply voltage.
Under this condition, execution comprises that the correction of distortion of signal that is used to suppress mains switch 3 output of dc component can cause problem, and promptly audio signal output (output-voltage levels) can reduce when power source voltage Vcc reduction as described below fast when any distortion not taking place.
For example, when power source voltage Vcc is arranged on 13.2V, it is the design centre point in the D class A amplifier A that comprises correcting circuit 2 shown in Figure 22, have at the pulse-modulated signal ei of pulse modulator 1 under the situation of 50% duty ratio (below be called " non-modulation situation "), suppose that the gain of regulating correcting circuit 2 makes that from the dc current potential of D class A amplifier A output (below be called " amplifier output dc current potential ") be 6.6V, it is a half value of power source voltage Vcc.
In this case, even when power source voltage Vcc changes between 11V and 16V, proofread and correct even carry out under the non-modulation situation of pulse-modulated signal ei, amplifier output dc current potential also remains on approximate 6.6V.
In addition, in above-mentioned gain-adjusted, when power source voltage Vcc is arranged on 13.2V as the design centre, have at the pulse-modulated signal ei of pulse modulator 1 output under the situation of approximate 80% pulsating load ratio, amplifier output dc current potential reaches 11V.Yet, when power source voltage Vcc when 13.2V reduces to 11V, above-mentioned correction allow the pulse-modulated signal ei of pulse modulator 1 output have approximate 80% pulsating load than the time amplifier output dc current potential reach 11V.This means, when power source voltage Vcc is reduced to 11V, exceeds about 80% o'clock amplifier at the pulsating load ratio of pulse-modulated signal ei and output to saturation condition.
Figure 23 is an oscillogram of explaining this state.In Figure 23, reference marker 420 expression is when power source voltage Vcc is reduced to 11V, the waveform (sinusoidal waveform) of output voltage when non-modulation situation amplifier is exported the dc current potential and is arranged on 6.6V.This output-voltage levels is (11-6.6) * 2=8.8Vpp.
Here, the amplifier output dc current potential under the non-modulation situation of conversion is to 5.5V, and this moment, it was a half value of supply voltage (11V), and the waveform of output voltage (sinusoidal waveform) is by 421 expressions.In this case, output-voltage levels is increased to 5.5V * 2=11Vpp.
As described, when supply voltage fluctuates in a big relatively scope, preferably should change according to the fluctuation of power source voltage Vcc than the consistent amplifier output dc current potential of determining, with high as far as possible increase output-voltage levels with the pulsating load of pulse-modulated signal ei.Particularly, preferably change the setting of reponse system, so that amplifier is exported a dc current potential half value of power source voltage Vcc always under non-modulation situation.
Therefore, provide dc output reference signal generator 410 and arithmetic element 411. these parts as shown in figure 22 to illustrate below in the present embodiment.
Dc output reference signal generator 410 produces a reference signal, is used under non-modulation situation hold amplifier output dc current potential in a desired value.The desired value that under non-modulation situation amplifier output dc current potential has been described provides the half value to the power source voltage Vcc of mains switch 3.
Therefore, in the present embodiment, with respect to the gain 1/K of feedback circuit 5, the level regulator 402 with 1/ (2K) gain of dc output reference signal generator 410 by providing at this makes input voltage, and promptly power source voltage Vcc decays to Vcc/ (2K).
Because the ac fluctuation part that is included in the power source voltage Vcc is deleted by the LPF401 in the dc output reference signal generator 410, the reference signal of generation can not be subjected to the ac influence of fluctuations.
In arithmetic element 411, subtracter 403 deducts the dc output reference signal that dc output reference signal generator 410 is exported from the output voltage of feedback circuit 5, and adder 404 adds a fixed potential Vsig/2 by in the voltage that subtraction obtained.Here, represented to be included in the dc component in the output of mains switch 3 by Vsw, the output voltage V fb of arithmetic element 411 can followingly represent:
Vfb＝Vsw/K-Vcc/(2·K)+Vsig/2......(5)
Here, under the non-modulation situation of pulse-modulated signal ei, correcting circuit 2 is that the mode that Vfb equals to be included in the low frequency part Vsig/2 of pulse-modulated signal ei is operated with the low frequency part that is included in feedback signal ef.
Therefore, bring equation Vfb=Vsig/2 in expression formula (5), Vsw equals Vcc/2.Because the output that LPF4 produces the D class A amplifier A from dc component Vsw, said structure allow audio signal to be input in the loud speaker 406 that will reproduce.
Although present embodiment has been described in arithmetic element 411 subtraction and the additive process that the dc current potential of the signal by feedback circuit 5 is carried out, these processes can be directly to the output execution of mains switch 3 and can the result be decayed by feedback circuit 5.
In this case, needless to say, the signal that subtract should have the Vcc/2 value, and the signal that add should have the KVsig/2 value.Subtraction or additive process can be carried out with reverse order.
The 4th preferred embodiment
Figure 24 is the structured flowchart of expression according to the D class A amplifier A of the 4th preferred embodiment of the present invention.Present embodiment is different from the 3rd preferred embodiment part and is not provide arithmetic element 411 and replaces correcting circuit 2 with correcting circuit 430.
Correcting circuit 430 is carried out and correcting circuit 2 identical operations of describing in the 3rd preferred embodiment basically, but comprises that the output of response dc output reference signal generator 410 is controlled at the function of the amplifier output dc current potential under the non-modulation situation. and the circuit structure of correcting circuit 430 is shown in Figure 25.
As shown in figure 25, correcting circuit 430 structurally is different from correcting circuit 2, wherein the output of dc output reference signal generator 410 is applied to non-return ("+") input of the imported operational amplifier 57 of difference that constitutes second integral device 24 by resistance 440, this non-return ("+") input is connected to a fixed potential point, provides a dc fixed potential Vc1 at this point through resistance 441.
At integrator (second integral device 24) when circuit structure is inserted into feedback signal ef along the path of its transmission as shown, determine that by feedback operation a dc operating point mainly determined by the operation of integrator (second integral device 24) with big dc gain.
Particularly, the second integral device 24 that is made of operational amplifier 57 has the gain of operational amplifier 57 under the state that the dc signal feeds back hardly.Therefore, though because the dc output potential of feedback operation operational amplifier 56 changes, the difference that causes between the difference input of operational amplifier 57 is very little. in other words, determine at the dc of relevant position operating point so that satisfy this condition.
In fact the impedance that is increased in "-" input of operational amplifier 57 and "+" input is to the resistance 52 that is sufficiently more than "-" input that is connected to operational amplifier 57,53 and be connected to the resistance 440 of "+" input of operational amplifier 57,441 resistance, and reduce by resistance 52,53,440 and 441 provide the output impedance of the parts of signal to arrive enough less than resistance 52,53, and 440 and 441 resistance is not difficult.
For ease of explanation, resistance 52 will have identical resistance R3 with 441, and resistance 53 will have identical resistance R4 with 440, and will be Vcc/ (2K) according to the output of the 3rd preferred embodiment dc output reference signal generator 410.Then, in the following expression of dc current potential Vp of "+" input of operational amplifier 57:
Vp＝(Vc1·R4+Vcc·R3/(2·K)/(R3+R4)......(6)
In addition, represent the dc current potential of "-" input of operational amplifier 57 by Vn, electric capacity 55 has with respect to the infinitely-great impedance of dc component, can ignore its influence.Therefore, have following formula:
Vn＝(Vt0·R4+Vfb·R3)/(R3+R4)......(7)
Wherein Vt0 is the dc current potential from operational amplifier 56 outputs, and Vfb is the dc current potential from feedback circuit 5 outputs.
If the dc current potential is Vsig/2 under the non-modulation situation of input pulse modulation signal ei, and this moment, Vt0 had the Vt00 value, then had following formula:
Vt00＝Vc0·(R1+R2)/R1-Vsig·R2/(2·R1)......(8)
Wherein R1 is the resistance of resistance 50, and R2 is the resistance of resistance 51.
As shown in this expression formula, the Vt00 value is a fixed value of being determined by the resistance of the dc current potential of pulse-modulated signal ei, the fixed potential Vc0 of "+" input that offers operational amplifier 56 and resistance 50 and 51.Therefore, fixed potential Vc1 is set to equaling Vt00, expression formula (6) and (7) become on the right first and equate under the non-modulation situation of pulse-modulated signal ei.
As previously mentioned, when the correcting circuit normal running, the Vp and the Vn of expression almost are equal to each other in expression formula (6) and (7).Therefore in this case, expression formula (6) and (7) become on the right second and equate.That is, has expression formula Vfb=Vcc/ (2K).This expression is carried out feedback operation and is made the dc current potential of feedback signal ef become to equal dc output reference signal under the non-modulation situation of pulse-modulated signal ei.
In addition, if mains switch 3 has dc component Vsw and has expression formula Vfb=Vsw/K, then carry out feedback operation so that Vsw becomes a half value of power source voltage Vcc.
According to condition is set, all resistance R1 can be set equal identical resistance R to R4, fixed potential Vc0 is arranged on Vsig/2.This allows to have expression formula Vt00=Vsig/2 and Vc1=Vsig/2, and it has simplified circuit structure.
Although resistance 52 and 441 has identical resistance in the foregoing circuit structure, and resistance 53 and 440 has identical resistance, also can obtain identical effect even resistance has different resistances mutually.
In addition, although it is complicated a little, can use and be provided for making the output signal of dc output reference signal generator 410 oppositely to replace sending the structure of this output signal to the device of "+" input of the operational amplifier 57 that constitutes second integral device 24, promptly be used for making the reverse device of direction of dc output reference signal potential increase/minimizing, wherein provide a reverse signal "-" input to operational amplifier 57 by a resistance with respect to increase/minimizing power source voltage Vcc.In this case, be fixed at the current potential of operational amplifier 57 "+" input, this can improve the stability of operational amplifier 57.
Although the output stage of D class A amplifier A is described according to single-ended in the above description, the present invention is not limited thereto, have two so-called BTL structures that are used for the output stage of 180 ° of output audio signals of mutual out-phase but can also be applied to.That is to say, use correcting circuit each output stage in the BTL structure according to the present invention in addition, can obtain the effect of correcting distortion.
The 5th preferred embodiment
Figure 26 is the structured flowchart of explanation according to fifth preferred embodiment of the invention D class A amplifier A.
The D class A amplifier A comprises pulse modulator 1, level adjustment circuit 450, correcting circuit 2, mains switch 3, feedback circuit 5, LPF4, loud speaker 406, the first constant-voltage control circuits, 407, the second constant-voltage control circuits 408 and level reference signal generator 460.Provide power source voltage Vcc to the D class A amplifier A by power terminal 409 from an external source.
Pulse modulator 1 comprises delta sigma modulation circuit 1a, is used for the modulation of digital audio and video signals delta sigma, and pulse-modulated signal conversion equipment 1b, be used for the delta sigma audio signal modulation is converted to pulse-modulated signal.Level reference signal generator 460 comprises a LPF451 and an attenuator 452.
Pulse modulator 1 produces a pulse-modulated signal, and it is the binary signal that obtains by the pulse width modulation to audio signal.Mains switch 3 is carried out handover operation according to the logical value of correction signal, the level of this correction signal (amplitude) is regulated by level adjustment circuit 450, its pulse duration is further proofreaied and correct by correcting circuit 2, makes the power supply supply company receive the load of the output of D class A amplifier A (loud speaker 406).Handover operation is by carrying out as describing in the switching device shifter that is disposed at first preferred embodiment.
LPF4 from the signal of mains switch 3 output the deletion HFS with to will offering the audio signal demodulation of loud speaker 406, thereby reproduce sound. feedback circuit 5 makes from the suitable level and provide (feedback) output signal to correcting circuit 2 of the amplitude attenuation to of the signal of mains switch 3 outputs.
First constant-voltage control circuit 407 makes the power source voltage Vcc that provides by power terminal 409 from external source be stabilized in a certain value, and provides this stable power voltage vcc to pulse modulator 1.Second constant-voltage control circuit 408 makes the power source voltage Vcc that provides by power terminal 409 from external source be stabilized in a certain value, and provides this stable power voltage vcc to correcting circuit 2.
Although be to be directly connected to terminal 409 in Figure 26, in fact mains switch 3 generally is connected to terminal 409 by a low pass filter that is made of an inductance and electric capacity.
Yet this low pass filter will be eliminated the high-frequency noise that is included in the power source voltage Vcc, and it is different from constant- voltage control circuit 407 and 408, can not fully produce the effect of the voltage fluctuation that suppresses audio band medium and low frequency part.
This be because be used to make offer need powerful relatively mains switch 3 the stable constant-voltage control circuit of voltage can bring unfavorablely, promptly in constant-voltage control circuit bigger power loss can take place, and can increase the cost of installation constant-voltage control circuit.In current embodiment, the correcting circuit 2 that is used for the correction feedback signal replaces constant-voltage control circuits and uses.
Correcting circuit 2 has identical as shown in Figure 2 internal structure, and the repetitive description thereof will be omitted at this.In addition, in correcting circuit 2 signal waveform of each point with shown in Figure 3 identical.In the present embodiment, impulse waveform 30 is the pulse-modulated signals (below be called the level adjustment pulse-modulated signal) from level adjustment circuit 450 output.
As described, carry out a series of feedback operation, wherein correcting circuit 2 is according to from relatively generation and output calibration signal Vc between the level adjustment pulse-modulated signal of level adjustment circuit 450 outputs and the feedback signal of exporting and be input to by feedback circuit 5 correcting circuit 2 from mains switch 3, and this correction signal Vc becomes feedback signal after by mains switch 3 and feedback circuit 5.
As described in first preferred embodiment, Fig. 3 show when when any distortion does not take place in mains switch 3 at the waveform of each point of correcting circuit 2.When distortion takes place mains switch 3, described as first preferred embodiment, change shown in Fig. 8-11 by the correcting circuit waveform.
Correcting circuit 2 is carried out incoming level regulating impulse modulation signal according to the difference of low frequency part between level adjustment pulse-modulated signal and feedback signal basically and is proofreaied and correct (pulse duration correction), thereby exports a correction signal.Therefore, change even offer the power source voltage Vcc of mains switch 3, if power source voltage Vcc changes the high-quality audio signal that also can obtain not have distortion in a certain scope.
Yet, when the power source voltage Vcc that offers mains switch 3 exceeded above-mentioned range, following point can appear.
Just, when increasing, the power source voltage Vcc that offers mains switch 3 surpasses an appropriate value, and it is bigger when a lot of than the amplitude of impulse waveform 30 that the amplitude of impulse waveform 34 becomes, the upper edge of the integrated waveform 32 that forms based on impulse waveform 34 has the inclination more sharp-pointed than the integrated waveform that forms based on impulse waveform 30 31, as shown in figure 27.Therefore, the time cycle of integrated waveform 31 and 32 before a half period that arrives impulse waveform 30 intersects mutually, makes correcting circuit 2 and mains switch 3 export new pulse.Below, will be called waveform and cut apart phenomenon.
Even taking place, this waveform cuts apart phenomenon, for example the distortion meeting is very little to the influence of sound quality, because the low frequency part of paired pulses modulation signal has been carried out correction, yet, the electromagnetic wave that the quantity of the on/off operation of increase mains switch 3 can increase power loss unfriendly and be produced by mains switch 3.
In the present embodiment, provide level adjustment circuit 450 so that stop this waveform of generation to cut apart phenomenon.The operation of level adjustment circuit 450 will be described below.
Level adjustment circuit 450 can be a structure as shown in figure 28.In Figure 28, analog switch 450a carries out and selects and export an input crest voltage or an earthy handover operation according to the logic level of the pulse-modulated signal that provides from pulse modulator 1.
Have the pulse duration identical with pulse-modulated signal from the signal of level adjustment circuit 450 outputs, its amplitude is adjusted to the value of the crest voltage that equals to import separately.
In level reference signal generator 460, the fluctuation that comprises the relative high frequency of audio signal frequency band partly decays from the power source voltage Vcc of LPF451, and further decays to approximate 1/K in attenuator 452.Resulting voltage is output to level adjustment circuit 450 as crest voltage (level reference signal).
As a result, the amplitude (peak value) that is input to the level adjustment pulse-modulated signal of correcting circuit 2 becomes approximate Vcc/K.This value is no better than from the amplitude of the feedback signal of feedback circuit 5 output, as previously mentioned.
Waveform takes place cut apart the reason of phenomenon and be, compare with the amplitude of pulse-modulated signal, the amplitude of feedback signal is increased to and has exceeded restriction.In this preferred embodiment, the amplitude that level adjustment circuit 450 is regulated the level adjustment pulse-modulated signal that is input to correcting circuit 2 is to the amplitude of feedback signal no better than, and this can be avoided that waveform takes place and cuts apart phenomenon.
The amplitude of regulating impulse modulation signal also can make the audio signal parts that is included in the pulse-modulated signal change on level.As previously mentioned, the audio signal parts that is included in the feedback signal of operation correcting circuit 2 so that make is consistent with audio signal parts in being included in pulse-modulated signal.
Therefore, in the present embodiment, output audio signal changes level according to the fluctuation of carrying out with level adjustment circuit 450 that is adjusted on the power source voltage Vcc, makes calibration result partial loss in the correcting circuit 2.
Yet, cut-off frequency by LPF451 in the level reference signal generator 460 is set is an enough low value, might avoid appearing at the rapid fluctuations (fluctuation in high relatively frequency band) of the power source voltage Vcc that takes place in the output of level reference signal generator 460, and avoid 450 pairs of this rapid fluctuations of level adjustment circuit to carry out and regulate.
With this set, by correcting circuit 2 correction pulse width so that fully avoid the distortion that takes place with respect to the fast relatively fluctuation of power source voltage Vcc in the audio band, and with respect to the fluctuation of relative mitigation, i.e. bigger fluctuation in low relatively frequency, the stop pulse width is proofreaied and correct, the amplitude of regulating impulse modulation signal is cut apart phenomenon to avoid that waveform takes place, and this can be avoided efficient to reduce and electromagnetic interference.
Although present embodiment has been described the feedback circuit 5 with 1/K gain, and the attenuator 452 that also has the 1/K attenuation rate, it is the process situation of carrying out at by correcting circuit 2 paired pulses modulation signals and feedback signal much at one, and usually, these processes needn't be equal to each other.
The 6th preferred embodiment
In the 5th preferred embodiment, fluctuation with respect to power source voltage Vcc in a relatively low frequency band, level adjustment circuit 450 is regulated the level of the level adjustment pulse-modulated signal that is input to correcting circuit 2, yet cut apart phenomenon thereby avoid taking place waveform., change from the fluctuation of the Audio Meter of D class A amplifier A output, caused the fluctuations in volume of loud speaker 406 according to power source voltage Vcc.
According to following the 6th preferred embodiment that will describe, the D class A amplifier A has can avoid waveform to cut apart the generation of phenomenon, avoids taking place according to the fluctuation of power source voltage Vcc the structure of fluctuations in volume simultaneously.
Figure 29 is the structured flowchart of explanation according to the D class A amplifier A of current embodiment. the same with the 5th preferred embodiment, the D class A amplifier A of present embodiment also comprises pulse modulator 1, level adjustment circuit 450, correcting circuit 2, mains switch 3, feedback circuit 5, LPF4, loud speaker 406, the first constant-voltage control circuits, 407, the second constant-voltage control circuits 408 and level reference signal generator 460. provide power source voltage Vcc to the D class A amplifier A by power terminal 409 from an external source.
The D class A amplifier A of present embodiment is different from the 5th preferred embodiment part and is to have increased A/D converter 500, and pulse modulator 1 further has a multiplication constant generator 510 and a multiplier 520.
In the present embodiment, level reference signal generator 460 is also as the modulation index control-signals generator, and being used to produce a control will provide by the signal of the multiplication constant of multiplier 520 (being modulation index) (below be called the modulation index control signal).A/D converter 500 conversions are numerical data from the modulation index control signal of level reference signal generator 460 outputs, and provide digitized modulation index control signal to the multiplication constant generator 510 in the pulse modulator 1.Multiplication constant generator 510 turns to 1 with input specification when equaling a predetermined value when power source voltage Vcc, and obtains the inverse of a normalized input, is used to provide multiplier 520 these inverses that obtain as multiplication constant, i.e. a modulation index.Multiplier 520 constitutes a modulation index control circuit with a given audio signal and multiply each other .A/D transducer 500, multiplication constant generator 510 and a multiplier 520 of this multiplication constant as numerical data.
Therefore, to the given adjusting of an audio signal, this can be avoided taking place the fluctuations in volume of loud speaker 406 in level adjustment circuit 450 in cancellation.This operation will be described below.
When power source voltage Vcc is a predetermined value, the output of level reference signal generator 460 will be turned to 1 by specification, and when power source voltage Vcc is a predetermined value, the audio signal parts that is included in level adjustment circuit 450 outputs will be represented by e1.Then, represent the 460 normalized outputs of level reference signal generator by G, the audio signal parts that is included in the output of level adjustment circuit 450 can be expressed as Ge1.
In addition, indicate to be input to the audio signal data of delta sigma modulating device by e0, have expression formula e1=Me0, wherein M is the conversion gain among delta sigma modulating device 1a and the pulse-modulated signal conversion equipment 1b.
Here, when power source voltage Vcc is a predetermined value, the output specification of AD converter 500 is turned to 1, normalized output of AD converter 500 above-mentioned no better than G that becomes.Multiplication constant generator 510 calculates the inverse of this normalized output, and its output almost becomes 1/G.
Here, the audio signal by e00 represents to be input to multiplier 520 has expression formula e0=e00/G.
Therefore, the audio signal parts that is included in the output of level adjustment circuit 450 is expressed as GM=Me00, does not wherein comprise G.This shows that the audio signal parts that will input to correcting circuit 2 is not worth the influence that G fluctuates owing to the fluctuation of power source voltage Vcc, also just fluctuations in volume can not take place.
The 7th preferred embodiment
The 6th preferred embodiment is carried out predefined procedure to input audio signal in pulse modulator 1, cut apart phenomenon to avoid that waveform takes place, and avoids simultaneously taking place owing to fluctuations in volume takes place in the power source voltage Vcc fluctuation.
Yet, increase and to be used for can causing shortcoming such as cost increases inevitably AD converter 500, multiplication constant maker 510 and multiplier 520 that input audio signal is carried out these prior defined procedures.
Consider that waveform cuts apart phenomenon and in fact occur in power source voltage Vcc and be increased to and exceed a certain limit value, present embodiment provides a kind of like this structure, wherein the level adjustment pulse-modulated signal from level adjustment circuit 450 outputs had a certain amplitude before reaching this limit value, and and if only if power source voltage Vcc increase when exceeding this limit value, the amplitude that increases the level adjustment pulse-modulated signal is cut apart phenomenon to avoid that waveform takes place.Comparing this with the 6th preferred embodiment simplifies the structure.
Present embodiment provides the structure identical with the 5th preferred embodiment, and be different from the 5th embodiment part and be that level reference signal generator 460 has internal structure shown in Figure 30, rather than structure shown in Figure 26.
In Figure 30, provide "+" input that power source voltage Vcc arrives comparator 453 and an input of switch 454 by LPF451 and attenuator 452.
Provide a fixed voltage V0 "-" input from a fixed voltage source 455 to comparator 453.This fixed voltage V0 also offers other input of switch 454.
The output of comparator 453 offers the control input end of switch 454.When an output voltage of attenuator 452 during greater than fixed voltage V0, switch 454 is selected this output voltage of attenuators 452, when the output voltage of attenuator 452 is equal to or less than fixed voltage V0, selects fixed voltage V0.
Therefore, in current embodiment, change along solid line shown in Figure 31 601 from the fluctuation of the level reference signal of level reference signal generator 460 output according to power source voltage Vcc.Reference symbol Vcc1 represents the supply voltage value when the output voltage of attenuator 452 equals fixed voltage V0.
In current embodiment, level adjustment circuit 450 is carried out level adjustment so that make the magnitude of voltage of the level adjustment pulse-modulated signal of output be fixed on the interior V0 of a frequency band, increase less than the increase according to the power source voltage Vcc in a frequency band of the amplitude of Vcc1 and the level adjustment pulse-modulated signal that will export in this frequency band power source voltage Vcc, supply voltage is equal to or greater than Vcc1. in this frequency band
Aforesaid operations is cut apart in the frequency band of phenomenon constant by keeping at one waveform can not taking place from the amplitude of the level adjustment pulse-modulated signal of level adjustment circuit 450 outputs, can obtain to avoid to change at level with the fluctuation of power source voltage Vcc, simultaneously by cutting apart the increase of power source voltage Vcc in the frequency band of phenomenon and increase from the amplitude of the level adjustment pulse-modulated signal of level adjustment circuit 450 outputs and also can obtain to avoid taking place the function that waveform is cut apart phenomenon according to waveform taking place at one from the audio signal level of D class A amplifier A output.
Dotted line shown in Figure 31 600 expressions are as described in the 5th preferred embodiment, and attenuator 452 provides the characteristic under the situation of decay of approximate 1/K in whole frequency band.Characteristic when chain-dotted line 602 shown in Figure 31 is illustrated in frequency band of expansion, in this frequency band, be increased to the voltage of Vcc2 at this voltage from Vcc1 to the output that surpasses 1/K and change level reference signal generator 460 by the attenuation that increases in the attenuator 452, the signal of exporting from level reference signal generator 460 remains on a certain level.Should be noted that and to expand a frequency band that in this frequency band, the signal of exporting from level reference signal generator 460 remains on a certain level by the fixed voltage V0 that increases fixed voltage source 455.
Be increased in AD converter 500, multiplication constant maker 510 and the multiplier 520 described in the 6th preferred embodiment in the structure of present embodiment, can carry out the process that is used to avoid the fluctuations in volume that in a frequency band, takes place, increase in the fluctuation of the amplitude of this frequency band level adjustment pulse-modulated signal according to power source voltage Vcc.
Although the output stage that illustrates in above preferred embodiment is according to single-ended description, the present invention can also be applied to so-called two the BTL structures that are used for the output stage of 180 ° of output audio signals of mutual out-phase that have.Just, apply the present invention to each output stage in the BTL structure in addition, also can obtain the effect of above-mentioned correcting distortion.
The invention provides a kind of efficient D class A amplifier A, compare with traditional D class A amplifier A, can effectively reduce the distortion that produces by the fluctuation of the supply voltage that offers mains switch in the output signal, when even this amplifier fluctuates in a big relatively scope at supply voltage, use that also can be without a doubt, in this amplifier, when any distortion not taking place, audio signal output level reduces hardly.
Though the present invention is had been described in detail and describes, above-mentioned explanation all is schematically illustrating of carrying out in all its bearings, be not restrictive. therefore it should be understood that without departing from the present invention and can make multiple modification and change it.

Claims (5)

1. D class A amplifier A comprises:
A mains switch is used for according to pulse width modulating signal the power supply that supply voltage is provided being carried out on/off switch;
A correcting circuit is used for proofreading and correct being transfused to the pulse duration of the described pulse width modulating signal of mains switch according to the amplitude of described mains switch output signal;
A level reference signal generator is used for generating a level reference signal from described supply voltage; And
A level adjustment circuit is used for regulating being transfused to the amplitude of the described pulse width modulating signal of described correcting circuit according to described level reference signal value,
Wherein said level reference signal generator in described supply voltage is lower than the frequency band of preset value one of output with fixed value as described level reference voltage of signals, and one of output will surpass the numerical value of described fixed value as described level reference voltage of signals along with the increase of described supply voltage in described supply voltage is equal to or higher than the frequency band of described preset value.
2. D class A amplifier A according to claim 1, wherein
Described level reference signal generator comprises:
A low pass filter is used for removing HFS from described supply voltage; And
An attenuator is used for the output of described low pass filter is decayed.
3. D class A amplifier A according to claim 1 further comprises:
A pulse modulator is used for the pulse duration of input signal is modulated to export a pulse width modulating signal;
A modulation index control-signals generator is used for generating a modulation index control signal from described supply voltage; And
A modulation index regulating circuit is used for according to the value of described modulation index control signal the modulation index in the described pulse modulator being regulated.
4. D class A amplifier A according to claim 3, wherein
Described level reference signal generator can also be used as described modulation index control-signals generator.
5. D class A amplifier A according to claim 3, wherein
Described modulation index regulating circuit comprises:
An A/D converter is used for described modulation index control signal is converted to a digital signal;
A multiplication constant maker is used for generating a multiplication constant according to the value of the described digital signal of described A/D converter output; And
A multiplier is used for making described multiplication constant and the signal multiplication that inputs to described pulse modulator.
CNB2005101317185A 2002-10-03 2003-09-30 Class d amplifier Expired - Fee Related CN100486111C (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
JP291195/2002 2002-10-03
JP2002291195A JP3776392B2 (en) 2002-10-03 2002-10-03 Class D amplifier
JP333412/2002 2002-11-18
JP52385/2003 2003-02-28

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
CNB031648886A Division CN1277351C (en) 2002-10-03 2003-09-30 Class D amplifier

Publications (2)

Publication Number Publication Date
CN1790898A CN1790898A (en) 2006-06-21
CN100486111C true CN100486111C (en) 2009-05-06

Family

ID=32282849

Family Applications (2)

Application Number Title Priority Date Filing Date
CNB2005101317185A Expired - Fee Related CN100486111C (en) 2002-10-03 2003-09-30 Class d amplifier
CNB200510131719XA Expired - Fee Related CN100461626C (en) 2002-10-03 2003-09-30 Class D amplifier

Family Applications After (1)

Application Number Title Priority Date Filing Date
CNB200510131719XA Expired - Fee Related CN100461626C (en) 2002-10-03 2003-09-30 Class D amplifier

Country Status (2)

Country Link
JP (1) JP3776392B2 (en)
CN (2) CN100486111C (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102484458A (en) * 2009-08-26 2012-05-30 松下电器产业株式会社 Sensor device

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1429455A1 (en) * 2002-12-11 2004-06-16 Dialog Semiconductor GmbH Linearization of a PDM class-D amplifier
JP4759050B2 (en) * 2006-05-15 2011-08-31 旭化成エレクトロニクス株式会社 Drive device
CN101443999B (en) * 2006-05-15 2012-07-25 旭化成电子材料元件株式会社 Drive device, drive method, and information device
JP4859644B2 (en) * 2006-12-01 2012-01-25 新日本無線株式会社 Class D amplification method and class D amplifier
CN113029207B (en) * 2021-03-17 2022-06-28 上海睿奈电子科技有限公司 High-sensitivity and configurable sensor driving and signal processing integrated circuit

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5410592A (en) * 1993-06-04 1995-04-25 Harris Corporation Class `D` audio speaker amplifier circuit with state variable feedback control
US5672998A (en) * 1995-08-09 1997-09-30 Harris Corporation Class D amplifier and method
WO1998019391A2 (en) * 1996-10-31 1998-05-07 Bang & Olufsen A/S Pulse modulation power amplifier with enhanced cascade control method
JP3346581B2 (en) * 1997-04-02 2002-11-18 バング アンド オルフセン・パワーハウス・エイ／エス Pulse reference control method and system for enhanced power amplification of pulse modulated signal
DE69818631T2 (en) * 1997-08-12 2004-08-12 Koninklijke Philips Electronics N.V. REINFORCING ARRANGEMENT FOR DIGITAL SIGNALS
US5959501A (en) * 1998-01-14 1999-09-28 Harris Corporation Class D amplifier with scaled clock and related methods
FI107659B (en) * 1999-11-26 2001-09-14 Nokia Mobile Phones Ltd Signal quality improvement
JP4387601B2 (en) * 2000-02-01 2009-12-16 アルパイン株式会社 Digital amplifier
US6466087B2 (en) * 2000-12-28 2002-10-15 Nokia Mobile Phones, Ltd. Method and apparatus providing digital error correction for a class D power stage

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102484458A (en) * 2009-08-26 2012-05-30 松下电器产业株式会社 Sensor device
CN102484458B (en) * 2009-08-26 2014-12-24 松下电器产业株式会社 Sensor device
US9287837B2 (en) 2009-08-26 2016-03-15 Panasonic Intellectual Property Management Co., Ltd. Sensor device

Also Published As

Publication number Publication date
CN1790899A (en) 2006-06-21
CN100461626C (en) 2009-02-11
JP2004128958A (en) 2004-04-22
CN1790898A (en) 2006-06-21
JP3776392B2 (en) 2006-05-17

Similar Documents

Publication Publication Date Title
US6924700B2 (en) Class D amplifier
CN101459411B (en) Class-D amplifier
US11431310B2 (en) Tracking and correcting gain of open-loop driver in a multi-path processing system
US20060103362A1 (en) Frequency stabilization technique for self oscillating modulator
CN101150281A (en) Method and device for generating impulse bandwidth modulation signals and its application
CN101238424A (en) Switching regulator with variable slope compensation
CN101267188A (en) Digital input class-D amplifier
WO2020147637A1 (en) Reference voltage generation circuit and switched-mode power supply
US6067327A (en) Data transmitter and method therefor
CN100486111C (en) Class d amplifier
CN107707210B (en) Self-zeroing high-gain differential amplifying circuit
CN104868701A (en) Hybrid compensating circuit of power supply converter
WO2022170645A1 (en) Low-distortion class d power amplifier applied to high-power range
CN212305266U (en) Filter capacitor current feedback self-oscillation audio digital power amplifier
JP2000009409A (en) Circuit for detecting variation of inductance
TW405299B (en) A cmos soft clipper
US9634683B1 (en) Low power sigma-delta modulator architecture capable of correcting dynamic range automatically, method for implementing low power circuit thereof, and method for correcting and extending dynamic range of sigma-delta modulator automatically
CN112135390A (en) Dimming circuit and power supply chip
CN103683929A (en) Self-adaptive loop compensating method, compensating circuit and switching power supply with compensating circuit
TWI762258B (en) Analog front end with pulse width modulation current compensation
WO2022021505A1 (en) Digital power amplifier for current feedback self-excited oscillation audio of filter capacitor
JP2012186715A (en) Signal converter and signal conversion method
CN110081991A (en) A kind of small several times signal amplifying apparatus can be used for temperature sensor and method
JP2004179998A (en) Preamplifier
CN104365018A (en) Switched amplifier for variable supply voltage

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20090506

Termination date: 20140930

EXPY Termination of patent right or utility model