EP2597889A1 - Headphones with non-adaptive active noise control - Google Patents

Abstract

The headset has a mixer circuit i.e. summing circuit (46), receiving inputs signals delivered by a feed forward filter (32), a feedback filter (42) and a stabilizer filter (44). The stabilizer filter is connected with a feedback branch for receiving input signal delivered by an internal microphone (36), and delivering output signal as an input to a combining circuit. The mixing circuit delivers a linear combination of the output signals delivered by the three filters together with a fraction of to-be played back audio signal (S) with respective weighting of gains applied to the signals.

Description

The present invention relates to an audio headset comprising an "active noise control" system.

Such a headset can be used for listening to an audio source (music for example) from a device such as an MP3 player, radio, smart-phone, etc. to which it is connected by a wired connection or by a wireless link, in particular a connection of the Bluetooth type (trademark of the Bluetooth SIG).

If it is equipped with a microphone set capable of capturing the voice of the wearer of the headset, it is also possible to use this headset for communication functions such as "hands-free" telephony functions, in addition to the listening to the audio source. The headset transducer then reproduces the voice of the remote speaker with which the wearer of the headset is in conversation.

The helmet includes two headphones joined by a bow. Each earpiece comprises a closed shell housing a sound reproduction transducer (hereinafter simply referred to as a "transducer") and intended to be applied around the user's ear with the interposition of a circumaural pad isolating the ear of the ear. external sound environment. When the headset is used in a noisy environment (metro, street, train, airplane, etc.) the wearer is partially protected from noise by the headphones headphones, which isolate it with the closed shell and the circumaural pad.

However, this purely passive protection is only partial, part of the sounds, especially in the lower part of the frequency spectrum, can be transmitted to the ear through the shell of the headphones, or through the cranial box of the carrier.

It is for this reason that techniques have been developed called "active noise control" or ANC ( Active Noise Control ), the principle of which is to capture the incident noise component by means of a microphone placed on the hull helmet earphones and temporally and spatially superimposing on this noise component an acoustic wave which is ideally the inverted copy of the pressure wave of the noise component. This is to create in this manner a destructive interference with the noise component and reduce, ideally neutralize, the pressure variations of the parasitic acoustic wave.

The implementation of this principle involves solving a large number of difficulties, which has led to proposals of a very varied nature, which can be grouped into two categories.

A first category is that of ANC methods using adaptive filters, that is to say filters whose transfer function is dynamically changed, continuously, by a real-time analysis algorithm of the signal. These treatments could be implemented in particular through the development of digitization techniques and signal processing by specialized processors, programmed to implement real-time algorithms.

The DE 37 33 132 A1 is a typical example of an ANC processing implementing such adaptive filters. Other examples of ANC methods involving adaptive filters are described in particular in the US 6,041,126 A , US 2003/0228019 A1 and WO 2005/112849 A2 .

These techniques can be effective in terms of noise reduction, but they have the disadvantage of being necessarily digital and require a relatively high computing power, resulting in a relatively complex design and a relatively high realization cost .

In addition, the digital processing introduces significant delays in the compensation signal, and the adaptive character implies a minimum convergence time of the algorithms. All this hinders the responsiveness of the system, especially with regard to irregular noises. As a result, denoising is especially effective against essentially periodic noise in narrow bands.

The second category of ANC processes - to which belongs the technique of the invention - is that of non-adaptive fixed filtering systems, that is to say where the parameters of the various filters used are predetermined.

These ANC systems combine fixed feedback, closed-loop, and feedforward, open-loop filtering. The feedback filtering channel is based on a signal collected by a microphone disposed within the acoustic cavity (hereinafter "front cavity") delimited by the shell of the earpiece, the circumaural pad and the transducer. In other words, this microphone is arranged close to the ear of the user, and receives mainly the signal produced by the transducer and the residual noise signal, unneutralized, still noticeable in the front cavity. The signal from this microphone, from which the audio signal of the music source to be reproduced by the transducer is subtracted, constitutes an error signal for the feedback loop of the ANC system. The feedforward filtering channel , on the other hand, uses the signal picked up by the external microphone collecting the parasitic noise prevailing in the immediate environment of the helmet wearer.

Such a system is described in particular by the US 2010/0272276 A1 which, in addition to the feedback and feedforward filtering channels , provides a third filtering channel, which processes the audio signal from the music source to be reproduced. The output signals of the three filter channels are combined and applied to the transducer to reproduce the signal of the musical source associated with a surrounding noise canceling signal.

Since the parameters of the various filters are fixed, fixed filtering techniques can be implemented in both analog and digital technology, in a way that is less demanding in terms of resources than for adaptive filter techniques.

However, fixed filtering methods have limitations and disadvantages.

A first disadvantage is the relatively high sensitivity to the variations of the electroacoustic paths between the transducer and the error microphone, that is to say the internal microphone placed in the front cavity. The electro-acoustic response between these two elements can indeed be modified because of the volume variations of the front cavity and its sealing with respect to the outside. The main factors likely to vary this electro-acoustic response are the positioning of the helmet on the head, the shape of the ear of the user, the different tightenings of the helmet on the head, and the presence of hair in the place where the circumaural pads come to bear. Other variations may be due to the electronic components used (resistors, capacitors, transducer and microphone), whose electrical characteristics may fluctuate over time.

These variations in acoustic response are likely to produce an undesirable effect called waterbed effect : beyond the frequency band noise, the noise will be amplified in a relatively narrow frequency band, usually around 1 kHz, perfectly noticeable and of course harmful. If it is too important this phenomenon can even generate a Larsen effect, a phenomenon that can be seen for many helmets when the pad is accidentally removed.

Another factor to take into account is the volume of the front cavity, since a reduced front volume increases the variability of the electroacoustic response between the transducer and the error microphone, since in this case the relative volume variation between the normal listening position and the transition position where the user approaches the helmet on his head will be more important.

A small front cavity volume is therefore an additional factor of loss of stability of the feedback loop, with the same consequences as those described above. However, in practice it is desirable to make earphones of relatively small volume, both for reasons of comfort and weight, which goes against the requirement of stability of the ANC system.

Specifically, the settings of the various filter channels are adjusted to produce performance corresponding to a given electroacoustic response, for gain and phase margins to ensure sufficient stability and maximum performance. In this respect, it is estimated that a closed-loop system must generally have a phase margin greater than 45 ° and a gain margin of at least 10 dB. But these theoretical margins are often insufficient, because of the great variability of electroacoustic responses encountered in practice in the field of active noise control headphones.

The problem of the invention is, in such a non-adaptive ANC system, to fight against the risks of instability, with increased gain and phase margins allowing, despite a small front cavity volume, to avoid any appearance of effect waterbed or Larsen despite variations in the introduction of the helmet on the head, clamping headphones, sealing more or less good circumaural pads.

This improvement in stability will of course have to be obtained without degradation of the noise-canceling performance of the ANC system, that is to say that the latter will have to retain all its effectiveness of neutralizing the components of parasitic noise, whatever their nature more or less. less periodic and their frequency spectrum.

Of course, the audio signal from the music source (or the voice of the remote speaker, in a telephony application) should not be distorted and its spectrum should not be reduced by the ANC processing, although the signal of neutralization of noise and the audio signal to be reproduced are amplified by the same channel and reproduced by the same transducer.

The basic idea of the invention is to reduce the bandwidth of the feedback filter in the upper spectrum, that is to say in the unstable frequency zone, so as to reduce or even eliminate the risks of effect. waterbed or Larsen. This limitation of the bandwidth can provide, as will be seen, an increase in the gain margin of at least 15 dB, preferably at least 17 dB, and the phase margin of at least 45 °, preferably at least 60 °.

At the same time, the feedforward filter will compensate for the loss of performance in the higher frequencies of the noise spectrum to be eliminated (that is, around 1 kHz).

Finally, a stabilizing filter is arranged in parallel with the feedback filter . This stabilizing filter makes it possible to increase the phase margin of the feedback filter by increasing the phase in the critical zone of the waterbed effect: to compensate for the phase decrease due to the acoustics, in particular because of the path propagation of the sound of the transducer to the error microphone, a limited resonance is created by the stabilizing filter in order to raise the phase and thus increase the phase margin.

These three channels ( feedback, feedforward and stabilizer) are arranged in parallel, and the signals delivered at the output of the filters are combined with each other and with the audio signal to be reproduced by means of a combiner delivering a linear combination of these different signals, for amplification and reproduction by the transducer.

• une première branche feedforward en boucle ouverte, avec un premier filtre passe-bande recevant en entrée un signal délivré par un microphone externe apte à capter un bruit acoustique régnant dans l'environnement du casque ;
• une deuxième branche feedback en boucle fermée, avec un deuxième filtre passe-bande recevant en entrée un signal d'erreur délivré par un microphone interne à la cavité ;
• une troisième branche avec un troisième filtre ; et
• un circuit de mixage, recevant en entrée les signaux délivrés par les premier, deuxième et troisième filtres ainsi que le signal audio à reproduire, et délivrant en sortie un signal apte, après amplification, à piloter le transducteur.

More specifically, the invention relates to a helmet comprising an active noise control system comprising, in a manner known in itself from the US 2010/0272276 A1 above, two earphones joined by a holding bar and each comprising a sound reproduction transducer of an audio signal to reproduce, transducer housed in an acoustic cavity defined by a shell provided with a circumaural pad. This headset features an active noise control system with:

• a first open loop feedforward branch, with a first bandpass filter receiving as input a signal delivered by an external microphone adapted to pick up acoustic noise prevailing in the helmet environment;
• a second closed-loop feedback branch, with a second band-pass filter receiving as input an error signal delivered by a microphone internal to the cavity;
• a third branch with a third filter; and
• a mixing circuit, receiving as input the signals delivered by the first, second and third filters and the audio signal to be reproduced, and outputting a signal adapted, after amplification, to drive the transducer.

• le contrôle actif de bruit est un contrôle non adaptatif, les paramètres des premier, deuxième et troisième filtres étant des paramètres prédéterminés ;
• le troisième filtre est un filtre passe-bande stabilisateur agencé en parallèle de la deuxième branche feedback, recevant en entrée le signal délivré par le microphone interne, et délivrant en sortie un signal appliqué en entrée du circuit combineur, ce troisième filtre étant apte à augmenter localement la phase de la fonction de transfert du deuxième filtre dans un zone d'instabilité prédéterminée ;
• les première, deuxième et troisième branches sont agencées en parallèle, et le circuit de mixage est un circuit sommateur délivrant en sortie une combinaison linéaire des signaux délivrés par les premier, deuxième et troisième filtres ainsi que d'au moins une fraction du signal audio à reproduire, avec pondération des gains respectifs appliqués à ces signaux.

Characteristically, the invention

• the active noise control is a non-adaptive control, the parameters of the first, second and third filters being predetermined parameters;
• the third filter is a stabilizing band-pass filter arranged in parallel with the second feedback branch , receiving as input the signal delivered by the internal microphone, and outputting a signal applied to the input of the combiner circuit, this third filter being able to increase locally the phase of the transfer function of the second filter in a predetermined instability zone;
• the first, second and third branches are arranged in parallel, and the mixing circuit is a summing circuit outputting a linear combination of the signals delivered by the first, second and third filters as well as at least a fraction of the audio signal to reproduce, with weighting of the respective gains applied to these signals.

The predetermined instability zone in question is notably a zone of effect waterbed around a frequency of 1 kHz.

The upper cut-off frequency of the second filter is preferably less than 150 Hz, preferably less than 120 Hz, and its bandwidth less than 65 Hz, preferably less than 55 Hz.

The gain margin of the feedback branch of the active noise control is advantageously at least 15 dB, preferably at least 17 dB, and the phase margin of at least 45 °, preferably at least 60 °.

The audio signal to be reproduced is preferably inputted to both the second filter and the summing circuit, the second filter receiving as input a signal obtained by combining said error signal delivered by the internal microphone with at least a fraction of the signal audio to reproduce, and it is not applied to the third filter.

• La Figure 1 illustre de façon générale un casque audio reposant sur la tête d'un utilisateur.
• La Figure 2 est une représentation schématique montrant les différents signaux acoustiques et électriques ainsi que les divers blocs fonctionnels impliqués dans le fonctionnement d'un casque audio à contrôle actif de bruit.
• La Figure 3 est une coupe en élévation de l'un des écouteurs du casque selon l'invention, montrant la configuration des divers éléments mécaniques et organes électromécaniques de celui-ci.
• La Figure 4 est une vue de face de l'écouteur de la Figure 3.
• La Figure 5 est une vue de dos de l'écouteur des Figures 3 et 4.
• La Figure 6 est une vue de dessous de l'écouteur des Figures 3 à 5.
• La Figure 7 est une vue générale, sous forme de diagramme fonctionnel, des divers éléments du système de contrôle actif de bruit du casque de l'invention.
• La Figure 8 illustre un exemple de réalisation, sous forme analogique, du filtre feedforward de la Figure 7.
• La Figure 9 illustre un exemple de réalisation, sous forme analogique, du filtre feedback de la Figure 7.
• La Figure 10 illustre un exemple de réalisation, sous forme analogique, du filtre stabilisateur de la Figure 7.
• La Figure 11 est une caractéristique illustrant l'atténuation introduite par la coque de l'écouteur, rapportée à l'atténuation interne de la cavité avant de l'écouteur.
• La Figure 12 représente, en amplitude et en phase, le diagramme de Bode de la fonction de transfert du filtre feedforward du circuit de la Figure 7.
• La Figure 13 représente le lieu de Black du système de contrôle actif de bruit selon l'invention, avec et sans action du filtre stabilisateur.
• La Figure 14 représente le module de la fonction de transfert du filtre feedback du circuit de la Figure 7 pour différentes configurations (avec bande passante complète, avec bande passante réduite, avec et sans filtre stabilisateur).
• La Figure 15 représente la phase de la fonction de transfert du filtre feedback du circuit de la Figure 7, également pour différentes configurations.
• La Figure 16 est le tracé de Nyquist du circuit de la Figure 7, également pour différentes configurations.
• La Figure 17 représente la caractéristique d'atténuation, en boucle fermée, du circuit de la Figure 7, également pour différentes configurations.

An embodiment of the device of the invention will now be described with reference to the appended drawings in which the same reference numerals designate identical or functionally similar elements from one figure to another.

• The Figure 1 generally illustrates an audio headset resting on the head of a user.
• The Figure 2 is a schematic representation showing the different acoustic and electrical signals as well as the various functional blocks involved in the operation of an active noise control headphones.
• The Figure 3 is an elevational section of one of the earphones of the helmet according to the invention, showing the configuration of the various mechanical elements and electromechanical members thereof.
• The Figure 4 is a front view of the earpiece of the Figure 3 .
• The Figure 5 is a back view of the earpiece of the Figures 3 and 4 .
• The Figure 6 is a bottom view of the listener Figures 3 to 5 .
• The Figure 7 is a general view, in the form of a functional diagram, of the various elements of the active noise control system of the helmet of the invention.
• The Figure 8 illustrates an example embodiment, in analog form, of the feedforward filter of the Figure 7 .
• The Figure 9 illustrates an exemplary embodiment, in analog form, of the feedback filter of the Figure 7 .
• The Figure 10 illustrates an exemplary embodiment, in analog form, of the stabilizing filter of the Figure 7 .
• The Figure 11 is a characteristic illustrating the attenuation introduced by the shell of the earphone, related to the internal attenuation of the front cavity of the earpiece.
• The Figure 12 represents, in amplitude and in phase, the Bode diagram of the transfer function of the feedforward filter of the circuit of the Figure 7 .
• The Figure 13 represents the Black locus of the active noise control system according to the invention, with and without the action of the stabilizing filter.
• The Figure 14 represents the module of the transfer function of the feedback filter of the circuit of the Figure 7 for different configurations (with full bandwidth, reduced bandwidth, with and without stabilizing filter).
• The Figure 15 represents the phase of the transfer function of the feedback filter of the circuit of the Figure 7 , also for different configurations.
• The Figure 16 is the Nyquist route of the circuit of the Figure 7 , also for different configurations.
• The Figure 17 represents the attenuation characteristic, in a closed loop, of the circuit of the Figure 7 , also for different configurations.

On the Figure 1 , there is shown a headphone placed on the head of his user. This headset comprises, in a conventional manner, two earphones 10, 10 'joined by a holding bar 12. Each of the earphones 10 comprises an outer shell 14 which is applied to the contour of the ear of the user. , with interposition between the shell 14 and the periphery of the ear of a flexible circumaural pad 16 for providing a sound seal, acoustically, between the region of the ear and the external sound environment.

The Figure 2 is a schematic representation showing the different acoustic and electrical signals as well as the various functional blocks involved in the operation of an active noise control headphones.

The earphone 10 encloses a sound reproduction transducer 18, hereinafter referred to simply as a "transducer", carried by a partition 20 defining two cavities, namely a front cavity 22 on the side of the ear and a rear cavity 24 on the opposite side. .

The front cavity 22 is defined by the inner partition 18, the wall 14 of the earpiece, the pad 16 and the outer face of the user's head in the region of the ear. This cavity is a closed cavity, with the exception of the inevitable acoustic leaks in the region of contact of the pad 16. The rear cavity 24 is a closed cavity, with the exception of an acoustic vent 26 making it possible to obtain a reinforcement low frequencies in the front cavity 22 of the earpiece. Such acoustic reinforcement is more advantageous than electric amplification, because it allows an improvement of the effect of suppression of ambient noise by the active control system, without saturation and with less electrical noise.

For active noise control, the earpiece 10 carries an external microphone 28 for sensing the surrounding noise outside the earpiece, schematized by the wave 30. The signal collected by this external microphone 28 is applied to a microphone. feedforward 32 filter stage of active noise control system.

Each earphone 10, 10 'is provided with its own active noise control system, the respective external microphones 28, 28' ( Figure 1 ) being independent of each other.

The helmet may possibly wear, as shown Figure 1 , another external microphone 34 for communication functions, for example if the headset is provided with "hands-free" telephony functions. This additional external microphone 34 is intended to capture the voice of the wearer of the helmet, it does not intervene in the active control of the noise and in the following we will only consider as external microphone that the microphone (s) 28 dedicated (s) ) active noise control.

The headset is also provided with an internal microphone 36 disposed as close as possible to the auditory canal of the ear, to capture the residual noise present in the internal cavity 22, noise that will be perceived by the user.

• du bruit résiduel 30 provenant de la transmission du bruit externe environnant 30 au travers de la coque 14 de l'écouteur, et
• d'une onde sonore 40 générée par le transducteur 18, qui est, idéalement, la copie inversée du bruit 38, c'est-à-dire du bruit à supprimer au point d'écoute, selon le principe des interférences destructives.

By ignoring the audio signal of the music source reproduced by the transducer (or the voice of the remote speaker, in a telephony application), the acoustic signal picked up by this internal microphone 36 is a combination:

• residual noise 30 from the transmission of the surrounding external noise 30 through the shell 14 of the earphone, and
• a sound wave 40 generated by the transducer 18, which is, ideally, the inverted copy of the noise 38, that is to say the noise to be removed at the point of listening, according to the principle of destructive interference.

Since the neutralization of the noise by the sound wave 40 is never perfect, the internal microphone 36 collects a residual signal which is used as an error signal e applied to a feedback filter branch 42 in a closed loop and to a stabilizing branch 44 (Specific to the invention) whose signals are combined at 46 with the signal of the open loop feedforward branch 32 to drive the transducer 18.

In addition, the transducer 18 receives an audio signal to be reproduced from a musical source (walkman, radio, etc.), or the voice of the remote speaker, in a telephony application. As this signal experiences the effects of the closed loop distorting it, it will be preprocessed upstream by equalization in a digital signal processor, so as to present the desired transfer function, determined by the gain of the open loop and the response. target without active control.

The Figures 3 to 6 illustrate, according to several angles of view, an embodiment of the different mechanical and electroacoustic elements schematically represented on the Figure 2 , for one of the headphones 10 (the other headphone 10 'being made identically).

It contains the partition 20 dividing the inside of the shell 14 into a front cavity 22 and a rear cavity 24 with, mounted on this partition, the transducer 18 and the internal microphone 36 carried by a grid 48 maintaining it nearby. of the ear canal of the user. On the Figures 5 and 6 , there is also shown the external microphone 28 dedicated to the active noise control and the additional microphone 34 for the communication functions "hands-free", and the vent 26, consisting for example of a series of small holes covered with an acoustically resistive plastic gate.

The Figure 7 illustrates in block diagram form the active noise control circuit according to the invention, with the electrical and acoustic transfer functions involved in the operation of this circuit.

The circuit essentially comprises three branches arranged in parallel, with a feedforward filter 32, a feedback filter 42 and a stabilizing filter 44.

The signal collected by the external microphone 28 is preamplified by a gain G1 (for example G1 = +8 dB), then is applied to the feedforward filter 32.

The signal collected by the internal microphone 36 is applied concurrently to the stabilizing filter 44 and the feedback filter 42, with application of respective gains G2 (eg G2 = 0 dB), and G3 (eg G3 = +9 dB).

The signals coming in parallel from the filters 32, 44 and 42 are combined with each other by an adder 46 with respective gains G5, G6 and G7 (for example G5 = -6 dB for the signal from the feedforward filter 32, G6 = + 6 dB for the signal from the stabilizing filter 44 and G7 = 0 dB for the signal from the feedback filter 42 ) .

The audio signal S (signal line-in ) from the music source (MP3 player, radio, etc.) or telephony circuits is the subject of digital processing (decoding, equalization, audio effects such as spatialization, etc. .) by a digital signal processor or DSP 50. Moreover, as this signal is affected by the closed loop which distorts it, it is preprocessed upstream in the DSP 50 by appropriate equalization, so as to have the function of desired transfer, determined by the gain of the open loop and the target response without active control.

• avec application d'un gain G4 (par exemple G4 = -14 dB), au filtre feedback 42, et
• avec application d'un gain G8 (par exemple G4 = -6 dB), à un sommateur 52 qui combinera ce signal à celui capté par le microphone interne 36 après que ce dernier ait été préamplifié du gain G3, pour l'appliquer en entrée du filtre feedback 42.

The audio signal at the output of the DSP 50 is applied to the active control circuit in two places, respectively:

• with application of a gain G4 (for example G4 = -14 dB), to the feedback filter 42, and
• with application of a gain G8 (for example G4 = -6 dB), to a summator 52 which will combine this signal with that sensed by the internal microphone 36 after the latter has been preamplified G3 gain, to apply it to the input of the feedback filter 42 .

The injection of the audio signal to be reproduced S in two different places of the circuit makes it possible to obtain a balanced equalization between the low frequencies and the high frequencies. Indeed, the portion of the signal injected at the input of the general adder 46 will undergo attenuation of the active control, which will give acute frequency components; in contrast, the part of the signal injected on the adder 52 at the input of the feedback filter 42 will undergo the low-pass filtering of the circuit, giving serious frequency components. The respective gains G8 and G4 applied to these two signal parts can balance the bass and treble spectrum of the signal to reproduce.

It will be noted that the audio signal to be reproduced is injected only at the input of the feedback filter 42 (via the adder 52), but is not input into the branch of the stabilizing filter 44, which makes it possible to adjust this stabilizing filtering without disturbing the equalization of the music to be reproduced: in fact, the stabilizing filter 44 receives only the sound picked up by the internal microphone 36, excluding the audio signal to be reproduced, which does not interfere with the stabilization function.

Finally, the output signal of the general adder 46, which is a linear combination of the signals from the three feedforward, feedback and stabilization filtering channels as well as the audio signal to be reproduced, is applied to the transducer 18 after amplification by a power stage. 54.

The Figures 8, 9 and 10 illustrate examples of embodiments, in analog technology, of the feedforward 32, feedback 42 and stabilizer 44 filters respectively. In these figures, V _i and V _o indicate the input and output voltages respectively of the filters, and V _mid indicates the voltage middle between the positive and negative terminals of the power supply of the operational amplifier used by the filter. We will describe in more detail, with reference to Figures 12 to 17 , the respective transfer functions of these different filters and the way in particular the stabilizing filter 44 to change the response of the feedback filter 42 so as to increase the overall performance of the active noise control system.

As can be seen, these three filters can be made with a minimum of components, so with a very low material cost.

Moreover, in the examples illustrated, the feedforward 32 and feedback 42 filters are made in the form of first-order low-pass filters, but it is possible without difficulty to obtain second-order bandpass filters by modifying the elements. resistive and capacitive.

We will now describe the general operation of the active noise control system according to the invention, which has just presented the general architecture.

• H_c : fonction de transfert entre le signal reçu par le microphone externe 28 et le signal reçu par le microphone interne 36, représentative de la fraction du bruit extérieur qui a traversé la coque de l'écouteur du casque ;
• H_o : fonction de transfert entre le signal reproduit par le transducteur 18 et le signal reçu par le microphone externe 28, représentative de la fraction de signal acoustique qui a été transmise par la coque de l'écouteur jusqu'au microphone externe ;
• H_a : fonction de transfert entre le signal produit par le transducteur 18 et le signal reçu par le microphone interne 36 ;
• d : signal de bruit environnant (signal de bruit à atténuer, idéalement à neutraliser par le contrôle actif) ;
• e : signal d'erreur délivré par le microphone interne 36 (signal que l'on souhaite minimiser),
• H_FF : fonction de transfert du filtre feedforward 32 (qui est une fonction fixe, non adaptative),
• H_FB : fonction de transfert du filtre feedback 42 (qui est également une fonction fixe), le cas échéant modifiée par la mise en oeuvre du filtre stabilisateur 44.

The following notations will be used:

• H _c : transfer function between the signal received by the external microphone 28 and the signal received by the internal microphone 36, representative of the fraction of the external noise that has passed through the shell of the earphone of the helmet;
• H _o : transfer function between the signal reproduced by the transducer 18 and the signal received by the external microphone 28, representative of the acoustic signal fraction that has been transmitted by the shell of the earphone to the external microphone;
• H _a : transfer function between the signal produced by the transducer 18 and the signal received by the internal microphone 36;
• d : surrounding noise signal (noise signal to be attenuated, ideally to be neutralized by the active control);
• e : error signal delivered by the internal microphone 36 (signal that it is desired to minimize),
• H _FF : feedforward filter transfer function 32 (which is a fixed, non-adaptive function),
• H _FB : transfer function of the feedback filter 42 (which is also a fixed function), possibly modified by the implementation of the stabilizing filter 44.

If it is desired to represent the error signal e as a function of the noise signal d, the following transfer function of the external microphone 28 is obtained to the internal microphone 38 (this microphone, located closest to the ear canal of the user, representing the perceived signal at the listening point): $$\frac{e}{d}=\frac{H_{\mathrm{vs}}+H_{\mathit{FF}}\left(H_{\mathrm{at}}+H_o\right)+\epsilon }{1-H_{\mathit{FB}}\left(H_{\mathrm{at}}+H_o\right)}$$

The term ε represents all loopbacks of orders greater than or equal to 2; concretely, this term is negligible in the other terms of the numerator and will be ignored. By the way | H _o | << | H _a |, because H _o contains the additional attenuation due to the shell of the earpiece.

The Figure 8 illustrates an example of a survey as a function of the frequency of the H _o / H _a module , which therefore represents the attenuation of the shell of the earpiece relative to the internal attenuation of the cavity.

If we make the approximation | H _o | << | H _a | , the noise transfer function can therefore be simplified to give: $$\frac{e}{d}=\frac{H_{\mathrm{vs}}+H_{\mathit{FF}}{H}_{\mathrm{at}}}{1-H_{\mathit{FB}}{H}_{\mathrm{at}}}$$

In order for the noise collected by the internal microphone to be low, ie to minimize the error signal, it is necessary that: $$\frac{e}{H_{\mathrm{vs}}d}=\frac{1+\frac{H_{\mathrm{at}}{H}_{\mathit{FF}}}{H_{\mathrm{vs}}}}{1-H_{\mathrm{at}}{H}_{\mathit{FB}}}$$

From a stability point of view, the stability of the feedforward H _FF is greater than that of the H _FB feedback due to the absence of a feedback loop (the feedforward operates in an open loop).

On the other hand, as explained in the introduction, feedforward and feedback tend to produce an undesirable effect of noise amplification in a small band of frequencies beyond the noise suppression band, usually around 1 kHz ( "waterbed effect" ). In addition, with the feedback of the feedback filter , this effect can quickly run out and turn into a feedback effect.

However, although the feedforward is more stable, it can not be used without feedback, because the noise suppression it provides alone is less effective. Indeed, to have a perfect suppression, it would be necessary that H _FF = H _c / H _has , which is difficult to achieve because H _c and H _a are very variable for the reasons mentioned above: variable and reduced front volume, position and tightening of the helmet, etc. In practice, the noise suppression by a feedforward filter alone is typically close to 10 dB, while with a feedback it is possible to reach 20 dB.

• 1°) de réduire la bande de fréquences du filtre feedback, de manière à augmenter les marges de gain et de phase (typiquement à au moins 15 dB et 60°), notamment dans la zone fréquentielle où se situent les risques d'instabilité incontrôlée ;
• 2°) de compenser par le filtre feedforward la baisse corrélative de performances dans les plus hautes fréquences (jusqu'à 1 kHz) ; et
• 3°) de réduire l'effet waterbed par un filtre stabilisateur associé au filtre feedback, ce qui permet de diminuer, voire de supprimer, les risques d'effet Larsen.

The invention makes it possible to overcome the disadvantages that have just been described. Essentially, the invention proposes:

• 1 °) to reduce the frequency band of the feedback filter , so as to increase the gain and phase margins (typically at least 15 dB and 60 °), especially in the frequency zone where the risks of uncontrolled instability are located ;
• 2 °) to compensate by the feedforward filter the correlative decrease of performances in the highest frequencies (up to 1 kHz); and
• 3 °) reduce the waterbed effect by a stabilizing filter associated with the feedback filter , which reduces or even eliminate the risk of feedback.

It should be noted that the increase of the gain and phase margins by a reduction of the bandwidth was chosen in preference to a reduction of the open-loop gain (which would also have made it possible to increase these margins): a reduction of the gain in An open loop has the disadvantage of reducing the maximum performance of the active noise control, as opposed to a reduction in the bandwidth, which only reduces the attenuation frequency band of the noise control circuit. Therefore, in order not to reduce the maximum attenuation of the noise, the reduction of the bandwidth has been chosen in preference to a reduction of the open loop gain of the feedback filter .

As far as the feedforward filter is concerned , it is more stable because it operates in an open loop. It can therefore be used in the highest frequencies (up to 1 kHz) to compensate for the bandwidth loss of the feedback filter . This feedforward filter has a low gain and a low quality factor compared to the feedback filter , and its performance is adjusted to cover a wide band of frequencies.

The Figure 12 illustrates the Bode (amplitude and phase versus frequency) diagram of this feedforward 32 filter.

Regarding the stabilizing filter 44, its paralleling with the feedback filter 42 increases the phase margin of the latter in particular in the critical zone of the waterbed effect . And to compensate for the phase decrease due to acoustics, in particular because of the sound propagation path of the sound of the transducer to the error microphone (transfer function Ha ) the stabilizing filter creates a local resonance in this area allowing increase the phase and thus increase the phase margin.

These different aspects are visible in particular on the examples of diagrams of Figures 13 to 17 .

The Figure 13 illustrates the place of Black of the system, that is to say the Cartesian representation of the module of the open loop ( H _to H _FB ) according to its phase, by varying the frequency from 0 Hz to infinity. By drawing this place of Black, we can easily read the gain and phase margins, which are given by the intersection of the place with the two axes passing through the point of instability O, located at 0 dB and 0 °.

On the Figure 13 , the place of Black was dotted with the feedback filter alone before reducing the bandwidth, and in continuous line this same filter but with the reduced bandwidth (but without the stabilizer). Initially, the gain and phase margins ΔM and Δφ are respectively -12 dB and 25 °, and we see that the reduction in the bandwidth makes it possible to increase these values to respectively +18 dB and more than 60 ° .

• Figure 14 : le module de la fonction de transfert du filtre feedback ;
• Figure 15 : la phase de la fonction de transfert du filtre feedback ;
• Figure 16 : le tracé de Nyquist ;
• Figure 17 : la caractéristique d'atténuation, en boucle fermée.

The Figures 14 to 17 represent, pure the circuit of the figure 7 :

• Figure 14 : the module of the transfer function of the feedback filter;
• Figure 15 : the phase of the transfer function of the feedback filter;
• Figure 16 : the Nyquist route;
• Figure 17 : the attenuation characteristic, in a closed loop.

• en A la caractéristique correspondant au filtre feedback originel avec sa préamplification G3, avant réduction de la bande passante ;
• en B cette même caractéristique A, mais après réduction de la bande passante ;et
• en C la caractéristique finale, c'est-à-dire la caractéristique B après adjonction du filtre stabilisateur 44 avec sa préamplification G2.

We have shown in these figures:

• in A the characteristic corresponding to the original feedback filter with its preamplification G3, before reducing the bandwidth;
• at B this same characteristic A, but after reduction of the bandwidth, and
• in C the final characteristic, that is to say the characteristic B after addition of the stabilizing filter 44 with its preamplification G2.

As can be seen by comparing on the Figure 14 characteristics A and B (or C), the original bandwidth of the filter which was 80-160 Hz, a width of 80 Hz (characteristic A) was reduced to 65-115 Hz, a width of 50 Hz, reduced (characteristic B or C).

This reduction in bandwidth makes it possible, as we had seen in the examination of the Figure 13 , significantly increase the gain and phase margins and thus contribute to increased system stability.

The examination of Figure 15 shows that the use of the stabilizing filter 44 substantially increases, from about 30 to 35 °, the phase in the unstable zone of the waterbed effect , around 1 kHz.

We see on the Figure 16 that this phase increase significantly removes the open loop of the instability zone. This figure is a Nyquist plot on which the noise amplification zone N has been indicated in broken lines. As can be seen, in none of the three configurations A, B or C does the system surround the point of instability O: these systems are theoretically all stable. However, reducing the bandwidth of the feedback filter (switching from A to B) and its association with a stabilizing filter (switching from B to C) each time moves the trace of the point of instability, thus contributing to a better stability. overall system.

The theoretical attenuation curve of the Figure 17 shows that the area of the waterbed effect at 1 kHz is decreased. It can be noted that the waterbed effect area at 6 kHz is degraded but does not exceed 4 dB, like that at 1 kHz. This figure illustrates the simulated attenuation of the closed loop, where we can see a reduction in the depth of the waterbed effect in the 1 kHz zone from system A to system B (4 dB improvement) and the system B to system C (+3 dB improvement). The attenuation loss of about -5 dB found in the 100-800 Hz zone between system A and system B or C will be offset by the active control provided by the fixed feedforward filter 32.

Claims (8)

An audio headset, comprising two earphones (10) joined by a holding bar (12) and each comprising a sound reproduction transducer (18) of an audio signal to be reproduced, this transducer being housed in an acoustic cavity delimited by a shell (14) provided with a circumaural pad (16),
this headset having an active noise control system comprising: a first open-loop feedforward branch, with a first band-pass filter (32) receiving as input a signal delivered by an external microphone (28) adapted to pick up an acoustic noise (30) prevailing in the helmet environment; a second closed-loop feedback branch, with a second band-pass filter (42) receiving as input an error signal (e) delivered by a microphone (36) internal to the cavity; a third branch with a third filter (44); and a mixing circuit (46), receiving as input the signals delivered by the first, second and third filters as well as the audio signal to be reproduced (S), and outputting a signal capable, after amplification (54), of driving the transducer (18), characterized in that the active noise control is a non-adaptive control, the parameters of the first, second and third filters (32, 42, 44) being predetermined parameters; the third filter (44) is a stabilizing band-pass filter arranged in parallel with the second feedback branch , receiving as input the signal delivered by the internal microphone, and outputting a signal applied at the input of the combiner circuit,
this third filter being able to locally increase the phase of the transfer function of the second filter in a predetermined instability zone; the first, second and third branches are arranged in parallel, and the mixing circuit is a summing circuit (46) outputting a linear combination of the signals delivered by the first, second and third filters (32, 42, 44) and than at least a fraction of the audio signal to be reproduced (S), with weighting of the respective gains (G5-G8) applied to these signals. The headset of claim 1, wherein said predetermined instability zone is a waterbed effect zone around a frequency of 1 kHz. The headset of claim 1, wherein the upper cutoff frequency of the second filter is less than 150 Hz, preferably less than 120 Hz. The headset of claim 1, wherein the bandwidth of the second filter is less than 65 Hz, preferably less than 55 Hz. The headset of claim 1, wherein the gain margin of the feedback branch of the active noise control is at least 15 dB, preferably at least 17 dB. The headset of claim 1, wherein the phase margin of the feedback branch of the active noise control is at least 45 °, preferably at least 60 °. The headset of claim 1, wherein the audio signal to be reproduced (S) is input to both the second filter (42) and the summing circuit (46), the second filter receiving as input a signal obtained by combining (52) said error signal (e) delivered by the internal microphone (36) with at least a fraction of the audio signal to be reproduced. The headset of claim 7, wherein the audio signal to be reproduced (S) is not applied to the third filter (44).

Images