JP2023054207A - Acoustic device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an acoustic device for effectively suppressing howling even with a simple configuration.

SOLUTION: An acoustic device includes: a voice input section for converting voice into an electric signal; an amplification section for amplifying the electric signal converted by the voice input section; a voice output section for converting an amplification signal amplified by the amplification section into voice; a detection section for detecting timing immediately before a howling loop oscillation when the level of the amplification signal exceeds a predetermined value for a distinction between the voice and the howling; and a condition change section for executing signal processing to disturb a howling generation condition when the detection section detects the timing immediately before the howling loop oscillation.

SELECTED DRAWING: Figure 1

COPYRIGHT: (C)2023,JPO&INPIT

Description

TECHNICAL FIELD The present invention relates to an acoustic device, and more particularly to an acoustic device capable of effectively suppressing howling with a simple configuration.

A hearing aid is an acoustic device that includes a voice input unit (microphone), an amplifier unit that amplifies the voice captured by the voice input unit, and a voice output unit (earphones, etc.) that outputs the amplified voice.

In such an acoustic device, when the amplification degree (gain) is increased (the volume is increased), howling may occur. It is generally known that howling occurs when sound output from a speaker or an earphone enters a microphone at a volume higher than a certain level. Among acoustic devices, the gain of loudspeakers is less than 20 dB, while the gain of hearing aids is relatively high. In particular, the gain of hearing aids for people with mild to moderate hearing loss is 30 dB to 80 dB, and it is difficult to suppress howling only by volume control.

In hearing aids for people with moderate to severe hearing loss, in order to isolate the microphone from the earphone in the ear, the gap between the ear and the hearing aid is filled with silicon rubber or the like to shield and isolate the sound of the earphone from entering the microphone. are doing. However, since even a small amount of sound leakage causes howling, it cannot be a perfect countermeasure.

Oscillation conditions are satisfied when the amplifier gain exceeds a certain level. Here, when the howling starts to occur, the howling stops by manually lowering the amplifier gain. After the howling stops, the amplifier gain is restored. Even if the howling is stopped manually, the howling immediately occurs, and the manual operation becomes an area where the howling cannot be done in time. In systems with high gain, processing delay time becomes a problem.

In Patent Document 1, when howling is detected, the phase is shifted such as by reversing the phase. In Patent Document 2, the amount of phase shift of a phase shift circuit is changed when howling oscillation occurs, and the phase change state at that time is maintained when howling oscillation ceases. In Patent Document 3, an input section, a conversion section, an output section, and a control section are provided, and inversion and non-inversion are alternately performed within the same path.

Here, the howling generation mechanism is considered to occur due to the multiple feedback of signals such as audio (howling signal + audio signal: the same shall apply hereinafter). As a method for suppressing such howling, a concept equivalent to that of a so-called noise canceller is used in which the howling is eliminated by inverting and adding signals such as voices.

For example, as an electrical howling prevention technique, there is a method of processing an audio signal, etc., before a delay time due to a processing circuit, and adding them together in an electronic circuit. In this addition, the phase of the howling signal portion of the voice signal is shifted by 180 degrees and then added.

A 180-degree shift of the howling signal is generated by a PLL (Phase Locked Loop) or a phase shifter (the amount of shift is a function of the frequency). That is, it is a method of eliminating howling by making the phases opposite to each other. In this case, processing delay time and howling continue to occur. Moreover, a waveform equivalent to the howling signal currently occurring cannot be created. Also, the howling frequency cannot be specified without measurement. In the phase shifter method, a delay time of at least half a cycle (180 degrees) occurs. As a result, even if the addition of the opposite phases is in the best condition, there will always be an unerased portion. This unerased component triggers new howling generation in a high-gain system. Current countermeasures set various conditions to alleviate these conditions. All of them are countermeasures using digital circuits, and it is difficult to realize them with analog circuits.

However, many of the frequency components of the human voice are formant frequency F1 components, and even when viewed from the fundamental frequencies of musical instruments, vocal music, etc., the portion with a large amplitude is a portion of 1 kHz (kilohertz) or less. If howling is caused by multiple feedback of signals such as voice, howling at a frequency lower than 1 kHz should be frequently observed. Empirically, the howling frequency is 1 kHz or more. In other words, the howling signal is considered to be different from the audio signal.

Howling is considered to be loop oscillation that is fed back with a phase difference of 0 between the output signal So fed back in the sound field space (transfer function β having filter characteristics) and the input signal Si (see FIG. 9).
In this case, the loop (howling) oscillation open-loop transfer function is
So/Si=1+βA+(βA) ² + . . . +(βA) ⁿ + .
In this function, the amplitude theoretically becomes infinite when βA≧1. This is oscillation. Actually, it becomes the allowable maximum amplitude of the elements in the system. By the way, howling occurs even if a limiter is inserted into the hearing aid circuit. Also, the howling amplitude becomes the amplitude of the limiter level.

There are no filters with selectivity within the voice band of the sound system. Therefore, it is considered that the sound field space (sound transmission path) functions as a loop filter and determines the howling frequency. For example, by setting the distance between the microphone and the headphone to about several tens of centimeters, the gain of the amplifier is increased to cause howling. At this time, it can be confirmed that the tone of the howling changes or stops when the person moves his/her hand or moves. Howling does not occur because the sound of headphones enters the microphone directly. A howling signal is a correlated specific frequency signal. It is considered that the sound from the headphones forms a standing wave of sound, including multiple reflections from objects in the sound field space, and functions as a howling frequency filter.

JP-A-2000-341787 JP-A-61-108288 JP-A-64-082891

In the technique described in Patent Document 1, the howling is suppressed by inverting the phase of the signal when the howling is detected. That is, the howling suppression technique described in Cited Document 1 is a noise canceller system itself, and the audio signal is also canceled by the phase inversion. In addition, since howling is limited by automatic gain control (AGC), it is difficult to make a sharp distinction between an audio signal and howling. Moreover, howling cannot be effectively suppressed due to the delay time due to the AGC time constant.

Further, in the technique described in Patent Document 2, the phase shift amount is changed when howling oscillation occurs, and the phase change state in which howling oscillation is eliminated is maintained. With these techniques, new howling may occur due to the delay between when howling is suppressed and when the next howling occurs.

Further, in the technique described in Patent Document 3, the signal phase is alternately inverted and non-inverted at a predetermined timing regardless of the presence or absence of howling. However, with this technique, the phase of the signal is constantly inverted and non-inverted. Therefore, if the timing of the generation of howling and the inversion of the signal phase do not match, the howling may not be suppressed.

In particular, hearing aids tend to generate howling because of their high gain, and since the earphones are worn in the ear, the howling sound is transmitted directly and is very unpleasant.

SUMMARY OF THE INVENTION An object of the present invention is to provide an acoustic device capable of effectively suppressing howling even with a simple configuration.

A hearing aid according to an aspect of the present invention includes an audio input unit that converts audio into an electrical signal, an amplifier that amplifies the electrical signal converted by the audio input unit, and an amplified signal amplified by the amplifier that is converted into audio. an audio output unit, a howling detection unit that detects the occurrence of howling when the level of the amplified signal exceeds a preset value that distinguishes the audio from the howling, and an amplification unit that detects the occurrence of howling when the howling detection unit detects the occurrence of howling. at least one of changing the phase of the howling signal that has passed through and inputting it to the audio output unit, lowering the level of the howling signal, and inputting a dummy signal instead of the howling signal to the audio output unit. and a condition changing unit that performs processing. In this hearing aid, the condition changing unit detects occurrence of howling in the howling detection unit and performs change processing, and then ends the change processing at a predetermined timing.

Here, howling is loop oscillation in which a sound field space (sound transmission path) having specific filter characteristics serves as a feedback path. The conditions for loop oscillation are (C1) that the feedback signal at the feedback point at its peak frequency is in phase with the input signal, (C2) that the gain in the open loop is 1 or more, and (C3) that the loop oscillation is Loop oscillation occurs when three conditions are met: the presence of a frequency selective element in the path. In other words, loop oscillation does not occur unless even one of these three conditions is satisfied.

The condition changing unit performs change processing to break at least one of the above three conditions (C1 to C3). That is, changing the phase of the howling signal and inputting it to the audio output unit violates the above condition (C1), and lowering the level of the howling signal violates the above condition (C2). Inputting a dummy signal instead of a howling signal to the audio output unit violates the above condition (C3).

In the configuration of the hearing aid described above, when the level of the amplified signal exceeds a preset value that distinguishes between voice and howling, the occurrence of howling is detected, and change processing for suppressing howling is immediately performed. Therefore, howling can be suppressed before the occurrence of howling is noticed.

In the hearing aid described above, when the condition changing section changes the phase of the howling signal as the changing process, as a simple configuration, an opposite-phase signal obtained by inverting the phase of the howling signal may be input to the audio output section. In this hearing aid, the condition changing section may repeat switching between phase inversion and non-inversion each time howling occurrence is detected by the howling detection section. It is difficult for humans to distinguish between inversion and non-inversion of audio signals. Therefore, by repeating the switching between the inversion and non-inversion of the audio signal each time howling is detected, it is possible to suppress the howling without causing a sense of incongruity in the audio output.

In the above hearing aid, when the level of the howling signal is lowered as the change processing by the condition changing section, the level of the amplified signal may be set to zero for a preset time. As a result, the gain in the one-loop open loop that causes howling is less than 1, and howling can be suppressed.

In the above hearing aid, when the condition changing unit inputs a dummy signal instead of a howling signal to the audio output unit for a preset time as change processing, a dummy signal having a frequency other than an audible frequency is input as the frequency of the dummy signal. do it. As a result, a dummy signal that is inaudible to humans is output from the audio output unit, and the filter characteristics of the sound field space that causes howling are changed, so that howling can be suppressed.

In the hearing aid, the audio input section, amplifier section, audio output section, howling detection section, and condition changing section can be configured by analog circuits. As a result, even an analog hearing aid configured with an analog circuit can have a howling suppressing function.

ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to provide the hearing aid which can suppress howling effectively, even if it is a simple structure.

1 is a block diagram illustrating the basic configuration of a hearing aid according to this embodiment; FIG. (a) and (b) are diagrams for explaining a level comparator included in the detection unit. (a) to (e) are schematic diagrams showing examples of waveforms. It is a schematic diagram explaining switching of a normal phase and a reverse phase. FIG. 11 is a block diagram showing a second example of a condition changer; FIG. (a) and (b) are diagrams for explaining an example of a howling detection unit for restoring the original phase at the timing when a preset time has elapsed. FIG. 11 is a block diagram showing a third example of a condition changer; FIG. 11 is a block diagram showing a fourth example of a condition changing unit; 1 is a block diagram of an electronic circuit and sound field space; FIG.

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings. In the following description, the same members are denoted by the same reference numerals, and the description of members that have already been described will be omitted as appropriate.

(Hearing aid configuration)
FIG. 1 is a block diagram illustrating the configuration of a hearing aid according to this embodiment.
A hearing aid 1 according to this embodiment includes an audio input unit 10 , an amplifier unit 20 , an audio output unit 30 , a howling detection unit 50 and a condition change unit 60 .

The hearing aid 1 is used by inserting an audio output unit 30 such as an earphone or a speaker into a person's ear, unlike a loudspeaker-type acoustic device that is placed at a fixed position. In addition, the hearing aid 1 is often carried around by people, and the conditions of use and the space in which it is used fluctuate greatly. Furthermore, the gain of audio amplification is higher than that of loudspeaker-type acoustic devices. In general, the gain of a loudspeaker-type acoustic device is 20 dB or less, while the gain of the hearing aid 1 is 30 dB or more (the gain of the hearing aid 1 for a person with mild to moderate hearing loss is about 30 dB to 80 dB). Since the hearing aid 1 has such special features, howling is more likely to occur than in loudspeaker-type acoustic devices, and when howling occurs, the user feels a strong sense of discomfort.
In view of such peculiarities of the hearing aid 1, the present embodiment has a configuration capable of suppressing the occurrence of howling within a short period of time so that the howling cannot be detected by the ear.

In the hearing aid 1 according to this embodiment, the sound input section 10 is a microphone that converts sound into an electric signal. In addition, in the present embodiment, voice includes not only human voice but also various sounds such as surrounding sounds, noise, music, and instrumental sounds. The amplification section 20 amplifies the electrical signal converted by the audio input section 10 . In this embodiment, the amplifier 20 is a part corresponding to a microphone amplifier and a preamplifier.

The audio output section 30 converts the amplified signal amplified by the amplification section 20 into audio. The audio output unit 30 is a speaker, earphone, headphone, or the like. In this embodiment, an output amplifier 35 is provided in front of the audio output section 30 . The output amplifier 35 is provided with a volume 37 so that the gain of the output amplifier 35 can be adjusted by the volume 37 .

The howling detection unit 50 detects occurrence of howling when the level of the amplified signal amplified by the amplification unit 20 exceeds a preset value (threshold level Th) that distinguishes between voice and howling. The howling detector 50 has a level comparator 51, for example. 2(a) and 2(b) are diagrams for explaining the level comparator. As shown in FIG. 2A, the level comparator 51 compares the amplified signal (input) amplified by the amplifier 20 with a preset reference value (threshold level Th). The level comparator 51 outputs a high detection signal CS, for example, when the amplified signal exceeds a reference value (threshold level Th) (see FIG. 2B).

The condition change unit 60 performs change processing when the howling detection unit 50 detects the occurrence of howling (when the detection signal CS is output). The change processing includes (A1) changing the phase of the howling signal that has passed through the amplification unit 20 and inputting it to the audio output unit 30, (A2) lowering the level of the howling signal, and (A3) replacing the howling signal. and inputting a dummy signal to the audio output unit 30. After the howling detection unit 50 detects the occurrence of howling and performs the change processing, the condition change unit 60 terminates the change processing at a predetermined timing.

In such a hearing aid 1, the conditions for howling to occur can be broken by performing change processing by the condition changing unit 60. FIG. That is, when the level of the amplified signal amplified by the amplifying unit 20 exceeds a preset value (threshold level Th), the occurrence of howling is detected, and the change processing for suppressing the howling is immediately performed. Therefore, howling can be suppressed before the occurrence of howling is noticed. After detecting occurrence of howling and performing change processing, the change processing is terminated at a predetermined timing.

Here, the threshold level Th for detecting occurrence of howling will be described.
The sound pressure level defines 0 decibel (dB) as the lowest level of sound that humans can hear (20 micropascals: μP). A whisper is 20 dB, and hearing below 25 dB is considered normal. The sound pressure of normal conversation is about 60 dB, which is comfortable to hear.

As an example, if the maximum rating of the headphone is 16Ω and 50mW, it is limited at 2.5Vpp. If the specification rating of the microphone is 60 dB and 1.57 mVpp, the gain is 2.5 Vpp/1.57 mVpp=64 dB (1592 times). Therefore, this microphone and head fin configuration is suitable for people with mild to moderate hearing loss.

With such a configuration suitable for mild-to-moderate hearing loss, the gain of the system is approximately 60 dB. Therefore, an example of determining the howling detection level (threshold level Th) based on the gain of this system will be described.

In the waveform after the 40 dB microphone amplifier, 0.25 Vpp (designed maximum level) is clipped at the 20 dB output amplifier (60 dB total). Assuming 0.75 Vpp, which is approximately three times the clipped 0.25 Vpp, the threshold level Th is obtained by adding 1/2 of this to the headphone limit of 2.5 Vpp. That is, the design power supply voltage VCC: 5V, and the threshold level Th should be set to 2.5+0.75/2=2.86V. As a result, the threshold level Th for reliably distinguishing between voice and howling is set.

In the above example, the threshold level Th was calculated with a system gain of about 60 dB. However, even when the system gain is about 40 dB, howling can be detected by the same threshold level Th as in the case of 60 dB.

The inventor of the present application conducted an experiment on howling suppression. In the experiment, a circuit was constructed in which a mechanical toggle switch was attached to the positive-phase output and negative-phase output of the amplifier output so that either positive-phase or negative-phase output could be selected. In this circuit, the amplifier gain is set to cause howling. Howling stops when the toggle switch is switched when howling occurs. Howling occurs again after a while. Also, when the toggle switch is switched, the howling stops. Raise the amplifier gain and repeat the same operation. It can be seen that the howling occurrence interval shortens as the gain increases. Based on the results of this experiment, the inventor of the present application has created a configuration that can effectively suppress howling while having a simple configuration using the condition changing unit 60 .

(First example of condition modification part)
A first example of the condition changer 60 is to invert the phase of the howling signal as change processing (included in change processing (A1)). As shown in FIG. 1, the amplified signal output from the amplifying section 20 includes a positive-phase signal Sφ1 and a negative-phase signal Sφ2. The positive-phase signal Sφ1 and the negative-phase signal Sφ2 are signals that are 180 degrees out of phase with each other.

The condition changing unit 60 has a toggle switch 61 and a flip-flop 65 that sequentially switch between the positive phase signal Sφ1 and the negative phase signal Sφ2. When the flip-flop 65 receives the rise of the high detection signal CS output from the level comparator 51, the flip-flop 65 switches between high and low of the trigger signal TS and outputs the trigger signal TS. In other words, the trigger signal TS is inverted each time the occurrence of howling is detected. The toggle switch 61 sequentially switches between the positive-phase signal Sφ1 and the negative-phase signal Sφ2 according to the trigger signal TS output from the flip-flop 65 .

For example, when the toggle switch 61 of the condition changing unit 60 receives a High signal as the trigger signal TS output from the flip-flop 65 when the occurrence of howling is detected by the howling detecting unit 50, the positive phase signal Sφ1 is selected. When it receives a Low signal as the trigger signal TS, it selects the opposite phase signal Sφ2. That is, the toggle switch 61 is switched according to High/Low of the trigger signal TS, and either the positive phase signal Sφ1 or the negative phase signal Sφ2 is sent to the output amplifier 35 .

As a result, switching between inversion and non-inversion of the phase of the howling signal is repeated each time howling detection section 50 detects the occurrence of howling. By switching between the positive phase signal Sφ1 and the negative phase signal Sφ2 in the audio signal every time howling occurrence is detected, one of the loop oscillation conditions (C1) "A feedback signal at a feedback point at that peak frequency is input. "Being in phase with the signal" is lost, and as a result, the occurrence of howling can be made unrecognizable. It is difficult for humans to distinguish between the inversion and non-inversion of the audio signal accompanying the phase inversion of the howling signal.

Next, the waveform change in the change processing (A1) will be described.
FIGS. 3A to 3E are schematic diagrams showing examples of waveforms.
FIG. 4 is a schematic diagram illustrating switching between normal phase and reverse phase.
FIG. 3(a) shows an example of the positive-phase signal S.phi.1, and FIG. 3(b) shows an example of the negative-phase signal S.phi.2. For convenience of explanation, howling sine waves are shown in FIGS. 3(a) and 3(b). Also, although the waveform of the audio signal actually contains harmonic components, it is simplified to schematically represent only the formant F1 or fundamental wave portion.

As shown in FIGS. 3(a) and 3(b), howling waveforms are different from voice signal waveforms. That is, when howling occurs, not the audio signal itself but the howling waveform is fed back to the microphone.

FIG. 3(c) shows the waveform of the amplified signal (output signal waveform from the condition changing section 60) when howling is detected. 3(d) shows the output of the level comparator 51 (detection signal CS), and FIG. 3(e) shows the output of the flip-flop 65 (trigger signal TS).

As shown in FIG. 3, when howling occurs while an audio signal (amplified signal) is being output, the level of the amplified signal (positive phase signal Sφ1 and negative phase signal Sφ2) sharply increases. By setting a value higher than the normal level of the audio signal as the reference value of the level comparator 51, it is possible to detect a level change when howling occurs.

For example, when the level of the amplified signal (for example, the positive phase signal Sφ1) exceeds the reference value of the level comparator 51 at timing t (see the arrow in the figure), the level comparator 51 outputs a high detection signal CS (see the arrow in the figure). 3(d)). When the flip-flop 65 receives this High detection signal CS, the flip-flop 65 inverts the output signal (trigger signal TS) (switches from High to Low in the example shown in FIG. 3(e)).

The toggle switch 61 of the condition changing unit 60 switches the selection from the positive phase signal Sφ1 to the negative phase signal Sφ2 in response to the trigger signal TS output from the flip-flop 65 switching from High to Low. As a result, as shown in FIG. 3(c), the waveform of the output signal from the condition changing section 60 changes to the opposite phase signal Sφ2 after timing t.

When the signal is switched to the reversed-phase signal Sφ2 at the timing when the howling occurs, the reversed-phase signal Sφ2 of the howling is sent to the audio output section 30 only for that moment. This reverse-phase signal Sφ2 of the howling is output as a sound by the sound output unit 30 and returned to the sound input unit 10, which is one of the loop oscillation conditions (C1) "The feedback signal at the feedback point at the peak frequency is Being in phase with the input signal collapses, and howling is instantly suppressed.

As shown in FIG. 4, when this change processing (A1) is repeated, for example, when howling occurs during audio output based on the positive phase signal Sφ1, the level comparator 51 outputs a high detection signal CS, and the flip-flop 65 is inverted (from High to Low in this example), and the toggle switch 61 of the condition changing section 60 selects the opposite phase signal Sφ2. Howling is suppressed at the moment when the positive-phase signal Sφ1 is switched to the negative-phase signal Sφ2.

If the sound output based on the reversed-phase signal Sφ2 continues, the reversed-phase signal Sφ2 may eventually generate howling. When the howling level is reached, the level comparator 51 outputs a High detection signal CS at that instant, the output of the flip-flop 65 is inverted (in this example, from Low to High), and the toggle switch 61 of the condition changing unit 60 is turned on. selects the positive phase signal Sφ1. Howling is suppressed at the moment when the reverse phase signal Sφ2 is switched to the normal phase signal Sφ1.

By repeating this change processing (A1), the normal-phase signal Sφ1 and the reverse-phase signal Sφ2 are actually switched at the timing when howling starts to occur, and no recognizable howling occurs.

Here, the configuration of the hearing aid 1 according to the present embodiment (sound input unit 10, amplifier unit 20, sound output unit 30, howling detection unit 50, and condition change unit 60) is configured by analog circuits. The positive-phase signal Sφ1 and the negative-phase signal Sφ2 output from the amplifying section 20 can be easily generated by an analog circuit. The level comparator 51 and the flip-flop 65 can also be realized with simple analog circuits. Since there is no D/A converter or A/D converter in the circuit system, there is no need to provide a complicated circuit configuration. In this embodiment, since it can be realized with a simple configuration such as the level comparator 51, the flip-flop 65, and the toggle switch 61, even the analog hearing aid 1 can have a howling suppressing function.

Further, the circuit system of the hearing aid 1 according to this embodiment does not include a delay circuit (time constant circuit) such as an AGC (automatic gain control unit). Since such a delay circuit (time constant circuit) is not included, the time from occurrence of howling to completion of the change processing can be shortened to a negligible extent. Therefore, the change processing is executed immediately after howling begins to occur, and the occurrence of recognizable howling can be prevented.

The inventors of the present application constructed the hearing aid 1 with an analog circuit without including a delay circuit, and conducted an experiment of howling prevention by the change processing (A1). As a result, the time from when the High signal was output from the level comparator 51 to when the signal was switched by the toggle switch 61 of the condition changing section 60 was about several hundred ns (nanoseconds). For example, assuming howling at 2 kHz, its period is 0.5 ms (milliseconds). Since the switching time of the toggle switch 61 is several 100 ns, the howling is suppressed by the change processing and completed within less than one cycle of the howling. In an experiment, it was confirmed that howling can be prevented without being recognized by such instantaneous switching.

In a device with a gain of 30 dB or more, such as the hearing aid 1, there is a high possibility that howling will reoccur immediately after performing the change processing. Therefore, when a delay circuit is included, it takes a delay time from detection of occurrence of howling to execution of suppression processing, and there is a possibility that howling may be recognized during this time. Like the hearing aid 1 according to the present embodiment, it does not include a delay circuit, and the level comparator 51 detects howling and then instantly performs change processing, thereby effectively suppressing howling with a simple analog circuit. becomes possible.
Even if the hearing aid 1 according to the present embodiment is a digital hearing aid, howling can be prevented by applying the same algorithm as described above.

(Second example of condition modification part)
FIG. 5 is a block diagram showing a second example of the condition changing unit.
In a second example of the condition changing unit 60, as the change processing, in addition to inverting the phase of the howling signal (included in the change processing (A1)), the inverted phase of the howling signal is changed after the howling is suppressed. I am in the process of restoring it.

As shown in FIG. 5, the howling detector 50 determines that howling occurs and outputs the trigger signal TS when the level of the amplified signal (in this case, the positive phase signal Sφ1) exceeds a predetermined reference value. When the howling detector 50 is a normal level comparator 51, the detection signal CS output from the level comparator 51 becomes the trigger signal TS. The condition changing unit 60 switches the toggle switch 61 according to High/Low of the trigger signal TS to send either the positive phase signal Sφ1 or the negative phase signal Sφ2 to the output amplifier 35 . Howling is suppressed by inverting the phase of the howling signal.

When the howling is suppressed and the level of the amplified signal becomes equal to or lower than a predetermined reference value, the detection signal CS (trigger signal TS) output from the howling detection section 50 becomes Low. When the trigger signal TS is sent to the condition changing section 60, the toggle switch 61 is switched to switch between the positive phase signal Sφ1 and the negative phase signal Sφ2. As a result, after the howling is suppressed, the original phase of the audio signal is restored. When the howling stops and the filter characteristics of the sound field space disappear, it takes some time until the next howling occurs. Therefore, even if the output is returned to the normal phase signal Sφ1, howling does not occur immediately.

In the above example, the original phase is returned to the original phase at the timing (detection signal CS) at which howling is suppressed. You may make it perform the process which returns to.
FIGS. 6A and 6B are diagrams for explaining an example of a howling detector for restoring the original phase at the timing when a preset time has elapsed.
As shown in FIG. 6A, the howling detector 50 has a circuit configuration (including a feedback resistor 511 and a capacitor 512) that delays the timing at which the output of the level comparator 51 falls from High to Low. . As a result, as shown in FIG. 6(b), even when the input signal falls below a predetermined level, the output of the level comparator 51 falls with a delay of a predetermined time due to hysteresis.

In the above description, the toggle switch 61 is switched using the detection signal CS as the trigger signal TS. BS may be output. When the return signal BS is sent to the condition changing unit 60, the toggle switch 61 is switched to switch from the reverse phase signal Sφ2 to the normal phase signal Sφ1. As a result, after howling is suppressed, the signal is returned to the original positive phase signal Sφ1.

(Third example of condition changing part)
FIG. 7 is a block diagram showing a third example of the condition changing section.
A third example of the condition changer 60 is to lower the level of the howling signal as change processing (included in change processing (A2)).

As shown in FIG. 7, the toggle switch 61 of the condition changing section 60 is connected to the positive phase signal Sφ1 and the ground line GL which is the ground potential. When the level of the amplified signal (in this case, the positive phase signal Sφ1) exceeds a predetermined reference value, the howling detector 50 determines that howling has occurred and outputs the trigger signal TS. When the howling detector 50 is a normal level comparator 51, the detection signal CS output from the level comparator 51 becomes the trigger signal TS.

The condition changing unit 60 switches the toggle switch 61 according to High/Low of the trigger signal TS. In this example, the toggle switch 61 selects the amplified signal (positive phase signal Sφ1) in the normal state, and the positive phase signal Sφ1 is sent to the output amplifier 35 . When howling occurs and the trigger signal TS is sent to the condition changing unit 60, the toggle switch 61 is switched to select the ground line GL. As a result, the signal level sent to the output amplifier 35 becomes the ground potential (zero).

When the signal level sent to the output amplifier 35 instantaneously becomes the ground potential upon detection of howling occurrence, one of the conditions for loop oscillation (C2) "the gain in the open loop must be 1 or more" collapses. , howling is suppressed.

When the howling is suppressed and the level of the amplified signal becomes equal to or lower than a predetermined reference value, the howling detector 50 outputs the return signal BS. When the return signal BS is sent to the condition changing unit 60, the toggle switch 61 is switched to switch from the ground line GL to the positive phase signal Sφ1. As a result, after howling is suppressed, the signal is returned to the original positive phase signal Sφ1.

In this way, after the occurrence of howling is detected and change processing is performed, after howling is suppressed, the audio output of the original phase is immediately restored. It doesn't make you feel uncomfortable. In the above example, the process of returning to the original phase is performed at the timing when howling is suppressed, but after performing the change process, the process of returning to the original phase is performed at the timing when a preset time has elapsed. You may do so.

(Fourth example of condition modification part)
FIG. 8 is a block diagram showing a fourth example of the condition changing section.
A fourth example of the condition changing section 60 is to input the dummy signal DS instead of the howling signal to the audio output section 30 as the changing process (included in the changing process (A3)).

As shown in FIG. 8, condition changing section 60 includes switch 62 . The signal generator 63 is connected to the input terminal of the switch 62 , and the output terminal is connected to the line of the positive phase signal Sφ 1 output from the amplifier section 20 . A signal generator 63 generates a dummy signal DS. The frequency of the dummy signal DS should be a frequency other than the audible frequency. When the level of the amplified signal (positive phase signal Sφ1) exceeds a predetermined reference value, the howling detector 50 determines that howling has occurred and outputs the trigger signal TS. When the howling detector 50 is a normal level comparator 51, the detection signal CS output from the level comparator 51 becomes the trigger signal TS.

The switch 62 of the condition changing section 60 is opened and closed by the trigger signal TS output from the howling detection section 50 . In this example, the switch 62 is normally open and the dummy signal DS is not sent to the output amplifier 35 . That is, in a normal state, only the positive phase signal Sφ1 is sent from the output amplifier 35 to the audio output section 30. FIG.

When howling occurs and the trigger signal TS is sent to the condition changing section 60 , the switch 62 is closed and the dummy signal DS of the signal generating section 63 is sent to the output amplifier 35 . When the occurrence of howling is detected and the dummy signal DS is sent to the output amplifier 35, the sound output unit 30 outputs sound based on the sound signal in which the dummy signal DS is superimposed on the normal phase signal Sφ1. When the sound superimposed with the dummy signal DS is output, one of the conditions for loop oscillation (C3) "there is a frequency selection element in the path of loop oscillation" is broken, and howling is suppressed. .

If the frequency of the dummy signal DS is other than the audible frequency, even if the sound with the dummy signal DS superimposed thereon is output, the sound of the dummy signal DS will not be noticed. The frequency of the dummy signal DS may be an audible frequency, but in this case, it is preferable that the duration is short enough to be unrecognizable by humans. Such superimposition of the dummy signal DS changes the filter characteristics of the sound field space that causes the howling, and the howling can be suppressed.

When the howling is suppressed and the level of the amplified signal becomes equal to or lower than a predetermined reference value, the howling detector 50 outputs the return signal BS. When the return signal BS is sent to the condition changing section 60, the switch 62 is opened and the superposition of the dummy signal DS is cancelled. As a result, after howling is suppressed, the signal is returned to the original positive phase signal Sφ1.

(program processing)
The hearing aid 1 according to this embodiment can also be realized by a mobile terminal and program processing executed thereon. A microphone provided in the mobile terminal is used as an audio input unit 10 , and a speaker provided in the mobile terminal and earphones and headphones connected to the mobile terminal in a wired or wireless manner are used as an audio output unit 30 . The amplifier unit 20 is incorporated as a circuit of the mobile terminal. The howling detection unit 50 and the condition change unit 60 can be realized as program processing (application program) executed by the portable terminal.

In the program processing, a step of detecting the occurrence of howling (howling detection step) when the level of the amplified signal exceeds a preset value for distinguishing between voice and howling; changing the phase of the howling signal passing through the amplifying unit 20 and inputting it to the audio output unit 30; lowering the level of the howling signal; and inputting a dummy signal to the audio output unit 30 instead of the howling signal. A step of performing at least one of the change processes (change process step) is executed by a computer (CPU).
This change processing step includes ending the change processing at a predetermined timing after the occurrence of howling is detected in the howling detection step and the change processing is performed.

As described above, according to the hearing aid 1 according to the present embodiment, it is possible to provide the hearing aid 1 that can effectively suppress howling even with a simple configuration.

Although the present embodiment and other examples have been described above, the present invention is not limited to these examples. For example, in the above example, the positive-phase signal Sφ1 and the negative-phase signal Sφ2 are used as the amplified signals, but any signal out of phase with respect to the positive-phase signal Sφ1 can be similarly applied. In addition, additions, deletions, and design changes of constituent elements to the above-described embodiments and other examples by those skilled in the art as appropriate, and combinations of features of the embodiments as appropriate may also be included in the scope of the present invention. As long as it has the gist, it is included in the scope of the present invention.

The configuration of the present invention can be used in an existing sound system for public address without remodeling the inside of the system. In other words, what can be used is an audio device system that is divided into a microphone amplifier device and an output amplifier device. good.

DESCRIPTION OF SYMBOLS 1... Hearing aid 10... Voice input part 20... Amplifier part 30... Voice output part 35... Output amplifier 37... Volume 50... Howling detection part 51... Level comparator 60... Condition change part 61... Toggle switch 62... Switch 63... Signal generation part 65 Flip-flop 511 Feedback resistor 512 Capacitor BS Return signal CS Detection signal DS Dummy signal GL Ground line Sφ1 Positive phase signal for microphone input Sφ2 Negative phase signal for microphone input TS Trigger signal

Claims (2)

an audio input unit that converts audio into an electrical signal;
an amplifier that amplifies the electrical signal converted by the audio input unit;
an audio output unit that converts an amplified signal amplified by the amplification unit into audio;
a detection unit that detects just before loop oscillation of howling when the level of the amplified signal exceeds a preset value that distinguishes between voice and howling;
and a condition changing unit that performs signal processing for breaking a howling occurrence condition when the detecting unit detects just before loop oscillation of howling. 2. The acoustic device according to claim 1, wherein said condition changing section starts said signal processing when said detection section detects just before loop oscillation of said howling.

Images