JP5266830B2 - Self-excited class D amplifier - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a clip detecting circuit capable of detecting the occurrence of a clip in a self-exciting class D amplifier with a simple circuit configuration. <P>SOLUTION: A clip detecting part includes: a positive pulse width measuring part 20 including a capacitor 24 to be charged when a PWM signal output from the switch circuit of the amplifier is on a high side; a positive clip detection signal generating part (a comparator) 21 for comparing the output voltage Vp of the positive pulse width measuring part 20 with a threshold voltage Vtp; a negative pulse width measuring part 30 including a capacitor 34 to be charged when the PWM signal is on a low side; and a negative clip detection signal generating part (a comparator) 31 for comparing the output voltage Vm of the negative pulse width measuring part 30 with a threshold voltage Vtm. The threshold voltages Vtp, Vtm are set in prescribed time within the range of 10 to 50 times as large as a period of a self-exciting oscillation frequency in a pulse width modulation signal when the level of an analog acoustic signal input to the amplifier is zero. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

The present invention relates to a self-excited class D amplifiers, and more particularly, about the technology for detecting Oite clip to the self-excited class D amplifier.

The self-excited class D amplifier is a device that amplifies power by pulse width modulation of an input analog acoustic signal. As a conventional self-excited class D amplifier, for example, an integration circuit, a comparator, a switch circuit, and a low-pass filter, a first feedback circuit that feeds back an output of the low-pass filter to an input of the comparator, and a low-pass There has been a self-excited class D amplifier provided with a second feedback circuit that feeds back the output of the filter to the input of the integrator (see Patent Document 1 below). In the self-excited class D amplifier disclosed in Patent Document 1, the first feedback circuit mainly determines the self-excited oscillation frequency, and the second feedback circuit determines the gain and frequency characteristics of the entire amplifier. By providing the two feedback circuits of these first and second feedback circuits, the influence and distortion components due to fluctuations in the power supply voltage can be suppressed.
Japanese Examined Patent Publication No. 61-021007

  In an amplifier for an acoustic signal, there is a problem that “clip (distortion)” occurs in a signal after power amplification. “Clip” refers to the signal (output signal) after power amplification when the level of the input acoustic signal is too high and the level of the amplified signal is greater than the power supply voltage used for amplification. This is a phenomenon in which the upper and lower peaks of the waveform are limited to the power supply voltage and the upper and lower peaks of the waveform are crushed (the output signal “sticks” in the vicinity of the power supply voltage). When the clip is generated, for example, the sound quality of the acoustic signal output from the speaker is deteriorated, and there is a problem that a risk of generating an unpleasant sound is generated.

As a clip prevention technique applied to class D amplifiers, for example, based on a pulse width modulation signal (PWM signal) output from a comparator that performs pulse width modulation, the occurrence of a clip is detected in the pulse width modulation, and the clip is detected. A technology for switching the integration constant of the integration means provided in the preceding stage of the comparator to a constant having a lower order than the normal time (see Patent Document 2 below), and the maximum level allowed as an input signal based on the power supply voltage There has been a technique (see Patent Document 3 below) that performs control to attenuate the level of the input acoustic signal when the determined and input acoustic signal is greater than the determined maximum level.
JP 2006-50589 A JP 2006-180199 A

  In the above-mentioned Patent Document 2, the generation of a clip is determined based on a comparison between the period of a triangular wave used for pulse width modulation and the pulse width of a pulse width modulation result signal (PWM signal) as a configuration for determining the occurrence of a clip. A configuration for determining is disclosed. However, this clip determination configuration is premised on application to a “separately excited class D amplifier” having a clock source for generating a triangular wave, and is applied to a self-excited class D amplifier having no clock source. I couldn't.

  In the technique described in Patent Document 3, a power supply voltage detection process, a process for determining a maximum level based on the detected power supply voltage, and a process for comparing the determined maximum level with the level of an input signal For example, various arithmetic processes are required. Therefore, a configuration for performing these calculations, that is, components such as a signal processing device (DSP) and an AD converter must be installed, and the device configuration becomes complicated. It is unreasonable from the viewpoint of cost, circuit scale, etc., to install a DSP, an AD converter, or the like for clip detection in an amplifier such as an audio amplifier that does not necessarily include a DSP.

  Further, the above-mentioned Patent Document 1 has a problem that the self-excited class D amplifier has a problem that the duty ratio of the PWM signal changes due to the change of the output level and the self-excited oscillation frequency changes. Since the self-excited class D amplifier has the above characteristics, when the output level rises, the self-excited oscillation frequency decreases, and in a state where the level rises as clipping occurs in the output level, the peak of the output level It has been pointed out that there is a problem that the self-excited oscillation frequency decreases to zero (direct current). Patent Document 1 proposes the idea of limiting the fluctuation range of the self-excited oscillation frequency by limiting the output level of the integrator in the input stage as a solution to the problem. However, here, there is no criterion for determining the lower limit of the self-excited oscillation frequency, in other words, the criterion for determining the occurrence of clipping. Further, a specific circuit configuration for making the determination is not described.

The present invention has been made in view of the above, and aims to provide a simple circuit configuration, a self-excited D-class power amplifying apparatus that can be detected clips generated in self-excited D-class power amplifier To do.

The present invention generates a pulse width modulated signal to the input signal by pulse width modulation, the self-excited class D amplifier to obtain an output signal of the pulse width modulated signal to the switching amplifier, one of the self-excited class D amplifier Measuring means for inputting a pulse width modulation signal extracted from the stage and measuring the time until the input pulse width modulation signal is inverted from at least one of the high side and the low side to the other; and the measuring means Generating means for generating a clip detection signal indicating the occurrence of a clip when the time measured by the above exceeds a predetermined threshold, wherein the predetermined threshold is the pulse when the level of the input signal is zero and generating means is set to a predetermined time in the self-oscillation range 10 times to 50 times the period of the frequency of the width modulated signal, during the duration of the clip detection signal It is a self-excited class D amplification device characterized in that it comprises a damping means for attenuating the input signal by the attenuation amount corresponding to the length.

According to the above configuration, the pulse width modulated signal taken from one of the stages of the self-excited class D amplifier measures the time from at least one of high-side and low-side until inverted to the other by meter measuring means, When the measured time exceeds a predetermined time (predetermined threshold) within a range of 10 to 50 times the period of the self-excited oscillation frequency of the pulse width modulation signal when the level of the input signal is zero, A clip detection signal indicating the occurrence can be generated, and the input signal can be attenuated with a larger attenuation amount as the duration of the clip detection signal is longer.

Also, the self-excited class D amplification device of the present invention, the in generating means, the predetermined threshold, at the level of the input signal over 10 times the period of self-oscillation frequency of the pulse width modulated signal at zero And can be configured to be set to a predetermined time of 50 μsec or less.

According to the above configuration, the time until the pulse width modulation signal is inverted from at least one of the high side and the low side to the other is the self-oscillation frequency of the pulse width modulation signal when the level of the input signal is zero. When a predetermined time within a range of 10 times or more and 50 μsec or less is exceeded, it is determined that clipping has occurred in the output signal or that the output signal has become close to the clip.

Further, Oite the self-excited class D amplifier circuit of the present invention, the measuring means, the period until the pulse width modulation signal is inverted from at least one of high-side and low-side to the other, it is charged with a predetermined charging voltage A measuring unit that includes a capacitor and outputs a voltage across the terminals of the capacitor, and the generating unit includes a comparator that compares the output voltage of the measuring unit with a threshold voltage corresponding to the predetermined threshold, It may be a generating means for generating the clip detection signal based on the result.

  According to the above configuration, the measuring unit may be configured by a charging circuit including a capacitor that is charged with a predetermined charging voltage for a period until the pulse width modulation signal is inverted from at least one of the high side and the low side to the other. In addition, the generation means can be constituted by a voltage comparison circuit including a comparator that compares the output voltage of the measurement means with a threshold voltage corresponding to the predetermined threshold. During the period in which one of the high-side or low-side pulses continues, the capacitor of the measuring unit is charged, and when the period in which one of the high-side or low-side pulses continues exceeds a predetermined threshold, the output voltage of the measuring unit is When the threshold voltage is exceeded, the generating means generates a clip detection signal.

Also, the self-excited class D amplification device of the present invention, from a subsequent stage of the switching amplifier can be configured to retrieve the input to the clip detection circuit.

According to this inventions, the self-in-excited D-class amplifier, a predetermined time (a predetermined threshold value in the range 10 to 50 times of the period of self-oscillation frequency of the pulse width modulation signal when the level of the input signal of zero ) Is used as a criterion for detection, a clip in the output signal of the amplifier or a state close thereto is detected. Here, the self-excited oscillation frequency when the level of the input signal is zero is the upper limit of the fluctuation range of the self-excited oscillation frequency, and the lower limit of the fluctuation range of the self-excited oscillation frequency is for detecting the clip. It is defined by setting a predetermined threshold value that is a criterion. The fluctuation range of the self-excited oscillation frequency corresponds to the dynamic range of the input signal of the self-excited class D amplifier. Therefore, in accordance with the present invention, by the level of the input signal for a predetermined time in the self-oscillation range 10 times to 50 times the period of the frequency at the zero, it is set as the threshold value of the clip discovery, self-excited class D While securing a sufficient dynamic range of the input signal to the amplifier, it is possible to detect an excessive input as a clip or a state close to it, and suppress the clip by attenuating the input signal with an attenuation amount according to the duration of the clip detection signal There is an excellent effect of being able to.

  Further, if the threshold condition for determining clip detection is further limited, a predetermined time that is 10 times or more of the period of the self-excited oscillation frequency when the level of the input signal is zero and 50 μsec or less is set to the threshold value. By setting as, in addition to the above-described effect, an excellent effect can be obtained that the self-oscillation frequency can be prevented from falling to the audible band.

Further, the measuring means is constituted by a charging circuit including a capacitor, and the generating means is constituted by a comparison circuit including a comparator, whereby the threshold value is determined by a combination of the charging time constant of the capacitor, the charging voltage, and the threshold voltage of the comparator. It can be determined, an excellent effect that becomes very simple circuit configuration.

Also, by configuring so as to take out the pulse width modulation signal input from the subsequent switching amplifier, an excellent effect that there is little adverse effect on the self-oscillation frequency.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

Figure 1 is a block diagram showing an outline of a configuration example of a self-excited class D amplifier that written to the present invention. In FIG. 1, a class D amplifier includes an input stage amplifier 1, a comparator 2, a switch circuit 3, a low-pass filter 4, a first feedback circuit 5 that negatively feeds back the output of the low-pass filter 4 to the comparator 2, and a low-pass filter. A second feedback circuit 6 that negatively feeds back the output of the filter 4 to the input stage amplifier 1, a clip detection unit 7, and a compressor 8 is configured.

  The clip detection unit 7 is a clip detection circuit for detecting a clip of the self-excited class D amplifier. The clip detection unit 7 measures the time until the pulse width modulation signal extracted from the subsequent stage of the switch circuit 3 is inverted from at least one of the high side and the low side to the other, and the measured time is set to a predetermined threshold value. When exceeded, a clip detection pulse signal indicating the occurrence of a clip is output. As will be described in detail later, the present invention is characterized in that the clip detection unit 7 is configured by a simple circuit such as a charge time constant and a voltage comparison, and a threshold value used as a criterion for determining the occurrence of a clip in the clip detection unit 7 There is a feature in the setting.

  Reference numerals 1 to 6 are main components of the self-excited class D amplifier. The point of operation of the class D amplifier is that an analog acoustic signal (input signal) input from the input terminal 10 is converted into a pulse width modulation signal by the comparator 2, and the switch circuit 3 is controlled on / off by the pulse width modulation signal. The PWM signal output from the switch circuit 3 is demodulated into an analog acoustic signal by the low-pass filter 4, and the analog acoustic signal (output signal) output from the low-pass filter 4 is output to a load such as a speaker. .

  The input stage amplifier 1 includes a differential amplifier 9 and a capacitor C1 inserted between the output terminal and the negative input terminal of the differential amplifier 9. The positive input terminal of the differential amplifier 9 is connected to the output terminal of the compressor 8 inserted before the input stage amplifier 1, and the negative input terminal of the differential amplifier 9 is connected to the low-pass filter via the second feedback circuit 6. 4 output terminals.

  An input stage output signal (input signal) output from the input stage amplifier 1 is input to the positive input terminal of the comparator 2. A feedback signal extracted from the output of the low-pass filter 4 is input to the negative input terminal of the comparator 2 via the first feedback circuit 5. The comparator 2 compares the input stage output signal of the input stage amplifier 1 input from the positive input terminal and the feedback signal input from the negative input terminal, and generates a pulse width modulation signal (PWM signal) based on the comparison result. To do. The comparator 2 outputs a high-side (positive voltage value) pulse when the input stage output signal is larger than the feedback signal, and outputs a low-side (negative voltage value) pulse when the input stage output signal is smaller than the feedback signal. Output. By converting the input signal into a PWM signal by the comparator 2, the level information represented by the amplitude of the input signal is converted into time information represented by the pulse width of the PWM signal.

  The comparator 2 is configured as a hysteresis comparator having hysteresis characteristics by feeding back the signal at the output terminal to the positive input terminal (by applying positive feedback). Therefore, a voltage difference is generated by an offset Δ between the voltage value when the pulse of the PWM signal output from the comparator 2 is inverted from positive to negative and the voltage value when the pulse is inverted from negative to positive. Specifically, for example, assuming a state in which a high-side pulse (positive voltage value) is output from the comparator 2, the input signal> the feedback signal in that state. From this state, when the voltage value of the feedback signal at the negative input terminal gradually increases, the PWM signal output from the comparator 2 becomes low-side when the voltage value of the feedback signal at the negative input terminal exceeds a certain value. Invert to pulse. Thereafter, when the voltage value of the feedback signal at the negative input terminal gradually decreases, the output of the comparator 2 is inverted to a high-side pulse when the voltage value of the feedback signal at the negative input terminal falls below a certain value. . At this time (when the PWM signal is inverted from the low side to the high side), the voltage value of the feedback signal is lower by the offset Δ than the voltage value at the time when the PWM signal is inverted from the high side to the low side first. Value. As is well known, it is possible to prevent the output from becoming unstable due to the output signal of the comparator 2 being inverted due to a slight voltage difference such as noise.

  The PWM signal output from the comparator 2 is input to the switch circuit 3. The switch circuit 3 in the output stage includes two switching elements having a push-pull configuration and a positive power source and a negative power source (not shown) that supply a power source voltage to this stage. The switching element is generally composed of a transistor or FET. The switch circuit 3 alternately turns on and off the two switching elements at intervals according to the pulse width of the input PWM signal, thereby outputting a PWM signal that has been power-amplified using the power supplied by the power supply.

  A low-pass filter (LC filter) 4 including an inductor L 1 and a capacitor C 2 is connected to the subsequent stage of the switch circuit 3, and the PWM signal output from the switch circuit 3 is input to the low-pass filter 4. The low-pass filter 4 removes the high frequency component of the PWM signal after power amplification output from the switch circuit 3, and demodulates the analog acoustic signal from the PWM signal. When the level of the input signal of the amplifier is zero, the duty ratio of the PWM signal (ratio of the pulse width to one cycle) is 50%, and the level of the output signal of the low-pass filter 4 is also zero. When an acoustic signal is input to the amplifier, the pulse width of the PWM signal changes according to the level of the input acoustic signal. As a result, the level of the output signal of the low-pass filter 4 is also proportional to the time of the pulse width. Change. If the high-side pulse width is long with respect to one period of the PWM signal, the output signal of the low-pass filter 4 will swing to the positive side, and if the low-side pulse width is long with respect to one period of the PWM signal, the low-pass filter 4 The output signal of the signal amplitudes to the negative side. The output signal (analog acoustic signal) of the low-pass filter 4 is output to a load such as a speaker connected to the output terminal 11.

  The output signal of the low-pass filter 4 is input to the negative input terminal of the comparator 2 via the first feedback circuit 5. The first feedback circuit 5 includes a delay circuit including a resistor R1 and a capacitor C3. The first feedback circuit 5 functions as negative feedback for the PWM modulation in the self-excited oscillation method in the comparator 2. The self-excited oscillation frequency of the PWM signal is determined by the time delay of the first feedback circuit 5.

  The output signal of the low-pass filter 4 is input to the negative input terminal of the differential amplifier 9 via the second feedback circuit 6. The second feedback circuit 6 includes resistors R2 and R3. That is, a speaker output signal (output signal of the low-pass filter 4) and a voltage obtained by dividing the ground by the resistors R2 and R3 are input to the negative input terminal of the differential amplifier 9. The second feedback circuit 6 determines the gain and frequency characteristics of the entire class D amplifier including the feedback circuits 5 and 6 from the input stage amplifier 1. The second feedback circuit 6 functions as negative feedback for making the output voltage to the speaker proportional to the input voltage to the positive input terminal of the differential amplifier 9.

  The configuration of the self-excited class D amplifier including the input stage amplifier 1, the comparator 2, the switch circuit 3, the low-pass filter 4, the first feedback circuit 5, and the second feedback circuit 6 is a conventionally known configuration. For example, it is disclosed by the said patent document 1.

  In the self-excited class D amplifier of FIG. 1, a compressor 8 is inserted before the input stage amplifier 1. The compressor 8 is a circuit that suppresses the dynamic range of the input acoustic signal, and has an effect of adjusting the level of the input acoustic signal based on a dynamically changing current gain (gain attenuation effect). .

  As an example, the compressor 8 has five types of parameters: a gain parameter, a threshold parameter, a ratio parameter, an attack parameter, and a release parameter. The gain parameter is a parameter indicating the gain value of the compressor 8 when the gain attenuation effect of the signal level output from the compressor 8 is not performed. The threshold parameter is a parameter indicating a level serving as a threshold value at which the gain attenuation effect by the compressor 8 starts to be applied. The ratio parameter is a parameter indicating the rate of gain attenuation applied to the level of the input signal exceeding the threshold. The attack parameter is a parameter indicating how much time delay the gain attenuation effect starts after the level of the input signal exceeds the threshold value. The release parameter is a parameter indicating how much time delay the gain attenuation effect is terminated from the time when the level of the input signal becomes smaller than the threshold value.

  The compressor 8 amplifies the level of the input acoustic signal with a current gain corresponding to the set value of the gain parameter during normal times (when no gain attenuation effect is applied). When the level of the input signal exceeds the threshold parameter value (threshold), the value of the current gain is attenuated by the amount by which the level exceeds the threshold multiplied by the ratio parameter setting value. The level of the output signal of the compressor 8 is suppressed. The reaction speed from when the level of the input signal exceeds the threshold value until the gain attenuation effect corresponding to the amount exceeding the threshold value occurs depends on the set value of the attack parameter. When the level of the input signal becomes smaller than the threshold value, the gain attenuation effect is terminated and the normal operation is restored. The reaction speed until the gain attenuation effect of the current gain ends (returns to the normal gain corresponding to the gain parameter) after the level falls below the threshold value depends on the set value of the release parameter. Note that, according to the general parameter setting of the compressor 8, the setting value of the attack parameter is set to a value larger than the setting value of the release parameter.

  The compressor 8 provided in the self-excited class D amplifier in FIG. 1 represents the pulse width of the clip detection pulse signal when the clip detection pulse signal is input from the clip detection unit 7 in addition to the above general operation. Control is performed to attenuate the value of the current gain according to the time length. That is, the compressor 8 functions to suppress a clip generated at the output stage by reducing the signal level at the input stage when the clip detection unit 7 detects the occurrence of a clip. Specifically, while the clip detection pulse signal indicates the occurrence of a clip, the current gain of the compressor is attenuated at a predetermined first rate. The current gain is returned to the normal gain (gain parameter value) at the second rate. The first rate may be the same as the attack rate described above, and the second rate may be the same as the release rate described above.

  FIG. 2 is a block diagram illustrating an example of a detailed configuration of the clip detection unit 7. The clip detection unit 7 is roughly divided into a positive side pulse width measurement unit 20 and a positive side clip detection that detects the occurrence of a clip based on the measurement result of the positive side pulse width measurement unit 20 and outputs a positive side clip detection signal. Signal generation unit 21, negative side pulse width measurement unit 30, and negative side clip detection signal generation unit that detects the occurrence of a clip based on the measurement result of negative side pulse width measurement unit 30 and outputs a negative side clip detection signal 31.

  As shown in FIG. 1, in this embodiment, the extraction point of the signal input to the clip detection unit 7 is the output point of the switch circuit 3 (immediately before the low-pass filter 4). The PWM signal output from the switch circuit 3 is input to the input terminal 40 of the clip detection unit 7. In the self-excited class D amplifier, depending on the extraction point of the signal input to the clip detection unit 7, the signal extraction affects the delay time at that point, and the self-oscillation frequency changes accordingly. There is a risk. A configuration in which a signal is extracted immediately before the low-pass filter 4 as in this embodiment is preferable in that the adverse effect on the self-excited oscillation frequency is small.

  The positive pulse width measuring unit 20 includes a first diode 22, a first transistor 23, a first capacitor 24, and a negative power source 25 connected via a pull-up resistor R7. The clip detecting unit 7 The pulse width of the high-side pulse (positive voltage value) of the PWM signal (the output signal of the switch circuit 3) input to is measured. The PWM signal input terminal 40 is connected to the anode terminal of the first diode 22 via the resistor R4, and the cathode terminal of the first diode 22 is connected to the ground serving as the reference voltage (0 V). The first diode 22 constitutes a circuit for extracting a low-side pulse (negative voltage value) of the PWM signal.

  The first transistor 23 is a PNP transistor. The base terminal of the first transistor 23 is connected to the anode terminal of the diode 22 via a base resistor R5, and a low-side pulse of the PWM signal is input from the base terminal. The collector terminal of the first transistor 23 is connected to the power supply line of the negative power supply 25, and the emitter terminal of the transistor 23 is connected to the ground. The first transistor 23 is a switching element that is turned on / off according to the PWM signal. The first transistor 23 is turned on when the low-side pulse extracted by the first diode 22 is input to the base terminal, and the PWM signal is high. In the case of a side pulse, it is off.

  The first capacitor 24 is inserted between the collector and the emitter of the first transistor 23, one terminal is connected to the ground, and the other terminal is connected to the power supply line of the negative power supply 25. A resistor R6 is connected to the first capacitor 24 in parallel. The negative power supply 25 supplies a charging voltage for the first capacitor 24. The inter-terminal voltage Vp of the first capacitor 24 is within a predetermined negative voltage value range that the negative power source 25 supplies from the ground level (0 V) according to the length of time that the first capacitor 24 is charged by the negative power source 25. It changes with. This inter-terminal voltage Vp corresponds to the output voltage of the positive pulse width measuring unit 20 (inter-terminal voltage of the resistor R6).

  The positive clip detection signal generator 21 is configured by a comparator connected to the subsequent stage of the first capacitor 24. The negative input terminal of the comparator 21 is connected to the first capacitor 24 at the connection point P, and the output voltage Vp extracted from the connection point P is input to the negative input terminal. A predetermined threshold voltage Vtp is input to the positive input terminal of the comparator 21. The threshold voltage Vtp is a negative voltage of a predetermined level given from a negative power supply (not shown), and is one element that determines a threshold serving as a reference for determining the occurrence of clipping. The comparator 21 compares the output voltage Vp input from the negative input terminal with the threshold voltage Vtp input from the positive input terminal, and outputs a pulse signal corresponding to the comparison result. That is, when the output voltage Vp is lower than the threshold voltage Vtp (when the absolute value Vp is larger than the absolute value Vtp), it is determined that the occurrence of clipping in the high-side pulse of the PWM signal is detected. A high-side pulse (“1” in binary notation) corresponding to the side clip detection signal is output. On the other hand, when the output voltage Vp is higher than the threshold voltage Vtp (when the absolute value Vp is smaller than the absolute value Vtp), a low-side pulse (“0” in binary notation) is output.

  The negative side pulse width measuring unit 30 includes a second power source 32 connected via a second diode 32, a second transistor 33, a second capacitor 34, and a pull-up resistor R11. The pulse width of the low-side pulse (negative voltage value) of the PWM signal (the output signal of the switch circuit 3) input to is measured. The PWM signal input terminal 40 is connected to the cathode terminal of the second diode 32 via the resistor R8, and the anode terminal of the second diode 32 is connected to the ground. The second diode 32 constitutes a circuit that extracts a high-side pulse (positive voltage value) of the PWM signal.

  The second transistor 33 is composed of an NPN transistor. The base terminal of the first transistor 23 is connected to the cathode terminal of the diode 22 via a base resistor R9, and a high-side pulse of the PWM signal is input from the base terminal. The collector terminal of the second transistor 33 is connected to the power supply line of the positive power supply 35, and the emitter terminal of the second transistor 33 is connected to the ground. The second transistor 33 is a switching element that is switched on / off according to the PWM signal. The second transistor 33 is turned on when the high-side pulse extracted by the second diode 32 is input to the base terminal, and the PWM signal is In the low-side pulse state, it is in the off state.

  The second capacitor 34 is inserted between the collector and the emitter of the second transistor 33, one terminal is connected in a round, and the other terminal is connected to the power supply line of the positive power supply 35. A resistor R10 is connected to the second capacitor 34 in parallel. The positive power source 35 supplies a charging voltage for the second capacitor 34. The voltage Vm between the terminals of the second capacitor 34 is within a predetermined positive voltage value range that the positive power supply 35 supplies from the ground level (0 V) according to the length of time that the second capacitor 34 is charged by the positive power supply 35. It changes with. This inter-terminal voltage Vm corresponds to the output voltage of the negative side pulse width measuring unit 30 (inter-terminal voltage of the resistor R10).

  The negative clip detection signal generation unit 31 is configured by a comparator connected to the subsequent stage of the second capacitor 34. The positive input terminal of the comparator 31 is connected to the second capacitor 33 at the connection point M, and the output voltage Vm extracted from the connection point M is input to the positive input terminal. A predetermined threshold voltage Vtm is input to the negative input terminal of the comparator 31. The threshold voltage Vtm is a positive voltage of a predetermined level given from a positive power source (not shown), and is one element that determines a threshold value serving as a reference for determining the occurrence of clipping. The comparator 31 compares the output voltage Vm input from the positive input terminal with the threshold voltage Vtm input from the negative input terminal, and outputs a pulse signal corresponding to the comparison result. That is, when the output voltage Vm is higher than the threshold voltage Vtm (when the absolute value Vm is larger than the absolute value Vtm), it is determined that the occurrence of clipping in the low-side pulse of the PWM signal is detected, and the negative side A high-side pulse (“1” in binary notation) corresponding to the clip detection signal is output. On the other hand, when the output voltage Vm is lower than the threshold voltage Vtm (when the absolute value Vm is smaller than the absolute value Vtm), a low-side pulse (“0” in binary notation) is output.

  The output terminals of the comparator 21 and the comparator 31 are connected to the two input terminals of the OR circuit 41, respectively. The OR circuit 41 outputs a voltage value (“1” in binary notation) corresponding to the clip detection pulse signal to the compressor 8 (see FIG. 1) during the period when the output of either the comparator 21 or the comparator 31 is “1”. In other cases, a voltage value corresponding to “0” is output in binary.

  In the clip detection unit 7 configured as described above, when the input PWM signal is on the high side, the second transistor 33 is turned on and the first transistor 23 is turned off. In this state, the first capacitor 24 of the positive pulse width measuring unit 20 is charged with a negative power supply voltage supplied by the negative power supply 25. The first capacitor 24 is charged by the negative power source 25 during a period from when the high-side pulse of the PWM signal rises to when the pulse is inverted to the low side. Therefore, during the period in which the high-side pulse of the PWM signal continues, the level of the output voltage Vp, which is the voltage between the terminals of the first capacitor 24, decreases to the negative side (the absolute value of the output voltage Vp increases). If the period during which the high-side pulse of the PWM signal lasts continues for a certain time or more, the output voltage Vp becomes lower than the threshold voltage Vtp (the absolute value Vp exceeds the absolute value Vtp). Therefore, when the high-side pulse of the PWM signal continues for a certain period of time corresponding to the threshold voltage Vtp, the comparator 21 (positive clip detection signal generation unit) causes the high-side pulse (in binary notation, “1”). )) Is output. On the other hand, in this state, since the potential of the second capacitor 34 is equal to the ground, the output voltage Vm at the connection point M is at the ground level (0 V).

  When the input PWM signal is on the low side, the first transistor 23 is turned on and the second transistor 33 is turned off. In this state, the second capacitor 34 of the negative side pulse width measuring unit 30 is charged by the positive power supply voltage supplied from the positive power supply 35. The second capacitor 34 is charged by the positive power source 35 during a period from when the low-side pulse of the PWM signal falls to when the pulse is inverted to the high side. Therefore, during the period in which the low-side pulse of the PWM signal continues, the level of the output voltage Vm that is the voltage across the second capacitor 34 increases to the positive side (the absolute value of the output voltage Vm increases). If the duration of the low-side pulse of the PWM signal continues for a certain time or longer, the output voltage Vm becomes higher than the threshold voltage Vtp (the absolute value Vm exceeds the absolute value Vtp). Therefore, when the low-side pulse of the PWM signal continues for a certain fixed time corresponding to the threshold voltage Vtm, the comparator 31 (negative-side clip detection signal generation unit) causes the high-side voltage value (in binary notation, “1”). )) Is output. On the other hand, in this state, since the potential of the first capacitor 24 is equal to the ground, the output voltage Vp at the output point p is at the ground level (0 V).

  While the high or low pulse of the PWM signal output from the switch circuit 3 continues for “a certain time” or longer, the positive clip detection signal generator (comparator) 21 or the negative clip detection signal generator (comparator) 31. Outputs “1”, so that the OR circuit 41 outputs the clip detection pulse signal (“1” in binary notation) while the high or low pulse of the PWM signal continues for a certain period of time. Output to the compressor 8. The compressor 8 reduces the level of the input signal by performing control to attenuate the current gain while the clip detection pulse signal is input from the clip detection unit 7. The longer the time represented by the pulse width of the clip detection pulse signal is (that is, the higher the level of the speaker output signal), the more the compressor 8 attenuates the current gain.

  In a normal state where no clipping occurs in the PWM signal, the pulse of the PWM signal starts from the high side before the output voltage Vp input to the positive clip detection signal generation unit (comparator) 21 falls below the threshold value Vtp. Inverts to the low side, and also reverses from the low side to the high side before the absolute value of the output voltage Vm input to the negative side clip detection signal generator (comparator) 31 exceeds the threshold value Vtm. The output of either the unit (comparator) 21 or the negative clip detection signal generation unit (comparator) 31 remains low. Therefore, the OR circuit 41 outputs “0”.

  In the clip detection unit 7, the “certain time” that is a criterion for detecting the occurrence of clip is the charging time constant of the capacitors 24, 34, the charging voltage supplied from the power sources 25, 35, and the threshold voltages Vtp, Vtm. It depends on the combination. The “predetermined time” of the clip generation determination criterion is used to determine how long at least one of the high side and the low side of the PWM signal output from the switch circuit 3 lasts will cause a clip or a state close thereto. It is an indicator.

  In the self-excited class D amplifier of FIG. 1, the self-excited oscillation frequency depends on the voltage value of the speaker output signal (acoustic signal). That is, when the duty cycle of the PWM signal output from the switch circuit 3 changes due to a change in the level of the acoustic signal input to the self-excited class D amplifier, the self-excited oscillation frequency also changes. When the level of the speaker output signal increases, the self-excited oscillation frequency decreases, and in a state where clipping occurs in the speaker output signal, the oscillation frequency approaches zero (direct current). In other words, the state where clipping occurs (the state where the level of the acoustic signal is excessive) is the upper limit of the dynamic range of the input signal of the self-excited class D amplifier, and from another point of view, the self-excited oscillation frequency is It is in a state where it has fallen to the lower limit of the allowable fluctuation range. In the self-excited class D amplifier, the self-excited oscillation frequency when the input signal level is zero is a high frequency of several hundred kHz to several MHz. From the above, it can be understood that the fluctuation range of the self-excited oscillation frequency corresponds to the dynamic range of the input signal of the self-excited class D amplifier.

  In the self-excited class D amplifier having the characteristics as described above, the clip detecting unit 7 of the present invention serves as a clip generation determination criterion for the clip detecting unit 7 so as to ensure a sufficient dynamic range of the input signal of the self-excited class D amplifier. One feature is to set a threshold value (“fixed time”). As a result of the experiment, the threshold of the clip generation criterion satisfying the requirement of sufficiently securing the dynamic range of the input signal is 10 to 10 of the period of the self-excited oscillation frequency when the level of the acoustic signal input to the amplifier is zero. It was found that a time of about 30 times is suitable. Here, as described above, the fluctuation range of the self-excited oscillation frequency defined by the above conditions corresponds to the dynamic range of the input signal of the self-excited class D amplifier. It can be secured. The threshold according to this condition is 25 μs to 75 μs. From the viewpoint of obtaining a large dynamic range, the magnification with respect to the period of the self-excited oscillation frequency when the level of the acoustic signal is zero is set to be larger than the above condition, and the time is about 50 times the period of the self-excited oscillation frequency. May be set as a threshold value for the criterion for determining the occurrence of a clip. In other words, the clip detection unit is configured so that the threshold value used as a criterion for determining the occurrence of clip in the clip detection unit 7 is a predetermined time within a range of 10 to 50 times the period of the self-excited oscillation frequency when the level of the acoustic signal is zero. The combination of the charge time constant of 7, the charge voltage, and the threshold voltage is determined.

  By the way, if the threshold value for determining the occurrence of clip is set longer than 50 μsec, the lower limit of the self-excited oscillation frequency is the human audible band in the self-excited class D amplifier of FIG. There is a problem in that there is a risk of falling inside. In view of this point, more preferably, the threshold value of the criterion for determining the occurrence of clipping is any value within the range of 10 times or more of the period of the self-excited oscillation frequency when the level of the input signal is zero and 50 μsec or less. It can be said that it is better to set the time. In order to further improve safety, it is preferable to secure a margin between the upper limit of the audible band and set a time within a range of 25 μsec to 30 μsec as a threshold. When the “certain time” is set to 30 μsec, the lower limit of the fluctuation range of the self-excited oscillation frequency is 33 kHz, so there is no risk that the lower limit of the self-excited oscillation frequency falls to the human audible band. Also, under this condition, since the period of the self-excited oscillation frequency when the level of the input signal is zero is secured 10 times or more, the dynamic range of the input signal can be secured.

  FIG. 3 is a waveform diagram for explaining the operation of the clip detection unit 7 and the operation of the compressor 8. In the figure, a is an output signal (analog acoustic signal) output from the amplifier, b is a PWM signal output from the switch circuit 3, c is a voltage (output voltage) Vp between terminals of the first capacitor 24, and d is a first signal. Terminal voltages (output voltages) Vm and e of the two capacitors 34 are clip detection pulse signals (output signals of the clip detection unit 7). Further, f is an input signal (analog acoustic signal) input to the amplifier. The PWM signal indicated by b has a wider pulse width as the level of the acoustic signal (symbols a and f) increases positively or negatively, and becomes narrower as the level of the acoustic signal is closer to zero. In the waveform of FIG. 3, the clip detection pulse signal indicated by e is a signal whose low-side falling pulse (negative voltage value) indicates the occurrence of clipping (in binary notation “1”), and the high-side This pulse (positive voltage value) is represented in binary notation and corresponds to “0”.

  In FIG. 3, for example, there is a location where a clip occurs at the positive level of the speaker output signal a (location indicated by reference numeral 42). As the high-side pulse width of the PWM signal b increases, it can be seen that the output voltage Vp indicated by the symbol c increases to the negative side. When the high-side pulse width of the PWM signal b exceeds the threshold value (“fixed time”) for determining the occurrence of clip, the output voltage Vp exceeds the threshold voltage Vtp, and the falling pulse (2 When expressed in decimal, "1") is output. Since the output of the low-side pulse of the clip detection pulse signal continues during the period in which the high-side pulse width of the PWM signal b exceeds the threshold value, the compressor 8 performs control to attenuate the current gain with respect to the input signal f. When the falling pulse of the PWM signal b enters from this state, the output voltage Vp becomes zero, so that the clip detection pulse signal e is inverted to a high-side pulse. At this time, the compressor 8 stops the current gain attenuation control based on the clip detection pulse signal e and performs control to return to the normal gain (current gain corresponding to the value of the gain parameter).

  Further, at a portion where clipping occurs at the negative level of the speaker output signal a (for example, a portion indicated by reference numeral 43), the output voltage Vm indicated by reference sign d increases as the low-side pulse width of the PWM signal b increases. Increase to the side. When the low-side pulse width of the PWM signal b exceeds the “certain time”, the output voltage Vm exceeds the threshold voltage Vtm, and a falling pulse of the clip detection pulse signal indicated by e is output. Since the output of the low-side pulse of the clip detection pulse signal continues during the period when the low-side pulse width of the PWM signal b exceeds the “certain time”, the compressor 8 controls to attenuate the current gain with respect to the input signal f. Do. When the rising pulse of the PWM signal b enters from this state, the output voltage Vm becomes zero, so that the clip detection pulse signal e is inverted to a high-side pulse. At this time, the compressor 8 stops the current gain attenuation control based on the clip detection pulse signal e and performs control to return to the normal gain.

  As a result of the current gain attenuation control by the compressor 8 in response to the above clip detection, the level of the input signal f is suppressed and the clip of the speaker output signal a in the output stage is corrected as shown by reference numerals 44 and 45 in FIG. Appears in place. In a state in which clipping does not occur in the speaker output signal a (normal state), the pulse of the PWM signal b is inverted before the output voltages Vtp and Vtm exceed the threshold voltages Vtp and Vtm. The signal e is on the high side (“1” in binary notation).

  As described above, according to the clip detection unit 7 of this embodiment, the self-excited class D amplifier has a simple circuit configuration including the charging time constants of the capacitors 24 and 34 and the comparators 21 and 31 that perform voltage comparison. It is possible to detect a clip that is generated when the level of the acoustic signal handled by is excessively increased. By setting the threshold used by the clip detection unit 7 to determine the occurrence of a clip to any predetermined time within a range of 10 to 50 times the period of the self-excited oscillation frequency when the level of the input signal is zero The dynamic range of the acoustic signal handled by the self-excited class D amplifier can be sufficiently secured. Further, the threshold for determining the occurrence of clipping is a value of 50 μsec or less within a range of 10 to 50 times the period of the self-excited oscillation frequency when the level of the input signal is zero, that is, the self-excited oscillation frequency. In addition to the above effects, the self-oscillation frequency falls to the audible band when the level of the acoustic signal rises excessively, by setting the value to 10 times or more and 50 μsec or less of the period of Can be prevented. Further, according to the configuration of the self-excited class D amplifier of FIG. 1, the extraction point of the PWM signal input to the clip detection unit 7 is the output point of the switch circuit 3 (immediately before the low-pass filter 4). Thus, the configuration in which the input signal to the clip detection unit 7 is extracted from the output stage of the self-excited class D amplifier is preferable in that the adverse effect on the self-excited oscillation frequency due to the extraction of the PWM signal is small.

  Although FIG. 2 shows an example in which the positive pulse width measuring unit 20 of the clip detecting unit 7 is configured with the negative power source 25, the positive side pulse width measuring unit 20 may be configured with a positive power source. FIG. 4 shows a configuration example when the positive pulse width measuring unit 20 is configured with a positive power source. As shown in FIG. 4, a diode 50 is inserted between the PWM signal input terminal 40 and the base terminal of the NPN transistor 51, the input terminal 40 is connected to the cathode terminal of the diode 50, and the anode of the diode 50 is also connected. Since the terminal is connected to the ground, the high-side pulse of the PWM signal is input to the base terminal of the transistor 51. The collector terminal of the transistor 51 is connected to the power supply line of the positive power supply 52, and the emitter terminal of the transistor 51 is connected to the ground. The base terminal of another NPN transistor 53 is connected to the collector terminal of the transistor 51. The collector terminal of the transistor 53 is connected to the power supply line of the positive power supply 54, and its emitter terminal is connected to the ground. A capacitor 55 is inserted between the collector and emitter of the transistor 53, and a resistor R12 is connected in parallel with the capacitor 55. The output voltage Vp corresponding to the voltage across the capacitor 55 is input to the positive input terminal of the comparator 56, and the threshold voltage Vtp is input to the negative input terminal of the comparator 56. The threshold voltage Vtp is a predetermined positive voltage value.

  According to the configuration of FIG. 4, when the high-side pulse of the PWM signal is input, the transistor 51 is turned on. At this time, the current of the positive power supply 52 flows between the collector and the emitter of the transistor 51 and does not flow to the base terminal of the transistor 53. Therefore, the subsequent transistor 53 is turned off. Therefore, when the transistor 53 is in an off state (that is, a period in which the pulse of the PWM signal is high side), the capacitor 55 is charged by the positive voltage supplied from the positive power supply 54, and the high side pulse of the PWM signal corresponds to the threshold voltage Vtp. When the comparator 56 continues for a certain period of time, the comparator 56 outputs a high-side voltage value corresponding to a clip detection signal indicating that the positive side clip occurrence has been detected.

  In the above embodiment, the pulse width is measured on both the positive and negative sides (high side and low side) of the PWM signal, and the clip is detected on both the positive and negative sides. However, in the steady state, the acoustic signal has a high side and a low side. Are often symmetric, so that the clip may be detected by measuring the pulse width only on either the high side or the low side.

  In the above-described embodiment, as one example of using the clip detection pulse signal output from the clip detection unit 7, the clip detection pulse signal is input to the compressor 8, and the compressor 8 generates an acoustic signal according to the clip detection pulse signal. Although an example in which the signal level is attenuated has been described, the usage form of the clip detection pulse signal is not limited to the above. For example, an LED lamp that informs the user of the occurrence of a clip according to the clip detection pulse signal may be lit according to the clip detection pulse signal, or as another example, a self-excited class D The amplifier may be configured to be controlled by a microcomputer, and the clip generation status notified by the clip detection pulse signal may be configured to be recorded in the operation log.

  Further, the configuration of the self-excited class D amplifier to which the clip detection unit 7 of the present invention is applied is not limited to the configuration shown in FIG. 1, and may be an appropriate configuration conventionally known. As another configuration example of the self-excited class D amplifier to which the clip detection unit 7 is applied, for example, the configuration disclosed in JP-A-2003-101357, the configuration disclosed in JP-A-2003-115730, or There is a configuration shown in 2005-123949. The self-excited class D amplifier basically includes an integrator, a comparator, an output stage (switch circuit), and a low-pass filter, and has a feedback circuit for self-excitation. In the configuration of the self-excited class D amplifier shown in the above embodiment, the configuration in which the switch circuit 3 and the low-pass filter 4 are included in the first feedback circuit 5 is shown, but the configuration of the self-excited class D amplifier is the low-pass filter 4 or The output stage (switch circuit 3) may not be included in the feedback circuit.

  In the above-described embodiment, the description has been made mainly on the detection of the clip state of the self-excited class D amplifier using the clip detection unit 7 according to the present invention. That is, the “predetermined time” (charging time constant, charging voltage, and threshold voltage), which is a threshold value for determining the occurrence of a clip set in the clip detection unit 7, detects the clip of the amplifier or a state close thereto, and inputs the amplifier It was set for the purpose of ensuring the dynamic range of the signal. However, the clip detection unit 7 according to the present invention is not limited to being used for detecting the clip state of the self-excited class D amplifier or a state close thereto, but by varying the “certain time” as a threshold value. Also, it can be used as a threshold detection of a limiter or compressor (effector having a level suppressing function) that operates with a certain output level as a threshold (threshold).

  In the above embodiment, the control for attenuating the gain value in the compressor 8 when the clip detection pulse signal is output is configured to directly control the current gain according to the clip detection pulse signal. Instead of directly controlling the gain, the threshold value of the compressor 8 may be controlled according to the clip detection pulse signal. Specifically, while the clip detection pulse signal indicates the occurrence of a clip, the threshold value of the compressor 8 is moved from a value corresponding to the threshold parameter to a lower value at a predetermined first rate to indicate no occurrence of a clip. The threshold value may be returned to a value corresponding to the threshold parameter at a predetermined second rate slower than the first rate. Even with this threshold control, the function of suppressing clipping can be achieved.

The block diagram which shows the structural example of the self-excited class D amplifier to which the clip detection part which concerns on one Example of this invention is applied. The block diagram which shows an example of a detailed structure of the clip detection part of FIG. The wave form diagram for demonstrating operation | movement of a clip detection part. FIG. 3 is a block diagram illustrating another configuration example of the clip detection unit in FIG. 2.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Integrator, 2 Comparator, 3 Switch circuit, 4 Low pass filter, 5 1st feedback circuit, 6 2nd feedback circuit, 7 Clip detection part, 8 Compressor, 9 Differential amplifier, 10 Output terminal, 20 Positive side pulse width measurement 21, positive clip detection signal generator (comparator), 22 diode, 23 transistor, 24 capacitor, 25 negative power supply, 30 negative pulse width measurement unit, 31 negative clip detection signal generator (comparator), 32 diode, 33 transistor, 34 capacitor, 35 positive power supply



Four

Claims (4)

In a self-excited class D amplifier that generates a pulse width modulation signal by performing pulse width modulation on an input signal and obtains an output signal by switching amplification of the pulse width modulation signal .
The pulse width modulation signal extracted from one of the stages of the self-excited class D amplifier is input, and the time until the input pulse width modulation signal is inverted from at least one of the high side and the low side to the other is determined. Measuring means for measuring;
When the time measured by the measuring unit exceeds a predetermined threshold, the generating unit generates a clip detection signal indicating the occurrence of a clip, and the predetermined threshold is zero when the level of the input signal is zero and generating means is set to a predetermined time of said pulse width within the self-oscillating range 10 times to 50 times the period of the frequency modulated signals,
A self-excited class D amplifier comprising attenuation means for attenuating the input signal by an attenuation amount corresponding to the duration of the clip detection signal . In the generating means, the predetermined threshold value is set to a predetermined time which is not less than 10 times the period of the self-excited oscillation frequency of the pulse width modulation signal when the level of the input signal is zero and is not more than 50 μsec. The self-excited class D amplifier according to claim 1 . The measuring means includes a capacitor charged with a predetermined charging voltage for a period until the pulse width modulation signal is inverted from at least one of the high side and the low side to the other, and outputs a voltage between terminals of the capacitor Because
The generating unit includes a comparator that compares an output voltage of the measuring unit and a threshold voltage corresponding to the predetermined threshold, and is a generating unit that generates the clip detection signal based on the comparison result. The self-excited class D amplifier according to claim 1 or 2 . Self-excited class D amplifier according to any one of claims 1 to 3, characterized in that from a subsequent stage of the switching amplifier, takes out a pre-Symbol pulse width modulated signal.

Images