US1832431A - Electric wave transmission system - Google Patents

Description

Nov. 17, 1931. L..l J. slvlAN 1,832,431

ELECTRIC WAVE TRANSMISSION SYSTEM Filed Aug. 29. 1930 TERM/NAL A FIG l I reen/NM a 3 sou n G -so F G. 2

w M g couuEnc/AL (ass/r .sm/vo) rnA/vsmrren k una PEA/r vou-4c: 2 q CURVE B' AVERAGE rorAL voLucs '3 E l., o I33oa Q 2 -lo x Y :zot u Q ,5 cunvc A: Loss m PEAK val. uns EQUAL/zen i lo- 'a' Q 'noz Y m c I I I l I I I I l l I I l l l l I I I l I zo so noo soo |,ooo 5.ooo

FREQUENCY, C YCLLZS` PER J'ECOND FIG. 3

TERM/NAL A' rERM/NAL a' I l 33 /2\ /3 I4 l5] 9( I I I l [Ti-gi S rm/uw/rrsn nece/vsn l? b i z u F/G. 4 $1; canori/:ER rRANsu/rren S s m S Q E s 2s. E c t l l l L zo so 50o |,ooo 5,000

|00 FREQUENCY, CYCLES PER SECOND [A rro /vEr Patented Nov. 17, 1931 y UNITED STATES PATENT ori-ICE LEON J'. SIVIAN, OF EAST ORANGE, NEW JERSY, ASSIGNOR TO BELL TELEPHONE LABORATORIES, INCORPORATED, OE' NEW YORK, N. Y., A CORPORATION 0F NEW YORK ELECTRIC `WAVE TRANSMISSON SYSTEM Application filed August 29, 1930. Serial No. 478,654.

This invention relates to Wave transmission systems, as for example, telephone systems.

An object of the invention is to decrease deleterious effects of disturbing energy.

Signals must often be transmitted through a noise or intereference Zone, that is, through ,a region or through apparatus in which noise .region were attenuated sufficiently to reduce the signal energy level at the receiver to the desired value and to correspondingly reduce the noise energy level at the receiver. By this method, if the signal amplification (and therefore the margin of energy in favor of the signal) be sufficiently great, the noise energy can be reduced to a negligibly small value Without eflacing the signal. That is, the effect of the disturbance can be made negligibleif sufficient energy is used at the sending end of the system and sufficient uniform or distortionless attenuation at thereceiving end.

However, a limitation on the practice of such method is that restrictions are. imposed von the sending energy. In many systems a sending end amplifier or other transmitting apparatus 1s operated at its maximum permissible load (for the shapes of the waves Vit is transmitting) regardless of disturbing 'energy that may enter or be set up in the system between the `sending end and the receiving end. In such cases, if greater amplication of each frequency component be attempted, it is at the expense of distortion, and the system is then said to have become overloaded. That is, upon the increase of the power of each frequency component at the sending end. the amplifying apparatus at that end and in many cases other overloaded apparatus in the' system) itself introduces distortion or noise. This distortion limits the advantage to be gained fromincreasing the sending end power of every frequency component with the given apparatus. In other words it is of'no avail to reduce the disturbing eects of a giveninterfering energy if in doing so the Vsystem becomes so overloaded that, for example, because of harmonic generation (and other distortion frequenqy or distortion component generation), the signal suffers a degradation that more than offsets the advantage of the reduction of the effects of the given interfering energy.

However, it has also been pointed out heretofore (Hartley Patent V1,7 87,843, December 3, 1929) that, even when the load limit of the sending apparatus has been'reached (for the shapes of the Waves it is transmitting), the deleterious effects of noise upon the received signal can in many or most cases still be reduced, by changing the wave shapes. The'frequency distribution of signal energy or of time-averaged signal power sent from a transmitting'station generally is not uniform. F or instance, as indicated by the curve showing the energy` frequency 4 distribution of average speech, page 7 9 of H. Fletchers Speech and Hearing, D. Van Nostrand Company, New York, 1929, 'the frequency spectrum of the energy of normal speech is not uniform; nor is it uniform after modification by the microphone or other sending apparatus. To the contrary, the time averaged value of the power in speech is greater for some frequencies than for others. That is, the frequency spectrum of the signal contains frequencies of comparatively low energy level and frequencies of comparatively high energy level. (A high energy level occurring at any particular frequency region does not necessarily implyhigh sound intensities inthe region, but may result from either a large number of speech sounds having low intensities in the region or a few sounds having components of high intensities inthe region.) Moreover, it isusually the case that the frequency distribution of energy of the noise is different from that of the signal, and that when the frequency distribution of the noise energy is not uniform over the utilized frequency range, the maxihaving mum value of single frequency time averaged power or of single frequency energy occurs at one frequency for the signal and at a different frequency for the noise. A method heretofore proposed for so changing the wave shape of signals as to increase the amount that noise effects can be reduced, is to change the frequency distribution of the signal energy (as for example, speech energy) before the signal is sent through the sending end amplifier or other apparatus that otherwise might be too greatly overloaded.

According to this method heretofore proposed, this change in the energy-frequency relations in the speech before transmission may be effected by the use of a network having a transmission-frequency characteristic such that the network discriminatcs against the transmission of components in the frequency regions of certain signal energy levels and in favor of the components in the frequency regions of certain other signal energy levels. For example, it has been proposed that under certain conditions the frequency zones discriminated againstbe those `in which the interference or noise energy to which the line is subjected is small and those frequency zones which the network favors be those in which the interference energy is relatively high. A. network at the receiver a characteristic complementary to that of the transmitter network then restores the speech to its original energy distribution and at the same time attenuates the noise. lt is found in practice, that an increase in intelligibility by the use of such complementary networks can be effected if the networks are given the proper characteristics. nstead of making the networks at the transmitter and receiver exactly complementary, they may be so designed as to compensate for distortion in the system such as that due to unequal attenuation of the different signal frequencies. For example if desired the overall system attenuation obtained by adding the individual attenuations of the transmitting and receiving terminal apparatus to that of the line, and considering amplification as a negative attenuation, may be made the same for all frequencies in the entire frequency range which it 1s desired to transmit. The response-frequency characteristic or loss-frequency che` acteristic of the transmitter or of some other element or elements of the system wit-hout the receiving equalizer may correct or neutralize to a considerable extent the distortion introduced by the sending equalizer, and in some cases no receiving equalizer is needed. A particular proposal heretofore has been that, when the signal, before entering the interference zone, passes through vacuum tube amplifiers or other apparatus having a limited load carrying capacity, the predistortion of the signal (that is, its distortion before it is transmitted to the interference zone, or to the amplifiers or the other apparatus just mentioned) be made such to obtain approximately the same energy level for all of its frequencies, in order to bring the energy level of each of its frequencies up to the maximum uniform energy level for the apparatus included in the system. (It has also been proposed, that to override a noise having a narrow frequency range, the predistortion of the signal be made such as to provide a particularly high signal energy gain for the interfering frequency range.)

As indicated above, then, it has already been proposed to override interference by increasing the sending energy of the signal, and moreover', make the increase selective with respect to frequency so that overloading will be less than were the sending energy merely amplified uniformly over the utilized frequency' range; and as also indicated above, it has heretofore been particularly proposed to give the sending energy for speech transmission a distribution uniform with respect to frequency (or in other words to equalize the sending energy to a flat energy-frequency charzu'teristic), when the limiting factor dctermining` the sending energy is overloading of amplifiers or other apparatus.

in accordance with the present invention instantaneous maximum amplitudes (i. e. peak amplitudes) at different frequencies in the output voltage of a (sustained) speech or other .signal wave as delivered (by a subscribers circuit, for example,) to a predistorting network or equalizer, are equalized to a substantially uniform value (or other desired relative values) before the speech or signal is sent through amplifiers, or other apparatus susceptible to overloading, into an interference or noise zone; or where from the standpoint of providing` apparatus overload capacity it is nneconomical to provide for all peaks, such equalization of the peak amplitudes is apprrmched to some pres .fihed degree, as for example by so equalizing a chosen percentage, for instance 75% of all of the peak powers occurring in a. succession of short time periods. In the case of speech, for example, the transmission capacity required to take care of all of the peak powers is much larger than that required to take care of the lower 75% of the peak powers occurring in a succession of short time periods (for instance periods of /Q second duration).

rl`he equalization of the peak amplitudes or powers, rather than equalization of long,F time average values of power or energv. is highly advantageous in many cases. Most types of telephone apparatus as now used are relatively indifferent to this longaverage. ln general, it is maximum or high instantaneous values of voltage or power that cause lill lio

overloading' of lsuch apparatus and consequent limiting values of signal distortion.

' This is the case, for example, for vacuum tube ainpliiiers. Moreover, it has been found that in the case of normal speech applied to a high quality telephone transmitter or to a com- "m-ercial subset transmitter for instance, the "characteristic of peak amplitudes vs. frequency in the output is markedly different in form from the characteristic of long time average values of energy vs. frequency. I '-In'receiving signals transmitted over the y'exposed region (i. e. interference region) in general there is employed a rcceiving'equalizer network having characteristic of trans- 4' mission loss or etliciency vs. frequency such as to 'compensate for the response-frequency or loss-frequency characteristic of any or all of the other elements'of the system, as desired. For example, to give the system the of the sending equalizer and the exposed portion combined; or to compensate for the loss-frequency characteristic of the sending equalizer, the exposed portion and the telephone receiver, the receiving lequalizer has a-.loss varying with frequency in the inverse manner of the sum of the losses in those elements; or to give the system a flat air-to-air characteristic, the receiving equalizer has a transmission-frequency characteristic which is the inverse of that of the remainder of the system, i. e. a transmission-frequency characterist-ic such that the sending and acoustic correction is removed and the 'rest of the system from the sending end to the listener is corrected. In some cases no receiving equalizcr is needed, particularly when the transyinitter or any portion of the system gives 'considerable signal distortion which tends to neutralize the signal distortion introduced bythesending end equalizer.

ln one specific aspect the invention is a radio telephonesystem in which an equal- -iziiig Vnetwork is connected between atelcphone circuit and a radio transmitter, and the equaliziiig network is designed to have a. given loss versus frequency characteristic such as to change the amplitude frequency distribution of normal speech waves received by it to an amplitude frequency distribution in which the peak amplitudes, (or anyxchosen fraction of the .peak amplitudes occurring in a succession of small time periods, which are lower than the remaining peak amplitudes).

have equal values throughout the utilized frequency spectrum. The signal level ini- `pressed on the radio transmitter is such as .initting antenna are received by an antenna which feeds a radio receiver including demodulating or detecting means that yields the speech signal. The detected signal is transmitted to a receiver, through a receiving equalizer network if desired.` The receiving` equalizer, if used, can have any suitable characteristic as indicated above. The principle of the invention is not limited to application to radio transmission systems nor to transmission of speech signals, but is of general application to transmission systems and signals of various types as for example loaded submarine cable systems.

Other objects and aspects of the invention will be apparent from the following description and claims.

Fig. 1 shows schematically a voice irequency telephone system, embodying a forni Y of the invention Fig. 2 shows curves for facilitating explanation of the form of the invention shown in Fig. 1; Y

Fig. 3 shows a radio telephone system embodving 'a second form ofthe invention; and

Fig. 4 shows curves for facilitating explanation of the latter 'form of the invention.

In Fig. l at terminal A, a telephone transmitter 1, which is the usual transmitter of a commercial deskstand subset, represents a transmitting source of speech frequency signals. The output of the transinitter'l is fed through the subscribers loop to a predistorting equalizing network 3. The output of network 3 is fed to amplifier 4 which in turn is connected with line circuit 5, having therein a plurality of amplifiers such as Gand 7, the purpose of which is to'maintain the level of the transmitted signal Well above the noise level of the line, thereby rendering the signal free from interference produced b v the noise. The line circuit 5 is connected at the'ieceiving terminal B with amplifier 8, which in turn is connected with a corrective network or receiving equalizer 9. The output of the network 9 is fed to receiver 11. The system has been shown as arranged for oneway signaling, to simplify descriptionV of the invention; but corresponding:V two-way signaling can be effected by well-known adaptations.

Curve A of Fig. Qshows approximately` the relative losses required in network' at `the different frequencies in order to prespeech is impressed on the transmitter, to a fiat or uniform peak voltage-frequency characteristic of speech signals for delivery to the amplifier 4. This curve'A is derived from the staircase curve B of Fig. s. which is plotted from measurements yof peak voltage amplitudes of speech signals delivered by .a subset transmitter such as 1 in atypical Vcommercial subscribers circuit when normal speech is impressed onthetransmitter,l 'In making these` measurements, the output of the subscribers circuit was divided into frequency bands by, filters and the peak voltage amplitudes for vshort periods of time were measured for each frequency band. For the curve B, each of the time periods was of one second duration, and thus the peak value in each selected frequency band corresponds to the maximum voltage occurring in the selected band during-one second of sustained speech. Simultaneously with thesev peak measurements, thevalue of the integrated tot-als of voltages in one half (positive half or negative half) yof the ventire speech wave (for the whole frequency range, all of the bands) for fifteen seconds of sustained speech, also were observed. The curve B is plotted between frequency,.in cycles per second on a logarithmic'scale, and band peak voltage divided' by average total voltage, in decibels. Bv average total voltage is meant the timeintegrated voltage of the total frequency range divided'by the total time of the integration. l

The curve B is-a composite'cuiwe for four male voices, and is for'the Western Electric #337 deskstand transmitter. For each caller approximately sixty peak values were observed in each band. All this was then repeat-ed for the next band, and so on until the entire range had been covered. The curve A is drawn somewhat arbitrarily as a smooth curve corresponding to the staircase curve B. the smooth curve including approximately the same area below as above the staircase curve.

The transmission characteristic or lossfrequency characteristic of network 9 is not shown.. as it may be of any suitable character, as indicated above. For example, it'may be complementary to that of network 3 i. e. the inverse-of that of network 3)' in case'it is not desired to have the network 9 compensate for distortion in portions of the svstem other than network 3, or it may be designed to take care also of the distortion introduced 'in any or all of the other portions of the system` notably the receiver. In some cases the network 9 is not needed. and mav be omitted, particularly when the transmitter'l or another portion of the system gives considerable signal distortion which tends to neutralize the signal distortion introduced b v equalizer 3.

It willbe obvious that the line circuit 5 may, if desired, be one of the channels of. a carrier telephone system, including the necessary apparatus for modulation, demodulation. etc.

In Fig. 3 transmitter 1 at terminal A reprcscntsl a source of speech frequency signals connected with a predistorting network 3, which in turn is connected with a radio frequency transmitter 12. In this case, the telephone transmitter is assumed -to "be of the high-quality type, as for example a condenser transmitter, producing substantially no distortion in transforming the air waves comprising its input into electrical variations. The output of the radio transmitter is connected with the antenna 13, which radiates the high frequency signal to a receiving antenna 14 associated with terminal B. The noise level through which the radio transmission course occurs is ordinarily fairly uniform over the utilized radio frequency range. At station B the antenna 14 is connected with a radio receiver 15, which detects the signal and transmits it through a corrective or equalizing network 9 to receiver l1.

vNetwork 3 may have the loss-frequency characteristic A shown in Fig. 4 which is of such form that when normal speech signals are impressed on transmitter 1 the peak amplitudes of the network output voltage will be substantially equal for all the transmitted frequencies. This curve A is derived from the staircase curve B of Fig. 4 as curve A is derived from curve B. Curves A and B are curves for a distortionless transmitter which correspond to the curves A and B respec-tively for the transmitter of Fig. 1. The curve B is plotted from measurements of peak voltage amplitudes of speech signals delivered by a transmitter such as 1 when normal speech is impresse-d on the transmitter. These measurements were made as described above in connection with curve B, except that curve B is a composite curve for tive male voices instead of four, and is for the Western Electric 394; type condenser transmitter alone, rather than in a subscribers circuit of a commercial type.

Corresponding to case of network 9, the transmission-frequency characteristic or lossfrequency characteristic of network 9 may be of any suitable type, as for example, complementary to that of the predistorting network 3, or such as to take care also of the distortion in any or all of the other portions of the system. In some cases the network 9 is not needed, and may be omitted.

lf desired the characteristics of the predistorting networks such as 3 and 3, instead of being such as to render the maximum peak amplitudes of the signals uniform or of any other predetermined relative values over the utilized frequency range, can be such as to give such uniform or other predetermined rela-tive values to any chosen percentage of Y the peak amplitudes occurring in a succession of short time periods, which are lower than the remaining peak amplitudes. For example, the loss-frequency characteristic of equalizer 3 instead of being derived from curve B can be correspondingly derived from the solid line curve of the lower half of Fig. 1li of my paper on Speech power and its measurement, published in the Bell System Technical Journal, October 1929.

The latter curve is a composite curve for three male voices.V It shows the relative values of the lower of thepeakamplitudes, for periods of one eight second duration, of signal pressure (or voltage) in selected frequency bands of the utilized frequency range, for normal speech. Coinijiarisonyof that curve with the solid line curve in the upper 'half of the same figure, whichA is the corresponding curve of relative values vof thek maximum peaks in the `various bands, shows how much larger the transmission capacity of the system must beto take care of the highest 25% of the peaks. For the peak amplitudes equalized, the amplifiers 4, 6, 7 and@ in Fig. l and some portion of the radio transmitter in Fig. 3, for example a power amplilier employed as usual to raise the power of the radio frequency signal waves beforeapplying themr to the antenna 1,3, operate at their maximum input amplitude capacity, i. e. at the maximum input amplitudes that they can handle without producing substantial signal distortion as a result of overloading. The networks 3 and Svmay include amplifiers for raising the power level of the predistorted signal, if desired. If desired the characteristic of the predistorting networks such as 3 and 3 may be such as to give a fiat or other desired shape of peak amplitude versus frequency characteristic fora chosen type of signals'other than speech. For example, in the case of transmission or reproduction of pianor music, the network 3 mayhave as its loss-frequency characteristic a smooth curve corresponding in form to the staircase curve A or B of Fig. l5 of my paper, mentioned above, in the same manner in which curve A. corresponds in form to curve B. The curves A and B of that paper give peak amplitude-frequency characteristics for sounds representing a piano selection .instead of for speech, but otherwise correspond` respectively Vto the curves of the upper and lower halves of 14 of that paper.

In practicing the invention, in general the predistorting networlrp (such as 3 or 3') changes the peak amplitude-frequency characteristic of the power delivered to the network in such manner as to cause the peak powers of the networkv output to have any predetermined frequency distribution or frequency characteristic, as for` example, to

- cause the peak powers of the network output to have their maximum permissible values at the various utilized frequencies. For instance, overloading limits are in general` a function offrequency and the network may change the frequency distribution of the peak powers of the given signal delivered to it in such manner as to cause apparatus subj ect t0 overloading to be operated at its maximum permissible values of `peak powers (for the given output ofthe predistortng network) at the various utilized frequencies; or, in other words, the predistorting network may change the given signal delivered to it in such manner that the peak power-frequency char-y acteristic ofthe output of the predistorting network conforms to (i. e., is Vsubstantially the same as) the peak power-frequency overload characteristic (for the given output of thepredistorting network) of the apparatus subject to overloading. Although the peak powerfrequency overload characteristic of the apparatus subject to overloading depends to some extent upon the form of the peak power-frequency characteristic of the output to that apparatus, for practical purposes ordinarilythis dependence can be neglected.

lt is desired to emphasize the fact that, as indicated above, the predetermined frequen-f cy distribution or frequency characteristic which it is desired to give the peak powers of the output of the predistorting network is not in all cases flat. The desired characteristic may be determined by limiting factors other than overloading in the most restricted sense of that term. For example, when the limiting factor iscross-talk into other circuits the desired characteristic ordinarily is not fiat, but .slopes in many cases in a way to indicate that higher values of peak power can be transmitted at the low frequencies in the utilized frequency range than at thev higher frequencies. In the case of eddy current shielding, however, the slope usually is the reverse. Moreover, even when the maximum permissible values of peak power at the various utilized frequencies in the output of the predistorting network are determined by overloading in a more restricted sense, the desired peak power versus frequency characteristicof the network output may be nonuniform with respect to frequency (i. e. may be not fiat) because the peak power versus frequency overload characteristic of the apparatus subject to overloading is not in all' cases fiat. `For example, in the case of a radio transmitter such as that of Fig. 3, the impedance into which/the power amplifier subject to overloading works may vary with frequency, and then the maximum output power that the amplifier can deliver without introducing objectionable distortion is less for some frequencies than for others. Consequently, the output of the network such as 3 should in'that case have a peak power versus frequency characteristic conforming to the non-uniform peak power versus frequency overload characteristic of the power amplifier. By way of further example, if desired the peak power versus frequency4 overload characteristic `of the ,apparatus subject to overloading can be considered to be such as Ato take into account the variation with frequency in disturbing eect of a given peak value ofasingle frequency disturbing power (such as` modulation power). In other words, the peak powers of interference, (such as modulation products) relative to the peak speech powers may be made to vary considerably over the frequency range. Moreover,

y characteristic required to give the network output any predetermined characteristic of peak power versus frequency will be apparent from the network transmission characteristie (for example A. or A) required to give the network output a flatcharacteristic of f peak power versus frequency; for the network transmission characteristic required to produce any given output characteristic of form other than flatis obtained by algebraically subtracting at each frequency from the network transmission characteristic required for the flat output characteristic the difference between the given output characteristic and the flat output characteristic.

It is true that, as indicated above, in general, the predistorting network such as 8 or 3 is given a transmission characteristic such that its output has a predetermined peak power versus frequency characteristic for a given peak power versus frequency characteristic of its input. That is, it is true that, in general, the transmission characteristic of the network is chosen to take into account both (l) the original peak power versus frequency characteristic of a given type of undistorted signal (speech, for example) and (2) the transmission characteristic of the portion of the system ahead of the network (for example, the response-frequency charv acteristic of the transmitter and circuit which feeds the network). lowever, in the case of a system such as that of Fig. 3, employing the condenser transmitter l', the latter characteristic is substanti ally flat, so that it has little effect upon the transmission characteristie required for the predistorting network relative to the corresponding effect in the case of a system such as that of Fig. l. On the other hand, the latter effect may be of major importance and, therefore, if desired the frequency distribution of peak power in the original undistorted signal can be neglected (regarding as uniform with respect to frequency) in determining the transmission characteristics of the predistorting network. To do this the difference between the characteristics A and A may be taken to give the (transmitting) effective peak response-frequency characteristic of the commercial subscribers circuit (i. e. the relative responses of the subscribers circuit to equal input amplitudes at the different frequencies in the utilized frequency range). Then, in the system using the commercial transmitter, as for example the system of Fig. l, the sending end equalizer can be given a transmission-frequency characteristic complementary to the (transmitting) response-frequency characteristic of the subscribers circuit, to equalize or compensate for merely the (transmitter) distortion of the subscribers circuit instead of for both that distortion and the peak amplitude versus frequency characteristic of the speech or other signal. Such choice of the transmission-frequency characteristic of the sending end equalizer might be warranted especially in case the same transmitter is to be used for transmitting types of signals differing considerably in the frequency distribution of their peak powers. The compensation for the subscribers circuit may be made only partly, if desired, since the part compensation may be more effective than complete compensation in case, for example,

the line noise in the exposed portion of the system is highest at the frequencies at which the transmitter characteristic is highest. )Vhen the sending equalizer is made to compensate for the characteristic of only the subscribers circuit, and not for the peak amplitude versus frequency characteristic of the signal, then no receiving end equalization for the sending end equalizer is needed, since the sending end equalizer has then merely compensated for signal distortion.

While the invention has been disclosed as embodied in particular forms, it is capable of embodiment in other forms without departing from the spirit and scope of the appended claims.

)What is claimed is:

l. The method which comprises changing the differences between the maximum instantaneous powers at the dierent frequencies, in a signal of substantial time duration, by amounts such as to give the maximum instantaneous powers of the signal at the different frequencies magnitudes having predetermined relative values.

Q. The method which comprises changing the differences between the maximum instantaneous powers at the different frequencies in a signal of substantial time duration by amounts such as to give the signal a predetermined characteristic of peak power versus frequency more nearly fiat than the corresponding characteristic of the signal before said change.

3. The method of operating on a signal in a signaling system including signal transmitting apparatus that has overloading limits in generale function of frequency, which comprises changing the peak amplitudes of the signal at its different frequencies to the maximum values that can be supplied to the aparatus without exceeding its overloading imits.

4. The method which comprises reducing the dierences between the maximum instantaneous powers at the different frequencies, in a signal of substantial time duration, by such amount as to render the relative values of said powers of the different frequencies substantially equal.

5. The method of reducing deleterious effects of noise on a signal of substantial time duration in a signaling system including signal transmitting apparat-us subject to overload primarily by instantaneous amplitude values of the signal, which comprises reducing the differences between the maximum instantaneous values of signal power that occur respectively in a number of mutually exclusive frequency bands of the signal which together constitute the signal, by amounts such as to render the relative values of the band peak signal powers substantially equal, and transmitting the resulting form of the signal through said apparatus to a noise zone, whereby said apparatus is loaded substantially equally by said different frequency bands of the signal passing into the noise zone.

6. In combination. a signal source, signal transmitting means subject to overloading by the signal from said source, and signal distorting means connectedv between said source and said signal transmitting means, said signal distorting means having transmission eiiciency independent of the amplitude of waves impressed thereon but varying with frequency in such manner as to change the differences between the peak amplitudes of the signal at its different frequencies by predetermined relative amounts.

7. A speech signal transmitting system comprising a portion subject to overloading and a corrective network preceding said portion in the system, said network having its characteristic of transmission eihciency versus frequency such as to change the original amplitude-frequency distribution of the signal into one in which the peak speech powers have predetermined relative values at the dierent frequencies.

8. A speech signal transmitting system comprising a portion subject to overloading and a corrective network precedingsaid portion in the system, said network having its characteristic of transmission efficiency versus frequency such as to change the original amplitude-frequency distribution of the signal into one in which the peak speech powers at the various frequencies have predetermined relative values and to impress the changed speech signal on said portion of the system at the maximum power level consistent with distortionless transmission of the signal.

9. A signal-transmitting system comprising a radlo transmitter subject to overloading and a corrective network preceding said radio transmitter in the system, said network having its characteristic of transmission efficiency versus frequency such as to change the original amplitude-frequency dis tribution of the signal into one in which the signal peak powers at the different frequencies have values that difer from each other by prescribed relative amounts.

l0. A signal transmitting system comprising a radio transmitter subject to overloading and a corrective network preceding said radio transmitter in the system, said network having its characteristic of transmission efficiency versus frequency such as to change the original amplitude-frequency distribution of the signal into one in which the peak signal powers at the different frequencies have equal values throughout the utilized frequency range and' toimpress the changed form of the signal on said radio transmitter at the maximum power level consistent with distortionless transmission of the signal by said radio transmitter.

ll. The method which comprises changing to predetermined relative values the differences between a chosen fraction of the peak powers of a signal at its various frequencies which occur in a succession of short time periods and which are lower than the remaining peak powers occurring in those periods.

19,. The method of operating on a signal in a signalling system including signal transmitting apparatus that has overloading limits in general a function of frequency, which comprises changing to the maximum magnitudes that can be supplied to the apparatus without exceeding its overloading limits a chosen fraction of the peak powers of a number of mutually exclusive frequency bands of the signal which together constitute the signal, which peak powers occur in a succession of short time periods and are lower than the remaining peak powers occurring in those periods.-

In witness whereof, I hereunto subscribe my name this 28 day of August, 1930.

LEON J. SIVIAN.

Cil

Images