DESCRIPTION INTERCOM DEVICE

TECHNICAL FIELD The present invention relates to an improvement in the howling prevention function of an intercom device.

BACKGROUND ART

In the past, an intercom system has been widely used, which is .mainly composed of an outdoor intercom device installed at a house entrance and an indoor intercom device disposed inside the house so as to enable a dweller to talk with a visitor at the house entrance. The outdoor intercom device has a speaker unit for outputting the dweller's voice, and a microphone for receiving the visitor's voice, and converting it into corresponding electric signals. Similarly, the indoor intercom device has a speaker unit for outputting the visitor's voice, and a microphone for receiving the dweller's voice, and converting it into corresponding electric signals. In these intercom devices, since the speaker unit is located near the microphone, it is known that a howling phenomenon easily occurs when the audio output of the speaker unit is picked up by the microphone. Therefore, various means for preventing the howling phenomenon have been proposed.

For example, Japanese Patent Early Publication No. 11-41342 discloses a hands-free automobile telephone system capable of preventing the howling phenomenon with a relatively simple circuit configuration. As shown in FIG. 39, this system comprises a telephone device IIP, an outside speaker 13P disposed in an automobile to provide an audio output (S2) corresponding to an electric signal from the telephone device, an outside microphone 12P disposed in the automobile to mainly collect a voice (Sl) of a driver (DR), and a second microphone 15P located near the outside speaker 13P to receive the audio output (S2) of the outside speaker. In this regard, the audio output (S2) of the

outside speaker 13P is also picked up by the outside microphone 12P. To prevent the howling phenomenon, this system also has, a delay circuit 16P for delaying an output signal (S2) of the second microphone 15P by an audio propagation time corresponding to a distance between the outside speaker 13P and the outside microphone 12P, an inverting circuit 17P for inverting an output (S2 ⁷ ) of the delay circuit 16P, and an adder circuit 18P for performing an operation of adding an output (-82^ of the inverting circuit 17P to the output signal (S 1+82^ of the outside microphone 12P. By this operation, the signal component corresponding to the audio output of the outside speaker 13P can be cancelled from the output signal of the outside microphone 12P to prevent the howling phenomenon.

In addition, Japanese Patent publication No. 3226121 discloses an intercom device capable of providing a comfortable two-way communication. As shown in FIG. 4OA, this intercom device is mainly composed of a loudspeaker 3 IS disposed in a housing 4OS, a first microphone IIS for receiving a voice of a speaking person 1OS, a second microphone 2 IS having directional characteristics with low sensibility to the voice of the speaking person to preferentially receive an audio output of the loudspeaker 3 IS. In this intercom device, since the audio output of the loudspeaker 3 IS as well as the voice of the speaking person 1OS are received by the first microphone IIS, a signal component corresponding to the audio output of the loudspeaker 3 IS is mixed as noise in an output signal of the first microphone IIS.

To deal with this inconvenience, an adaptive noise reduction process is performed. That is, as shown in FIG. 4OB, an adaptive filter circuit 24S performs an adaptive control such that an output signal of the second microphone 2 IS is approximated to the signal component corresponding to the audio output of the loudspeaker 3 IS in the output signal of the first microphone I IS. Then, a synthesizing circuit 15S such as a subtraction circuit synthesizes an output signal of the adaptive filter circuit 24S with the output signal of the first microphone IIS to remove or reduce the unwanted signal component from the

output signal of the first microphone. An output of the synthesizing circuit 15S is transmitted to another intercom device through a telephone line L. This process is also useful to prevent the howling phenomenon caused by an increase in loop gain during the communication state. By the way, in the above-described conventional devices, the output of one of the microphones is delayed by the delay time determined according to sound velocity and the distance between the speaker and the other microphone, so that the unwanted signal component is cancelled to prevent the howling phenomenon. This delay time is constant without depending on the frequency of the audio output of the speaker under ideal conditions (e.g., point sound source, sealed housing structure, and no variation in CR of circuit constructions). However, in fact, since it is difficult to satisfy the ideal conditions, the delay time changes depending on the frequency of the audio output of the speaker. Therefore, the conventional devices can effectively cancel the unwanted signal component with respect to a relatively narrow frequency range of the audio output of the speaker. However, it is not enough to cancel the unwanted signal component over a wider frequency range of the audio output of the speaker. Thus, there is room for further improvements in preventing the howling phenomenon and achieving a comfortable communication by the intercom device.

SUMMARY OF THE INVENTION

Therefore, a primary concern of the present invention is to provide an intercom device, which has the capability of canceling unwanted signal components with respect to plural frequency bands of an audio output of a speaker to achieve an improvement in howling prevention function. That is, the intercom device of the present invention comprises: a speaker unit accommodated in a housing; a first microphone spaced from the speaker unit by a first distance, and configured to receive an audio output of the speaker unit;

a second microphone spaced from the speaker unit by a second distance different from the first distance, and facing outside of the housing; a signal processing unit configured to, when the audio output of the speaker unit is picked up by the second microphone, remove a signal component corresponding to the audio output of the speaker unit from an output of the second microphone by using an output of the first microphone and at least two delay times, which are determined from a difference between the first distance and the second distance with respect to different frequencies of the audio output of the speaker unit. As a preferred embodiment of the above intercom device, the signal processing unit comprises: a first A/D converter configured to perform an A/ D conversion to an output of the first microphone; a second A/D converter configured to perform an A/D conversion to an output of the second microphone; a timing control portion configured to control timings of performing the A/D conversions at the first A/D converter and the second A/D converter such that the timing of performing the A/D conversion at the second A/D converter is different from the timing of performing the A/D conversion at the first A/D converter by a first delay time, which is determined from the difference between the first distance and the second distance with respect to a first frequency of the audio output of the speaker unit; an estimating portion configured to estimate, from the output of the second A/D converter, a signal obtained when the output of the second microphone is A/D converted at the timing different from the timing of performing the A/D conversion at the first A/D converter by a second delay time, which is determined from the difference between the first distance and the second distance with respect to a second frequency of the audio output of the speaker unit;

a first level adjusting portion configured to perform an output level adjustment to at least one of outputs of the first A/D converter and the second A/D converter; a second level adjusting portion configured to perform an output level adjustment to at least one of outputs of the first A/D converter and the estimating portion; and an operating portion configured to determine a first difference between the outputs of the first A/D converter and the second A/D converter after the output level adjustment and a second difference between the outputs of the first A/D converter and the estimating portion after the output level adjustment. In place of the estimating portion, a third A/D converter may be used, which is configured to perform an A/D conversion to the output of the second microphone at a timing controlled by the timing control portion to be different from the timing of performing the A/D conversion at the first A/D converter by a second delay time, which is determined from the difference between the first distance and the second distance with respect to a second frequency of the audio output of the speaker unit. In this case, the second level adjusting portion is configured to perform an output level adjustment to at least one of outputs of the first A/D converter and the third A/D converter, and the operating portion is configured to determine a first difference between the outputs of the first A/D converter and the second A/D converter after the output level adjustment and a second difference between the outputs of the first A/D converter and the third A/D converter after the output level adjustment.

According to the present invention, despite a relatively simple circuit configuration, it is possible to cancel unwanted signal components, which are mixed in an output of the first microphone when the audio output of the speaker unit is picked up by the first microphone, over a wide frequency range of plural frequencies of the audio output of the speaker unit, and reliably provide a comfortable communication to a user of the intercom device. In addition, since an expensive A/D converter with a high-speed processing capability is not

needed in the intercom device of the present invention, there is a further advantage of achieving an improvement in cost performance.

When the signal processing unit has the estimating portion, it is preferred that the second A/D converter performs the A/D conversion to the output of the second microphone provided at every conversion period, and the estimating portion estimates the signal by use of two outputs successively provided from the second A/D converter according to a linear prediction analysis method.

In addition, it is preferred that the signal processing unit further comprises a first filtering portion disposed between the first A/D converter and the operating portion, a second filtering portion disposed between the second A/D converter and the operating portion, and a third filtering portion disposed between the estimating portion (or the third A/D converter) and the operating portion.

It is also preferred that the operating portion outputs a sum of the first difference and the second difference or an average of the first difference and the second difference.

As a further preferred embodiment of the intercom device, the second microphone is disposed such that the second distance is larger than the first distance, and each of the first level adjusting portion and the second level adjusting portion attenuates the output of the first A/D converter. In this case, it is particularly preferred that the first microphone is disposed such that its sound receiving portion faces a diaphragm of the speaker unit, and the second microphone is disposed to receive audio information from outside of the intercom device.

Another concern of the present invention is to provide the intercom device , which is preferably used in a specific wiring system described below. That is, the wiring system comprises at least one transmission line installed in a building structure to transmit electric power and information signals, and a plurality of base units each adapted in use to be mounted in a wall surface of the building structure, and connected to the transmission line. The intercom device is characterized by having a first connector detachably connectable to a second

connector formed in each of the base units. When the first connector is connected to the second connector, the information signals provided from the transmission line through the base unit are output as audio information from the speaker unit of the intercom device, and an output of the signal processing unit of the intercom device is sent out to the transmission line through the base unit.

In addition, it is preferred that the intercom device for the above wiring system further comprises a third connector, which is detachable connectable to a fourth connector formed in a function unit configured to provide at least one of functions for supplying electric power provided from the transmission line, outputting information signals provided from the transmission line, and sending out information signals to the transmission line.

Furthermore, the intercom device of the present invention may be used in the wiring system with a function unit having a third connector detachably connectable to the second connector of the base unit, and a fourth connector detachable connectable to the first connector of the intercom device. In this case, when the third connector is connected to the second connector, and the first connector is connected to the fourth connector, the function unit provides at least one of functions for supplying electric power provided from the transmission line through the base unit, outputting information signals provided from the transmission line through the base unit, and sending out information signals to the transmission line through the base unit, and the intercom device operates such that the information signals provided from the transmission line through the base unit and the function unit are output as audio information from the speaker unit, and the output of the signal processing unit is sent out to the transmission line through the base unit and the function unit.

In the intercom device for the wiring system of the present invention, it is preferred that the first connector performs an electric power transmission with the base unit by means of electromagnetic coupling, and/ or the first connector performs a signal transmission with the second connector by means of optical coupling.

Further characteristics of the present invention and advantages brought thereby will be clearly understood from the best mode for carrying out the invention described below.

BRIEF EXPLANATION OF THE DRAWINGS

FIG. 1 is a schematic perspective view of an intercom device according to a first embodiment of the present invention;

FIG. 2 is a block diagram of the intercom device;

FIG. 3 is a cross-sectional view of the intercom device; FIG. 4 is a cross-sectional view of a microphone substrate mounted in the intercom device;

FIG. 5 is a cross-sectional view of a bare chip;

FIG. 6A is a plan view of the microphone substrate, and FIG. 6B is a schematic circuit diagram of the microphone substrate; FIG. 7 is a circuit diagram of an impedance converting circuit;

FIG. 8 is a block diagram of a signal processing unit of the intercom device;

FIGS. 9A to 9F are explanatory diagrams showing operations of the signal processing unit;

FIG. 10 is an explanatory diagram showing a linear interpolation method; FIGS. 1 IA and 1 IB show phasor-vector diagrams used to explain the linear interpolation method;

FIGS. 12A and 12B show signal waveforms in a phase adjustment step of the signal processing unit;

FIGS. 13A and 13B show signal waveforms in an output-level adjustment step of the signal processing unit;

FIGS. 14A and 14B show signal waveforms in an operation step of the signal processing unit;

FIG. 15 is a diagram showing a relationship between cancel amount and frequency;

FIG. 16 is a block diagram of a signal processing unit of the intercom device according to a modification of the first embodiment;

FIG. 17 is a diagram showing a relationship between cancel amount and frequency, which is obtained by actual measurements; FIG. 18 is a diagram showing a relationship between cancel amount and frequency, which is obtained by using the signal processing unit of this modification;

FIG. 19 is a block diagram of a signal processing unit of an intercom device according to a second embodiment of the present invention; FIGS. 2OA to 2OH are explanatory diagrams showing operations of the signal processing unit of the second embodiment;

FIG. 21 is a phasor-vector diagram showing a residual signal of a speaker unit;

FIG. 22A is a diagram showing a relationship between amplitude ratio "a(ω)" and frequency, and FIG. 22B is a diagram showing a relationship between phase difference "θ(ω) "and frequency;

FIGS. 23A and 23B are diagrams respectively showing theoretical and measured cancel amounts according to a conventional method;

FIGS. 24A and 24B are diagrams respectively showing theoretical and measured cancel amounts according to the present invention; FIG. 25 is a block diagram of a signal processing unit of the intercom device according to a modification of the second embodiment;

FIG. 26 is a schematic diagram of a wiring system using an intercom device according to a third embodiment of the present invention;

FIG. 27 is a schematic circuit diagram of a base unit of the wiring system; FIG. 28 is a rear perspective view of the base unit;

FIG. 29A is a perspective view showing a gate housing and a main housing of the base unit, and FIG. 29B is a plan view of a module port of the gate housing;

FIG. 30 is a plan view of an attachment plate for mounting the base unit on a switch box; FIG. 31 is a schematic circuit diagram of the function unit;

FIG. 32 is a perspective view of the intercom device for the wiring system;

FIG. 33 is a schematic circuit diagram of the intercom device for the wiring system;

FIG. 34 is a perspective view of the intercom device connected to the function unit or the base unit;

FIG. 35 is an exploded perspective view of the base unit;

FIG. 36A is a front view of the function unit, FIG. 36B is a side view of the function unit, FIG. 36C is an exploded side view of the function unit, and FIG.

36D is a perspective view of a joining member; FIG. 37 is a perspective view illustrating how to mechanically couple the function unit to the base unit;

FIG. 38 is a schematic circuit diagram of an intercom device for a power line carrier type wiring system;

FIG. 39 is a block diagram of a conventional automobile telephone system; and FIGS. 4OA and 4OB are schematic diagram and block diagram of a conventional intercom device.

BEST MODE FOR CARRYING OUT THE INVENTION

Referring to the attached drawings, an intercom device and a wiring system using the intercom device of the present invention are explained in detail according to preferred embodiments. (FIRST EMBODIMENT)

As shown in FIGS. 1 and 2, an intercom device 1 of the present embodiment has a housing 10, a speaker unit 2 accommodated in the housing, a first microphone 3 disposed such that its sound receiving portion faces a diaphragm

20 of the speaker unit 2, a second microphone 4 disposed such that its sound receiving surface faces outside of the housing, and an audio processor 5 for preventing howling phenomenon. The housing 10 is formed with a case 11 with a rear opening and a cover 12 for closing the rear opening, as shown in FIG. 3. In FIG. 1, "SW" designates a switch button for activating the intercom device in a

communication state, "L" designates a signal transmission line connecting between intercom devices.

The audio processor 5 includes a signal processing unit 50, a communication unit 51, echo cancellation units (52, 53), and amplifying unit 54. Electric signals provided from the signal transmission line "L" are sent to the speaker unit 2 through the communication unit 51, the echo cancellation unit 53 and the amplifying unit 54, so that audio information corresponding to the electric signals is output from the speaker unit 2. On the other hand, output signals of the first and second microphones (3, 4) are sent to the signal transmission line "L" through the signal processing unit 50, the echo cancellation unit 52 and the communication unit 51. < Speaker unit>

The speaker unit 2 comprises a yoke 20 formed in a cylindrical shape with an aperture at its one end, and a circular supporting member 21 extending around the aperture of the yoke. The yoke 20 is preferably made of a ferrous material having a thickness of about 0.8 mm. As the ferrous material, for example, it is possible to use a cold-reduced carbon steel sheet (SPCC or SPCEN based on JIS G3141) or a soft magnetic iron (SUY based on JIS C2504).

A columnar permanent magnet 22 is placed in the yoke 20. It is preferred that the permanent magnet 22 has a remanent flux density of 1.39T to 1.43T. A dome-like diaphragm 23 is connected at its outer peripheral edge to the circular supporting member 21. For example, the permanent magnet 22 can be made of neodymium. In addition, the diaphragm 23 can be made of a thermoplastic resin such as polyethyleneterephthalate (PET) and polyetherimide (PEI). It is also preferred that a thickness of the diaphragm 23 is in a range of 12 to 50 μm. A tubular bobbin 24 is connected to a rear surface of the diaphragm 23. A voice coil 25 is formed at a rear end of the bobbin by winding a polyurethane enameled copper wire (e.g., φ 0.05 mm) around a kraft paper tube. The bobbin 24 and the voice coil 25 can be vibrated back and forth in the aperture of the yoke 20.

When an electric current flows in the polyurethane enameled copper wire of the voice coil 25, an electromagnetic force is generated in the voice coil 25 by a magnetic field of the permanent magnet 22, so that the bobbin 24 vibrates back and forth together with the diaphragm 23. At this time, the diaphragm 23 outputs audio information corresponding to the electric current. Thus, the speaker unit 2 of the intercom device of the present embodiment is an electro- kinetic type speaker.

The speaker unit 2 is mounted in the housing 10. The housing 10 has a circular rib 15 formed on its front inner surface. The speaker unit 2 is disposed in the housing 10 such that the outer peripheral portion of the circular supporting member 21 contacts the circular rib 15, and the diaphragm 23 faces the front inner surface of the housing. A space surrounded by a front surface (i.e., the diaphragm-side surface) of the speaker unit 2 and the front inner surface of the housing 10 is defined as a front air chamber Cl. In addition, a space surrounded by a rear surface (i.e., the yoke-side surface) of the speaker unit 2 and rear and side inner surfaces of the housing 10 is defined as a rear air chamber C2. The front air chamber Cl is communicated with the outside of the housing 10 through a plurality of sound holes 13 formed in the front surface of the housing. The rear air chamber C2 is isolated from the front air chamber Cl by allowing the circular supporting member 21 to tightly contact the circular rib 15. In addition, the rear air chamber C2 is provided by a hermetically sealed space, which is isolated from the outside of the housing 10 by closing the opening of the case 11 with the cover 12 in an air-tight manner.

Thus, since the rear air chamber C2 facing the rear surface of the speaker unit 2 is the hermetically sealed space in the housing, a sound radiated from the rear surface of the speaker unit 2 is hard to leak from the rear air chamber C2. Therefore, it is possible to prevent that the sound radiated from the rear surface of the speaker unit 2 is picked up by the second microphone 4 positioned at a location other than the rear air chamber C2 in the housing 10, and reduce acoustic coupling between the speaker unit 2 and the second microphone 4.

In addition, the sound radiated from the rear surface (i.e., the rear surface of the diaphragm) of the speaker unit 2 is in reverse in phase to the sound radiated from the front surface (i.e., the front surface of the diaphragm) of the speaker unit 2. Therefore, when the sound radiated from the rear surface of the speaker unit 2 is interfered with the sound output from the front surface of the speaker unit 2, they are cancelled to each other, so that a reduction in sound pressure of the speaker unit may occur. However, as described above, since the sound radiated from the rear surface of the speaker unit is hard to leak to the outside of the housing 10, it is possible to prevent that a reduction in sound pressure of the speaker unit is caused by the interference between the sounds radiated from the front and rear surfaces of the speaker unit 2. < First microphone and Second microphone >

As shown in FIG. 4, the first microphone 3 is formed with a bare chip 30, an integrated circuit (IC) 31, and a shield case 32. Similarly, the second microphone 4 is formed with a bare chip 40, an integrated circuit (IC) 41, and a shield case 42. The first and second microphones (3, 4) are formed on a module substrate 6. That is, the bare chips (30, 40) and the integrated circuits (31, 41) are mounted on predetermined positions of the module substrate 6, and then wirings W are formed between the bare chip 30 and the integrated circuit 31, between the bare chip 40 and the integrated circuit 41, and between the integrated circuits (31, 41) and circuit patterns (not shown) formed on the module substrate 6 by wire bonding. Finally, each of the shield cases (32, 42) is attached to the module substrate 6 to accommodate the bare chips (30, 40) and the integrated circuits (31, 41) therein. The bare chip (30, 40) of each of the first and second microphones (3, 4) provides a capacitor-type silicon microphone, as shown in FIG.5, which is mainly formed with a silicon thin film 61 formed on a top surface of a silicon substrate 60 such that an opening 62 of the silicon substrate is closed by the silicon thin film 61, an electrode 63 formed directly above the silicon thin film 61 through an air gap 64, and pads 65 for outputting electric signals as a microphone output.

For example, when the second microphone 4 receives audio information, the silicon thin film 61 of the second microphone 4 is vibrated, so that changes in electric capacitance (amounts of electric charges) between the silicon thin film 61 and the electrode 63 occur. At this time, the electric signals are output from the pads 65 according to the changes in electric capacitance.

With respect to the bare chip 30 of the first microphone 3, the silicon substrate 60 is mounted on the module substrate 6 by die bonding. The shield case 32 has a through hole 33, through which audio information is supplied to the bare chip 30. On the other hand, with respect to the second microphone 4, the bare chip 40 is mounted on the module substrate 6 such that the silicon thin film 61 faces a sound hole 43 formed in the module substrate 6. The shield case 42 of the second microphone 4 has no sound hole. Therefore, the second microphone 4 is used to receive audio information through the sound hole 43. In brief, the first microphone 3 has a sound receiving portion for receiving the audio information from the upper side of FIG. 4 through the through hole 33 of the shield case 32, and the second microphone 4 has a sound receiving portion for receiving the audio information from the lower side of FIG. 4 through the sound hole 43 of the module substrate 6. In this embodiment, since both of the first and second microphones (3, 4) are mounted on the same (top) surface of the module substrate 6, the module substrate with a reduced thickness is available.

As shown in FIG. 6A, the module substrate 6 is configured in a substantially T-shape, which is a small rectangular portion for mounting the first microphone 3, a large rectangular portion for mounting the second microphone 4, and a coupling portion extending therebetween. In addition, a negative pad Pl, positive pad P2, first output pad P3, and a second output pad P4 are formed on the large rectangular portion of the module substrate 6. As shown in FIG. 6B, a negative voltage is applied from an outside power supply to the negative pad Pl, and a positive voltage is applied from the outside power supply to the positive pad P2. The applied electric power is supplied to the first and second microphones (3, 4) through wiring patterns on the module substrate 6.

When the first and second microphones (3, 4) receive audio information, an electric current is generated from each of the bare chips (30, 40) according to the audio information, and then sent to the corresponding integrated circuit (31, 41). Each of the integrated circuits (31, 41) performs an impedance conversion to the electric current, and then converts into a voltage signal. The voltage signals of the integrated circuits (31, 41) are output from the first and second output pads (P3, P4). In brief, the first output pad P3 outputs the voltage signal corresponding to the audio information received by the first microphone 3, the second output pad P4 outputs the voltage signal corresponding to the audio information received by the second microphone 4. Since the pair of negative and positive pads (Pl, P2) are commonly used to supply the electric power to both of the first and second microphones (3, 4), and the negative pad Pl is used as a common ground for the voltage signals output from the first and second output pads (P3, P4), there is an advantage that a reduction in cost and size of the intercom device is achieved by a refined circuit design with a reduced number of pads.

As shown in FIG. 7, the integrated circuit (31, 41) of each of the first and second microphones (3, 4) has a constant voltage circuit of converting a supply voltage +V (e.g., 5V) supplied through the positive and negative pads (Pl, P2) into a constant voltage Vr (e.g., 12V). The constant voltage Vr is applied to a series circuit of a resistance Rl and the bare chip (30, 40). A midpoint of connection between the resistance Rl and the bare chip (30, 40) is connected to a gate terminal of a J-FET (Junction-type FET) device T through a capacitor Q. A drain terminal of the J-FET device T is connected to the power supply +V, and a source terminal of the J-FET device T is connected to the negative side of the supply voltage through a resistance R2. In this regard, the J-FET device T is used for impedance conversion. A voltage of the source terminal of the J-FET device T is the microphone output. The impedance conversion circuit of the integrated circuit (31, 41) is not limited to the above. For example, a circuit having a function of a source follower circuit by an operational amplifier may be used. In

addition, an amplifying circuit may be formed in the integrated circuit (31, 41), if necessary.

The module substrate 6 having the first and second microphones (3, 4) is mounted on a front outer surface of the housing such that the sound receiving portion (i.e., the through hole 33 of the shield case 32) of the first microphone is exposed to the front air chamber Cl, and also the sound receiving portion of the second microphone 4 faces the outside of the housing 10 through the sound hole 43. Therefore, the first microphone 3 is used to receive an audio output of the speaker unit 2, and the second microphone 4 is used to receive the audio information from the outside of the housing 10. To efficiently receive the audio output of the speaker unit 2, it is particularly preferred that the sound receiving portion of the first microphone 3 is positioned to face the diaphragm 23 of the speaker unit 2 in the front air chamber Cl.

For example, when the intercom device 1 is placed at a house entrance, the first microphone 3 receives dweller's voice information as the audio output of the speaker unit 2 with higher sensitivity than a voice of a speaking person (e.g., visitor's voice information) at the house entrance. On the other hand, the second microphone receives the visitor's voice information with higher sensitivity than the dweller's voice information provided from the speaker unit 2. In brief, the output of the first microphone 3 has large amplitude with respect to the audio output of the speaker unit 2, and the output of the second microphone 4 has large amplitude with respect to the audio information provided from the output of the housing 10.

In addition, since the audio output of the speaker unit 2 is provided outside through the sound holes 13 formed in the front surface of the housing 10, there is a possibility that the audio output of the speaker unit 2 is picked up by the second microphone 4. In this regard, as shown in FIG. 3, the first microphone 3 is spaced from a center of the speaker unit 2 by a first distance Xl, and the second microphone 4 is spaced from the center of the speaker unit 2 by a second distance X2 (XKX2) larger than the first distance. Therefore, a timing where

the second microphone 4 receives the audio output of the speaker unit 2 is delayed than the timing where the first microphone 3 receives the same audio output of the speaker unit 2 by a delay time Td. As a result, a phase difference occurs between outputs of the first and second microphones (3, 4) in the case of receiving the audio output of the speaker unit. In general, the delay time Td is calculated by the following equation:

Td = (X2-X1) /Vs wherein Vs is sound velocity. Since the first distance Xl and the second distance X2 are predetermined, the delay time Td can be regarded as constant under ideal conditions (e.g., point sound source, sealed housing structure, and no variation in CR of circuit constructions). However, in fact, it is difficult to satisfy these ideal conditions, the delay time Td changes depending on frequency of the audio output of the speaker unit 2.

On the other hand, when the first and second microphones (3, 4) receive audio information provided from the outside of the housing 10, for example, the visitor's voice is received by, it can be regarded that the output of the second microphone 4 has substantially the same phase as the output of the first microphone 3 because a distance between the visitor and the first microphone 3 is substantially equal to the distance between the visitor and the second microphone 4.

Moreover, when the first and second microphones (3, 4) receive the audio output of the speaker unit 2, an amplitude of a signal component corresponding to the audio output of the speaker unit in each of the output signals of the first and second microphones are constant under the ideal conditions. However, in the fact, the amplitude changes depending on the frequency of the audio output of the speaker unit 2. < Signal processing unit>

In these circumferences, the present invention proposes more effectively preventing the howling phenomenon by considering the frequency dependency of the delay time, and provides the signal processing unit 50 configured to, when

the audio output of the speaker unit 2 is picked up by the second microphone 4, remove a signal component corresponding to the audio output of the speaker unit 2 from an output of the second microphone 4 by using an output of the first microphone 3 and at least two delay times, which are determined from a difference between the first distance Xl and the second distance X2 with respect to different frequency bands of the audio output of the speaker unit 2.

As shown in FIG. 8, the signal processing unit 50 of this embodiment, in which the outputs of the first and second microphones (3, 4) are input, is characterized by comprising a first A/D converting circuit 70 connected to the first microphone 3, a second A/ D converting circuit 71 connected to the second microphone 4, a timing control circuit 72 for determining each of timings of performing A/D conversions at the first and second A/D converting circuits (70, 71), an estimating circuit 73A connected to the second A/D converting circuit 71, bandpass filters (74A, 74B) connected to the first A/D converting circuit 70, bandpass filters (75A, 75B) connected to the second A/D converting circuit 71 and the estimating circuit 73A, attenuating circuits (76A, 76B) connected to bandpass filters (74A, 74B), and an operation circuit 77. Operations of these components of the signal processing unit 50 are explained below in detail.

The first A/D converting circuit 70 converts the output signal (an analog signal) of the first microphone 3 into a digital signal. The second A/D converting circuit 71 converts the output signal (an analog signal) of the second microphone 4 into a digital signal. The timing of performing the A/D conversion at the first A/D converting circuit 70 is synchronized with a rising edge of a control signal ScI output from the timing control circuit 72. On the other hand, the timing of performing the A/D conversion at the second A/D converting circuit 71 is synchronized with a rising edge of a control signal Sc2 output from the timing control circuit 72. In this embodiment, each of the control signals has a frequency of 8 to 16 KHz, and the A/D conversion is performed at every predetermined conversion period Tc in a range of 62.5 to 125 μsec in each of the first and second A/D converting circuits (70, 71).

When the first and second microphones (3, 4) receive the audio output of the speaker unit, the phase of the output signal of the second microphone is delayed than the phase of the output signal of the first microphone by a delay time TdI with respect to a signal component of a first frequency fl (e.g., IKHz) in the audio output of the speaker unit. Therefore, the timing control circuit 72 outputs the control signal Sc2 after the elapse of the delay time TdI from the output of the control signal ScI, as shown in FIGS. 9C and 9D. In this regard, FIG. 9 A shows a signal waveform Sl of the output of the first microphone 3, and FIG. 9B shows a signal waveform S2 of the output of the second microphone 4. In addition, FIGS. 9E and 9F show digital values obtained by the A/D conversions at the first and second A/D converting circuits (70, 71), respectively. Thus, by considering the delay time TdI, it is possible to perform the A/D conversions to the output signals of the first and second microphones (3, 4) under the same phase condition with respect to the signal component of the first frequency fl .

In addition, since the first and second A/D converting circuits (70, 71) of this embodiment use a sigma-delta (σ δ) method, the A/D conversion can be performed with a high S/ N ratio by a noise shaping where frequency transfer characteristics are changed with respect to only noise components such that the noise components are distributed in a higher frequency region than the signal components. In addition, by digitalizing most of treatments, it becomes easy to design an integrated circuit for the signal processing unit 50. A resolving power of the A/D conversion can be selected from one of 16 bits, 14 bits and 12 bits.

As described above, the first A/D converting circuit 70 outputs the digital value (al, bl, cl ) at every conversion period Tc. On the other hand, the second A/D converting circuit 71 outputs the digital value (a2, b2, c2 ) at every conversion period Tc. With respect to the first frequency fl, the digital value of the first microphone 3 is the same in phase as the digital value of the second microphone 4. For example, the digital value al of the first microphone 3 is the same in phase as the digital value a2 of the second microphone 4.

Similarly, the digital value bl of the first microphone 3 is the same in phase as the digital value b2 of the second microphone 4. However, since the delay time changes depending on the frequency of the audio output of the speaker unit 2, the digital value of the first microphone 3 is not the same in phase as the digital value of the second microphone 4 with respect to a second frequency (e.g., 3KHz) different from the first frequency fl. Therefore, it is needed to separately carry out the phase adjustment by considering a delay time in the case of the second frequency.

That is, in this embodiment, the output signal of the first A/D converting circuit 70 is sent to the bandpass filter 74A through a main signal channel, and also sent to the bandpass filter 74B through a branch signal channel branched from the main signal channel. On the other hand, the output of the second A/D converting circuit 71 is sent to the bandpass filter 75A through a main signal path, and also input in the estimating circuit 73A through a branch signal channel branched from the main signal path.

The estimating circuit 73A estimates, from the output signal of the second A/D converting circuit 71, a signal obtained when the output of the second microphone 4 is A/D converted at a timing different from the timing of performing the A/D conversion at the first A/D converting circuit 70 by a delay time Td2, which is different from the delay time TdI, and is determined from the difference between the first distance and the second distance with respect to the second frequency. In this embodiment, the estimating circuit 73A estimates the signal according to a linear interpolation method, as explained below.

For example, the digital values (a2, b2) output from the second A/D converting circuit 71 are positioned on the waveform S2 of the output signal of the second microphone 4, as shown in FIG. 10. The digital value a2 is obtained by performing the A/D conversion to the output signal of the second A/D converting circuit 71 at a timing tl. The digital value b2 is obtained by performing the A/D conversion to the output signal of the second A/D converting circuit 71 at a timing t2 after the elapse of the conversion period Tc from the

timing tl. In this situation, the estimating circuit 73A estimates a digital value obtained by performing the A/ D conversion to the output signal of the second microphone 4 at a timing t3 earlier than the timing t2 by a time lag Th2. In this regard, the time lag Th2 corresponds to a phase lag between the signal components of the first and second frequencies of the audio output of the speaker unit 2. In fact, the digital value to be estimated is on a position a3 of the signal waveform S2. However, since it is difficult to determine the digital value a3 by calculations, the estimating circuit 73A determines a digital value a3* as an approximate value of the digital value a3 according to the linear interpolation method. That is, as shown in FIG. 10, the approximate digital value a3 is found at a position corresponding to the timing t3 on a straight line S2' connecting between the digital values a2 and b2.

That is, in the waveform (FIG. 10) of the output signal of the second microphone 4, the linear interpolation method comprises the steps of drawing a straight line (e.g., S2') between two digital values (e.g., (a2, b2) or (b2, c2)) successively provided at every conversion period (e.g., Tc) from the second A/ D converting circuit 71, estimating a timing (e.g., t3) by use of the time lag (e.g., Th2) previously measured with respect to the second frequency, and determining an intersection point (e.g., a3') between the straight line (e.g., S2') and a vertical line (e.g., the dotted line Lv in FIG. 10) corresponding to the estimated timing (t3) This intersection point can be regarded as an approximate value of the actual digital value, which is obtained by performing the A/ D conversion to the output signal of the second microphone 4 at the estimated timing with respect to the second frequency. Therefore, by performing this estimation method at every conversion period, the estimating circuit 73A can provide the approximate digital value (a3', b3', c3'....) in place of the actual digital value (a3, b3, c3....) without using an additional A/ D converting circuit for performing the A/ D conversion to the output signal of the second microphone 4 with respect to the second frequency. In the above explanation, the timing t3 is earlier than the timing t2

by the time lag Th2. However, according to the magnitude relation between the first and second frequencies, the timing t3 may be later than the timing t2.

Thus, with respect to the first frequency component of the audio output of the speaker unit 2, the output signal of the first A/ D converting circuit is sent to the bandpass filter 74A to extract signal components of a first frequency band Gl including the first frequency fl therefrom. On the other hand, the output signal of the second A/ D converting circuit, which is obtained by performing the A/ D conversion to the output of the second microphone 4 at the different timing from the timing of performing the A/ D conversion at the first A/ D converting circuit 7OA by the delay time TdI, is sent to the bandpass filter 75A to extract signal components of the first frequency band Gl including the first frequency fl therefrom.

In addition, with respect to the second frequency component of the audio output of the speaker unit 2, the output signal of the first A/ D converting circuit is sent to the bandpass filter 74B to extract signal components of a second frequency band G2 including the second frequency therefrom. On the other hand, the output signal of the estimating circuit 73A, which is an approximate value of the digital value obtained by performing the A/ D conversion to the output of the second microphone 4 at the different timing from the timing of performing the A/ D conversion at the first A/ D converting circuit 7OA by the delay time Td2, is sent to the bandpass filter 75B to extract signal components of the second frequency band G2 including the second frequency therefrom.

As described above, the intercom device of this embodiment has the estimating circuit 73A for performing the phase adjustment with respect to the second frequency. If necessary, two or more of the estimating circuits may be added to the signal processing unit 50 shown in FIG. 8. For example, as shown by the dotted line in FIG. 8, a second estimating circuit 73B for performing the phase adjustment with respect to a third frequency different from the first and second frequencies can be added. In this case, the output of the second A/D converting circuit 71 is also input in the second estimating circuit 73B through

an extended branch signal channel. In addition, the branch signal channel is extended to send the output signal of the first A/ D converting circuit 70 to a bandpass filter 74C. According to the same manner as the estimating circuit 73A, the second estimating circuit 73B estimates a digital value obtained by performing the A/ D conversion to the output signal of the second microphone 4 at a timing t4 earlier or later than the timing t2 by a time lag Th3, which corresponds to a phase lag between the signal components of the first and third frequencies of the audio output of the speaker unit 2.

In addition, when the second estimating circuit 73B is formed in the signal processing unit 50 of FIG. 8 with respect to the third frequency component of the audio output of the speaker unit 2, the output signal of the first A/ D converting circuit is sent to the bandpass filter 74C to extract signal components of a third frequency band G3 including the third frequency therefrom. On the other hand, the output signal of the second estimating circuit 73B, which is an approximate value of the digital value obtained by performing the A/ D conversion to the output of the second microphone 4 at the different timing from the timing of performing the A/ D conversion at the first A/ D converting circuit 7OA by a delay time Td3, is sent to the bandpass filter 75C to extract signal components of the third frequency band G2 including the third frequency therefrom. Next, the linear interpolation method described above is explained by use of mathematical formulas. In the case of FIG. 10, the approximate value a3' is determined by the following equation (1): a3' = b2 + (a2-b2) -Th2/Tc - - - ( D

When the audio output of the speaker unit 2 is represented by "exp (jωt) ", the above equation can be generalized as below. ah = exp(jωt) -{( 1 + (exp(-j ωt) - 1 ) -Thm/To) } ^{■ ■ ■} (2) In this regard, "ah" is a general symbol of the approximate value, and "Thm (m=2 ~ n)" is a time log previously measured with respect to each of different frequencies (f2~fh) of the audio output of the speaker unit. For example, in the case of the estimating circuit 73A, the time log is Th2 (m=2). The time log Th2 is

previously measured with respect to the second frequency of the audio output of the speaker unit 2. In addition, when the digital value a2 is provided by "exp (jωt) ", the digital value b2 corresponds to "exp (jω(t+Tc)) ".

Therefore, in the appropriate value ah, amplitude and phase change by the term of { ( 1 + (exp (—jωt) — 1 ) ^• Thm/Tc) . An amplitude characteristic Ah (ω) is represented by the equation (3). In addition, a phase characteristic θh (ω) is represented by the equation (4).

Ah ( ω ) - " O)

— Thnrvsin ( ωt) θ h ( ω ) = arctan (4)

{Tc+ (cos ( ωt- 1 )-Thm}

As an example, relations among the appropriate value (i.e., estimated digital value) a3', the actual digital value a3, and the digital value b2 in FIG. 10 are shown by use of phasor vectors in FIGS. 1 IA (Thm<Tc/2) and 1 IB (Thm>Tc/2). According to the estimating circuit 73A using the linear interpolation method, it is possible to easily perform the phase adjustment by use of the output signal of the second microphone 4 with respect to each of different frequency components of the audio output of the speaker unit 2. In addition, even when the number of the estimating circuits is increased, it is enough to use only two (first and second) A/D converting circuits (70, 71) because the digital values provided from the second A/D converting circuit 71 are commonly used in the estimating circuits. Therefore, there are advantages of reducing the circuit components of the intercom device, and consequently achieving an improvement in cost performance of the intercom device. An output of the bandpass filter 74A connected to the first A/D converting circuit 70 is the same phase as the output of the bandpass filter 75A connected to the second A/D converting circuit 71 with respect to the first frequency component of the audio output of the speaker unit 2. By these bandpass filters (74A, 75A), the signal components for the first frequency band Gl including the

first frequency fl is extracted. On the other hand, as shown in FIGS. 12A and 12B, an output of the bandpass filter 74B connected to the first A/D converting circuit 70 is the same phase as the output of the bandpass filter 75B connected to the estimating circuit 73A with respect to the second frequency component of the audio output of the speaker unit 2. By these bandpass filters (74B, 75B), the signal components for the second frequency band G2 including the second frequency f2 is extracted. In addition, when the intercom device has the second estimating circuit 73B, an output of the bandpass filter 74C connected to the first A/ D converting circuit 70 is the same phase as the output of the bandpass filter 75C connected to the second estimating circuit 73B with respect to the third frequency component of the audio output of the speaker unit 2. By these bandpass filters (74C, 75C), the signal components for the third frequency band G3 including the third frequency is extracted. In fact, digital signals provided from the first and second A/D converting circuits (70, 71) are temporarily stored in a memory, and then a subsequent digital treatment is performed. For purposes of explanation, FIGS. 12A and 12B, FIGS. 13A and 13B and FIGS. 14A and 14B are shown by analog waveforms.

The output of each of the bandpass filters (74A, 74B, 74C) connected to the first A/D converting circuit 70 is attenuated by the attenuating circuit (76A, 76B, 76C) to perform an output level adjustment. In this embodiment, since the first microphone 3 is disposed to face the diaphragm 23 of the speaker unit 2, the output signal of the first microphone 3 has larger amplitude with respect to the audio output of the speaker unit 2 than the output signal of the second microphone 4. In addition, the amplitude difference is also caused by a difference in sensitivity between the first and second microphones (3, 4). Therefore, by forming the attenuating circuit (76A, 76B, 76C) at the downstream side of the first microphone 3, the output level of the bandpass filter (74A, 74B, 74C) connected to the first A/D converting circuit 70 can be substantially matched with the output level of the bandpass filter (75A, 75B, 75C) with

respect to each of the frequency bands (Gl, G2, G3), as shown in FIGS. 13A and 13B.

In the operation circuit 77, a difference between the output of the bandpass filter 75A and the output of the attenuating circuit 76A is calculated, so that the signal component for the frequency band Gl of the audio output of the speaker unit 2 is cancelled from the output of the second microphone 4. In addition, a difference between the output of the bandpass filter 75B and an output of the attenuating circuit 76B are calculated, so that the signal component for the frequency band G2 of the audio output of the speaker unit 2 is cancelled from the output of the second microphone 4, as shown in FIGS. 14A and 14B. Furthermore, when the second estimating circuit 73B is used, a difference between the output of the bandpass filter 75C and an output of the attenuating circuit 76C are calculated, so that the signal component for the frequency band G3 of the audio output of the speaker unit 2 is cancelled from the output of the second microphone 4. In FIG. 14A, each of the outputs of the attenuating circuits (76A, 76B, 76C) is inversed, and then added to the output of the bandpass filter (75A, 75B, 75C) to determine the difference therebetween. Next, a total of the differences for all of the frequency bands (Gl, G2, G3) is determined as an output of the signal processing unit 50. In the output of the signal processing unit 50, the signal components for a wide frequency band (Gl, G2, G3) of the audio output of the speaker unit 2 are removed from the output of the second microphone 4.

Next, the howling preventing effect of the signal processing unit 50 is theoretically explained by using mathematical formulas. When the audio output of the speaker unit 2 is represented by "exp (jωt) ", and a difference between the outputs of the first and second microphones (3, 4) is determined without performing the phase adjustment therebetween, a signal component corresponding to the audio output of the speaker unit 2 remaining in the output of the signal processing unit 50 is represented by the following equation (5). In

addition, the amplitude characteristic Ul is represented by the following equation (6).

exp(jωt)-exp( — j ω -Thm) — exp(jωt)

= exp(jωt)-{exp(— jω-Thm) — 1}

= exp(jωt) -{(cos(ω-Thm) — 1)— j(sin(ω -Thm))} "-(S)

U1 = {(cos(ω-Thm)-D-j(sin(ω-Thm))} ^■•■ (6)

On the other hand, when the difference between the outputs of the first and second microphones (3, 4) is determined under the condition that the phase and the amplitude of the output signal of the second microphone 4 are matched with them of the output signal of the first microphone 3 by using the estimating circuits using the linear interpolation method described above, the signal component corresponding to the audio output of the speaker unit 2 remaining in the output of the signal processing unit 50 is represented by the following equation (7). In addition, the amplitude characteristic U2 is represented by the following equation (8) .

exp(jωt)-exp(— jω -Thm)— exp(jωt) -{1 + (exp(— jωt) — 1 ) -ThInXTcJ = exp (j ωt) ^■ [{(cos ( ω -Thrn) — Thm ^■ cos ( ω -Tc) /Tc) —

(1 —ThmXTc)}—j(sin(ω-Thm)—Thm-sin(ω -Tc)XTc)] ^■■■ (7)

U2 = [{(cos (ω -Thm)— Thm-cos(ω -Tc)XTc)-(I-THmXTc)I

— j (sin (ω -ThIm)-ThITT sin (ω -Tc) XTc)] ^{■ ■ ■} (8) The time lag Thm and the A/ D conversion period Tc are the same as them shown by the above equations (1) to (4). Therefore, duplicate explanation is omitted. In addition, [the amplitude characteristic I Ul | ² — the amplitude characteristic | U2 | ² ] is represented by the following equation (9). In this regard, D = Thm/Tc.

|U1| ² — |U2| ² = 2D (1 -D) (1+cos( OJ-Tc)) +

2D{cos(ω -TcX2) -cos{ω ((TcX2) -Thm) }} ^{■ ■ ■} (9)

The first term of the right side member of the equation (9), i.e., [2D (1— D)

(1+ cos (coTc) ) ] is positive under the condition of 0<ωTc< π . In addition, the second term of the right side member of the equation (9), i.e., [2D {cos

(ωTc/2) cos {ω ( (Tc/2) — Thm) } } ] is positive under the conditions of 0 <ωTc< π and Thm< (Tc/2) . Therefore, [ | Ul | ² — | U2 | ² ] is positive. It shows that the signal component corresponding to the audio output of the speaker unit 2 remaining in the output of the signal processing unit 50 is decreased in the case of performing the phase adjustment by use of the estimating circuits, as compared with the case of not performing the phase adjustment. As a result, an improvement in howling preventing effect can be achieved by the present invention. The above explanation is based on the relation of Thm<Tc/2. However, the same effect is obtained in the case of Thm >Tc/2.

By the way, when the intercom device 1 is used at the house entrance, the audio information (e.g., visitor's voice) provided from the outside of the intercom device is received by the second microphone 4 with high sensitivity. On the other hand, the visitor's voice may be picked up by the first microphone 3 for receiving the audio output of the speaker unit 2 with high sensitivity. In this case, the output signal of the second microphone 4 is much larger in amplitude with respect to the visitor's voice than that of the first microphone 3. Moreover, as described above, since the output signal of the first microphone 3 is attenuated by the attenuating circuit (76A, 76B, 76C), the output of the bandpass filter (75A, 75B, 75C) connected to the second microphone 4 relatively increases in amplitude with respect to the visitor's voice than that of the attenuating circuit (76A, 76B, 76C). Therefore, even when the subtraction treatment is performed in the operation circuit 77, the signal component corresponding to the visitor's voice remains in the output of the signal processing unit 50 with sufficient magnitude.

In the output of the signal processing unit 50, the signal component corresponding to the audio output of the speaker unit 2 is uniformly reduced

over a wide frequency range as a sum of the plural frequency bands (Gl, G2, G3), and the signal component corresponding to the visitor's voice is maintained with sufficient amplitude. As a result, it is possible to effectively prevent the howling phenomenon, which is caused when the audio output of the speaker unit 2 is picked up by the second microphone 4, while maintaining a comfortable communication environment.

As an example, FIG. 15 shows results where the signal component corresponding to the audio output of the speaker unit 2 was cancelled by use of the signal processing unit 50 of this embodiment. In this example, the signal component was cancelled with respect to two frequency bands (Gl, G2). That is, the signal component corresponding to the audio output of the speaker unit was cancelled with respect to the first frequency band Gl of 2KHz or less including the first frequency of 1 KHz from the output of the second microphone 4. In addition, the signal component corresponding to the audio output of the speaker unit was cancelled with respect to the second frequency band G2 of more than 2KHz including the second frequency of 3KHz from the output of the second microphone 4. In FIG. 15, "Yl" shows the cancel amounts obtained by the signal processing unit 50, and "Y2" is an approximated curve of "Yl". The approximated curve Y2 has two peaks at the frequencies of about IKHz and about 3KHz. Thus, it has been demonstrated that the signal component corresponding to the audio output of the speaker unit 2 can be cancelled from the output of the second microphone 4 with respect to each of the different frequency bands by the signal processing unit 50 of the present invention.

An output of the signal processing unit 50 is sent to the communication unit 51 through the echo cancellation unit 52, and the communication unit 51 transmits an electric signal corresponding to the audio information of the speaking person (e.g., visitor's voice) to another intercom device for a receiving person (e.g., dweller) through the information line L. On the other hand, when the communication unit 51 receives the electric signal corresponding to the voice of the receiving person (e.g., the dweller's voice) from another intercom device, it

is output as the audio information from the speaker unit 2 through the echo communication unit 53 and an amplifying unit 54. In this regard, the echo cancellation unit 52 receives an output of the echo cancellation unit 53 as a reference signal, and performs an operation to the output of the signal processing unit 50 to further cancel unwanted signal component resulting from the audio output of the speaker unit picked up by the second microphone. On the other hand, the echo cancellation unit 53 receives the output of the echo cancellation unit 52 as the reference signal, and performs an operation to the output of the communication unit 51 to cancel unwanted signal component resulting from the audio output of the speaker unit picked up by the second microphone at the side of another intercom device.

Specifically, the intercom device 1 of this embodiment has a loop circuit comprised of speaker unit 2 → first and second microphones (3, 4) -* signal processing unit 50→echo cancellation unit 52→communication unit 51→echo cancellation unit 53→amplifying unit 54→speaker unit 2. The echo cancellation unit (52, 53) adjusts a loss amount of a variable attenuation means (not shown) formed in this loop circuit such that the loop gain becomes not larger than 1, in order to prevent the howling phenomena. In this regard, a small one of the sending signal and the receiving signal can be regarded as "not important", and the transmission loss of the variable attenuation means formed in the transmission channel having a smaller signal level is increased.

As a modification of the signal processing unit 50 described above, a signal processing unit 50' shown in FIG. 16 may be used. This signal processing unit 50' is characterized by not using the bandpass filters, and performing an average treatment in the operation circuit 77. Therefore, the same reference numerals of the above embodiment are applied to substantially the same components in this modification, and duplicate explanation thereof will be omitted.

In this modification, the output signal of the first A/ D converting circuit 70 is directly input in the respective attenuating circuits (76A, 76B, 76C), and then outputs of these attenuating circuits are sent to the operation circuit 77. On the

other hand, the output signal of the second A/ D converting circuit 71 is input in the operation circuit 77 and the respective estimating circuits (73A, 73B). Outputs of these estimating circuits are sent to the operation circuit 77.

In the operation circuit 77, a difference between the output of the second A/D converting circuit 71 and the output of the attenuating circuit 76A is calculated, so that the signal component corresponding to the audio output of the speaker unit 2 is cancelled with respect to the first frequency fl from the output of the second microphone 4. In addition, a difference between the output of the estimating circuit 73A and the output of the attenuating circuit 76B is calculated, so that the signal component corresponding to the audio output of the speaker unit 2 is cancelled with respect to the second frequency £2 from the output of the second microphone 4. When the intercom device has the second estimating circuit 73B, a difference between the output of the second estimating circuit 73B and the output of the attenuating circuit 76C is calculated, so that the signal component corresponding to the audio output of the speaker unit 2 is cancelled with respect to the third frequency f3 from the output of the second microphone 4.

Next, a total of the differences determined with respect to these frequencies (fl, f2, f3) is calculated. As the averaging treatment, the thus obtained total value is divided by the number "n" of the frequencies (n>2, in this case, n=3 (fl, f2, f3)) to obtain an output of the signal processing unit 50'. By performing this averaging treatment, it is possible to achieve a sufficient howling preventing effect without using the bandpass filters.

The averaging treatment of this modification is explained below in more detail. The following equation (10) is a mathematical formula for determining a cancel amount "Zmave" of the signal component corresponding to the audio output of the speaker unit 2 cancelled from the output of the second microphone

wherein "Zmk" is a cancel amount of the signal component corresponding to the audio output of the speaker unit cancelled with respect to a frequency fk from the output of the second microphone 4. For example, when fk= l, "ZmI" is the cancel amount determined from the difference between the output of the second A/ D converting circuit 71 and the output of the attenuating circuit 76A with respect to the frequency fl. In this regard, the output of the second A/ D converting circuit 71 is the same phase as the output of the attenuating circuit 76A with respect to the first frequency fl because the phase adjustment is performed to the output of the second microphone 4 by the second A/ D converting circuit 71. In addition, the output of the second A/ D converting circuit 71 is the same amplitude as the output of the attenuating circuit 76A with respect to the first frequency fl because the amplitude adjustment is performed to the output of the first microphone 3 by the attenuating circuit 76A. The frequency fl is a representative value of the first frequency band Gl. Similarly, when fk=2, "Zm2" is the cancel amount determined from the difference between the output of the estimating circuit 73A and the output of the attenuating circuit 76B with respect to the frequency f2. In this regard, the output of the estimating circuit 73A is the same phase as the output of the attenuating circuit 76B with respect to the first frequency £2 because the phase adjustment is performed to the output of the second A/ D converting circuit 71 by the estimating circuit 73A. In addition, the output of the estimating circuit 73A is the same amplitude as the output of the attenuating circuit 76B with respect to the first frequency f2 because the amplitude adjustment is performed to the output of the first microphone 3 by the attenuating circuit 76B. The frequency f2 is a representative value of the second frequency band G2.

FIG. 17 shows results of actual measurements of cancel amounts in the case of canceling the signal component corresponding to the audio output of the speaker unit 2 from the output of the second microphone 4. In FIG. 17, "CmI" is a measured value of the cancel amount with respect to the frequency fl of IKHz. In addition, "Cm2" is a measured value of the cancel amount with respect

to the frequency f2 of 3KHz. "Ct" is a theoretical value of the cancel amount obtained by determining a total of the measured values (CmI, Cm2) of these cancel amounts, and then dividing the thus obtained total value by the number "n" of the frequencies (n=2). On the other hand, FIG. 18 shows cancel amounts obtained by using the signal processing unit 50' according to this modification. In FIG. 18, "CsI" shows cancel amounts obtained by removing the signal component corresponding to the audio output of the speaker unit 2 from the output of the second microphone 4 with respect to the frequency fl of IKHz in the frequency band Gl of 2KHz or less, and removing the signal component corresponding to the audio output of the speaker unit 2 from the output of the second microphone 4 with respect to the second frequency f2 of 3KHz in the frequency band G2 of more than 2KHz, and then performing the above averaging treatment in the operation circuit 77. "Cs2" is an approximated curve of the cancel amounts CsI. This approximated curve Cs2 demonstrates that a sufficient canceling effect is uniformly obtained over the wide frequency band region corresponding to the frequency bands (Gl, G2).

As a further modification of this embodiment, instead of connecting the estimating circuits (73A, 73B) to the second A/ D converting circuit 71, the estimating circuits (73A, 73B) may be connected to the first A/ D converting circuit 70 to match the phase of the output signal of the first microphone 3 with the phase of the output signal of the second microphone 4 with respect to each of different frequencies according to the same manner as the above. Additionally, the intercom device 1 of this embodiment may have two or more of microphones as the first microphone 3. Similarly, two or more of microphones may be used as the second microphone 4. If necessary, bandpass filters may be disposed in the signal processing unit 50', as in the case of FIG. 8. (SECOND EMBODIMENT)

An intercom device of the second embodiment is substantially the same as the intercom device of the first embodiment except for using an alternative signal

processing unit 5OA described below. Therefore, the same reference numerals of the first embodiment are applied to substantially the same components in this embodiment, and duplicate explanation thereof will be omitted.

As in the first embodiment, a first microphone 3 is disposed such that its sound receiving portion faces a speaker unit 2. Therefore, when the intercom device is mounted at a house entrance, the first microphone 3 receives an audio output (e.g., the dweller's voice) of the speaker unit 2 with high sensitivity rather than the audio information (e.g., a visitor's voice) of a speaking person at the house entrance. On the other hand, a second microphone 4 is disposed such that its sound receiving surface faces outside of the intercom device. Therefore, the second microphone 4 can receive the audio information provided from the outside of the intercom device with high sensitivity than the audio output of the speaker unit 2. In brief, the output of the first microphone 3 has large amplitude with respect to the audio output of the speaker unit 2, and the output of the second microphone 4 has large amplitude with respect to the audio information provided from the outside of the intercom device.

In addition, the first microphone 3 is spaced from a center of the speaker unit 2 by a first distance Xl, and the second microphone 4 is spaced from the center of the speaker unit 2 by a second distance X2 (Xl <X2) larger than the first distance (ref. FIG. 3). Therefore, the timing where the second microphone 4 receives the audio output of the speaker unit 2 is delayed than the timing where the first microphone 3 receives the audio output of the speaker unit 2. As a result, a phase difference occurs between the outputs of the first and second microphones (3, 4) in the case of receiving the audio output of the speaker unit 2. In general, a time difference, i.e., delay time Td therebetween can be determined by the following equation: Td = (X2-X1) /Vs wherein Vs is sound velocity. Since the first distance Xl and the second distance X2 are predetermined, the delay time Td can be regarded as constant under ideal conditions (e.g., point sound source, sealed housing structure, and

no variation in CR of circuit constructions). However, in fact, it is difficult to satisfy these ideal conditions, the delay time Td changes depending on frequency of the audio output of the speaker unit 2.

Moreover, when the audio output of the speaker unit 2 is received by the first and second microphones (3, 4), signal components corresponding to the audio output of the speaker unit 2 in the output signals of these microphones are constant in amplitude without depending on frequency of the audio output of the speaker unit under the ideal conditions. However, under actual conditions, their amplitudes change depending on the frequency of the audio output of the speaker unit 2. On the other hand, when the audio information of the visitor is provided from the outside of the intercom device, it can be regarded that a distance between the visitor and the first microphone 2 is substantially equal to the distance between the visitor and the second microphone 4. Therefore, the output of the second microphone 4 has substantially the same phase as the output of the first microphone 3 in the case of receiving the audio information from the outside of the intercom device.

Under these conditions, the present embodiment provides a signal processing unit 5OA configured to, when the audio output of the speaker unit 2 is picked up by the second microphone 4, remove a signal component corresponding to the audio output of the speaker unit 2 from an output of the second microphone 4 by using an output of the first microphone 3 and at least two delay times, each of which is determined from a difference between the first distance Xl and the second distance X2 with respect to each of different frequency ranges of the audio output of the speaker unit 2. That is, as shown in FIG. 19, the signal processing unit 5OA, in which the outputs of the first and second microphones (3, 4) are input, is characterized by comprising a first A/ D converting circuit 70 connected to the first microphone 3, second and third A/ D converting circuits (71, 78A) each connected to the second microphone 4, a timing control circuit 72 for determining the timing of performing A/ D conversion at each of the first to third A/ D converting circuits

(70, 71, 78A), bandpass filters (74A, 74B) connected to the first A/D converting circuit 70, bandpass filter 75A connected to the second A/D converting circuit 71, bandpass filter 75B connected to the third A/D converting circuit 78A, attenuating circuits (76A, 76B) connected to bandpass filters (74A, 74B), and an operation circuit 77. Operations of these components of the signal processing unit 5OA are explained below in detail.

The first A/D converting circuit 70 converts the output signal (analog signal) of the first microphone 3 into a digital signal. Each of the second and third A/D converting circuits (71, 78A) converts the output signal (analog signal) of the second microphone 4 into a digital signal. The timing of performing the A/D conversion at the first A/D converting circuit 70 is synchronized with a rising edge of a control signal ScI output from the timing control circuit 72. On the other hand, the timing of performing the A/D conversion at the second A/D converting circuit 71 is synchronized with a rising edge of a control signal Sc2 output from the timing control circuit 72. Moreover, the timing of performing the A/D conversion at the third A/D converting circuit 78A is synchronized with a rising edge of a control signal Sc3 output from the timing control circuit 72. These timings are different from each other, which are determined with respect different frequency bands of the audio output of the speaker unit 2. In this embodiment, each of the control signals has a frequency 8 to 16 KHz. In each of the first to third A/D converting circuits (70, 71, 78A), the A/D conversion is performed at every conversion period Tc, which is, for example, determined in a range of 62.5 to 125 μsec.

FIGS. 2OA and 2OB respectively show waveforms obtained when the audio output of the speaker unit 2 is received by the first and second microphones (3, 4). FIGS. 2OC to 2OE show the timings of outputting the control signals (ScI, Sc3, Sc2) from the timing control circuit 72. FIGS. 2OF to 2OH show outputs of the first, third and second A/D converting circuits (70, 78A, 71).

With respect to a frequency fl in a frequency band Gl of the audio output of the speaker unit 2, since the phase of the output signal of the second

microphone 4 is delayed than the phase of the output signal of the first microphone 3 by a delay time TdI, the timing control circuit 72 determines the timing of outputting the control signal Sc2 such that the control signal Sc2 is output after the elapse of the delay time TdI from the output of the control signal ScI, as shown in FIGS. 2OC and 2OE. Thereby, it is possible to perform the A/D conversions to the output signals of the first and second microphones (3, 4) under the same phase condition with respect to the frequency fl. On the other hand, with respect to a frequency ff2 in a frequency band G2 (different from the frequency band Gl) of the audio output of the speaker unit 2, since the phase of the output signal of the second microphone 4 is delayed than the phase of the output signal of the first microphone 3 by a delay time Td2, the timing control circuit 72 determines the timing of outputting the control signal Sc3 such that the control signal Sc3 is output after the elapse of the delay time Td2 from the output of the control signal ScI, as shown in FIGS. 2OC and 2OD. Thereby, it is possible to perform the A/D conversions to the output signals of the first and second microphones (3, 4) under the same phase condition with respect to the frequency f2.

In this embodiment, the number of frequency bands to be considered is two (Gl, G2). If necessary, the number of frequency bands may be three or more. Therefore, when the number of frequency bands to be considered is V, which is a positive integer, it is needed to use nth A/D converters in the signal processing unit 5OA. For example, as shown by the dotted line in FIG. 19, a fourth A/D converting circuit 78B may be added to the signal processing unit 5OA with respect to a frequency f3 in a frequency band G3 different from the frequency bands (Gl, G2). In this case, the output of the second microphone 4 is also sent to the fourth A/D converting circuit 78B. The timing of performing the A/D conversion at the fourth A/D converting circuit 78B is synchronized with a rising edge of a control signal Sc4 output from the timing control circuit 72. In this regard, as lower the frequency band, the delay time increases. In this embodiment, since the frequency band Gl is a lower frequency band than the

frequency band G2, the delay time TdI is longer than the delay time Td2 (Td2< TdI).

Each of the first to third second A/D converting circuits (70, 71, 78A) of this embodiment uses a sigma-delta (σ δ) method. Therefore, the A/D conversion can be performed with a high S/ N ratio by a noise shaping where frequency transfer characteristics are changed with respect to only noise components such that the noise components are distributed in a higher frequency region than the signal components. In addition, by digitalizing most of treatments, it becomes easy to design an integrated circuit for the signal processing unit 5OA. A resolving power of the A/D conversion can be selected from one of 16 bit, 14 bit and 12 bit.

As shown in FIG. 2OF, the first A/D converting circuit 70 outputs a digital signal (al, bl, cl ) at every conversion period Tc. On the other hand, as shown in FIG. 2OH, the second A/D converting circuit 71 outputs a digital signal (a2, b2, c2 ) at every conversion period Tc. In addition, as shown in FIG. 2OG, the third A/D converting circuit 78A outputs a digital signal (a3, b3, c3 ) at every conversion period Tc. As described above, by controlling the timing of performing the A/D conversion at the second A/D converting circuit 71, the output signals of the first and second microphones (3, 4) can be A/D converted under the same phase condition with respect to the frequency fl (i.e., the frequency band Gl). Therefore, the digital signal (al, bl, cl) output from the first A/D converting circuit 70 is the same in phase as the digital signal (a2, b2, c2) output from the second A/D converting circuit 71. Similarly, by controlling the timing of performing the A/D conversion at the third A/D converting circuit 78A, the output signals of the first and second microphones (3, 4) can be A/D converted under the same phase condition with respect to the frequency £2 (i.e., the frequency band G2). Therefore, the digital signal (al, bl, cl) output from the first A/D converting circuit 70 is the same in phase as the digital signal (a3, b3, c3) output from the third A/D converting circuit 78A.

Thus, the output signal of the second microphone 4 is A/D converted at different timings by use of the plural A/D converting circuits (71, 78A), so that the phase of the output signal of the second microphone 4 is matched with the phase of the output signal of the first microphone 3 with respect to each of the frequencies (fl, £2). Therefore, each of the A/ D conversions can be performed at a relatively slow speed (e.g., 8 to 16 KHz). As a result, there is an advantage of achieving a cost reduction in the intercom device without using an expensive high-speed A/ D converter (e.g., 1 MHz).

The output signal of the first A/ D converting circuit 70 and the output signal of the second A/ D converting circuit 71 are sent to bandpass filters (74A, 75A) to extract only signal components concerning the frequency band Gl therefrom. In addition, the output signal of the first A/D converting circuit 70 and the output signal of the third A/D converting circuit 78A are sent to bandpass filters (74B, 75C) to extract the signal components therefrom with respect to the frequency band G2.

The output of each of the bandpass filters (74A, 74B) connected to the first A/D converting circuit 70 is attenuated by the attenuating circuit (76A, 76B) to perform an output level adjustment. In this embodiment, since the first microphone 3 is disposed near the diaphragm 23 of the speaker unit 2, the output of the first microphone 3 has larger amplitude than the output of the second microphone 4 with respect to of the audio output of the speaker unit 2. In addition, the amplitude difference is also caused by a difference in sensitivity between the first and second microphones (3, 4). Therefore, by adequately attenuating the amplitude of the output of the resulting from the first microphone 3, the output level of each of the bandpass filters (74A, 74B) connected to the first A/D converting circuit 70 can be substantially matched with the output level of each of the bandpass filters (75A, 75B) connected to the second and third A/D converting circuits (71, 78A).

When the fourth A/D converting circuit 78B is connected to the second microphone 4, the output signal of the first A/D converting circuit 70 and the

output signal of the fourth A/ D converting circuit 78B are sent to bandpass filters (74C, 75C) to extract only signal components concerning the frequency band G3 therefrom. According to the same manner described above, when the amplitude of the output of the resulting from the first microphone 3 is adequately attenuated by an attenuating circuit 76C, the output level of the bandpass filter 74C connected to the first A/ D converting circuit 70 can be substantially matched with the output level of the bandpass filters 75C connected to the fourth A/ D converting circuit 78B.

In the operation circuit 77, the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency band Gl is cancelled from the output of the second microphone 4 by subtracting the output of the attenuating circuit 76A from the output of the bandpass filter 75A. Similarly, the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency band G2 is cancelled from the output of the second microphone 4 by subtracting the output of the attenuating circuit 76B from the output of the bandpass filter 75B. When the signal processing unit 5OA has the fourth A/ D converting circuit 78B, the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency band G3 is cancelled from the output of the second microphone 4 by subtracting the output of the attenuating circuit 76C from the output of the bandpass filter 75C. In FIG. 19, the output of each of the attenuating circuits is inversed, and then added to the output of the corresponding bandpass filter. Alternatively, the output of each of the bandpass filters may be inversed, and then added to the output of the corresponding attenuating circuit. Subsequently, a total of differences determined with respect to the frequency bands (Gl, G2, G3) is calculated to obtain an output of the signal processing unit 5OA, in which the signal components corresponding to the audio output of the speaker unit 2 are removed from the output of the second microphone 4 over a wide frequency range, i.e., the frequency bands (Gl, G2, G3).

By the way, when the intercom device 1 is used at the house entrance, the audio information (e.g., visitor's voice) provided from the outside of the intercom device is received by the second microphone 4 with high sensitivity. On the other hand, the visitor's voice may be picked up by the first microphone 3 for receiving the audio output of the speaker unit 2 with high sensitivity. In this case, the output signal of the second microphone 4 is much larger in amplitude with respect to the visitor's voice than that of the first microphone 3. Moreover, as described above, since the output signal of the first microphone 3 is attenuated by the attenuating circuit (76A, 76B, 76C), the output of the bandpass filter (75A, 75B, 75C) connected to the second microphone 4 relatively increases in amplitude with respect to the visitor's voice than that of the attenuating circuit (76A, 76B, 76C). Therefore, even when the subtraction treatment is performed in the operation circuit 77, the signal component corresponding to the visitor's voice remains in the output of the signal processing unit 5OA with sufficient magnitude.

Therefore, according to the output of the signal processing unit 5OA, the unwanted signal component corresponding to the audio output of the speaker unit 2 is reduced or cancelled from the output of the second microphone 4 over the wide frequency range without decreasing a wanted signal component corresponding to the audio information (e.g., visitor's voice) provided from the outside of the intercom device. As a result, it is possible to effectively prevent the howling phenomenon, while achieving comfortable communication between the intercom devices.

Next, the howling preventing effect of the signal processing unit 5OA of this embodiment is theoretically explained by using mathematical formulas. When the audio output of the speaker unit 2 is received by the first microphone 3, and a voice of a speaking person H is received as the audio information provided from the outside of the intercom device by the second microphone 4, sound collecting characteristics (Ql(ω), Q2(ω)) of the first and second microphones (3, 4) are represented by the following equations (11) and (12) on the assumption that

differences in phase and amplitude between the outputs of the first and second microphones (3, 4) are not dependent on frequency. Q1(ω) = αexp(jωt) + /?exp(jω ₂ t) -"(1I) Q2'(ω) = α? _ie χpϋ(ωt— 0)}+βeχp(jω ₂ t) ---(12) In this regard, " a ", " a 1", " β" are constants. The first term of each of the equations (11) and (12) provides the signal component corresponding to the audio output of the speaker unit 2, and the second term thereof provides the signal component corresponding to the voice of the speaking person.

In addition, when performing a cancellation treatment for reducing the signal component corresponding to the audio output of the speaker unit 2 by subtracting "Q2'(ω)" from a value obtained by multiplying an attenuation ratio "a" by "Ql(ω)", without considering the frequency dependencies of the phase difference and the amplitude difference, a signal (Yean') after the cancellation treatment is represented by the following equation (13).

Yean' = Q2'(ω)-aQ1(ω)-exp(-jθ)

= βexp(jω ₂ t)-(1-aexp(-j0)) ---(1S)

According to this cancellation treatment, only the signal component corresponding to the voice of the speaking person H remains. However, in fact, since the differences in phase and amplitude between the outputs of the first and second microphones (3, 4) depend on frequency, the sound collecting characteristic "Q2(ω)" of the second microphone 4 is represented by the following equation (14) in consideration of the frequency dependency. Q2(ω) = Qfi(ω)eχp{j(ωt— θ ( ω ) ) } + _/ 3 exp (j ω ₂ t) ---(14)

When performing the cancellation treatment for reducing the signal component corresponding to the audio output of the speaker unit 2 by subtracting "Q2(ω)" from the value obtained by multiplying the attenuation ratio "a" by "Ql(ω)" in consideration of the frequency dependencies of the phase difference and the amplitude difference, a signal (Yean) after the cancellation treatment is represented by the following equation (15).

Yean = exp (j ωt) ^■ { a -, ( ω ) exp ( — j θ ( ω ) ) — Qf ₁ } +

£exp(jω ₂ t)-{1-aexp(-j0)} "-(15)

In this regard, the first term of the equation (15) provides a remaining signal component corresponding to the audio output of the speaker unit 2, and "θ(ω)" of the first term corresponds to the phase difference (second microphone 4— first microphone 3) in the case of receiving the audio output of the speaker unit 2 by the first and second microphones (3, 4). Hereinafter, this is called as the phase difference "θ(ω)" for the audio output of the speaker unit 2. The second term of the equation (15) is the signal component corresponding to the voice of the speaking person H.

Next, it is considered about " { a ■, (ω) e x p (— j θ _k (ω) ) — a _k } ", which is obtained by expanding 'Ma ₁ (ω) exp (— j θ (ω) ) — α i } " in the first term of the equation (15) with respect to each of the frequency bands. In this regard, the problem is how much the amplitude of the audio output of the speaker unit 2 is attenuated from "Qj ₁ ". When "a (ω) " and "a _k " are respectively values obtained by dividing " a ■, (ω) " and " a _k " by "a", the following relation (16) is obtained.

α?i (ω)exp{— j(0 _k (ω) — θ)} — a _k

= α|a(ω)exp{-j(e _k (ω)-e)}-a _k | -"(16) In this regard, "a ( ω ) " is an amplitude ratio (second microphone 4/first microphone 3) of the signal components obtained when the audio output of the speaker unit 2 is received by the first and second microphones (3, 4). Hereinafter, this is called as an amplitude ratio "a (ω) " for the audio output of the speaker unit 2. In addition, " | a (ω) exp {— j (θ _k (ω) — θ) } — a _k | " of the equation (16) is shown by a phasor vector "Ve" in FIG.21.

Next, results of actual measurements of the amplitude ratio "a (ω) " and the phase difference "θ (ω) " are shown in FIGS. 22A and 22B. As larger the frequency, the amplitude ratio "a (ω) " and the phase difference "θ (ω) "increase. For example, on the assumption that the amplitude ratio "a (ω) " linearly

changes from -20 dB (0.1) to -16 dB (0.16) with respect to a frequency range of from 400 Hz to 4000 Hz, and the phase difference "θ (ω) " linearly changes from -5° to -25° with respect to the frequency range of from 400 Hz to 4000 Hz, the following approximate expressions (17) and (18) can be derived. a ( ω ) = 2. 5 x 1 CT ⁶ X ω + 0. 1 - " ( 1 7) ^β ( ω ) = 1 . 2 x 10- ³ χ ω — 7. 2 " - ( 1 8)

In addition, the following is an equation (19) for determining a cancel amount "Zm" of the audio output of the speaker unit 2. Zm = 20Log ₁₀ (| a ( ω )exp{-j ( 0k ( ω )- θk) } -a _k |) - " (19) When the cancellation treatment is performed at each of IKHz and 3KHz by using an equation obtained by substituting the equations (17), (18) into the equation (19) without using the technical concept of the present invention of removing the signal component corresponding to the audio output of the speaker unit 2 with respect to plural different frequency bands, theoretical values (YtI, Yt2) of the cancel amounts at the frequencies of IKHz and 3KHz are respectively shown in FIG. 23A, and measured values (YsI, Ys2) of the cancel amounts at the frequencies of IKHz and 3KHz are respectively shown in FIG. 23B. In this regard, the cancel amount is defined as, when the audio output of the speaker unit 2 is picked up by the second microphone 4, an amount of the signal component corresponding to the audio output of the speaker unit 2 removed or cancelled from the output of the second microphone 4. These results show that a canceling effect is obtained over a relatively narrow range at the vicinity of IKHz in the case of performing the canceling treatment with respect to the frequency of IKHz, and a canceling effect is obtained over a relatively narrow range at the vicinity of 3KHz in the case of performing the canceling treatment with respect to the frequency of 3KHz.

On the other hand, when the cancellation treatment of the present invention is performed with respect to the frequency band Gl of 2KHz or less including the frequency fl (= 1 KHz) and the frequency band G2 of more than 2KHz including the frequency f2 (=3 KHz) by using an equation obtained by substituting the

equations (17), (18) into the equation (19) according to the technical concept of the present invention, a theoretical value Yt3 of the cancel amount is shown in FIG. 24A, and a measured value Ys3 of the cancel amount is shown in FIG. 24B. From these results, it is demonstrated that according to the canceling treatment of the present invention, a remarkable canceling effect is obtained over a wide frequency region having two peaks of the cancel amounts at the frequencies of IKHz and 3KHz.

On a comparison between the theoretical cancel amount Yt3 shown in FIG. 24A and the measured cancel amount Ys3 shown in FIG. 24B, the profile of the theoretical cancel amount is similar to that of the measured cancel amount. However, the theoretical cancel amount is larger in magnitude than the measured cancel amount. It is thought that this difference is caused by several factors that a S/ N ratio of the first and second microphones (3, 4) used in this embodiment is about 6OdB, a sound pressure of the audio output of the speaker unit 2 picked up by the second microphone 4 is about 8OdB under a background noise of about 40 dB, various noises occur during the operation of the intercom device, and there is a limitation in resolving power due to quantifying bit number for the A/ D conversion.

An output of the signal processing unit 5OA is sent to the communication unit 51 through the echo cancellation unit 52, and the communication unit 51 transmits an electric signal corresponding to the audio information of the speaking person (e.g., visitor's voice) to another intercom device for a receiving person (e.g., dweller) through the information line L. On the other hand, when the communication unit 51 receives the electric signal corresponding to the voice of the receiving person (e.g., the dweller's voice) from another intercom device, it is output as the audio information from the speaker unit 2 through the echo communication unit 53 and an amplifying unit 54. In this regard, the echo cancellation unit 52 receives an output of the echo cancellation unit 53 as a reference signal, and performs an operation to the output of the signal processing unit 5OA to further cancel unwanted signal component resulting from the audio

output of the speaker unit picked up by the second microphone. On the other hand, the echo cancellation unit 53 receives the output of the echo cancellation unit 52 as the reference signal, and performs an operation to the output of the communication unit 51 to cancel unwanted signal component resulting from the audio output of the speaker unit picked up by the second microphone at the side of another intercom device.

Specifically, the intercom device 1 of this embodiment has a loop circuit comprised of speaker unit 2 → first and second microphones (3, 4) → signal processing unit 50A→echo cancellation unit 52→communication unit 51 ^→ echo cancellation unit 53→amplifying unit 54→speaker unit 2. The echo cancellation unit (52, 53) adjusts a loss amount of a variable attenuation means (not shown) formed in this loop circuit such that the loop gain becomes not larger than 1, in order to prevent the howling phenomena. In this regard, a small one of the sending signal and the receiving signal can be regarded as "not important", and the transmission loss of the variable attenuation means formed in the transmission channel having a smaller signal level is increased.

As a modification of the signal processing unit 5OA of the present embodiment described above, a signal processing unit 5OA' shown in FIG. 25 may be used. This signal processing unit 5OA' is characterized by not using the bandpass filters, and performing an average treatment in the operation circuit 77. Therefore, the same reference numerals of the above embodiment are applied to substantially the same components in this modification, and duplicate explanation thereof will be omitted.

As described above, an output signal of the first A/ D converting circuit 70 is input in the respective attenuating circuits (76A, 76B) , and then outputs of the attenuating circuits are sent to the operation circuit 77. On the other hand, each of output signals of the second and third A/D converting circuits (71, 78A) is directly input in the operation circuit 77.

In the operation circuit 77, a difference between the output of the second A/D converting circuit 71 and the output of the attenuating circuit 76A is

calculated, so that the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency fl is cancelled from the output of the second microphone 4. Similarly, a difference between the output of the third A/D converting circuit 78A and the output of the attenuating circuit 76B is calculated, so that the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency f2 is cancelled from the output of the second microphone 4. Furthermore, when the signal processing unit 5OA' has the fourth A/D converting circuit 78B, a difference between the output of the fourth A/D converting circuit 78B and the output of the attenuating circuit 76C is calculated, so that the signal component corresponding to the audio output of the speaker unit 2 with respect to the frequency f3 is cancelled from the output of the second microphone 4.

After a total of the differences determined with respect to all of the frequencies (fl, f2, f3) is calculated, the averaging treatment is performed. That is, a resultant total value is divided by the number (n) of the frequencies used for the phase adjustment (n>2, in this case, n=3) to obtain an output of the signal processing unit 5OA', in which the signal components corresponding to the audio output of the speaker unit 2 are removed from the output of the second microphone 4 over a wide frequency range, i.e., the frequency bands (Gl, G2, G3) including the frequencies (fl, f2, f3). By performing the averaging treatment described above, it is possible to obtain a sufficient howling preventing effect without using the bandpass filters. In addition, there are advantages of reducing the circuit scale, and achieving a reduction in cost and size of the intercom device. As a further modification of this embodiment, instead of connecting the third and fourth A/D converting circuits (78A, 78B) to the second microphone 4, they may be connected to the first microphone 3 to match the phase of the output signal of the first microphone 3 with the phase of the output signal of the second microphone 4 with respect to plural different frequency bands. Additionally, in the intercom device of this embodiment, two or more of microphones may be

used as the first microphone 3. Similarly, two or more of microphones may be used as the second microphone 4. If necessary, bandpass filters may be disposed in the signal processing unit 5OA', as in the case of FIG. 19.

(THIRD EMBODIMENT) The present embodiment provides a wiring system using the intercom device

1 of the present invention. That is, this wiring system enables power and signal transmissions between electric devices spaced away from each other in a building structure.

As shown in FIG. 26, the wiring system comprises a power line Lp and an information line Ls installed in the building structure, which are connected to a commercial power source AC and the Internet network NT through a distribution board, switch boxes 102 embedded in wall surfaces of the building structure, base units 103 mounted on the switch boxes, and connected to the power line Lp and the information line Ls, and function units 104, each of which is detachably connected to a desired one of the base units 103, and has at least one of functions for supplying electric power from the power line Lp, outputting information from the information line Ls, and inputting information into the information line Ls in a connected state with the desired base unit. The intercom device of the present invention, preferably the intercom device 1 of the first or second embodiment described above can be regarded as one of the function units 104. In the present description, the wall surfaces of the building structure are not limited to surfaces of sidewalls standing between adjacent rooms. That is, the meaning of the wall surfaces includes outdoor and indoor wall surfaces, and the indoor wall surfaces comprise ceiling and floor surfaces as well as the sidewall surfaces. In the drawings, MB designates a main circuit breaker, BB designates a branch circuit breaker, and GW designates a gateway (e.g., a router or a built-in hub).

As shown in FIGS. 27 and 28, each of the base units 103 is formed with a gate housing 130 having terminals (131a, 132a, 131b, 132b) connected to the power supply line Lp and the information line Ls, and a main housing 135

detachably connected to the function unit 104. Theses housings can be made of a synthetic resin having electrical insulating characteristics (e.g., a noncrystalline plastic such as ABS resin). The terminals (131b, 132b) are used to extend the wirings. The gate housing 130 and the main housing 135 have a pair of a module port 134 and a module connector 142, which are detachably connected to each other to simultaneously establish both of supplying the electric power from the gate housing 130 to the main housing 135, and making a signal transmission therebetween. In addition, when the function unit 104 has the module connector 142, it can be connected to the gate housing 130 in place of the main housing 135. If necessary, the gate housing 130 and main housing 135 may be integrally formed.

Circuit components of the base unit 103 are designed to transmit the electric power and the information signal to the function unit 104. For example, the base unit 103 is provided with an AC/AC converter 160, DC power section 161, transceiver section 162, E/O converter 163, O/E converter 165, and a function portion 167.

The AC/AC converter 160 converts commercial AC voltage to a lower AC voltage having an increased frequency, and applies the lower AC voltage to a coil 172. The DC power section 161 generates an operating voltage of the internal circuit components from a stable DC voltage obtained by rectifying and smoothing the lower AC voltage. The transceiver section 162 transmits and receives the information signal for enabling the mutual communication through the information line Ls. The E/O converter 163 converts the information signal received from the information line Ls to an optical signal, and outputs the optical signal though a light emitting device (LED) 164. On the other hand, the O/E converter 165 receives the optical signal provided from the outside, e.g., the function unit 104 by a light receiving device (PD) 166, and converting the received optical signal into the information signal to transmit it to the transceiver section 162. The base unit 103 shown in FIG. 27 has an outlet tap as the function portion 167. Alternatively, a sensor device or a controller may be formed as the function portion 167.

The module port 134 formed at a front surface of the gate housing 130 is composed of an electric power port 134a for supplying the electric power and an information signal port 134b for accessing the information line Ls, as shown in FIG. 29B. Arrangement and shapes of the electric power port 134a and the information signal port 134b are standardized (normalized or stylized) in the wiring system of this embodiment. For example, as shown in FIG. 29B, each of the electric power port 134a and the information signal port 134b is configured in a substantially rectangular shape such that they are arranged in parallel to each other. On the other hand, the module connector 142 formed at a rear surface of the main housing 135 is composed of an electric power connector 142a and an information signal connector 142b, as shown in FIG. 29A. Arrangement and shapes of the electric power connector 142a and the information signal connector 142b are standardized (normalized or stylized) in the wiring system of this embodiment. For example, as shown in FIG. 29A, each of the electric power connector 142a and the information signal connector 142b is configured in a substantially rectangular shape such that they are arranged in parallel to each other.

In this embodiment, the module port 134 has a guide portion 133 such as a ring-like wall or a ring- like groove extending around the electric power port 134a and the information signal port 134b. This guide portion 133 is formed to be engageable to an engaging portion 145 such as a ring-like wall of the main housing 135, which is formed around the electric power connector 142a and the information signal connector 142b. By simply engaging the engaging portion 145 to the guide portion 133, the electric power connector 142a and the information signal connector 142b can be simultaneously connected to the electric power port 134a and the information signal port 134b. Therefore, there advantages that the main housing 135 can be easily attached to or removed from the gate housing 130 with improved connection reliability. A module connector having the same structure described above may be formed in the function unit

104. The module port 134 and the module connector 142 may be formed by female and male connectors.

In the present embodiment, the gate housing 130 is directly fixed to the switch box 102. If necessary, the gate housing 130 may be fixed to the switch box 102 through an attachment member 175, as shown in FIG. 30. For example, this attachment member 175 is formed by a rectangular frame 176 with a window hole 177. In this case, the gate housing 130 is mounted on the attachment member 175 by fitting a front portion of the gate housing into the window hole 177, and then engaging hooks (not shown) formed at both sides of the gate housing to the rectangular frame 176. The attachment member 175 mounting the gate housing 130 thereon is fixed to the switch box 102 by use of fixing screws. Alternatively, the gate housing 130 may be directly fixed to the wall surface by use of exclusive clamps without using the switch box 102.

The function unit 104 is designed to provide a desired function by using the electric power supplied to the function unit 104 through the base unit 103 or using the mutual communication of the information signal between the function unit 104 and the information line Ls through the base unit 103. For example, when the function unit 104 is connected to the base unit 103 mounted in the wall surface at a high position near the ceiling, it preferably has a receptacle function of receiving a plug with hook of a lighting apparatus, security function such as a motion sensor, temperature sensor, and monitoring camera, or a sound function such as speaker. In addition, when the function unit 104 is connected to the base unit 103 mounted in the wall surface at a middle height, at which the function unit 104 can be easily operated by the user, it preferably has a switch function of turning on/off the lighting apparatus, control function for electric appliances such as air-conditioning equipments, a display function such as liquid crystal display or a timer function. The intercom device of the present invention is preferably connected as one of the function units 104 to the base unit 103 at the middle height. Moreover, when the function unit 104 is connected to the base unit 103 mounted in the wall surface at a low position near the floor, it preferably has a receptacle function for receiving a plug of an

electric appliance such as electric vacuum cleaner, the sound function such as speaker, or a footlight function.

Specifically, as shown in FIG. 31, when a function section 181 of the function unit 104 is formed by a switch, operation data obtained by operating the switch is transmitted to a processing section 188 such as CPU through an

I/O interface 189. Then the processed data is sent to, for example, an infrared remote controller (not shown) through a transceiver section 187, so that an electric appliance to be controlled is turned on /off by receiving a remote control signal emitted from the infrared remote controller. In addition, when the function section 181 is formed by a sensor, data detected by the sensor is transmitted as the information signal to the information line Ls, and then informed to the user by a required communicator.

The coil 172 wound around a core 170 of the base unit 103 shown in FIG.

27 is used as a power transmission means for supplying electric power to the function unit 104 in a noncontact manner. That is, the coil 172 of the base unit

103 provides an electromagnetic coupling portion that works as a first side of a transformer. On the other hand, as shown in FIG. 31 , the function unit 104 has an electromagnetic coupling portion Xl comprised of a coil 182 wound around a core 180, which works as a second side of the transformer. Therefore, by forming electromagnetic coupling between the base unit 103 and the function unit 104, a low AC voltage can be induced in the coil 182 of the function unit

104 to supply of electric power from the base unit 103 to the function unit 104. Similarly, an electromagnetic coupling portion X2 is formed at the opposite side of the electromagnetic coupling portion Xl, and used to make a connection between the function units. In this embodiment, since the low AC voltage having the higher frequency than the commercial AC voltage is obtained by the AC /AC converter 160, there is an advantage that the electromagnetic coupling portions used as the transformer can be downsized.

In addition, the light emitting device (LED) 164 of the base unit 103 shown in FIG. 27 is used to transmit an optical signal as the information signal to the

function unit 104 in a noncontact manner. On the other hand, as shown in FIG. 31, a light receiving device (PD) 186 is disposed in the function unit 104 such that the light emitting device 164 is in a face-to-face relation with the light receiving device 186 when the function unit 104 is connected to the base unit 103. Similarly, to transmit the optical signal as the information signal from the function unit 104 to the base unit 103, the function unit 104 has a light emitting device (LED) 184, which is disposed in the face-to-face relation with the light receiving element (PD) 166 when the function unit 104 is connected to the base unit 103. Thus, each of the base unit 103 and the function unit 104 has the pair of the E/O converter (163, 183) and the O/E converter (165, 185) as an optical coupling portion to enable the mutual communication of the information signal therebetween in a noncontact manner.

As shown in FIG. 27, the electromagnetic coupling portion X used to supply the electric power and the optical coupling portion Y used for the mutual communication of the information signal are disposed at a side surface of the base unit 103 to be spaced from each other by a constant distance. Arrangements and shapes of the electromagnetic coupling portion X and the optical coupling portion Y are standardized (or stylized) to make the base unit

103 shareable among the function units 104. In addition, the pair of the electromagnetic coupling portion and the optical coupling portion may be formed at each of both sides of the function unit 104, as shown in FIG. 31. That is, the optical coupling portion Yl formed at one side (e.g., left side) of the function unit

104 is composed of the light receiving device 186 located at the upper side and the light emitting device 184 located at the lower side, and the optical coupling portion Y2 formed at the opposite side (e.g., right side) of the function unit 104 is composed of a light emitting device 194 located at the upper side and a light receiving device 196 located at the lower side. In this case, since the base unit 103 is connected to the one side of the function unit 104, and another function unit or the intercom device of the present invention can be connected to the

other side of the function unit 104, there is an advantage of further improving function expandability of the wiring system.

The intercom device 1 of this embodiment can be connected to the base unit 103, as shown in FIG. 32. In addition, internal components of the intercom device 1 of this embodiment are shown in FIG. 33. As understood from the comparison with the function unit 104 of FIG. 31, the intercom device 1 of FIG. 33 is substantially equivalent to a function unit having the components of the intercom device shown in FIG. 2 as the function section 181.

Since this intercom device 1 has the stylized connectors as well as the function unit 104 and base unit 103 described above, it can be used in a desired combination with the base unit 103 and the function unit 104. For example, as shown by the arrow (D in FIG. 34, the intercom device 1 can be connected to the function unit 104 having the timer function, which is connected to the base unit 103. Alternatively, as shown by the arrow (2) in FIG 34, the intercom device 1 may be connected at its one side to the base unit 103. At this time, the function unit 104 having the timer function can be connected to the other side of the intercom device 1. Thus, there is a remarkable advantage that a layout of the function units 104 having various functions and the intercom device 1 can be changed with a high degree of freedom according to the user's needs. In addition, when the intercom device 1 is operated in conjunction with at least one functions unit 104, the wiring system of this embodiment can provide a higher order function in addition to the communication function. For example, when the intercom device 1 is connected to one of the base units 103, and the function unit 104 having a sensor function (e.g., motion sensor or fire sensor) is connected to another one of the base units 103, which is disposed at a proper location, the intercom device 1 can output an alarm sound by receiving alarm signal from the function unit. Therefore, the intercom device 1 of this case has an alarm generating function for disaster or crime prevention as well as the communication function. In addition, since it is not needed to individually prepare a speaker for generating the alarm sound, there is a further advantage of achieving an improvement in cost performance of the wiring system as a whole.

Next, means for mechanically coupling between the base unit 103 and the function unit 104 (or the intercom device 1) is explained.

For example, the base unit 103 shown in FIG. 35 has a cosmetic cover 112, which is detachably attached to a front surface of the gate housing 135, and horizontal guide rails 114 formed at upper and lower portions of the gate housing. The numeral 115 designates a stopper wall formed in a substantially center position in the longitudinal direction of the guide rail 114. On the other hand, as shown in FIGS. 36A to 36C, the function unit 104 (or the intercom device 1) has concave portions 128 for accommodating joining members 150 therein, which are formed at both of its upper and lower sides, horizontal guide rails 124 extending in the concave portions 128, and cover members 126 pivotally supported at upper and lower ends of the function unit 104. In addition, as shown in FIG. 36D, the joining member 150 has a groove 152, in which the horizontal guide rail (114, 124) can be slidably fitted. In this case, a cosmetic cover 112 is firstly removed from the gate housing

135 of the base unit 103. Then, the electromagnetic coupling portion X and the optical coupling portion Y of the base unit 103 are connected to the electromagnetic coupling portion Xl and the optical coupling portion Yl of the function unit 104 (or the intercom device 1). Next, the cover member 126 is opened, and the joining member 150 is slid along the guide rail 124, as shown by an arrow in FIG. 37. The slide movement of the joining member 150 is performed until the joining member 150 contacts the stopper wall 115. As a result, the joining member 150 is engaged to the guide rail 114 of the base unit 103 over about half length of the joining member 150. On the other hand, the remaining portion of the joining member 150 is still engaged to the guide rail 124 of the function unit 104 (or the intercom device 1). Thus, by engaging the joining member 150 to both of the base unit 103 and the function unit 104, the base unit 103 can be mechanically coupled with the function unit 104 (or the intercom device 1). Finally, the cosmetic cover 112 is attached to the base unit 103, and the cover member 126 is closed.

According to this coupling means, since the joining member 150 is held between the cosmetic cover 112 and the gate housing 135, it is possible to prevent accidental falling of the joining member 150, and obtain the stable mechanical connection therebetween without spoiling the beauty of them. In addition, since the joining member 150 is always accommodated in the concave portion 128 of the function unit 104 (or the intercom device 1), there is no worry about loss of the joining member 150. This coupling means is also usual in the case of mechanically coupling between the function units 104, or between the function unit 104 and the intercom device 1. As an information- signal transmitting method available in the wiring system, a baseband transmission or a broadband transmission can be used. In addition, the protocol is not limited to a specific one. For example, sound and visual information signals may be transmitted and received according to JT-H232 packet to make the mutual communication between the intercom devices. In a control system, it is also preferred to use a routing protocol for a broadcast or a unicast that controlling can be performed at a control ratio of 1: 1 or 1:N according to operation data. Alternatively, the protocol used between the base units may be different from the protocol used in the function unit or the intercom device connected to the base unit. In this case, a protocol conversion is preferably performed by the base unit.

In addition, the wiring system of the above embodiment is essential to use both of the information line and the power line installed in the building structure. As a modification of this embodiment, a power-line carrier communication may be used in the wiring system. That is, the wiring system according to this modification is essential to install only the power line in the building structure. Therefore, the base unit is connected to only the power line. On the other hand, the base unit, the function unit and/or the intercom device need to have a transceiver unit for transmitting and receiving information signals in the power-line carrier communication manner. For example, when the transceiver unit is formed in the base unit, the electric power carrying the information signals thereon is separated to the

electric power transmission and the information signal transmission by the base unit. Therefore, the same function unit and the intercom device used in the wiring system of the above embodiment are also available in this power-line carrier type wiring system. Alternatively, the transceiver unit may be formed in the intercom device 1, as shown in FIG. 38. In this case, the intercom device is detachably connected to the base unit and/ or the function unit by a power-line carrier type connector Z. In addition, the intercom device also comprises a PLC modem 210 for receiving the information signals carried by the power line, and transmitting information signals through the power line, a processing section 220 connected to the PLC modem, an I/O interface 230 between the processing section, and a function section 240 for providing the intercom function. Similar circuit components may be formed in the function unit and/or the base unit. As a modulation method for the power-line carrier communication, for example, it is possible to use a broad spectrum diffusion model, multi carrier model or an orthogonal Frequency division multiplexing (OFDM) model.

INDUSTRIAL APPLICABILITY

As described above, according to the intercom device of the present invention, when the audio output of the speaker unit is picked up by the microphone for receiving audio information from outside of the intercom device, the signal component corresponding to the audio output of the speaker unit can be cancelled from the output of the microphone with respect to each of a plurality of frequency bands. Therefore, there is an advantage of reliably preventing the howling phenomenon over a wide frequency range. In addition, when the signal processing unit for preventing the howling phenomenon has the estimating circuit, it is possible to provide a compact intercom device with improved cost performance without using a high-speed, expensive A/D converter.

Moreover, according to the wiring system using the intercom device of the present invention, there is an advantage that a desired combination of the intercom function and another function(s) can be readily achieved depending on

the user's demand with high flexibility, as compared with a conventional intercom device without function expandability.

Thus, due to the advantages of achieving reduction in size and cost of the intercom device as well as improving the howling prevention function, the intercom device of the present invention is expected to be more widely used, as compared with the conventional ones. In addition, the wiring system using the intercom device of the present invention holds the promise of providing comfortable and convenient living and working environments for individual users in the information society.