DE19541563C2 - Voice switch (X radio) - Google Patents

Description

The invention relates to a voice switch according to the preamble of claim 1.

A voice switch (more precisely voice recognition switch) of this type is from Pa tentschrift DE 41 27 295 C2 known and has the disadvantage that the illustrated Switch under noise conditions only by means of pronounced voice instructions (Operating instructions) in which at least one pronounced spoken s-spec trum (SS spectrum) is contained, can be activated and therefore, z. B. in one Car, in business or work rooms, only considerably more difficult and therefore not useful is easy to use.

A disadvantage is also that pronounced in English with a sharp "s" Operating instructions, such as "service", in German at the beginning of the word with which "s" be articulated so that such a switch for English and German speakers Users often have to be interpreted differently and thus be economical seeks an unfavorable solution.

Another disadvantage arises from the fact that its evaluation principle is based on the duration of an SS spectrum (operating or operating spectrum) just to be finished to measure certain times, which leads to interference signals (unwanted em pfangene ss spectra), which occur with a high repetition frequency, e.g. B. in broadcasting voice and music programs, fall within these time intervals and wrong scarf effect.

In addition, long-term ss spectra (duration line noise signals) generates an intermittent supply voltage that just if incorrect switching results.

This device has the additional disadvantage that at ambient volumes up to Room volume, in the operating instructions used for this no ss- Spectra preceding vowel or umlaut can be used so that you are free selectable, label-related actuation options remain closed.  

Furthermore, voice switches are known which are used to exclude disturbing surroundings Use directional microphones in the immediate vicinity of the voice switch operator and are not omnidirectional, but only directional actable, d. H. are only of limited user-friendliness.

It is therefore the object of the invention to create an inexpensive voice switch according to the preamble of claim 1, which has the following features:

• - Omnidirectional operability under noise conditions due to unstressed actuation instructions that contain at least one SS spectrum
• - Availability of products to be developed, product-neutral and, regardless of this, be drawing-related operating instructions in which operating spectra in German and English are pronounced immediately
• - Switch position stability when receiving interference signals with a high repetition frequency and continuous wave effect
• - Use of a customizable circuit layout that meets the requirements
• - Use of existing, three-dimensional chip circuit structures for inexpensive manufacture of the voice switch.

This object is achieved by the characterizing features of claim 1. Advantageous further developments of the subject matter of the invention are in the Unteran sayings 2-9 marked,

The invention is described below with the necessary details. It will be shown

• - Functional concept of the voice switch
• - Fig. 1 application example (City Radio) in the form of a block diagram
• - Fig. 2 circuit diagram of the application example (audio amplifier and evaluation logic for this in
• - Fig. 3 shows the basic audio amplifier functions
• - - Fig. 4, the basic evaluation logic functions).

Functional concept of the voice switch

The solution to the central task is that the normally in rela tiv long spoken words, low-energy SS spectra, in short, especially before and after a consonant, comparatively higher frequency An parts and have higher sound power levels. Such ss spectra already pronounced without emphasis, typical ambient noise in living and working spaces penetrate fragments of their duration by an appropriately designed one direction received and evaluated for switching purposes.

An example of this is that under noise conditions for rest or to ge aimed to address internationally used word "psst".

The resulting overall concept with the load-bearing functional components Be operating instructions, operability, frequency selection and operating time measurement How switch position stability is described as follows.

Operating instructions

The task of independent switching functions through short consonant-containing, pro product-neutral and, regardless of this, designation-related operating instructions develop in which fields of activity in German and English Chen has on the one hand to the use of acronyms, d. H. of product neutral Initial forms such as LCX (control system lights X-controlled) or CRX (city or Car radio, X-controlled) and on the other hand for the use of vowels or umlauts in designation-related operating instructions, such as Radio X or lamp X, in which are preceded by an SS spectrum. Beyond that too the use of certain numbers, e.g. B. in combinations such as "Radio X6" possible.

Usability

The operability is reali by means of the audio amplifier and the evaluation logic siert. The operating spectra are amplified in the audio amplifier and here too set the upper limit frequency of their operating bandwidth. Further in the audio amplifier in a filter, amplifier and volume measurement arrangement with variable  adjustable bandwidth, sound frequencies such as vowels and umlauts with the exception the ss spectra, isolated and volume-wise divided into several levels the.

The lower limit frequency of the operating bandwidth is defined in the evaluation logic and to maintain the operability and switching corresponding to the requirements positional stability, at ambient volume levels above room level, in Dependence of the ambient volume measured in the audio amplifier, the operator bandwidth for ss spectra reduced, the set ss spectra duration (operating duration) is shortened and from an ambient over-volume level at which the switch can no longer identify the operating instructions, the operating bandwidth Reduced to zero so that the switch can no longer be operated and in its respective case current switch position persists.

Frequency selection and operating time measurement

The frequency selection and measurement of the operating time is based on that tone half-waves in the evaluation logic are time-analog in an amplitude-limited saw tooth voltages are converted and only ss spectra reduce the saw cause tooth amplitudes, their duration by means of a so-called false pulse detector tors (missing pulse detector, commercially available timer module) measured is that when the predetermined time value is reached, d. H. Reaching an im Block specified voltage, produces a voltage dip at its output, from which a positive control and a negative control pulse are derived. Falling below the specified time value (operating spectrum too short) on the other hand, does not produce a voltage drop.

Switch position stability

The neutralization of interference signals, i.e. H. of ss spectra due to high sequence frequency or continuous wave effect between the operating spectra and during pre coordinated measuring intervals (actuation intervals) are achieved,  that in the evaluation logic after the arrival of a first operating spectrum, the remaining distance to the predetermined next, continuously in the shortest possible time distance is measured by suitable means when a predetermined is exceeded level remaining distance, a generated control pulse that the subsequent time gate chain (actuation interval chain) consisting of monos (timer modules as above) normally triggers, is disabled. That means it will also be between the actuation intervals measured.

This means that ss spectra that are received at the correct time cause them to move between the Operating spectra occurring are accompanied, no switching operation.

In addition, control pulses trigger a monostable multivibrator every time their tipping time sets the operating bandwidth to zero, so that subsequent, random of interference signals or incorrect operation during the actuation interval generated tax impulses, remain without effect.

Fig. 1, application example (City Radio)

The basic operation of the voice switch according to the invention, as an integral part of the city radio using the example of a car radio (car radio), X-ge controls (CRX) is explained below with reference to the block diagram shown in FIG. 1.

To increase traffic safety, the ma Manual button actuation of the car radio for switching on and off, especially for Channel selection and for switching to cassette mode and vice versa by the inventor Functional voice switch as follows.

The sound waves (audio signals) converted into electrical voltages by the microphone (M) shown are amplified via the input ( 1 ) in the audio amplifier and differentiated according to volume in terms of vowel, umlaut or broadband into two level categories.

In a downstream evaluation logic, these audio signals are evaluated, taking into account the determined volume levels with regard to frequency, duration and chronological order, and generate, if they correspond to correspondingly specified values, control pulses that are generated by 3 toggle flip-flops (3 FF) and the the relays (R1-R3) connected to the outputs ( 33 ), ( 34 ) and ( 36 ) can be converted into the following functions.

With the control word

• - "CRX" is the car radio by means of two ss spectra following at a certain distance on and off
• - "X" .. ss is the switchover by means of an extended ss spec trums from station reception to cassette operation (e.g. playback) and vice versa rea lized
• - "X, X is in the position of station reception using two, compared to CRX in shorter distance following ss spectra of each spoken in normal length next (music) station set and in the position cassette mode, z. B switched from Play Deck A to B and vice versa.

Fig. 2, the circuit diagram of the application example (audio amplifier and evaluation logic) Audio amplifier

The voltages coupled out from the microphone (M) via the capacitor (C1) and present at the input ( 1 ) of the audio amplifier pass through a commercially available operational amplifier ( 1 ) over a broadband band, are fed in via ( 48 ), a coupling capacitor in ( 46 ), pass through a DC voltage amplifier (T1-T4), are fed in via ( 43 ), a coupling capacitor in ( 41 ) and are available in broadband with rectangular amplitudes at the collector of the (T5, output of the audio amplifier). The base resistance of the (T5) is used to determine the volume response threshold for SS spectra.

By designing the audio amplifier and determining the selection of the filter capacitor (Cs1) at ( 47 ), a constant amplification of the sound frequencies up to the upper cut-off frequency (oG) of the operating spectrum (approx. 10 kHz) is achieved, which also drops sharply ( Fig . 3, I).

In addition, the sound frequencies pass through the switch (S1), the capacitor (C2), the input ( 3 ), an operational amplifier ( 2 ) with coupling or ( 8 ) and ( 9 ) inserted between ( 6 ) and ( 7 ) or ( 8 ) and ( 9 ) Differentiating capacitor and downstream transistors (T6-T8), a filter, amplifier and volume measurement arrangement. By selecting the low-pass capacitor (Cs2) at ( 8 ), sound frequencies above the lower cut-off frequency (uG) of the operating spectrum, e.g. B. ge from about 8 kHz to ground ( Fig. 3, I), so that ss spectra themselves do not discharge the capacitors (Cp2) and (Cp1) on ( 10 ) and ( 45 ) ( Fig. 3, III ).

The switching elements (OP1-Cs2) connected between ( 3 ) with an upstream capacitor (C2) and ( 8 ) form a filter and amplifier unit, whose high-pass (C2) and low-pass capacitors (Cs2) can be dimensioned so that only individual vowels , Umlauts or a certain sequence thereof are filtered out of the audio signals ( Fig. 3, II).

So are according to the choice of capacitors (Cs1) and (C2), ie by pressing the switches (S1) and (S2), z. B. only o-signals to ( 8 ) are available, which are fed in via a differentiating capacitor (Cd) in ( 9 ) and are subsequently evaluated loudly.

The audio amplifier thus enables the selection of vowels and umlauts, so that in connection with ss spectra also label-related operating instructions can be realized in which the range of activities in English and German are pronounced immediately.

The sound frequencies fed into ( 9 ) via ( 8 ), due to the discharge diodes (D1) and (D2), generate a voltage reduction depending on the volume of the capacitors (Cp1) and (Cp2) charged to different potentials, the next being the Capacitor (Cp1), which is at its highest potential, is discharged and (Cp2) is discharged at even higher ambient volume levels ( Fig. 3, III), so that the transistors (T7) and (T8) are blocked one after the other and via the outputs ( 11 ), ( 12 ), ( 13 ) and ( 42 ) ( Fig. 3, IV and V) with upstream resistors and downstream decoupling diodes, circuits of the evaluation logic with a DC voltage, ie raise the lower operating limit frequency above ( 14 ) , shorten the operating time using ( 15 ) and ( 16 ) and reduce the operating bandwidth to zero using ( 39 ).

The charging time of (Cp1) set by means of (R1) serves to bypass the time, e.g. B. the letter sequence i, k, which arises when the letter X is pronounced and thus to determine the point in time, when charging the (Cp1) negative voltage jump at the collector (T7 or at 11 ), with which a certain difference capacitor via a dif Mono (actuation interval) in the evaluation logic can be triggered.

Evaluation logic

The sound signals routed to the input (based on T1) of the evaluation logic, they generate a control pulse (voltage pulse with positive rising edge) when certain conditions are met at ( 30 ), ( 31 ) and ( 38 ), which the downstream flip-flop via the Controls inputs ( 32 ), ( 35 ) and ( 37 ) and switches the relays connected to their output ( 33 , R1), ( 34 , R2) and ( 36 , R3) (see Fig. 1).

For frequency selection on the capacitor (C1 to 14 ) sawtooth voltages (time constant R1C1) are generated, the maximum amplitude (Umax) of which results from the addition of the base-emitter opening voltages of the transistors (T2) and (T3) and their maximum rise time plus switch-back time , the lower cut-off frequency (uG) of the operating spectrum is determined by FIG. 4, I).

Sound frequencies falling in the operating bandwidth thus produce a reduction the sawtooth amplitudes and blocking the transistors (T2) and (T3).  

The blocking of these transistors causes the blocking of the diodes (D1) and (D2) and thus the release of the charging of the capacitors (C2) and (C3) at ( 15 ) and ( 16 ), so that after reaching a predetermined voltage threshold, the corresponds to a time given differently by the charging time constants (R2C2) or (R3C3), a voltage drop at the respective output ( 3 ) of the fault pulse detector (MD1) or (MD2) arises ( Fig. 4, II), with which differentiation capacitors control and control pulses (positive and negative voltage pulses) are generated.

The duration of the operating spectra in the actuation instructions CRX and X, X is evaluated using (MD1) and that of the operating spectrum in X..ss using (MD2). The extended ss spectra in X..ss also generate control pulses on (MD1) ( Fig. 4, II, output 17 ), which are neutralized with the help of the (M1) function described below.

The components labeled (M2), (M3), (M4) are monos. While the (MD1) triggers the (M2) via a differentiating capacitor between ( 17 ) and ( 20 ), the (M3) becomes different from that of the, the flip-back flank of the (M2) via a differentiating capacitor between ( 22 ) and ( 23 ) generated drive pulse and that (M4) of that generated by (M3) via a differentiating capacitor and switch (S3) between ( 26 ) and ( 27 ). (M2) to (M4) thus form a time gate chain, where (M3) or (M4) is triggered before the arrival of the expected operating spectrum and their tilting duration corresponds to the length of the expected operating spectrum. (M2) is triggered by the start of the ss spectrum in CRX, X, X and X..ss, but its duration is shorter than the time interval resulting from the audio signals spoken between the 1st and 2nd operating spectrum X, X he gives.

Furthermore, the tipping time of the (M3) is dimensioned so that it is shorter than the time distance that is spoken by the between the 1st and 2nd operating spectrum in CRX chenen sound signals.  

If the 2nd operating spectrum generates a switching process in X, X or CRX, it will be the same timed via (R5, T5) or (R6, T5) the (M5) triggered and prevented, as below described that subsequently an operating spectrum sent by a too long or Interference signal generated, another control pulse, takes effect.

The tipped back (M2) cannot be triggered again during the tipping time of (M1), (M3) or (M4) by the blocking diodes (D4), (D6) or (D7) connected to the respective output ( 3 ) are and thus also neutralized, by the discharge diode (D11) connected to its output ( 3 ), the effect of an additional control pulse generated with X..ss on (MD2).

The operating spectrum in X..ss sent with a given duration is therefore only effective during the tilting time of (M2), so that an SS spectrum generated too long with X..ss does not affect the flip-flop ( 3 ) or (R3) can press several times.

A positive voltage flank generated at the output ( 3 ) of the (MD1) or at ( 17 ) during the flip-over time of (M3) or (M4) is applied to the inputs ( 25 ) and ( 29 ) as a control pulse via differentiating capacitors and, due to the blocking diodes (D8) or (D9) blocked during the tipping time of (M3) or (M4), via coupling diodes at ( 30 ) or ( 31 ) or at the control inputs ( 32 ) or ( 35 ) of the flip-flop ( 1 ) or ( 2 ), as described at the beginning of this chapter, are effective.

If the ambient volume exceeds the room volume, the volume measuring arrangement in the audio amplifier uses the blocked transistor (T7) and a coupling diode on ( 14 ) of the evaluation logic to set a DC voltage potential that shifts the lower limit frequency of the operating bandwidth upwards by a certain amount, thereby increasing the operating bandwidth reduced and the energy input lowers, so that the switch position stability conditions are maintained. Under these volume conditions (T7) also switches to inputs ( 15 ) and ( 16 ) to maintain operability, which causes a shortening of the originally defined operating time, so that the voice switch can still be operated with ss spectra pronounced without emphasis. If you choose z. B. to improve the memorability of operating instructions, instead of a product-neutral initial form such as CRX, a designation-related such as Radio X and tunes the filter, amplifier and volume measurement arrangement in the audio amplifier, as described, to "o", so the ( 11 ) after the end of the spoken "o", a negative voltage jump across a differentiating capacitor and by pressing the switch (S3) on ( 27 ) (ie Auftren voltage of the capacitor connection shown in ( 26 )) and the voice switch with the spoken below X of the radio X operating instructions, instead of CRX.

In order to prevent a first operating spectrum or interference signals with a high repetition frequency that are sent for too long from triggering the monos (M3 and M4), a mono (M1) is introduced, which generates a difference with the first one from (MD1) to ( 17 ) Decorative capacitor fed into ( 18 ), positive voltage pulse and triggered by the switching transistor (T4), thereby releasing the charging of the charging capacitor (C4) connected to ( 19 ) and switching its output voltage to (M1, pin 3) at maximum level ( Fig. 4, III), during this time by the resistance (R4) and upstream decoupling diode (D3), the charging time constant (R3C3) of the (MD1) was considerably reduced, the number of the ( 3 ) of the (MD1) and 17 he testified Voltage dips increased accordingly, thereby interrupting the charging of the (C4) again and again ( Fig. 4, III) and, when receiving too long transmitted operating spectra or interference signals, via ( 23 ) to trigger the (M3) via a differentiating capacitor Control pulse fed into ( 23 ) is neutralized by means of the blocking diode (D5), furthermore during this time, further triggering of the (M2) is prevented via the blocking diode (D4) and, in the absence of operating spectra, the (M1) according to the specified one Time, ie when it reaches its internally determined switching threshold (F), falls back into its stable position and switches its output voltage back to ground-to-ground ( Fig. 4, III).

A first operating spectrum or interference signal with a high repetition frequency that is sent for too long thus extends the tilt-back time of the (M1) according to its duration and thus prevents the mono (M3) and, depending on it, (M4) from being triggered. If only the ss spectra defined in the operating instructions are received, the resulting control pulses fed into ( 25 ) and ( 29 ) are triggered via resistors (R5) or (R6), a coupling diode and the transistor (T5), the mono ( M5), which by means of the resistor (R7) connected to its output ( 3 ) and downstream decoupling diode, reduces the operating bandwidth to zero, so that during its tipping time, which is greater than the time interval to the next actuation interval, one by one operating spectrum sent too long or the control signal generated by the interference signal cannot take effect.

Furthermore, the (M5) is put into its astable position over time at a considerable over-room volume in ( 39 ), and only falls back to its stable position after falling below this volume level after a specified time.

The (M5) thus prevents any operation when the room volume is high Availability of the device according to the invention and after actuation, for a specified time, another switching.

Claims (9)

1. Voice switch, e.g. B. for a radio, a lamp, a phone, a camera, its switching on, off, switching or triggering by a voice instruction, consisting of one or more words, in which or at least one be spoken sharply spoken s spectrum (SS spectrum) is included, up to ambient volume levels can be achieved in stable switch position with a microphone, downstream audio amplifier and downstream evaluation logic, which generates a pulse under certain conditions, which the above devices over a connected in series with an energy source Re controls relay or semiconductor switch, characterized in that the executed in microtechnology and z. B. integrated into a radio device in German and English, up to ambient volumes in excess volume by means of an unstressed, short voice instruction, in which at least one SS spectrum is preceded or followed by a consonant, such as. B. by means of the acronym CRX, which is fed through a microphone M into the input 1 of the downstream audio amplifier, at the output 43 the ss spectra generated corresponding voltages, which via a coupling capacitor, the input 41 , the output transistor T5 to the input of the Evaluation logic are routed and with the correct sequence and the correct duration, switch the relay R1 and thus the radio on and off at their output 33 , in the switched-on state of the radio, the voltages corresponding to the SS spectra generated by means of the voice instruction X, X. , actuate relay R3 at output 36 of the evaluation logic and switch from one transmitter to the next, and in addition operate the voltages corresponding to the ss spectra generated by means of the extended spoken voice instruction X..ss and operate relay R2 with the correct duration at output 34 of the evaluation logic and switch the radio from station reception to cassette mode and vice versa, continue in cassette mode ch the voice instruction X, X in turn is pressed R3 and now switches between playdecks A and B, furthermore the operability through a step-by-step reduction in the measurement duration of the SS spectra depending on the ambient volume by reducing the charging time constant R2C2 at output 15 and R3C3 at output 16 of the evaluation logic by connecting the resistors in parallel, which are connected upstream of the output 12 and 11 of the audio amplifier with a downstream decoupling diode, and under these volume conditions, maintaining the switching position stability by raising the lower operating frequency limit by reducing the RC filter R1C1 at output 14 of the evaluator Telogik by parallel connection of the resistor which is connected upstream of the output 13 of the audio amplifier with a downstream decoupling diode, is also achieved, interference signals that occur between the SS spectra, generate corresponding control pulses, which, via the input 18 of the evaluator ogik to the switching transistor T4, interrupt the charging of the capacitor C4 of the mono M1 at the output 19 , cause a correspondingly prolonged voltage at the output 3 of the M1, which prevents the triggering of the mono M3 via the downstream blocking diode D5 and which by means of the Coupling capacitor between output 22 and input 23 control pulse, neutralize, continue to actuate up to ambient volumes in room volume by voice instructions, in which the ss spectrum is preceded by at least one vowel or umlaut, which is achieved in the with the switches S1 and S2 variably adjustable filter, amplifier and volume measurement arrangement of the audio amplifier at output 11 generates a corresponding control pulse, which is conducted via a coupling capacitor connected in series, when switch S3 is actuated, in the evaluation logic to input 27 , which triggers the mono M4 for a predetermined time , so that de r generated during this time by means of an SS spectrum at the output 3 of the MD1 or output 17 , via a differentiating capacitor fed into the input 29 control pulse through which from the blocking diode M9 blocked during the tilting time from M4 to the output 31 and which Actuation of the relay R1 at the output 33 causes, in addition, at considerable ambient room volumes, at which the device no longer recognizes the voice instruction, the transistor T8 in the audio amplifier is blocked by the voltage reduction at the capacitor Cp2 at the output 10 , on the output 42 with an upstream Decoupling diode and resistor switches a DC voltage, which is passed through a galvanic connection to the input 39 of the evaluation logic and via the switching transistor T5 with an upstream Basisdi ode, the Mono M5 during this time in its unstable position and connected to its output 3 by means of Resistor R7 with upstream decoupling gsdiode, the charging time of the filter capacitor C1 at the output 14 is shortened so that ss spectra do not cause any voltage subtraction on it, so that the operating band is reduced to zero, so that the device can no longer be operated and remains in its respective switching position. 2. Voice switch according to claim 1, characterized in that for the sound frequencies fed into input 1 of the audio amplifier, by means of a downstream operational amplifier 1 , via the coupling capacitor inserted between output 48 and input 46 and downstream DC voltage amplifier T1-T4, one at output 43 broadband, constant gain up to the upper limit frequency of the ss spectrum, which is determined by the filter capacitor Cs1 at the output 47 , arises and, moreover, the gain drops sharply. 3. Voice switch according to claim 2, characterized in that in the audio amplifier the circuit between input 3 and output 8 by adding an operational amplifier 2 and inserted between the output 6 and input 7 coupling capacitor with a downstream amplifier stage T6, to egg ner filter and amplifier unit , whose high-pass capacitor C2 connected upstream of input 3 is reduced when the switch S1 connected upstream is actuated, or the low-pass capacitor Cs2 at output 8 is enlarged by actuating switch S2, so that it is below the lower limit frequency of the selected vowel or umlauts or above its upper limit frequency and act below the lower cut-off frequency of the ss spectrum. 4. Voice switch according to claim 3, characterized in that in the audio amplifier the circuit between the input 9 with upstream tem differentiating capacitor Cd and the outputs 11 , 12 , 13 and 42 , by means of the integration circuits R1Cp1 at the output 45 and R2Cp2 at the output 10 and the discharge diodes D1 and D2 or downstream transistors T7 and T8 to a volume measuring arrangement is designed. 5. Voice switch according to claim 4, characterized in that the positive amplitudes at the input of the evaluation logic at the base of the switching transistor T1, at the RC filter R1C1 connected to its collector and output 14 , insofar as they are below the lower limit frequency of the SS Spectrum lie, the maximum voltage amplitude and above this limit generate a voltage subtraction that the transistors T2 and T3 and the diodes D1 and D2 block and thus release the charging of the capacitors C2 and C3 of the false pulse detectors MD1 and MD2 at the outputs 15 and 16 . 6. Voice switch according to claim 5, characterized in that in the evaluation logic ss-spectra, at the connected to the outputs 15 and 16 charging capacitors C2 and C3 generate a voltage that short-circuit when exceeding a threshold defined in the fault pulse detector MD1 and MD2 is reduced, corresponds to the predetermined operating time and cause a voltage drop at MD1 and MD2, whose positive edge at outputs 17 and 38 with downstream differentiating capacitors are converted into control pulses, while its negative edge at MD1 and output 17 by means of downstream differentiating capacitor at output 20 as Control pulse triggers the Mono M2. 7. Voice switch according to claim 6, characterized in that the triggered in evaluation logic mono M2 for tilting back into its stable position, via the output 22 with a downstream differentiating capacitor transition in the A 23 supplies a drive pulse, which triggers the mono M3 at Zurückkip pen via the output 26 with a downstream differentiating capacitor and an unactuated switch S3 puts an actuation pulse on the input 27 which triggers the mono M4, so that during the tipping time of M3 or M4 by means of the blocked diodes D8 or D9 connected to the inputs 25 or 29 fed control pulses via coupling diodes at the outputs 30 or 31 switch the flip-flop 2 or 1 connected to it, continue to trigger the M2 during the tipping time of M3 or M4 via the blocking diodes D6 or D7 and during the tipping time of M4 via the blocking diode D10, which is prevented by the Mono M1. 8. Voice switch according to claim 7, characterized in that in the evaluation logic, during the astable state of the mono M1, by means of the resistor R4 and upstream decoupling diode D3, the charging time constant R3C3 at the output 16 is considerably reduced, the number of the fault pulse detector MD1 at the output 17th generated voltage drops correspondingly increased and thus, upon receipt of operating spectra sent too long, the control pulse fed into the input 23 via a differentiating capacitor for triggering the mono M3 is neutralized by means of the blocking diode D5, as when receiving interference signals, and also via the blocking diode D4 further triggering of the Mono M2 is prevented and if SS spectra remain off, the M1 falls back into its stable position after a specified time and its voltage at output 3 switches to ground potential. 9. Voice switch according to claim 8, characterized in that in the evaluation logic, which is fed via the inputs 25 or 29 , during the tilting times of mono M3 or M4 control pulses, the mono M5 via the resistors R5 or R6, a downstream coupling diode, trigger a switching transistor T5, during the flip-over time of M5, which is greater than the time from the next actuation interval, set the operating bandwidth to zero via resistor R7 with a downstream decoupling diode, and prevent an ss -Spectrum generated control pulse subsequently takes effect.