$005459792A
`115
`United States Patent
`5,459,792
`[11] Patent Number:
`Oct. 17, 1995
`[45] Date of Patent:
`Reichel et al.
`
`ACAAA
`
`[54] AUDIO INPUT CIRCUIT FOR VOICE
`RECOGNITION
`
`[75]
`
`Inventors: Ken Reichel, Hudson; Kelly Statham,
`Cleveland Heights, both of Ohio
`
`[73] Assignee: Audio-Technica U.S., Inc., Stow, Ohio
`
`[21] Appl. No.: 168,047
`
`{22]
`
`Filed:
`
`Dec. 15, 1993
`
`Tite C18 accccsccsssssssesscececcecccssnanssnenensesees HO4R 3/00
`E51]
`52] U.S. Che eceecsesssssesessresctescnssseronsssnssasesssenseesnvesses 381/111
`
`
`(58] Field of Search 0... 381/56, 57, 111,
`381/122, 112-115
`
`[56]
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`4,378,467
`5,208,866
`
`3/1983 Ferrantelli oo...cceeseseeseeeseeee 381/111
`5/1993 Kato et al.
`...cscsccssesssssceccereeee 381/57
`
`Assistant Examiner—Jerome Grant, I
`Attorney, Agent, or Firm—Welsh & Katz, Ltd.
`
`[57]
`
`ABSTRACT
`
`An audio input circuit for providing a transducer output
`signal from a soundtransducerto the input of a host device
`including a universal interface for coupling the transducer
`output signal to the host device. The universal interface
`includes a first signal input, a second signal input, a first
`output conductor, and a second output conductor. Addition-
`ally, the audio input circuit includes a switch mechanism
`coupled between the first and second signal
`input and
`between the first and second output conductors for selec-
`tively isolating the first output conductor from the second
`output conductorandforselectively coupling the first output
`conductor to the second output conductor. The audio input
`circuit conditions the output of an inexpensive microphone
`such that the output response of the microphoneis ideally
`suited to the input requirements of the voice recognition host
`device.
`
`Primary Examiner—Edward L. Coles, Sr.
`
`17 Claims, 4 Drawing Sheets
`
`
`
`||||| |
`
`||
`
`46
`NORMAL
`POSITION
`
`|II|| l
`
`||
`
`s
`
` |
`
`
`
`IgwieLo
`[ie
`=f DEVE
`
`| || |i1
`
`|
`
`Page 1 of 13
`
`GOOGLE EXHIBIT 1006
`
`Page 1 of 13
`
`GOOGLE EXHIBIT 1006
`
`

`

`OL
`
`LSOH
`
`SOIAAG
`
`WSYSAINN
`
`JOVANSLNI
`
`NIVS
`
`U.S. Patent
`
`Oct. 17, 1995
`
`Sheet 1 of 4
`
`5,459,792
`
`a1avisnrav
`
`Y311is/Y3ISalIdAV
`
`Page 2 of 13
`
`Page 2 of 13
`
`

`

`Oct. 17, 1995
`
`Sheet 2 of 4
`
`5,459,792
`
`U.S. Patent
`Ia|rt-f-~4|91PO7]||!P|Le|
`iiga_ta|i|y2il!bb(HOLIMSLada||Lo|:bey2812"2bbo}|Pr|||Op
`ieIAN|fl|o,||NOLLUISOd
`roy\iLiededt|~-|—!!|If1|yos["|ot17
`777A4HI
`jane],||7QP|189|tt|be|'dul
`
` pgp.||OMS4Ne7ast\!o|2szs4H2d|||idTootostid-1odLye+|gus!938103)|Tedsd]|!
`
`1Bral||||]|TWWHON
`
`Page 3 of 13
`
`Page 3 of 13
`
`

`

`5,459,792
`
`FREQUENCY
`
`(Hz) I
`
`Oct. 17, 1995
`
`Sheet 3 of 4
`
`>is
`
`> ©”
`
`n—I
`<Q
`
`US. Patent
`
`Page 4 of 13
`
`Page 4 of 13
`
`

`

`U.S. Patent
`
`Oct. 17, 1995
`
`Sheet 4 of 4
`
`5,459,792
`
`20K
`
`10
`
`IK
`
`500
`
`200
`
`loo
`
`50
`
`
`
`FREQUENCY(Hz)
`
`20 FIG.4
`
`
`
`SIGNALLEVEL(DB)
`
`Page 5 of 13
`
`Page 5 of 13
`
`

`

`5,459,792
`
`1
`AUDIO INPUT CIRCUIT FOR VOICE
`RECOGNITION
`
`BACKGROUND OF THE INVENTION
`
`This invention relates generally to audio inputcircuits for
`use with a sound transducer and more particularly to audio
`inputcircuits for interfacing with host devices such as voice
`recognition host devices. The invention provides a universal
`interface suchthat the audio inputcircuit may be attached to
`a variety of voice recognition host devices through various
`switch settings. Additionally, the audio input circuit condi-
`tions the output of an inexpensive microphone such that the
`output response of the microphoneis ideally suited to the
`input requirements of the voice recognition host device.
`Voice recognition systems are being used with increased
`regularity in industry, the home, and the office. Such systems
`are typically computer based and are either free standing
`units or are incorporated within a computer system. Cur-
`rently, there is no standardized format governing the elec-
`trical or physical connections between an input sound trans-
`ducer, such as a microphone,and differing and various voice
`recognition host devices. Some systems are self-contained
`and include a microphone while other systems connect to an
`externally supplied microphone. Users wishing to attach an
`external microphoneto a voice recognition host device may
`find that each individual system may require a unique
`interconnection format such as a differential input signal or
`a common mode input signal.
`Additionally, there are various configurations for supply-
`ing power to the microphone. Thus, the use of a particular
`voice recognition host device selected from a wide array of
`available systems is cumbersome and inconvenient due to
`the wide variety of input requirements and interconnection
`formats available.
`
`10
`
`15
`
`20
`
`30
`
`35
`
`2
`such microphonesreduces the marketability of many voice
`recognition systems. In addition, such expensive micro-
`phones typically cannot connect directly to more than one
`type of system. On the other hand, inexpensive microphones
`typically attenuate sharply below 500 Hz, and respond
`nonlinearly above 500 Hz exhibiting undesirable variations
`in amplitude. This nonlinear response of inexpensive micro-
`phones impairs voice recognition performance.
`Thus,it is an object of the present invention to provide an
`audio input circuit that substantially overcomes the above
`probiems.
`It is another object of the present invention to provide an
`audio input circuit that universally interfaces to a wide
`variety of voice recognition host devices.
`It is a further object of the present invention to provide an
`audio input circuit
`including a universal
`interface, and
`filtering to enhance the frequency response characteristics of
`an inexpensive microphone.
`It is an additional object of the present invention to
`provide an audio input circuit that includes a center-tap
`transformer for providing an inductive element as part of a
`filtering circuit.
`
`SUMMARY OF THE INVENTION
`
`The audio inputcircuit includes a microphone, an adjust-
`able gain control circuit, an amplifier/filter circuit, a univer-
`sal interface circuit, and a power sensing circuit. A micro-
`phoneoutput signal serves as the input to the adjustable gain
`control circuit and the output of the adjustable gain control
`circuit serves as the input to the amplifier/filter circuit. The
`universal
`interface circuit
`selectively couples
`signals
`between the amplifier/filter circuit and the voice recognition
`host device. The power sensing circuit supplies electrical
`powerto the amplifier/filter circuit and to various compo-
`nents of the audio inputcircuit generally through a plurality
`of wires.
`
`Typically, the host device receives audio signals directly
`from the microphoneorreceives the signals from a micro-
`phonecircuit. Often, a two or three wire connector and a
`plug are used to couple the audio signal to the host device.
`Additionally, the host device may supply an external DC
`bias voltage component ononeor more ofthe wires. The DC
`bias voltage may be used to supply powerto the microphone
`or microphonecircuit, or may unique to the host device. In
`one particular configuration, the external DC bias voltage
`may be referred to as phantom power whenit is used to
`supply electrical power to the audio input circuit
`in a
`balanced, differential mode format. However, the voltage
`level of the DC bias voltage may be inadequate for use by
`the audio input circuit. Hence, there is a need to be able to
`utilize or block the external DC bias voltage supplied by the
`host device.
`
`Page 6 of 13
`
`The universal interface may be configured to couple the
`audio input circuit to many commercially available voice
`recognition host devices by selectively activating various
`switches. Additionally, the audio inputcircuit automatically
`compensates for poor frequency response characteristics of
`an inexpensive microphone. The audio inputcircuit modifies
`the response of such a microphoneto be substantially linear
`in the frequency region most pertinent to voice recognition
`systems. Furthermore,
`low frequency noise components,
`such as ambient room noise and mechanical microphone
`handling noise, which contribute adversely to voice recog-
`nition performance,are substantially attenuated. Thus, voice
`recognition performance of the host system is significantly
`improved.
`Morespecifically, the present audio input circuit provides
`a transducer output signal from a sound transducer to an
`Voice recognition host devices accept input from a micro-
`input of the host device. The audio inputcircuit includes a
`phone or similar device and typically process the input
`universal interface that couples the transducer outputsignal
`signal in accordance with specialized algorithms and signal
`to the host device. The universal interface includesafirst
`processing hardware. However, these systems are typically
`signal input, a second signal input, a first output conductor,
`very sensitive to the frequency response characteristics of
`and a second output conductor. A plurality of switches
`the signal received. If the microphone signal received has a
`coupled between the first and second signal
`inputs and
`nonlinear frequency response, or falls off sharply in a
`between the first and second output conductors are selec-
`particular frequency band, voice recognition performance
`tively activated to selectively isolate the first output con-
`may be degraded. Typically, voice recognition host devices
`ductor from the second output conductor and maybeset to
`perform optimally when the input signal received is essen-
`selectively couple the first output conductor to the second
`tially linear or has a uniform energy spectrum throughoutthe
`output conductor.
`frequency range of approximately 200 Hz to 10,000 Hz.
`The audio inputcircuit also includesa first selectable DC
`Usually, high quality, expensive microphones provide this
`bias voltage blocking circuit for preventing an external DC
`optimal output response. However, the increased cost of
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`Page 6 of 13
`
`

`

`5,459,792
`
`3
`voltage present on the first output conductor from appearing
`on thefirst signal input. Similarly, a second selectable DC
`bias voltage blocking circuit prevents an external DC volt-
`age present on the second output conductor from appearing
`on the second signal input. Additionally, a selectable AC
`coupling circuit couples at least one of the first signal input
`or the second signal input to an electrical return path.
`The adjustable gain control circuit receives the transducer
`output signal and the amplifier/filter circuit receives the
`adjustable gain output for providing an amplified transducer
`outputsignal. Additionally,a filter operatively coupled to the
`amplifier/filter circuit removes low frequency signals from
`the amplified transducer output signal.
`The filter includes resistive, capacitive, and inductive
`elements configured as a tuned circuit and operatively
`couples to the amplifier wherein the inductive element is
`formed by a portion of a primary winding of a transformer.
`The transformer has a secondary winding with a first ter-
`minal operatively coupled to the first signal input, a second
`terminal operatively coupled to the second signal input, and
`a third terminal operatively coupled to an external power
`source. The third terminal of the transformeris a center-tap
`terminal. The amplified output signal, referred to as the
`secondary transducer output signal, is generated across the
`secondary winding of the transformer for connection to the
`universal interface circuit.
`
`The plurality of switches provided by the universal inter-
`face circuit allows the present invention to be connected to
`a variety of host devices. By selectively activating indi-
`vidual switches, the secondary transducer output signal may
`be routed to either the first output conductor, the second
`output conductor, or both. Additionally, by activating the
`appropriate switches, the selectable AC coupling circuit may
`couple eitherthe first signal input or the secondsignal input
`to an electrical return path. Thus, the secondary transducer
`output signal may be referenced to ground as may be
`required by the host device.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a block diagram depicting the major components
`of the audio input circuit in accordance with the invention;
`FIG.2 is a schematic diagram showing one embodiment
`of the audio input circuit in accordance with the invention;
`FIG. 3 is a graph depicting the frequency response of a
`typical inexpensive microphone without use of the present
`audio input circuit;
`FIG. 4 is a graph depicting the enhanced frequency
`response of an inexpensive microphone when coupled with
`the current audio input circuit in accordance with the inven-
`tion.
`
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENTS
`
`Referring now to FIG. 1, a block diagram of the audio
`input circuit 10 is shown. The audio inputcircuit 10 includes
`five basic circuit blocks: a microphone12,an adjustable gain
`control circuit 14, an amplifier/filter circuit 16, a universal
`interface circuit 18, and a power sensing circuit 20. A
`microphone output signal 24 serves as the input to the
`adjustable gain control circuit 14 and an adjusted output
`signal 26 serves as the input to the amplifier/filter circuit 16.
`The universal interface circuit 18 selectively couples signals
`28 and signals 30 between the amplifier/filter circuit 16 and
`the voice recognition host device.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`-
`65
`
`4
`The powersensing circuit 20 supplies electrical power to
`the amplifier/filter circuit 16 and to various components of
`the audio input circuit 10 generally through a plurality of
`wires 32. The power sensing circuit 20 supplies electrical
`powerfrom either an internal battery source 34 or receives
`external electrical power from the host device through the
`universal interface circuit 18. The universal interface circuit
`18 may receive power from the voice recognition host
`device through the plurality of wires 30 and selectively
`routes the power received, to the power sensing circuit 20
`through a plurality of wires 36. Hence, the universal inter-
`face circuit 18 selectively facilitates the transfer of output
`signals from the amplifier/filter circuit 16 to the host device
`and transfers signals from the host device to the power
`sensing circuit 20.
`Referring now to FIG.2, a schematic diagram of the audio
`input circuit 10 as connected to a voice recognition host
`device is generally shown. Such an arrangementfor a voice
`recognition card or voice recognition module may include
`use in an IBM®, Sun®, Macintosh®, Microsoft®, or other
`suitable computer system.
`Microphone
`An electret condenser microphone 12 produces an elec-
`trical transducer output signal 38 proportional to the sound
`pressure levels of an audio input signal 40. The microphone
`12 is preferably an electret unidirectional element
`that
`includes an FET input converter internal to the microphone
`as is well knownin the art. The FET front-end included in
`this type of microphoneserves as an impedance conversion
`stage.
`A load resistor R1 couples the microphone 12 to an
`electrical power node PWR,and functionsas a load resistor
`for the FET-type microphone. However, other microphones
`may be used which do not require the load resistor R1.
`Adjustable Gain Control
`The transducer output signal 38 serves as an input signal
`to the adjustable gain control circuit 14. More particularly,
`the adjustable gain control circuit 14 includes a variable
`resistive element, such as a potentiometer VR,to provide an
`adjustable gain. The adjustable gain control circuit 14
`includes the potentiometer VR in combination with a resis-
`tor R2 where R2 is shown as part of the amplifier/filter
`circuit 16 as will be described in greater detail below. The
`potentiometer VR, and the resistor R2, form a voltage
`divider such that the voltage of the signal received at the
`base of a transistor T1 is varied by changing the resistance
`of the potentiometer VR through manual adjustment of the
`potentiometer. The adjustable gain control circuit 14 allows
`the amplifier/filter circuit 16 to provide between approxi-
`mately 6 DB and 20 DBof gain. Whenthe potentiometer VR
`is set to 0 ohms, maximum gain is achieved, and whenset
`to approximately 10,000 ohms, a minimum gain is achieved.
`Transformer
`
`Acenter-tap transformer TRF provides a primary winding
`PRI, a secondary winding SEC,a primary center-tap termi-
`nal PCT, and a secondary center-tap terminal SCT. The
`inductance of the primary winding PRI is approximately 7
`Henries. A portion of the primary winding PRI provides an
`inductive element wherein an inductance L1 of approxi-
`mately 3.5 Henries is provided between the primary center-
`tap PCT and each terminal P1 and P2 of the primary winding
`PRI. The inductance L1 acts as the inductive element
`included in the amplifier/filter circuit 16 as will be described
`in greater detail below. The center-tap transformer TRF is
`preferably a Mouser Model TM021, however, other suitable
`transformers may be used.
`
`Page 7 of 13
`
`Page 7 of 13
`
`

`

`5
`
`5,459,792
`
`6
`shown,thesignal is attenuated by approximately 10 DB per
`octave below approximately 150 Hz. This attenuation effec-
`tively reduces ambient room noise and microphone handling
`noise. Additionally, the signal is essentially flat or linear
`from approximately 150 Hz through 10,000 Hz. Beyond
`10,000 Hz, the natural response limitations of the micro-
`phone 12 (FIG. 2) cause the signal to attenuate. Thus, the
`signal characteristics of a microphoneas shownin FIG.4 are
`ideally suited for connection to the input of voice recogni-
`tion host devices.
`
`Amplifier/Filter
`The amplifier/filter circuit 16 includes the transistor T1
`arranged in modified common-emitter configuration, as
`shown, to provide amplification. The transistor T1 is pref-
`erably a PNP-type 2N3906 transistor. However, other suit-
`able transistors may be used. The base ofthe transistor Tl
`receives an adjusted gain output signal 42 from the output of
`the adjustable gain control circuit 14. To provide amplifi-
`cation,
`the emitter-base junction of the transistor T1 is
`forward-biased and the base-collector junction is reverse-
`biased as is well known in the art. The emitter of the
`Universal Interface & Power Sensing
`transistor T1 is directly coupled to the electrical power node
`Referring back to FIG. 2, the universal interface circuit 18
`PWR while the collector connects to the first terminal P1 of
`includesa first signal input 50 connectedto the first terminal
`the primary winding PRI. The second terminal P2 of the
`S1 of the secondary winding SEC. Similarly, a second signal
`primary PRI winding connects to ground 46.
`input 52 connectsto the second terminal S2 of the secondary
`The resistor R2 in series with a capacitor C1 couples the
`winding SEC. Therefore, the first signal input 50 is electri-
`base of the transistor T1 with the primary center-tap PCT to
`cally identical to terminal $1 while the second signal input
`produce an RLC circuit combination. Additionally, a resistor
`§2is electrically identical to terminal S2.
`R3 in series with a capacitor C2 couples the base of the
`Whenthefirst signal input 50 and the second signal input
`transistor Tl with the primary center-tap PCT. The primary
`§2 are used as signal inputs to the universal interface circuit
`center-tap PCT couples the collector of the transistor T1
`18, they receive the secondary transducer output signal 48
`with R3 and C1. Thus, the combination of L1, R2, and C1,
`present across the secondary winding SEC. However, the
`provide a tuned resonancefilter.
`first signal input 50 and the second signal input 52 may also
`The adjusted gain outputsignal 42 received by the base of
`function as outputs for the universal interface 18 when an
`the transistor T1 is amplified by the amplifier/filter circuit
`external DC voltage is supplied by the voice recognition
`16. An amplified transducer output signal 44 provided by the
`host device as will be described later. In this situation,
`collectorof the transistor T1 is developed across the primary
`terminals S1 and S2 may receive the external DC voltage
`winding PRI of the transformer TRF. Accordingly, a sec-
`supplied to the universal input circuit 18 from the host
`ondary amplified transducer output signal 48 is induced
`device, through the first signal input 50 and the second
`across the secondary winding SEC of the transformer TRF
`signal input 52.
`and appears between a first
`terminal $1 and a second
`A first selectable DC bias voltage blocking circuit 60
`terminal $2 of the secondary winding SEC.
`couplesthefirst signal input 50 to an interface terminal, such
`The tuned resonancefilter including L1, R2, and C1, in
`as a ring conductor 62 of a mini-plug 63. Similarly, a second
`combination with the transistor T1 act as a low frequency
`35
`selectable DC bias voltage blocking circuit 64 couples the
`filter such that low frequencies, such as those below 150 Hz,
`second signal input 52 to an interface terminal, such asatip
`are attenuated. These frequencies are typical of ambient
`conductor 66 of the mini-plug 63. A grounded interface
`room noise and microphone handling noise.
`terminal servesas a shield terminal for a shield conductor 68
`Thus, the amplifier/filter circuit 16 with the center-tap
`of the mini-plug 63. The shield conductor 68 provides a
`transformer TRF compensates for undesirable transducer
`commonground connection between the audio inputcircuit
`output signal 38 characteristics by providing an electrical
`10 and the voice recognition host device. The mini-plug 63
`inversion of the microphone response characteristics. In
`is a standard molded connector used to connectthe present
`other words, whenthe transducer output signal 38 attenuates
`invention to the voice recognition host device.
`(drops-out), the amplifier/filter circuit 16 provides additional
`Conventional voice recognition host devices may provide
`gain to compensate for the drop-out. When the transducer
`an external DC bias voltage to the tip conductor 66 or to the
`output signal 38 response is elevated, the amplifier/filter
`ring conductor 62. Thus, terminal $1, terminal $2, the first
`circuit 16 provides attenuation. Thus,
`for
`frequencies
`signal input 50, and the second signal input 52 may also
`between approximately 120 Hz and 10,000 Hz, the second-
`receive the external DC bias voltage. The external DC bias
`ary transducer output signal 48 present at terminals $1 and
`voltage may be a voltage used to supply powerto the audio
`§2 is substantially linear. For frequencies above 1000 Hz the
`input circuit 10 and is known as phantom power in one
`natural response of the microphone 12 is substantially linear.
`particular configuration as will be described later. However,
`Frequency Response
`the external DC bias voltage may not be at appropriate
`voltage levels useful to the audio input circuit 10 and may
`Referring now to FIG. 3, a graph shows that the trans-
`require blocking so that the battery 34 can provide the
`ducer output signal 38 (FIG. 2) is highly nonlinear from
`required power. The first and second selectable DC bias
`approximately 40 Hz through 1,000 Hz. Many voice recog-
`voltage blocking circuits 60 and 64 serve to block the
`nition host devices achieve optimal performance when
`external DC bias voltage as will be discussed below in
`ambient room noise and microphone handling noise, typi-
`greater detail.
`cally present between approximately 50 Hz and 150 Hz,are
`sharply attenuated. Optimal performance of the voice rec-
`The secondary winding SEC of the transformer TRF
`ognition host device depends upon the linearity of the
`provides two functions.First, either terminal $1 or S2 ofthe
`transducer output signal 38 between 150 Hz and 10,000 Hz.
`secondary winding SEC mayreceive the external DC bias
`Thus, most voice recognition host devices perform poorly
`voltage through either the tip conductor 66 or the ring
`whena microphone with outputcharacteristics such as those
`conductor 62 depending upon the host device. Second, the
`depicted in FIG. 3 is used.
`secondary transducer output signal 48 present across the
`secondary winding SEC at terminals S1 and S2 is transmit-
`Referring now to FIG.4,a transducer output signal 38 is
`ted to the first and second signal inputs 50 and 52 of the
`shown when conditioned and optimized by the present
`universal interface circuit.
`invention as provided at terminals $1 and S2 (FIG. 2). As
`
`15
`
`20
`
`25
`
`30
`
`40
`
`30
`
`35
`
`60
`
`65
`
`Page 8 of 13
`
`Page 8 of 13
`
`

`

`5,459,792
`
`7
`As shown, the power sensing circuit 20 includes a phan-
`tom diode D1, a battery diode D2, a current limiting resistor
`R4, a double-pole/double (DPDT)switch 72, and the battery
`34. The secondary center-tap SCT of the transformer TRF is
`configured such that the external DC bias voltage applied to
`either terminal S1 or $2 is received by the secondary
`center-tap SCT andis coupled to the cathode of the phantom
`diode D1 through the current-limiting resistor R4. Diodes
`D1 and D2 are preferably a 1N5818, however, any suitable
`diode may be used. When the phantom diode D1 is forward-
`biased by the external DC bias voltage, the external DC bias
`voltage appearing at the electrical power node PWRsupplies
`electrical power to the audio input circuit 10. A noise
`filtering capacitor C5 couples the anode of the phantom
`diode D1 to ground 46 and reduces noise transients and
`ripple appearing on the electrical power node PWR.
`The battery 34 connects to the power node PWR through
`a pole P1 of the DPDT switch 72 and the diode D2. The
`DPDTswitch 72 essentially couples the battery 34 to pole
`P1 since the DPDTswitch 72 is configured such that pole P1
`and a pole P3 are connected when the DPDTswitch 72is in
`a normal position. A pole P2 and a pole P4 of the DPDT
`switch 72 are also connected when the DPDT switch 72 is
`in the normal position. Additionally, in the normal position,
`a pole P5 and a pole P6 are isolated from other poles in the
`DPDTswitch 72.
`
`When the voltage level present at the cathode of the
`phantom diode D1 (due to the external DC bias voltage)is
`greater than approximately 1.5 volts minus the voltage drop
`across D2,the battery diode D2 is reverse-biased and acts as
`an open circuit. Thus, with the DPDT switch 72 in the
`normal position and the battery diode D2 reverse-biased, the
`battery 34 is essentially out of the circuit with respect to the
`powernode PWR,and no current will flow from the battery.
`Therefore, the external DC bias voltage supplied from the
`secondary center-tap SCT, when greater than the voltage
`supplied by the battery 34, forces the battery diode D2 to
`disconnect the battery 34 so that the external DC bias
`voltage may supply all electrical power.
`Similarly, when the external DC bias voltageis either not
`presentoris less than the voltage suppliedby thebattery, the
`battery diode D2 is forward-biased and the phantom diode
`D1is reverse-biased. This allows the battery 34 to supply all
`of the electrical power to components of the audio input
`circuit 10 while preventing the battery current from leaking
`through the phantom diode D1. Thus, D1, D2, R4, and C5
`provide the power sensing circuit 20 for switchably provid-
`ing electrical power from either an internal battery 34 or
`from the externally supplied the external DC bias voltage.
`The DPDT switch 72 in the normal position allows the
`powersensing circuit 20 to connecteither the battery 34 or
`the external DC bias voltage to the power supply node PWR.
`However, when the DPDT switch 72 is in a mute position
`(opposite from the normal position), pole P3 and pole P5 are
`coupled while pole P4 and pole P6 are coupled. Theresistor
`R3 in series with the capacitor C2 couples the base of the
`transistor T1 with the primary center-tap PCT. Poles P4 and
`P6 couple to opposite ends of resistor R3, thus, when the
`poles are connected,the resistor R3 is short-circuited and the
`battery 34 is isolated. When the resistor R3 is short-cir-
`cuited, AC signals are essentially shunted through the
`capacitor C2 suchthat there exists an AC path from the base
`ofthe transistor T1 to the primary center tap PCT. Therefore,
`the transistor T1 operates with near 100% negative feedback
`and gain is reduced to substantially zero, regardless of the
`level of the adjusted gain output signal 42. Thus, no ampli-
`fied transducer output signal 44 is generated across the
`
`10
`
`15
`
`20
`
`25
`
`35
`
`40
`
`45
`
`50
`
`35
`
`60
`
`65
`
`8
`primary winding PRI and no secondary transducer output
`signal 48 exists on terminals S1 and S2. This essentially
`disables the audio input circuit 10 and provides a micro-
`phone mute function while simultaneously saving battery
`power.
`Thefirst selectable DC bias voltage blocking circuit 60
`includes a DC blocking capacitor C3 in parallel with a
`parallel DC bias switch SW1, the combination of which is
`in series with an in-line switch SW2. The secondselectable
`DC bias voltage blocking circuit 64 is similarly formed by
`a DC blocking capacitor C4 in parallel with a parallel DC
`bias switch SW3, the combination of whichis also in series
`with an in-line switch SW4. The selectable DC bias voltage
`blocking circuit 60 prevents an external DC bias voltage
`present on the ring conductor 62 from appearing onthefirst
`signal input 50. Similarly, the second selectable DC bias
`voltage blocking circuit 64 prevents the external DC bias
`voltage present on the tip conductor 66 from appearing on
`the second signal input 52.
`The first selectable DC bias voltage blocking circuit 60
`couples the first signal input 50 to the ring conductor 62,
`while the second selectable DC blocking voltage circuit 64
`similarly couples the second signal
`input 52 to the tip
`conductor 66.
`
`The secondary transducer output signal 48 present at
`terminals $1 and S2 floats with respect to ground reference.
`Thus, the secondary transducer output signal 48 present at
`terminal $1 will be equal in amplitude but opposite in phase
`and polarity to the secondary transducer output signal
`presentat 48 terminal S2. Since the terminal S1 is identical
`to the first signal input 50 and the terminal S2 is identical to
`the second signal input 52, the signal inputs 50 and 52 of the
`universal
`interface circuit 18 also float with respect
`to
`ground. The floating secondary transducer output signal 48
`presentat the first and second signal inputs 50 and 52 may
`be inappropriate for connectionto the voice recognition host
`device in somesituations, depending upon the requirements
`of the host device.
`
`Therefore, a selectable AC coupling circuit 84 provides a
`mechanism from which to form an electrical return path to
`groundto essentially groundthe signal inputs 50 or 52 of the
`universal interface circuit 18 when required by the voice
`recognition host device. A switch SW5 couples the first
`signal input 50 to an AC coupling capacitor C5. Addition-
`ally, a switch SW6 couples the secondsignal input 52 to the
`AC coupling capacitor C5. The selectable AC coupling
`circuit 84 provides two functions. First, by closing either of
`the switches SW5 or SW6, the correspondingsignal input 50
`or 52 is coupled to the AC coupling capacitor C5. This
`essentially creates an AC ground path between the respec-
`tive signal
`inputs 50 and 52 and ground 46. Thus,
`the
`opposite corresponding signal input 50 or 52 is then refer-
`enced to ground 46 as required by the voice recognition host
`device since the AC coupling capacitor essentially appears
`as a short-circuit to ground for AC signals.
`Second,if the signal inputs 50 or 52 were directly coupled
`to ground and the voice recognition host device provided the
`external DC biasvoltage on eitherthe tip conductor 66or the
`ring conductor 62, this bias voltage would essentially couple
`to ground causing an excessive current flow, possibly dam-
`aging the voice recognition device. Thus, the selectable AC
`coupling circuit 84 allows the voice recognition host device
`to provide the external DC bias voltage on either the tip
`conductor 66 or the ring conductor 62 without short-circuit-
`ing such a bias voltage to ground. Additionally, the select-
`able AC coupling circuit 84 provides a mechanism for
`
`Page 9 of 13
`
`Page 9 of 13
`
`

`

`5,459,792
`
`5
`
`10
`9
`terminal S1 across the secondary winding SEC to terminal
`selectively coupling least oneofthe first signal input 50 or
`S2 from appearing on the tip conductor 66. However,this
`the second signal input 52 of the universal interface circuit
`configuration still allows any AC signal to pass from the
`18 to an AC ground path.
`second signal input 52 to the tip conductor 66 since SW4is
`Thus, when the selectable AC coupling circuit 84 is
`closed. The tip conductor 66 transmits the secondary trans-
`:
`.
`:
`operative (SWS5 or SW6is closed), the signals appearing at
`ducer output signal 48 from the second signal input 52to the
`signal inputs 50 and 52 are common modesignals and vary
`voice recognition host device. Switch SWSis closed so that
`:
`.
`:
`:
`:
`from ground or zero volts to a maximum amplitude. When
`the first signal input 50 is AC-coupled to ground 46. Thus,
`the
`selectable AC
`ji
`ircuit 84 is
`not
`tive.
`th
`th