US2806173A - Signal amplitude limiting circuits - Google Patents

Description

Sept. 10, 1957 D. E. SUNSTEIN 2,806,173

SIGNAL AMPLITUDE LIMITING CIRCUITS Filed Jan. 9, 195] f/GWAL IOUKC! 3/ [/GNAA IOU/QC! 27 4 T 1- A- 7 A? A2 INVENTOR;

ited States Patent ()fiice SIGNAL AMPLITUDE LIMITING CIRCUITS David E. Sunstein, Cynwyd, Pa., assignor to Philco Corporation, Philadelphia, Pa., a corporation of Pennsyl- Vania Application January 9, 1951, Serial No. 205,106

3 Claims. Cl. 315-12 The present invention relates broadly to circuits for limiting the amplitude of electrical signals. More particularly, it relates to signal transducers adapted to be supplied with time varying signals and operative to confinethe maximum amplitude excursions of these signals between predetermined limits. It is a featured characteristic of such transducers that the aforesaid limits remain fixedly spaced from a given signal reference level even though this reference level may be changing, provided, only, that the period of the reference level changes is longer than the longest signal period. This differs radically from prior art amplitude limiting devices, such as overloading amplifiers and similar non-linear devices, whose limiting levels were fixed with respect to a constant reference level. V

The drawbacks of this prior art inflexibility of limiting levels, as well as the advantages of my novel circuit as hereinbefore characterized will become more readily apparent by reference to a system in which limiting circuits find frequent application.

In voice communication systems, for example, it is often desired to obtain more efiicient utilization of transmitter power than it is possible to attain by simply modulating a carrier wave with the periodic signals representative of speech information. It has been found that this efiiciency can be greatly enhanced by reducing the speech signals to a succession of square waves in which the discontinuities correspond to crossings of a zero signal reference level in the original speech signal. The square waves so produced were utilized to initiate pulses marking the instants at which the square wave changed from one amplitude to another, these pulses being transmitted by carrier and reconverted into square waves at the re ceiver.

It was found that the intelligibility of speech information represented by the reconstituted square waves depended principally upon the precise preservation of the frequency characteristic of the original speech signal, as

represented by the discontinuities of the square waves. These square waves, in turn, were heretofore derived from the original speech signal by the conventional process of confining the amplitude excursions of the signal between limits which were small compared to the unconfined amplitude of the signal, thereby obtaining signals having steep amplitude discontinuities or, in efliect, square waves. The reference level about which it was desired to carry out thelimiting process was the average value of the original signal. As long as this average value remained constant, prior art limiting circuits were well able to transform signal variations about this average value into square, waves of corresponding frequency. Unfortunately, signal components representing timber,

inflection and loudness of speech, which vary at rates Patented Sept. 10, 1957 applied to a prior art limiting circuit, the latter would still operate to limit their amplitude excursions to a nar row range about a constant reference level. As a result, many rapid amplitude variations represenative of speech information would fall completely outside the. limiting range and would not only fail to produce the desired corresponding square waves, but would not appear in the limiter output at all. During such intervals the needed signal frequency information would be lost and the intelligibility of the reproduced signal greatly impaired, if not altogether destroyed.

This difiiculty is not encountered when the limiting circuit is constructed in accordance with the invention, for its limiting levels will then automatically adjust themselves so as to maintain their predetermined spacing from the varying average signal value. Consequently, each speech signal variation will extend into the limiting range and produce the desired square wave indication of signal frequency.

It is, accordingly, a principal object of my invention to provide a limiting circuit adapted to confine the maximum amplitude excursions of time varying signals between limits which are fixedly spaced from a variable signal reference level. It is another object of the invention to provide a limiting circuit operative to confine the maximum amplitude excursions of time-varying signals between limits which remain fixedly spaced from a time-varying signal reference level, provided only the reference level varies more slowly than the signal amplitude.

It is still another object of the invention to provide an amplitude limiting circuit adapted to be supplied with a composite signal which includes high frequency and average value components, said circuit being operative to confine the amplitude excursions of the high frequency components between limits which are small compared to the average unlimited amplitude excursions and which are fixedly spaced from the envelope of the average value components.

It is a still further object of the invention to provide a limiting circuit operative to confine signal amplitude excursions between predetermined limits without producing severe distortion of the transmitted portion of the signal waveform.

To achieve the stated objects of the invention, as well as others which will appear, the signals to be limited, including their average value component, are utilized to deflect the beam of a cathode ray tube. This cathode ray tube is equipped with a screen which, in conjunction with apparatus external to the tube, produces an output that varies with beam deflection over a central boundary region of the screen, while being constant over the distal regions of the screen. From this output signal, the average value component is derived and utilized to deflect the electron beam in opposite sense to the deflection produced by the originally applied signals. By so doing, the deflecting effect of the average value component of the original signals is nullified, the beam deflection produced by the applied signals will always be centered about this average value and so will the amplitude excursions of the output signal. However, applied signal excursions which are of suflicient amplitude to deflect the beam beyond the variable output boundary region and into a region where it produces constant output will, of course, produce no additional change in this output so that the latter will remain at its constant maximum value until the deflection again decreased to the point where the beam returns to impingement upon the boundary region where it produces variable output. Thus, signal amplitude excursions from the average signal value will be reproduced in the output signal derived from the cathode raytube only gl n}; ine u e a ta et H 'ng the desired 'iEEl' liifiitffig' 'ffcfl The specific manner of construction and Operation of srrst ne amen ties the Pa e a het ia e re lent: n ed is expla ned the f subsequent gle; ile d descr pt on F 1, Q

eih a ireui fiee been briefiy' d esl t -e, s. zefet th i r vs: seriee s a eleth sie r Ore detailed reference may eqanet te t n ae ele atin a qde P I a Pair ff Qeflee:

.. Li! l I: rg-1 us: pt: V a nd 4 a heamntereent xe a e 2;? a, c on n'ected throilgh a resistor 17 to a source of post ye 122E? .r i ei n .vee leete e ee c 1.8 R "stor 19 connects avrnig N SP tive'lY sl si i a e y refe e e i 1 i an 5 a n lan a o t w i h d in N i d a ammatically rep e-v t Leslie-it eeai ted Que of hese target sections, as for example- ,15 is made of} a 'contfi'ssien ratio whi Whi e h ade of a conduc e material but h ess th n un y t H dial the d ferentiation i to e nt secondary veinjs ion ratios ne alf o 11 si stwi n granulated carbon on its beam confr o ng side.

uncgategl part Then 1f t e electron plane of sections elettens r aehing th to the total. number leavin exactly equal j tur. 01R the= invention, the rapidity with thlrst'orin'g potential appears i611 plate" depends on the 'time constantiofthe R-C networjk 'c'ons"isting"of-resistor 19 and a eeze Entire tar et may be con:

95! 951 .19 9 entran will 7 a r h e a. b am t te in eten a win 'ey'e'ntuallyappear on plate I4\yhicl '1'will compensate 4 V rter f'ie"2:"5'phTzi't oi- 'ztififr i6" diE- charge it, as 'the case may Be, to a potential equal to that of target 15, 16, which latter varies instantaneously in 5 response to variations in beam impingement position.

Assume now that the contemplated signal of time varying amplitude is applied't'o'ifipfit deflection plate 13 from rapidly yarying comp nent an; to 'the liig'li f uency initial a n et nef this si na th b m wi h flected, from its normally centerd position on the target, 15 a two signal components. 'As the beam departs from its e n l x on e other hand, co taut, d sgrep u a and ij 9f releeireli a X l electron arrival and pie This th ce iitial region Bfiyarying eleg tron emission intensity, hereinbefore briefly descrili edifn ay' he seen-to gig;

- t ei ht '1 dimens on 9f he of deflection: I l

i en s lzefiueen and' nm fl n t9 flee i2n 9f th b am eh 'delitional 9 t e al e s m of th ima e-p 1 th -tar et ,5 16, 11 t n fel owte. e e tial variati ns (the sea n thi t tnee i selec ed the def the lane speseh t nder 351 etwhi e. @11 er ct th V quently the e 'ee 915 01 f equen y Qqmppnents'applied V t9, deflection; plate: '13. .will be completely counteracted by correspondingpotentials, developed onfplate 14 so that thexwilkproduce n netbeamfdefiectiong to 7 their: instantaneous, amplitude iexcursions about men average Yalue component, and thes defiections twill give rise to prop ortidna'te 7o 7 output sinalg 'iyithin amplitude fore, always pro the central region ofvariable emission intensity.

' his then possible to's'et maximum arid-'rrii' riirnir'm-linii-ts V 75 for the speech frequency camponent amplitude excursions this timerconstant'is, the l signal QQPIQ 21- Assume u t e that h s isna h a components of speech information and a'jmore? slowly V in proportion td theiristantaneous algebraic sum of the e? ant is e theesteet termina i me Ya a etea- Pi d uetentiel e- 8 ee em e its Qt the app ie etcrmined by the time constant githe t eak 1 2 2 as heteinbefore'in eated, n ac t a t.

V Nopsnch connteraetion'is proyideid for. deflections produced by the speech frequency signalcomponents'i The latter will thereforeiproduce' beamdeflections proportional limits" defined. S eechfrequency-componentswill; there I 1 V lice'heam"dflectionsyyhicli extend without fear of .having some of these variations lost because of the superimposed deflecting effect of the aforesaid average value component.

In any specific case these limits are determined by the deflection sensitivity of the cathode ray tube and by the cross-sectional dimensions of the beam at its point of impingement on the target. As is well known, deflection sensitivity is defined as the ratio of actual beam deflection, at the target, to the deflection potential applied to the deflecting means. For any given beam impingement cross-section, a deflection equal to one half the beam dimension measured in the direction of deflection, will cause the beam to reach its limit of useful deflection, so far as the ability of the tube to reproduce input signal changes is concerned. If now the deflection sensitivity ofthe tube is high, a small change in input signal will drive the beam past this limit, so that the limits of transmission of the circuit are very narrow. As the deflection sensitivity is decreased, increasing values of signal fluctuations are required to drive the beam beyond the variable emission intensity region of the target and the transmission limits of the circuit increase.

Since cathode ray tubes are ordinarily designed to have a predetermined fixed deflection sensitivity, with resultant inflexibility of the transmission limits, it is often desirable to make the beam dimensions variable. For, if the crosssectional beam dimension measured in a direction parallel to the deflection plane is increased, then a greater signal variation will be required to deflect it completely out of the variable response region of the target, with the result th at the limits of amplitude transmission will be expanded. Fortunately, the beam size may be readily adjusted by simple electronic focussing controls, as for example by varying the value of the potential A applied to anode 12 of Figure 1.

Note that only the cross-sectional dimension measured parallel to the deflection plane of the beam affects the amplitude limits of the circuit. The cross-sectional dimension of the beam measured perpendicular to the deflection plane, on the other hand, determines the total current variation produced by a given beam deflection between the aforesaid limits and, consequently, determines the absolute amplitude of the limited output signal.

It remains to point out that, for perfect reproduction of input signal wave form, within the limits of transmissivity of the device, the beam should be rectangular in cross-section, so that the increase and decrease of output potential will be exactly proportional to the instantaneous beam deflection only. Should the beam have any other cross-sectional configuration, as, for example, the circular one encountered in most conventional cathode ray tubes, then a given large deflection from the normal centered beam position still within the variable output range will give a proportionately smaller change in output potential than a small deflection or, in other words, the output signal will be non-linearly related to the input signal in the region near the amplitude limits. The limiting action of the circuit will, of course, remain unimpaired.

It is apparent, from the foregoing discussion, that the limiting circuit illustrated in Figure 1 requires a specially constructed cathode ray tube, which is principally characterized by the provision of the unusual divided target plate 15, 16. It is well known that such custom built structures are very expensive and, if the cost of the circuit is to be held at a practical minimum, it will be preferable to use a conventional cathode ray tube in place of the special one illustrated in Figure l. A limiting circuit embodying all the concepts of my invention and using a cathode ray tube which is conventional in every respect is illustrated in Figure 2, to which more detailed reference may now be had.

The limiting circuit pictured therein features a cathode ray tube 23 which may be of any commercially available type. For example, it may include an electron emissive cathode 24, a beam intensity control grid 25, first and second focusing and accelerating anodes, respectively designated by reference numerals 26 and 27, and a conventional set of deflection plates 28, consisting of four individual deflection plates of which two are disposed in a vertical plane on either side of the electron beam 29 while the other two are placed in a horizontal plane on either side of the same electron beam. Of these latter, the upper plate 30 is connected to the output of a signal source 31 which latter may be similar, in every respect to signal source 21 of Figure l. The lower deflection plate 30a, on the other hand, is connected to the junction of a capacitor 32 and a resistor 33 which are arranged to form an R-C network similar to that consisting of resistor 19 and capacitor 20 of Figure 1. As is usual with cathode ray tubes, the one shown in Figure 2 is further provided with a fluorescent screen 34 deposited on the inside of the tube face. Conventional operating potentials are supplied to the various elements of the cathode ray tube. Thus, cathode 24 is connected to a suitable source of negative potential A-, the first anode 26 may be connected to a suitable source of positive potential A while the second anode 27 is connected to a suitable source of positive potential A++. Note, however, that of the four deflection plates 28, only one pair is externally connected, that being pair 30, 30a which is adapted to deflect the beam in a vertical plane upon application thereto of suitable deflection potentials. The pair of electrodes which is adapted for horizontal beam deflection is unconnected.

Note further, that beam intensity control grid 25 is connected to a point of constant potential, such as ground, thereby assuring the emission of an electron beam of constant intensity. Thus there is provided a beam of constant intensity, adapted to be deflected in a single plane.

Disposed in confronting relation to the face of the cathode ray tube and its fluorescent screen 34 is a photoelectric cell 35 which is thus adapted to intercept light emitted from the fluorescent screen and to produce internal electron emission whose intensity is proportional to the total amount of light falling on the photoelectric cell by virtue of its illumination by the fluorescent screen. The collector electrode 36 of the photoelectric cell is connected to a suitable source of positive potential B++, while the emitter electrode 37 is connected to a suitable source of lower positive potential B+, through a resistor 38. The emitter electrode 37 is further connected directly to the output terminal of the limiting circuit as a Whole, and to deflection plate 30a through resistor 33. This embodiment of my device is completed by the provision of an opaque barrier 39 disposed intermediate the lower half of fluorescent screen 34 and photoelectric cell 35. The upper edge of this barrier may be seen to bisect the fluorescent screen 34 in a horizontal plane, thereby preventing any light emitted by the lower half of the screen from reaching photoelectric cell 35.

Fundamentally, the operation of the device illustrated in Figure 2 is similar to that of the device shown in Figure l and described in connection therewith.

Thus, fluorescent screen 34, barrier 39 and photo cell 35 may be considered, for purposes of analysis, as oooperating to function as a beam interceptive member for the cathode ray tube with an over-all response characteristic similar to that of target 15, 16. Specifically, beam impingement on a part of screen 34 which is in full view of photoelectric cell 35 produces maximum illumination thereof and consequent maximum electron emission from the cell. On the other hand, beam impingement on an area of the screen which is entirely masked by barrier 39 causes no illumination of the cell and minimum electron emission therefrom. For intermediate impingement posi- Y tions of the beam, varying fractions of the luminous spot which it produces on the screen will be visible to the photoelectric cell depending on what fraction of the spot is hidden behind the opaque barrier. Therefore, such intermediate beam impingement positions will produce tential thereacross.

, 7 intgrmediate intensities of electron emission from the f' ijhere exists then again, as in the embodiment of Fig Lire l, a central region on the beam interceptive member which produceselectron emission intensities which are variable depending on beam impingement position, this region extending on either side of the top edge of barrier 39 to a height equal to the maximum cross sectional dimension of the beam measured in the direction of de'-- flection. This central region is again flanked by distal regions within which variations in beam impingement position produce no change in the electron emission intensity of the beam intercepting device.

As was the Case in the embodiment of Figure 1, itis again desired to have the undefiected beam centered within the afo'rcdescribed central region of varying electron emissivity. This will obtain when the luminous spot producedby the beam is bisected by the upper edge of barrier 39 and is thus half visible to, and half masked from the photoelectric cell 35. The cell will then be receiving one-half 'ofgits maximum illumination and will be producing one-half of the electron emission which it would produce in response to illumination by the entire spot. This electron emission will cause current flow through and potential drop across resistor 38 so that the potential on deflection plate 30a will be greater than the potential B+. By adjusting the source of potential B+ so that the sum of its potential and of the potential drop across resistor 38 during illumination of the cell by one-half the cathode ray spot is just equal to whatever potential is applied to deflection plate 30 in the absence of signals to be limited, this central beam position will be undisturbed, as there is no potential difference between plates 30 and 30a which would tend to deflect the beam.

It is now apparent, by analogy with the'embodiment of Figure 1, that application of the contemplated time varying signal from source 31 to deflection plate 30 will initially produce a beam deflection, from its normally cen- V tered position on the screen, which is proportional to the instantaneous amplitude of the signal. As the beam, and the luminous spot which it produces, depart from their centered position, the ratio of spot area which illuminates the photoelectric cell to spot area masked by barrier 39 will change, so will the total illumination of the photoelectric cell and, with it, the electron emission of the emitter electrode 37. In turn, this will change the total cell current flowing throughiresistor 38 and also the po- This change will be substantially proportional to the change indeflection signal applied to plate 30 until the beam is so far deflected from its centered position that it impinges entirely either on the masked or on the unmasked portion of the tube screen 34. Any additional change in applied deflection signal which would tend to increase the beam deflection beyond either of these points will, of course, have no additional efiect on the illumination of the photoelectric cell and consequently no further elfect on the potential drop across resistor 38. 7

As a result, there appears at the output terminal of the device, which is directly connected to the junction of photoelectric cell 35 and resistor 38, an output potential which varies in proportion to variations in the signal applied.v to deflection plate 30 between maximum and minimum values corresponding to maximum and minimum photocell illuminations.

Initially, the beam deflection will thus be instantaneous- 1y, proportional to the signal applied from source 31. As this signal is reproduced by potential variations across resistor 38, the potential across capacitor 32 will. tend to follow 'thesevariations, not instantaneously, butwith a rapidity determined, by the time constant of, the R-C. network 32,33; The operation of this network is identical to the corresponding'R-C' network 19., 20 of Figure 7 1 performingexactly-- the same function in exactly the' same way. Namely, suitable-choiceof its time constant enables the potentialacrosst'the capacitor 32 to fellciw variations in the low-frequency or average componentpf the reproduced signal which appears at the joutputjerminals, thereby vproviding a deflection potential for plate 30a which counteracts any deflecting influence exertedby the same averagecomponent of the original signal when applied to deflection plate 30. l r

Once again, deflections produced by thehigh frequency signal components are not counteracted and aretherefore efiective to deflect the beam into the limiting regions of the interceptive device. 7 I i y The same design principles which were presented in the description of Figure 1 also govern the setting of the actu al limits between which applied signal excursiohs are reproduced by the device of Figure 2. Thus, the signal amplitude limits are again a function'of the deflection sensitivity of the cathode ray tube, as well as of its beani' cross-section. Naturally, since most conventional cathode ray tubes feature a beaming having more or -less cii cular cross-section, it will not ordinarily be feasiblefto obtain the perfectly linear relation betwcen applied and reproduced signals within the signal transmission limits, which is characteristic of beams having rectangular "crosssection as was the case with the specially built'tubejof Figure l.- However, this limitation will not impair-the operation of the circuit to produce simple square waves from signals of fluctuating amplitude. V Y

Naturally, various modifications of the apparatus hereinbefore described will occur to those skilled in mean, 1

For example, it may on occasion be desirable to have the signal amplitude limits asymmetrically disposed with re? This can 'be done "very readily by suitably adjusting the B+ potential applied spect to the zero signal level.

to the beam interceptive means of my apparatus setha't the undeflected beam will no longer be centered within its region of varying electron emission intensity. Furthermore, my inventive concept contemplates :the' use of beam deflection systems other than those shown, such as conventional electromagnetic ones, for example, as Well as amplification, if necessary,- of the average potential fed back to the counteracting deflection plate.

whenever the corresponding component of output potential'happens to be too small to provide the desired counteraction of applied signal.

Accordingly, I desire my inventive concept to be limited only by the appended claims.

I claim: v V

l. A signal limiter comprising: a source of an electron beam; means responsive to impingement by said beam to produce an output which varies with the deflec-' V V tion-of said beam from one side to the other ofta predetermined boundary, said means comprising a first region on one side of said boundary responsive-to beam impingement thereon to produce a positive output proportional to the fraction of the electron beam impingement and lowerfrequencytb'ands to deflect said beam between" said regions and across said boundary; means for selecting, fromsaid combined outputs, frequency components in said lower frequency bandgand means responsive only: to saidsel'e'cted frequency components to deflect s'aid beam in a direction opposite to the directionoft deflec tion'producedl'by' said input signal.

2'. A signal'limiter comprising; a source of an electron beam; a. screenhaving two regions of different secondf ary electron emissivity, each of said regions extendingin a direction transverse to their common boundary an amount which is substantially greater than the maximum value of the projections of the cross-sectional dimensions of said beam upon an axis extending in the same said direction; means for directing said beam toward said screen; means responsive to electron beam impingement upon said screen to produce an output proportional to the algebraic sum of electrons arriving at said screen due to beam impingement and electrons leaving said screen due to secondary emission; means responsive to an input signal comprising components in upper and lower frequency bands for deflecting said beam from one of said screen regions to the other; means for selecting, from said output, frequency components in said lower frequency band; and mean responsive only to said selected frequency components to deflect said beam in a direction opposite to the deflection produced by said input signal.

3. Apparatus according to claim 2 and characterized in that the secondary emission ratios of said screen regions are greater than unity and less than unity, respectively.

References Cited in the file of this patent UNITED STATES PATENTS Schlesinger Apr. 19, Schlesinger Dec. 26, Ziebolz Mar. 16, Sears Mar. 18, Sears Jan. 11, Haynes Feb. 22, Hecht Mar. 8, Huifman Dec. 13, Szikai Feb. 7, Dome June 6, Sunstein Sept. 19, Hufiman Nov. 21, Van Overbeek et a1. Apr. 8, Muller Mar. 3, Nicoll Mar. 10,

Images